Re: Reversing a Power Transformer

[shiva]

"Chuck Harris" > wrote in message
...
> Now dim, you know the statute of limitations applies!
>
> I had to get the E-I laminations from someplace, you
> didn't expect me to cut them out of GOSS with a pair
> of snips, did you? That stuff is hard to cut.
>
> -Chuck
>

Oldtimers... Now, why didn't you just use a plasma cutter, like you did on
Patrick's roof?
-dim


> Shiva wrote:
> > "Chuck Harris" > wrote in message
> > ...
> >
> >>Hi Ian,
> >>
> >>I have no axe to grind, and I don't shout and squirm. If you
> >>feel that I do, perhaps it is your own prejudices at play. When
> >>I make mistakes, I admit to them, and try to correct them... if
> >>possible.
> >>
> >>I have read the whole thread on "Reversing a Power Transformer",
> >>from beginning to end, and I must have missed the point where it
> >>was stated (or implied) that core size is dependent on voltage.
> >>If I, or others, have stated that, it was an error, and needs to
> >>be corrected.
> >>
> >>In so far as I can recall, I have never believed that voltage
> >>determines core size. This isn't a recent revelation, I designed
> >>and constructed my first power transformer around 1974. It was a
> >>whopper that delivered 20V at 30A.
> >
> >
> > Just around that time someone stole my big battery charger, throwing the
> > carcass back, minus the transformer. Chuck? Anything to say?
> >
> > -dim
> >
> >
>

[BH]

William Sommerwerck wrote:
>>The transformer will probably hum loudly and get hot, and perhaps
>>burst into flames after a while. Eventually, a fuse, if fitted, will blow,
>>and when the insulation burns off the wires, a short will blow the
>>mains fuse. This is the result of saturation.
>
>
> But (as Red Green points out) that would at least as entertaining as your
> average "Matlock" episode. Probably more so.

And as Red Green would say, the handy man's secret weapon, duct
tape will fix it. Speaking of duct tape, I have a nephew who works
for 3M in Minnesota, he says they are coming out with a clear duct
tape. RG should really be excited about that.

Bill H.

[Patrick Turner]

Casino wrote:

> Hi All,
>
> I have a step-down power transformer with a 120-volt primary and
> 60-volt center tapped (30-0-30) secondary. I was wondering if I could
> reverse the transformer by plugging-in the 60-volt secondary into the
> AC outlet and use it as a step-up for 240-volts? Theoretically, it is
> possible, but I'm not sure if it's actually safe to do so.

DO NOT DO THIS.

The transformer will probably hum loudly, and get hot, and perhaps burst
into flame
after awhile. eventually, a thermal fus, if fitted, will blow, and
when the insulation burns off the wires, a short will blow a mains fuse.
This is the result of saturation..

>
>
> Also, when I travel Europe, can the 120-volt primary be plugged into
> 240-volts to supply the 120 (117) for North American appliances.
> Thanks,

ABSOLUTELY NOT!

Patrick Turner.

>
>
> C.W.

[William Sommerwerck]

> The transformer will probably hum loudly and get hot, and perhaps
> burst into flames after a while. Eventually, a fuse, if fitted, will blow,
> and when the insulation burns off the wires, a short will blow the
> mains fuse. This is the result of saturation.

But (as Red Green points out) that would at least as entertaining as your
average "Matlock" episode. Probably more so.

[Patrick Turner]

BFoelsch wrote:

> Perhaps we are slightly confused with terminology here.
>
> Saturation is a property of magnetic materials, and is dependent on flux
> density.
>
> I think that John Byrns meant to imply that the "degree of saturation," as
> caused by the magnetic flux, is independent of secondary current (or the
> reflected primary current). He is correct.
>
> An unloaded transformer draws a current based on its inductance, the applied
> frequency and voltage. The resulting magnetic induction is the product of
> the current in amperes multiplied by the number of turns. The transformer
> designer puts enough turns onto the primary to establish a flux density
> somewhat less than saturation. Failure to do this will result in saturation
> and loss of inductance, which will cause high current and overheating of the
> primary, even in a no-load condition.
>
> The kicker is that, once the transformer is loaded, the flux (and hence the
> flux density) is proportional to the primary ampere-turns MINUS the
> secondary ampere-turns. However, any secondary ampere-turns CREATE an equal
> number of primary ampere-turns ( by increasing the primary current!). As any
> number subtracted from itself is zero, the current drawn from the secondary
> of a transformer has almost no effect on the flux density within the core,
> and hence has almost no effect on the "degree of saturation." The flux is
> determined solely by the unloaded inductance, the applied voltage, and the
> frequency.
>
> The OP's issue of running a winding at 2X voltage will result in disaster in
> a commercial transformer design. Standard designs will not accomodate a 2X
> increase in flux density without saturation.

Well said.

Transformers get hot in two different ways.
First is due to iron losses, and if the primary has low inductance, then
the unloaded primary current is high, and the iron gets hot, even if the wire
size is large enough to prevent the copper getting hot.
Second is due to unloaded current, and loaded current heating the copper.

A normally designed transformer has perhaps a total of 20% of losses, ie, its
only 80% efficient,
so if 100 watts goes in, only 80 watts come out.
This tranny would need to be able to dissipate 20 watts of heat.
So the iron losses might be 10% of the total losses.
Therefore 10 watts are lost in the iron.
The 10 watts of heat lost in the core is constant, and present even with no
load.
:Let's assume there is no saturation, ie, harmonic distortion currents in the
winding
are less than 7%.


Therefore, if the input voltage is 120v, and there is no load then the
idle current is 10/120 = 0.083 amps.
Therfore the impedance of the primary is 120/0.083 = 1,440 ohms.
Therefore the inductance of the primary coil is 1,440/ ( 6.28 x 60Hz ) = 3.82
Henrys.
Now if we used twice the primary turns, the inductance would rise four times,
and ZLp would rise four times, and magnetising idle current would be a quater,
and iron
losses would be much smaller.
But the load current ability would be reduced dramatically.
So the VA of the tranny would reduce, but the efficiency might rise.

Or if we wanted to maintain the VA rating, and reduce iron losses, then simply
increasing core size will do it, so use a taller stack of laminations.
There are limits, since as we increase the iron stack, the length of turns and
winding resistance increases, so to increase VA, and to get more efficiency,
and to lower temperature, we must increase the window size, wire size, and
perhaps stack.

The use of GOSS reduces losses, but this material still saturates like all other
iron
products, and runs noisily if the B of the iron is too high.

I suggest all those who are confused go read the text books,
before deciding to use transformers for purposes for which they were not
deigned.

Patrick Turner.




>
>
> "William Sommerwerck" > wrote in message
> ...
> > I'm always open to learning new things, but since when is core saturation
> > independent of current?
> >
> > Obviously (?), all other things being equal, a transformer with more iron
> will
> > require a higher current to reach full magnetizing flux.
> >
> > By the way, no one (myself included) mentioned that a transformer designed
> for
> > 60Hz operation might not work very well on 50Hz -- and almost all European
> AC is
> > 50Hz.
> >
> >
> > > Core saturation is independent of current, and depends
> > > on the voltage and frequency.
> >

[Steve]

Sheesh!!

There are no absolutes here, in machine design industries
reverse connecting distribution dry type transformers is commonly
done. Of course allowing for certain limitations.

Here's a quote from Acme Transformers FAQ...

http://www.acmepowerdist.com/ael10.html

7) Can Acme transformers be reverse connected?

ACME dry type Distribution transformers can be reverse connected
without a loss of KVA rating, but there are certain limitations.
Transformers rated 1 KVA and larger single phase, 15 KVA and larger
three phase can be reverse connected without any adverse affects or
loss in KVA capacity. The reason for this limitation in KVA size is,
the turns ratio is the same as the voltage ratio. Example: A
transformer with a 480 volt input, 240 volt output - can have the
output connected to a 240 volt source and thereby become the primary
or input to the transformer, then the original 480 volt primary
winding will become the output or 480 volt secondary. On transformers
rated below 1 KVA single phase there is a turns ratio compensation on
the low voltage winding. This means the low voltage winding has a
greater voltage than the nameplate voltage indicates at no load. For
example, a small single phase transformer having a nameplate voltage
of 480 volts primary and 240 volts secondary, would actually have a no
load voltage of approximately 250 volts, and a full load voltage of
240 volts. If the 240 volt winding were connected to a 240 volt
source, then the output voltage would consequently be approximately
460 volts at no load and approximately 442 volts at full load. As the
KVA becomes smaller, the compensation is greater- resulting in lower
output voltages. When one attempts to use these transformers in
reverse the transformer will not be harmed; however, the output
voltage will be lower than is indicated by the nameplate.

So the correct answer without having a part number or specs on
the transformer is "maybe".

Steve
>

On Thu, 10 Jul 2003 11:00:34 +1000, Patrick Turner
> wrote:

>
>
>Casino wrote:
>
>> Hi All,
>>
>> I have a step-down power transformer with a 120-volt primary and
>> 60-volt center tapped (30-0-30) secondary. I was wondering if I could
>> reverse the transformer by plugging-in the 60-volt secondary into the
>> AC outlet and use it as a step-up for 240-volts? Theoretically, it is
>> possible, but I'm not sure if it's actually safe to do so.
>
>DO NOT DO THIS.
>
>The transformer will probably hum loudly, and get hot, and perhaps burst
>into flame
>after awhile. eventually, a thermal fus, if fitted, will blow, and
>when the insulation burns off the wires, a short will blow a mains fuse.
>This is the result of saturation..
>
>>
>>
>> Also, when I travel Europe, can the 120-volt primary be plugged into
>> 240-volts to supply the 120 (117) for North American appliances.
>> Thanks,
>
>ABSOLUTELY NOT!
>
>Patrick Turner.
>
>>
>>
>> C.W.

[Patrick Turner]

William Sommerwerck wrote:

> > The transformer will probably hum loudly and get hot, and perhaps
> > burst into flames after a while. Eventually, a fuse, if fitted, will blow,
> > and when the insulation burns off the wires, a short will blow the
> > mains fuse. This is the result of saturation.
>
> But (as Red Green points out) that would at least as entertaining as your
> average "Matlock" episode. Probably more so.

One doesn't leap with joy when one's house burns down,
or when you realise one hasn't renewed the fire insurance policy.....

Patrick Turner.

[Terry]

Patrick Turner wrote:
>
> William Sommerwerck wrote:
>
> > > The transformer will probably hum loudly and get hot, and perhaps
> > > burst into flames after a while. Eventually, a fuse, if fitted, will blow,
> > > and when the insulation burns off the wires, a short will blow the
> > > mains fuse. This is the result of saturation.
> >
> > But (as Red Green points out) that would at least as entertaining as your
> > average "Matlock" episode. Probably more so.
>
> One doesn't leap with joy when one's house burns down,
> or when you realise one hasn't renewed the fire insurance policy.....
>
> Patrick Turner.

Reversing a transformer:
Sorry to continue this backward transformer thing. And noting the
discussions about transformer saturation!
But from a practical viewpoint we have used a Hammond transformer
for some 35-40 years to step up 115 volts to 230 mainly to
operate a heavy electric drill; a 1953 Wolf SD4C. But also on low
wattage electronics.
The transformer was originally intended to step down 230 volts
across two phases of a 230 volt delta system, to provide 115
volts for some equipment. So we are using it backwards. It weighs
about 15-20 lbs. and rated at 500 watts. (Seems conservative to
me).
It is a complete isolation transformer with separate input and
output windings. (I've avoided saying primary and secondary cos
there's nothing on the transformer which indicates. Just 115 volt
on one side and 230 volt the other.)
Any way we have stalled the drill, used it for many minutes at a
time and under some pretty low voltage (long extension cord)
conditions and never a moments trouble! Or over heating.
I realize that these are not the same voltage requirements as for
electronics.
But the message seems to be? If the transformer is big enough for
the job it will, taking transformer losses and the practicalities
of its construction, work well in both directions. After all
power companies step up and down, granted with much more
specialized equipment throughout their systems.
Non isolating transformers:
A discussion about non-isolating auto, buck and boost, etc.
transformers is another subject I guess; but I've done that,
before I could afford a Variac! For example:
A typical front door 'bell' transformer, provided it's 12 volt
secondary insulation and ampere (wire gauge) capacity is OK will
allow one to boost or decrease the 115/120 input by approximately
10%. Two such transformers by +/- 20%. The secondaries being
connected in series with the mains input etc. Was a great aid in
a rural area where 40+ years ago long low voltage power lines,
increasing electricity demands, weak tubes and so on might cause
a customer to explain "She worked OK up to about 8 o'clock and
then she just quit. And every time the fridge or the water well
pump cuts in the sound goes off for a minute or two!"
Aha! Low voltage problem?
Have fun. Terry.

[Chuck Harris]

Hi Steve,

Somewhere here, I think you are misunderstanding what is being said.

The OP wanted to put 120V into a secondary that was rated for 60V
and hoped for 240V out of the original 120V primary. This is running
the transformer at 2x its rated voltage... and as has been said over
and over again, this is not a good idea, and will likely result
in the destruction of the transformer.

What the article you are quoting is saying is that transformers work
in a reciprocal way, which is always true... The transformer knows
nothing of which is its primary or secondary winding.

However, the designer of smaller transformers will often add turns
to secondary windings so as to compensate for the voltage drop due
to resistance in the primary and secondary windings... Something
that I stated, and got beat up over, in a thread a month or so ago,
when I stated that you shouldn't parallel dissimilar current,
filament windings.

The only "maybe" that exists in the OP's transformer is maybe it is
really a 240V to 110V transformer that was being used as a 120V to
60V transformer... maybe. A simple experiment is the easiest
way to tell. Put a 50W light bulb in series with the intended primary
and connect it to 120V and if the light bulb is dark, you are probably
fine. If the light bulb lights, forget about this idea.

There can be no new revelations in this thread. The only "aha!"s
will have to be those of readers taking the time to read, confirm,
and finally understand what has been said by the few of us here
who have actually learned the art of transformer design.

-Chuck


Steve wrote:
> Sheesh!!
>
> There are no absolutes here, in machine design industries
> reverse connecting distribution dry type transformers is commonly
> done. Of course allowing for certain limitations.
>
> Here's a quote from Acme Transformers FAQ...
>
> http://www.acmepowerdist.com/ael10.html
>
> 7) Can Acme transformers be reverse connected?
>
> ACME dry type Distribution transformers can be reverse connected
> without a loss of KVA rating, but there are certain limitations.
> Transformers rated 1 KVA and larger single phase, 15 KVA and larger
> three phase can be reverse connected without any adverse affects or
> loss in KVA capacity. The reason for this limitation in KVA size is,
> the turns ratio is the same as the voltage ratio. Example: A
> transformer with a 480 volt input, 240 volt output - can have the
> output connected to a 240 volt source and thereby become the primary
> or input to the transformer, then the original 480 volt primary
> winding will become the output or 480 volt secondary. On transformers
> rated below 1 KVA single phase there is a turns ratio compensation on
> the low voltage winding. This means the low voltage winding has a
> greater voltage than the nameplate voltage indicates at no load. For
> example, a small single phase transformer having a nameplate voltage
> of 480 volts primary and 240 volts secondary, would actually have a no
> load voltage of approximately 250 volts, and a full load voltage of
> 240 volts. If the 240 volt winding were connected to a 240 volt
> source, then the output voltage would consequently be approximately
> 460 volts at no load and approximately 442 volts at full load. As the
> KVA becomes smaller, the compensation is greater- resulting in lower
> output voltages. When one attempts to use these transformers in
> reverse the transformer will not be harmed; however, the output
> voltage will be lower than is indicated by the nameplate.
>
> So the correct answer without having a part number or specs on
> the transformer is "maybe".
>
> Steve
>
>
> On Thu, 10 Jul 2003 11:00:34 +1000, Patrick Turner
> > wrote:
>
>
>>
>>Casino wrote:
>>
>>
>>>Hi All,
>>>
>>>I have a step-down power transformer with a 120-volt primary and
>>>60-volt center tapped (30-0-30) secondary. I was wondering if I could
>>>reverse the transformer by plugging-in the 60-volt secondary into the
>>>AC outlet and use it as a step-up for 240-volts? Theoretically, it is
>>>possible, but I'm not sure if it's actually safe to do so.
>>
>>DO NOT DO THIS.
>>
>>The transformer will probably hum loudly, and get hot, and perhaps burst
>>into flame
>>after awhile. eventually, a thermal fus, if fitted, will blow, and
>>when the insulation burns off the wires, a short will blow a mains fuse.
>>This is the result of saturation..
>>
>>
>>>
>>>Also, when I travel Europe, can the 120-volt primary be plugged into
>>>240-volts to supply the 120 (117) for North American appliances.
>>>Thanks,
>>
>>ABSOLUTELY NOT!
>>
>>Patrick Turner.
>>
>>
>>>
>>>C.W.
>
>

[Chuck Harris]

Hi Terry,

I think you have missed the point of this thread. No one
has stated that you cannot run a 240 -> 120V transformer
as a 120V -> 240V transformer. Doing so will work just fine,
if you can stand the 240 being a little lower than you might
expect.

What we have said is the OP's 120V -> 60V transformer is not
likely to survive being used as a 120V -> 240V, or a 240V -> 120V
transformer.

This whole thread is about what happens when you apply 2x the rated
voltage to a winding on a usual commercially made transformer.

If you want to challenge the assertions using your Hammond 240<->120V
transformer as an example, hook the 120V side up to 240V, and use
it to provide 480V out of the 240V side.

It will not survive.

-Chuck

Terry wrote:

> Reversing a transformer:
> Sorry to continue this backward transformer thing. And noting the
> discussions about transformer saturation!
> But from a practical viewpoint we have used a Hammond transformer
> for some 35-40 years to step up 115 volts to 230 mainly to
> operate a heavy electric drill; a 1953 Wolf SD4C. But also on low
> wattage electronics.
> The transformer was originally intended to step down 230 volts
> across two phases of a 230 volt delta system, to provide 115
> volts for some equipment. So we are using it backwards. It weighs
> about 15-20 lbs. and rated at 500 watts. (Seems conservative to
> me).
> It is a complete isolation transformer with separate input and
> output windings. (I've avoided saying primary and secondary cos
> there's nothing on the transformer which indicates. Just 115 volt
> on one side and 230 volt the other.)
> Any way we have stalled the drill, used it for many minutes at a
> time and under some pretty low voltage (long extension cord)
> conditions and never a moments trouble! Or over heating.
> I realize that these are not the same voltage requirements as for
> electronics.
> But the message seems to be? If the transformer is big enough for
> the job it will, taking transformer losses and the practicalities
> of its construction, work well in both directions. After all
> power companies step up and down, granted with much more
> specialized equipment throughout their systems.

[Ian Iveson]

"John Byrns" > wrote in
> >
> > For mains trannies they require the voltage to be specified, but
it
> > has no effect on core size.
>
> The input voltage and frequency for a mains transformer are fixed
> quantities, so it is only natural to rate the windings in terms of
these.

No, John. The original question demonstrates that
input voltage is not a fixed quantity. That is what the pertinent
parts of the thread are about.

The core size has nothing to do with voltage for a given quality
transformer. I'm sure you know this really. You're just trying to
wind me up.

> > For OPT they don't even ask about
> > voltage as far as I can see... max AC primary voltage can be
derived
> > of course from power rating, impedance ratio, and speaker
impedance.
>
> Voltage is not a traditional measure for output transformer
> specifications, the impedance, power, and lowest frequency for
full power
> are the parameters generally specified. While Sowter may not
explicitly
> ask about the lowest frequency, they do make some implicit
assumptions
> about it, as I have had them wind some custom transformers for me.
I
> would also note that at no point in the discussions did they ever
ask
> about AC current which you consider so important.

I consider all current so important if it contributes to the flux.
That due to DC current in a PP transformer cancels, so it is not
important to saturation. That is why I said AC current. They can't
calculate DC current. Please, John, you know that. You would have
to tell them. They no doubt assume that, since bias current in PP
primaries is balanced, and since it must give rise to less power
dissipation in the transformer than full power, which they do know,
they won't care about bias current unless you want to involve them
more closely in your design.

As for AC current, I said they calculate it. You say that too,
below. That is why they do not ask.

> If I remember correctly
> they did ask about the DC current, which cancels in push pull
transformers
> of the sort I had them design, because of the heating effects,
although
> that is probably just for a safety check, because the primary wire
size
> may be controlled by their efficiency considerations, which
probably keeps
> the DC heating to a manageable level in most cases. So the
parameters for
> Sowter output transformers seem to be impedance, and power, plus
their
> implicit assumptions about lowest frequency, and efficiency.

As I said. Yes. For PP. But you are skirting round the killer
point aren't you? You have thought of DC current. You know about
valve transformers. Isn't something niggling in your mind?
Something about current? Something about saturation? Something
very obvious?

If you order a SE transformer do you specify bias voltage? No. Of
course not. You specify current. Why? Because current saturates,
not voltage.

There is only one kind of current. The B/H curve for the core
material does not distinguish AC from DC. The only reason
saturation voltage for a particular transformer varies with
frequency is that the impedance of the coil also varies with
frequency, and the voltage is *derived* via ohms law from the
current.

The unit of H is A/m. No voltage involved at all. Shame really,
because a pure inductor cannot have voltage across it at DC, being a
short, and so if voltage is what causes saturation you could have
infinite bias current and still not need a gapped core. But no, it
is number of turns and current that determines saturation for a
given core at a given temperature. Independent of frequency. That
is why I said it is directly related to current. Not to voltage.

> > Yes you could be perverse and insist on transposing all the
> > equations and isolate the least convenient variable and specify
> > using a different schema, but they would only have to translate
back
> > to something that makes actual practical sense to the winder.
>
> Voltage makes the most practical sense for power transformer
> specifications, as far as the core saturation point goes, from a
user
> point of view.

But the user must also specify power. Power determines the size of
the core. Voltage determines the number of turns, given a
particular quality of transformer.

The important point in answer to the original question is that the
transformer has the right amount of iron and copper for that quality
of transformer, and if Casino wishes to maintain the quality at
twice the voltage, all he has to do is take it apart, stretch the
windings, touch up the enamel, and rewind. They should be twice as
long, and half as thick, I guess. Then you would get half the
current at twice the voltage, because it would be four times the
inductance.

> Wouldn't that be at least four times as hot? If the impedance is
fixed,
> both the current and voltage will double, and the power will
quadruple.

You are right, of course. As it is, at double it's rated voltage,
no-load power will quadruple and not double as I said. We have done
this bit already. Sorry. Still small compared to when at full
power into a load. Plenty of margin. Care to bet? Found a
240V/120V so I could try it backwards. Cheapish so it won't be
GOSS. What do you think? Prepared to bet on less than a minute?

cheers, Ian

[Ian Iveson]

Now my ISP has switched me back on...

"Chuck Harris" > wrote:

>> [see below]

I take capitals as shouting. That is usual, I believe. You shout
sometimes.

It has been said that saturation depends on voltage. It has been
said that in order to avoid saturation the core should be made
bigger. You began yourself by explaining that the problem would be
related to skimpy margins of iron and copper, neither of which is
relevant and seems a bit like axe-grinding to me. For the same
quality transformer at twice the voltage Casino requires the same
quantity of both.

The impression of emphasis, and I did say emphasis, is because much
has been said about size of core, and virtually nothing has been
said about number of turns.

Yes, what you said about economy of materials is true, and yes you
can argue as has John that you merely meant that the symptom of
saturation appears in the core, rather than the cause.
Nevertheless, by not mentioning number of turns, which is the crux
of the matter, the impression was created. I was not the only one
who had difficulty following a thread that drifted towards the
moans of small producers in the face of the bean-counting
Juggernauts, blah.

The maths bits are fair enough but rather generalised with respect
to the original question of what would be the problem with that
particular transformer at that voltage. That left a gap, between
the reminiscences, the business interests, and the maths, where I
felt a simple answer might have been.

As far as "boning up" is concerned, I had failed previously to grasp
the point that currents due to load cancel, more or less, with
respect to saturation. Something I had not noticed in the fog of
knowledge
I already have. There is a reason for this that I have just been
wondering about. It explains why my spice transformer model still
has a linear core. I must have known everything else that is in the
model otherwise it wouldn't work, I suppose. Specified transformers
seem to work OK, so I guess I must have known something then...can't
really remember. Must have though because some of my transformers
are quite complicated and have quite a lot of wires coming out of
them.

And for audio-intended transformers you don't really need to know.
Saturation will be outside the full-power bandwidth. Er...I think
that follows anyway...yes it must.

On the face of it, as I have said somewhere in this thread, all I
need to do is regress the BH curves of GOSS and NOSS onto a
quadratic (or whatever) expression spice can understand. I
don't carry maths in my head. Never could. Always had to derive
from first principles every time until computers. Never completed a
maths exam in my life. Dunno how I passed so many, a currupt
education system I suspect. Anyway I have found that Xspice has two
complete transformer models with core, requiring a set of
co-ordinate pairs for B and H, which is encouraging, plus core
length and area. If I discover it in my own system I'll show you a
picture.

I am fundamentally uneasy about saturation. Coming from a
background with no practical electronic content, I still am
conscious of mapping mechanical dynamics onto electrical things.
But I have never needed to consider the saturation velocity of mass.
Or even its equivalent of the BH curve. Never been into rockets and
stuff, or really big things, or really small things moving very
quickly. Ballistics is about my limit.

What I say is related to my purpose here, which is not to
demonstrate what I know. I have never to my
knowledge been responsible for electrocution, and am careful to
frequently point out that I know virtually nothing about anything.

Your original comment to William:

>> Please don't give advice about things you don't understand.

Is pathetic and silly. This is a discussion group. I give opinions
on lots of things, and got sick of putting "IMO" in every sentence.
If there were only Chucks here, and no Williams, there would be no
point in coming. But William employs a similar strategy to me, it
seems, and it was his perseverance, despite your bullying, that
sorted the wheat from your chaff.

The rest of your first contribution was pure hyperbole, intended no
doubt to rub in your bullying with a bit of guilt-tripping.

>> If you try applying 120V, 60Hz to this 60V, 60Hz winding, you
>> will drive the transformer well into saturation. This will
cause
>> the primary current to approach that of the DC wire resistance.

>> In a word, You will let all the smoke out of your
>> transformer.

Only if he gets an arc. No poof. Much slower than that. Care to
bet? A current can't approach a resistance, BTW. Pedantry, yes,
but OTOH that sentence doesn't make sense anyway. Approach usually
implies that it gets close. We don't actually know how close it
would get, because we don't know for what proportion of the time the
core would saturate. There will be a lot of margin to play with, in
terms of power rating of the windings, so I would expect minutes,
not seconds, before destruction. You'd notice, wouldn't you? The
buzzing, the noise, etc. You wouldn't just walk away and let it
burn.

cheers, Ian

> Hi Ian,
>
> I have no axe to grind, and I don't shout and squirm. If you
> feel that I do, perhaps it is your own prejudices at play. When
> I make mistakes, I admit to them, and try to correct them... if
> possible.
>
> I have read the whole thread on "Reversing a Power Transformer",
> from beginning to end, and I must have missed the point where it
> was stated (or implied) that core size is dependent on voltage.
> If I, or others, have stated that, it was an error, and needs to
> be corrected.
>
> In so far as I can recall, I have never believed that voltage
> determines core size. This isn't a recent revelation, I designed
> and constructed my first power transformer around 1974. It was a
> whopper that delivered 20V at 30A. And stayed very cool and quiet
> at anywhere up to full loading. I used it for 15 or so years
> in a power supply of my own design. I have since sold the power
> supply, and as far as I know, it still is working. In the years
> that have passed since the design of that transformer, I have
> designed and built a forgotten number of linear and switching
> power transformers and inductors. The latest of which, was a
> 100KHz switcher that was part of a battery charger system I
> developed for the US Army.
>
> You are rapidly growing more correct in your statements about
> transformer design and characteristics. Your postings, up until
> this last couple, had what I interpreted as strongly misleading
> statements. I suspect that you have been boning up on the art
> of transformer design, and that is good.
>
> -Chuck
>
> Ian Iveson wrote:
> > "Chuck Harris" > wrote in message
> > ...
> >
> >>Ian,
> >>
> >>Please point out the article that "concentrates only on primary
> >
> > voltage
> >
> >>and core size".
> >
> >
> > Only if you promise not to shout and squirm. Is there anything
else
> > you think I have got wrong, or is the rest of what I have
written
> > here correct? You appear to wish to grind an axe rather than
> > clarify the truth.
> >
> > Ian
> >
> >
>

[Jeffrey D Angus]

Terry wrote:
> Reversing a transformer:
> It is a complete isolation transformer with separate input and
> output windings. (I've avoided saying primary and secondary cos
> there's nothing on the transformer which indicates. Just 115 volt
> on one side and 230 volt the other.)

The whole thread has been confusing in the subject title.
Reversing a transformer. They are bi-directional by nature.
The problem comes with putting more voltage across a winding
than what it was designed for. I.e., putting 120 Volts across
a 60 Volt winding. There in lies the problem.

Jeff

--
"They that can give up essential liberty to obtain a little temporary
safety deserve neither liberty nor safety." Benjamin Franklin
"A life lived in fear is a life half lived."
Tara Morice as Fran, from the movie "Strictly Ballroom"

[Casino]

Hey Everyone,

Once again, thanks to all those who have replied without getting
off-topic. For some strange reason, some of you guys think that the
transformer is a 240:120 step-down. It is not. It's a 120:60 step
down (which I wanted to use as a 240:120); in fact, it is actually a
30-0-30 and was part of a 30 volt DC power supply of some sold state
equipment.

C.W.

[John Byrns]

In article >, "Ian
Iveson" > wrote:

> "John Byrns" > wrote in
> > >
> > > For mains trannies they require the voltage to be specified, but
> it
> > > has no effect on core size.
> >
> > The input voltage and frequency for a mains transformer are fixed
> > quantities, so it is only natural to rate the windings in terms of
> these.
>
> No, John. The original question demonstrates that
> input voltage is not a fixed quantity. That is what the pertinent
> parts of the thread are about.

Yes, Ian, the input voltage is a fixed quantity, here in the US the
domestic supply voltage provided to residences is 120 volts at 60 Hz, in
Europe it is something different, but still fixed. The original posters
question was, can I hook the 60 volt winding of a power transformer to
that 120 volt supply without damaging the transformer? It is not
necessary to get the data on the core and calculate current to provide a
reasonably accurate to that question. My point is that I am not a
transformer designer, I am only a user. Sure the transformer designer is
going to do all the current calculations you talk about, but why should I
do that as a user, when I can much more easily determine approximately
what voltage a transformer winding can handle at a given frequency, based
on the original voltage and frequency ratings of the transformer, which in
commercial grade power transformers bear a close relationship to the
saturation point?

> The core size has nothing to do with voltage for a given quality
> transformer. I'm sure you know this really. You're just trying to
> wind me up.

Although I may be mistaken, I don't think I have made any claims in this
thread about core size vs. voltage, I don't think I have even mentioned
core size. I think you are confusing my comments with those of someone
else. You are the one that insists on injecting core size into the
discussion.

I'm not trying to wind you up, I'm simply saying that from the perspective
of a user of power transformers, the voltage spec. of a winding is the
most relevant for determining if a given voltage will saturate the core,
the current is only of interest to the transformer user in calculating the
power the transformer can pass without overheating. As a user,
calculating the saturation point based on current is a much more complex
undertaking than simply looking at the rated voltage and frequency.
Calculating saturation based on current requires me to have knowledge of
several characteristics of the core, and do relatively involved math, why
go to all that trouble when I know approximately how commercial power
transformers are designed with respect to saturation vs. voltage, and
frequency?


Regards,

John Byrns


Surf my web pages at, http://users.rcn.com/jbyrns/

[Chuck Harris]

Hi Ian,

When one writes on an ascii only medium, one must do something to
show emphasis. The only way I know to do that is write a word
or two in capital letters.

In the culture of email and news that I adopted, when I started
messing with DARPA back in the '70s, that is the way it was done.

For example:

I HATE YOU AND YOUR MOTHER'S NOSE! <- that would be shouting.

I think it would best if we REALLY REALLY try and concentrate....

^^^^^^^^^^^^^
Emphasis.

I am not shouting at you, I don't think I have ever done. However,
lacking a method of underlining or italicizing a word, I capitalize.
Please don't take that as shouting. You will find that a large
number of people do the same as I. They aren't shouting either.

Shouting is when you write whole angry sounding sentences in
capital letters.



Ian Iveson wrote:
> Now my ISP has switched me back on...
>
> "Chuck Harris" > wrote:
>
>
>>>[see below]
>
>
> I take capitals as shouting. That is usual, I believe. You shout
> sometimes.
>
> It has been said that saturation depends on voltage. It has been
> said that in order to avoid saturation the core should be made
> bigger. You began yourself by explaining that the problem would be
> related to skimpy margins of iron and copper, neither of which is
> relevant and seems a bit like axe-grinding to me. For the same
> quality transformer at twice the voltage Casino requires the same
> quantity of both.
>
> The impression of emphasis, and I did say emphasis, is because much
> has been said about size of core, and virtually nothing has been
> said about number of turns.

It really doesn't matter to which you attribute the saturation. The
OP only had the transformer he had. He wasn't planning on changing
it so that it would work. That being the case, one can explain the
reason for its failure to work properly at 2x its rated primary voltage
as:

1) the core is too small for the flux...
2) there aren't enough turns...
3) the frequency is too low...

[note, ... means I left out the boring part of the sentence]

Pick one. Whichever makes you happy. They are all true, and they all
explain the deficiencies of the OP's transformer when run at twice
its design voltage.

There is no implication in any of the above statements that the only
way to make a transformer that would handle a higher primary voltage
is to make one with a larger core... not at all! If you kept the
turns the same, that would be one solution. But not necessarily the
best solution.

For instance, I can make a quite dandy 110v to 240v transformer with
one turn of wire on the primary, and two turns on the secondary. I
can make this transformer, and it will run cool as a cucumber, all
day long at 25Hz! How would I do this???

Simple, I would need a core the size of a volkswagon for a silly
couple hundred VA transformer. (No, I didn't calculate this...
I am purposefully exagerating.)
>
> Yes, what you said about economy of materials is true, and yes you
> can argue as has John that you merely meant that the symptom of
> saturation appears in the core, rather than the cause.
> Nevertheless, by not mentioning number of turns, which is the crux
> of the matter, the impression was created. I was not the only one
> who had difficulty following a thread that drifted towards the
> moans of small producers in the face of the bean-counting
> Juggernauts, blah.

I am one of those bean counting Juggernauts! When I design a
transformer, or anything else, I have to make it as cheaply as will
meet the design requirements. That is what it means to be an
engineer. I have certain preferences when it comes to transformers
that I use for myself. I don't want to ever have to redo one,
so I would prefer it run at a low flux so that it will run cool
and have a minimum amount of vibration. Why? Because the insulation
will last longer that way. Because I don't like burning myself on
transformers that are operating "normally".

>
> The maths bits are fair enough but rather generalised with respect
> to the original question of what would be the problem with that
> particular transformer at that voltage. That left a gap, between
> the reminiscences, the business interests, and the maths, where I
> felt a simple answer might have been.

The answer was quite simple and was stated over and over again:

This transformer will saturate if run at 2x its rated voltage...
Don't do it!

The message was needlessly complicated when both you and William
came in and said ignore the above message, and go ahead. It'll
work just fine with reduced load current.

The message was further complicated by guys that tried to judge
the "book" by its cover, rather than its contents, and said the
"saturation" message is wrong, go ahead run the transformer backwards,
I do it all the time. Of course they meant they put 6V into a 6V
filament winding, or 110V into a 110V secondary of a 220->110V step
down transformer.
>
> As far as "boning up" is concerned, I had failed previously to grasp
> the point that currents due to load cancel, more or less, with
> respect to saturation. Something I had not noticed in the fog of
> knowledge
> I already have. There is a reason for this that I have just been
> wondering about. It explains why my spice transformer model still
> has a linear core. I must have known everything else that is in the
> model otherwise it wouldn't work, I suppose. Specified transformers
> seem to work OK, so I guess I must have known something then...can't
> really remember. Must have though because some of my transformers
> are quite complicated and have quite a lot of wires coming out of
> them.

Spice models for transformers are very simple if they don't take into
account core characteristics, leakage inductances, and capacitances.

Most of the time, an ideal transformer will adequately model the real
thing. It is cases like where you increase the primary voltage, or
drop the frequency so as to cause the core to head for the nonlinear
region that are not handled well by an ideal transformer model.
>
> And for audio-intended transformers you don't really need to know.
> Saturation will be outside the full-power bandwidth. Er...I think
> that follows anyway...yes it must.

When you take the typical audio output transformer, and head for the
lower end, 20Hz, or so, the transformer is quite likely to be starting
to saturate... That is, with the typical garden variety transformer.
Some of the transformers that top off in the 12 lbs range will do much
better.
>
> I am fundamentally uneasy about saturation. Coming from a
> background with no practical electronic content, I still am
> conscious of mapping mechanical dynamics onto electrical things.
> But I have never needed to consider the saturation velocity of mass.
> Or even its equivalent of the BH curve. Never been into rockets and
> stuff, or really big things, or really small things moving very
> quickly. Ballistics is about my limit.

Saturation is all about sudden nonlinearities. Some places you see
it in mechanical models are springs when they run out of range, (eg. a
compression spring that has its coils touching, a tension spring that
has its coils pulled all the way out... Or elastic deformation, a
spring that has been bent so much that it won't return.) There are
plenty of other examples, a shaft on a pot can be thought of as being
in mechanical saturation when it is turned against its stop.

>
> What I say is related to my purpose here, which is not to
> demonstrate what I know. I have never to my
> knowledge been responsible for electrocution, and am careful to
> frequently point out that I know virtually nothing about anything.
>
> Your original comment to William:
>
>
>>> Please don't give advice about things you don't understand.
>
>
> Is pathetic and silly. This is a discussion group. I give opinions
> on lots of things, and got sick of putting "IMO" in every sentence.
> If there were only Chucks here, and no Williams, there would be no
> point in coming. But William employs a similar strategy to me, it
> seems, and it was his perseverance, despite your bullying, that
> sorted the wheat from your chaff.

It puts my hackles up when someone like William tells the group to
ignore a person who provided a correct answer, and go with his
uneducated, and incorrect solution. That is where the bullying
occurred, not with me.

If William came in and stated that he didn't think this was right,
and gave his reasons, I would have entered this thread in a more
friendly manner. But he didn't. He flat out said the answer was
wrong! ... and he couldn't have been more wrong himself.


>
> The rest of your first contribution was pure hyperbole, intended no
> doubt to rub in your bullying with a bit of guilt-tripping.
>
>
>>> If you try applying 120V, 60Hz to this 60V, 60Hz winding, you
>>> will drive the transformer well into saturation. This will
>
> cause
>
>>> the primary current to approach that of the DC wire resistance.
>
>
>>> In a word, You will let all the smoke out of your
>>> transformer.

All of which is completely and verifiably true. Why do you call it
hyperbole? Just because you don't want it to be true isn't a good
reason.

As to your assertion that I believe current is measured in resistance,
do you really think I am that stupid? I left a couple of words
out in my haste to get the message done. Tell me you have never
done that.

"This will cause the primary current to approach that caused by the
DC resistance alone."

Does that sound better to you?
>
> Only if he gets an arc. No poof. Much slower than that. Care to
> bet? A current can't approach a resistance, BTW. Pedantry, yes,
> but OTOH that sentence doesn't make sense anyway. Approach usually
> implies that it gets close. We don't actually know how close it
> would get, because we don't know for what proportion of the time the
> core would saturate.

That is why I said "approach", rather than "will be". I have taken
the "I" laminations off of transformers and used them as demagnetizers,
so I do have a fair idea how long it would take to ruin the primary.
Even with a 50W light bulb in series, they get hot very fast.
It is not in my nature to needlessly destroy things. I would never
make a very good UL inspector.

The part that you are not aware of is that if you do this experiment
at 240V, the winding temperature will skyrocket immediately, and the
insulation will be damaged very quickly. No, a short won't
necessesarily happen right away, but the insulation will become
so brittle that it won't withstand any bending or vibration without
flaking off.

At saturation, a typical 1 ohm dc resistance 120V primary will be
sucking V*V/R watts at Vpeak. So, if you are sticking 240V into
the 120V primary, that will result in 240V*240V*2 = 115200Wpeak!

Do you think that might heat things up a bit?

Do you think it matters if it is only saturating 1/4 the time
at the peaks? (that drops the power to somewhere around a mere
28000W. I would have to integrate the waveforms to give you a
more exact answer.)

About the only thing that will keep these numbers from being
realized is the fact that the primary has some inductive reactance
even without a core, and the power cord you connect up with has
some resistance.

There will be a lot of margin to play with, in
> terms of power rating of the windings, so I would expect minutes,
> not seconds, before destruction. You'd notice, wouldn't you? The
> buzzing, the noise, etc. You wouldn't just walk away and let it
> burn.

Walk away from it? No, not me, I already know what will happen.
But some newbie might if folks such as you, and William,
convinced him that it would be ok as long as he just reduced
the load current.

-Chuck

[Ian Iveson]

The generally accepted code for emphasis is enclosure within
asterisks. Do what you like, but I will interpret capitals as
shouting coz I can't be bothered to remember you as an exception.

As for the rest, I have lost you somewhere it seems.

The problem with Casino's transformer is the windings, not the core.
You explain it how you like, you are just wriggling. What's the
difference between that transformer and one identical except for
double the voltage? Windings. That is all.

You have by no means demonstrated that William is wrong. You, OTOH,
most definitely *are* wrong.

It will not go "poof".

Nowhere near "poof".

As for squirming, which you denied, that is what you are doing now.
Anything other than proof of "poof" is squirming.

So put up or shut up and accept you are wrong. What's your bet?

cheers, Ian


"Chuck Harris" > wrote in message
...
> Hi Ian,
>
> When one writes on an ascii only medium, one must do something to
> show emphasis. The only way I know to do that is write a word
> or two in capital letters.
>
> In the culture of email and news that I adopted, when I started
> messing with DARPA back in the '70s, that is the way it was done.
>
> For example:
>
> I HATE YOU AND YOUR MOTHER'S NOSE! <- that would be shouting.
>
> I think it would best if we REALLY REALLY try and concentrate....
>
> ^^^^^^^^^^^^^
> Emphasis.
>
> I am not shouting at you, I don't think I have ever done.
However,
> lacking a method of underlining or italicizing a word, I
capitalize.
> Please don't take that as shouting. You will find that a large
> number of people do the same as I. They aren't shouting either.
>
> Shouting is when you write whole angry sounding sentences in
> capital letters.
>
>
>
> Ian Iveson wrote:
> > Now my ISP has switched me back on...
> >
> > "Chuck Harris" > wrote:
> >
> >
> >>>[see below]
> >
> >
> > I take capitals as shouting. That is usual, I believe. You
shout
> > sometimes.
> >
> > It has been said that saturation depends on voltage. It has
been
> > said that in order to avoid saturation the core should be made
> > bigger. You began yourself by explaining that the problem would
be
> > related to skimpy margins of iron and copper, neither of which
is
> > relevant and seems a bit like axe-grinding to me. For the same
> > quality transformer at twice the voltage Casino requires the
same
> > quantity of both.
> >
> > The impression of emphasis, and I did say emphasis, is because
much
> > has been said about size of core, and virtually nothing has been
> > said about number of turns.
>
> It really doesn't matter to which you attribute the saturation.
The
> OP only had the transformer he had. He wasn't planning on
changing
> it so that it would work. That being the case, one can explain
the
> reason for its failure to work properly at 2x its rated primary
voltage
> as:
>
> 1) the core is too small for the flux...
> 2) there aren't enough turns...
> 3) the frequency is too low...
>
> [note, ... means I left out the boring part of the sentence]
>
> Pick one. Whichever makes you happy. They are all true, and they
all
> explain the deficiencies of the OP's transformer when run at twice
> its design voltage.
>
> There is no implication in any of the above statements that the
only
> way to make a transformer that would handle a higher primary
voltage
> is to make one with a larger core... not at all! If you kept the
> turns the same, that would be one solution. But not necessarily
the
> best solution.
>
> For instance, I can make a quite dandy 110v to 240v transformer
with
> one turn of wire on the primary, and two turns on the secondary.
I
> can make this transformer, and it will run cool as a cucumber, all
> day long at 25Hz! How would I do this???
>
> Simple, I would need a core the size of a volkswagon for a silly
> couple hundred VA transformer. (No, I didn't calculate this...
> I am purposefully exagerating.)
> >
> > Yes, what you said about economy of materials is true, and yes
you
> > can argue as has John that you merely meant that the symptom of
> > saturation appears in the core, rather than the cause.
> > Nevertheless, by not mentioning number of turns, which is the
crux
> > of the matter, the impression was created. I was not the only
one
> > who had difficulty following a thread that drifted towards the
> > moans of small producers in the face of the bean-counting
> > Juggernauts, blah.
>
> I am one of those bean counting Juggernauts! When I design a
> transformer, or anything else, I have to make it as cheaply as
will
> meet the design requirements. That is what it means to be an
> engineer. I have certain preferences when it comes to
transformers
> that I use for myself. I don't want to ever have to redo one,
> so I would prefer it run at a low flux so that it will run cool
> and have a minimum amount of vibration. Why? Because the
insulation
> will last longer that way. Because I don't like burning myself on
> transformers that are operating "normally".
>
> >
> > The maths bits are fair enough but rather generalised with
respect
> > to the original question of what would be the problem with that
> > particular transformer at that voltage. That left a gap,
between
> > the reminiscences, the business interests, and the maths, where
I
> > felt a simple answer might have been.
>
> The answer was quite simple and was stated over and over again:
>
> This transformer will saturate if run at 2x its rated voltage...
> Don't do it!
>
> The message was needlessly complicated when both you and William
> came in and said ignore the above message, and go ahead. It'll
> work just fine with reduced load current.
>
> The message was further complicated by guys that tried to judge
> the "book" by its cover, rather than its contents, and said the
> "saturation" message is wrong, go ahead run the transformer
backwards,
> I do it all the time. Of course they meant they put 6V into a 6V
> filament winding, or 110V into a 110V secondary of a 220->110V
step
> down transformer.
> >
> > As far as "boning up" is concerned, I had failed previously to
grasp
> > the point that currents due to load cancel, more or less, with
> > respect to saturation. Something I had not noticed in the fog
of
> > knowledge
> > I already have. There is a reason for this that I have just been
> > wondering about. It explains why my spice transformer model
still
> > has a linear core. I must have known everything else that is in
the
> > model otherwise it wouldn't work, I suppose. Specified
transformers
> > seem to work OK, so I guess I must have known something
then...can't
> > really remember. Must have though because some of my
transformers
> > are quite complicated and have quite a lot of wires coming out
of
> > them.
>
> Spice models for transformers are very simple if they don't take
into
> account core characteristics, leakage inductances, and
capacitances.
>
> Most of the time, an ideal transformer will adequately model the
real
> thing. It is cases like where you increase the primary voltage,
or
> drop the frequency so as to cause the core to head for the
nonlinear
> region that are not handled well by an ideal transformer model.
> >
> > And for audio-intended transformers you don't really need to
know.
> > Saturation will be outside the full-power bandwidth. Er...I
think
> > that follows anyway...yes it must.
>
> When you take the typical audio output transformer, and head for
the
> lower end, 20Hz, or so, the transformer is quite likely to be
starting
> to saturate... That is, with the typical garden variety
transformer.
> Some of the transformers that top off in the 12 lbs range will do
much
> better.
> >
> > I am fundamentally uneasy about saturation. Coming from a
> > background with no practical electronic content, I still am
> > conscious of mapping mechanical dynamics onto electrical things.
> > But I have never needed to consider the saturation velocity of
mass.
> > Or even its equivalent of the BH curve. Never been into rockets
and
> > stuff, or really big things, or really small things moving very
> > quickly. Ballistics is about my limit.
>
> Saturation is all about sudden nonlinearities. Some places you
see
> it in mechanical models are springs when they run out of range,
(eg. a
> compression spring that has its coils touching, a tension spring
that
> has its coils pulled all the way out... Or elastic deformation, a
> spring that has been bent so much that it won't return.) There
are
> plenty of other examples, a shaft on a pot can be thought of as
being
> in mechanical saturation when it is turned against its stop.
>
> >
> > What I say is related to my purpose here, which is not to
> > demonstrate what I know. I have never to my
> > knowledge been responsible for electrocution, and am careful to
> > frequently point out that I know virtually nothing about
anything.
> >
> > Your original comment to William:
> >
> >
> >>> Please don't give advice about things you don't understand.
> >
> >
> > Is pathetic and silly. This is a discussion group. I give
opinions
> > on lots of things, and got sick of putting "IMO" in every
sentence.
> > If there were only Chucks here, and no Williams, there would be
no
> > point in coming. But William employs a similar strategy to me,
it
> > seems, and it was his perseverance, despite your bullying, that
> > sorted the wheat from your chaff.
>
> It puts my hackles up when someone like William tells the group to
> ignore a person who provided a correct answer, and go with his
> uneducated, and incorrect solution. That is where the bullying
> occurred, not with me.
>
> If William came in and stated that he didn't think this was right,
> and gave his reasons, I would have entered this thread in a more
> friendly manner. But he didn't. He flat out said the answer was
> wrong! ... and he couldn't have been more wrong himself.
>
>
> >
> > The rest of your first contribution was pure hyperbole, intended
no
> > doubt to rub in your bullying with a bit of guilt-tripping.
> >
> >
> >>> If you try applying 120V, 60Hz to this 60V, 60Hz winding, you
> >>> will drive the transformer well into saturation. This will
> >
> > cause
> >
> >>> the primary current to approach that of the DC wire
resistance.
> >
> >
> >>> In a word, You will let all the smoke out of your
> >>> transformer.
>
> All of which is completely and verifiably true. Why do you call
it
> hyperbole? Just because you don't want it to be true isn't a good
> reason.
>
> As to your assertion that I believe current is measured in
resistance,
> do you really think I am that stupid? I left a couple of words
> out in my haste to get the message done. Tell me you have never
> done that.
>
> "This will cause the primary current to approach that caused by
the
> DC resistance alone."
>
> Does that sound better to you?
> >
> > Only if he gets an arc. No poof. Much slower than that. Care
to
> > bet? A current can't approach a resistance, BTW. Pedantry,
yes,
> > but OTOH that sentence doesn't make sense anyway. Approach
usually
> > implies that it gets close. We don't actually know how close it
> > would get, because we don't know for what proportion of the time
the
> > core would saturate.
>
> That is why I said "approach", rather than "will be". I have
taken
> the "I" laminations off of transformers and used them as
demagnetizers,
> so I do have a fair idea how long it would take to ruin the
primary.
> Even with a 50W light bulb in series, they get hot very fast.
> It is not in my nature to needlessly destroy things. I would
never
> make a very good UL inspector.
>
> The part that you are not aware of is that if you do this
experiment
> at 240V, the winding temperature will skyrocket immediately, and
the
> insulation will be damaged very quickly. No, a short won't
> necessesarily happen right away, but the insulation will become
> so brittle that it won't withstand any bending or vibration
without
> flaking off.
>
> At saturation, a typical 1 ohm dc resistance 120V primary will be
> sucking V*V/R watts at Vpeak. So, if you are sticking 240V into
> the 120V primary, that will result in 240V*240V*2 = 115200Wpeak!
>
> Do you think that might heat things up a bit?
>
> Do you think it matters if it is only saturating 1/4 the time
> at the peaks? (that drops the power to somewhere around a mere
> 28000W. I would have to integrate the waveforms to give you a
> more exact answer.)
>
> About the only thing that will keep these numbers from being
> realized is the fact that the primary has some inductive reactance
> even without a core, and the power cord you connect up with has
> some resistance.
>
> There will be a lot of margin to play with, in
> > terms of power rating of the windings, so I would expect
minutes,
> > not seconds, before destruction. You'd notice, wouldn't you?
The
> > buzzing, the noise, etc. You wouldn't just walk away and let it
> > burn.
>
> Walk away from it? No, not me, I already know what will happen.
> But some newbie might if folks such as you, and William,
> convinced him that it would be ok as long as he just reduced
> the load current.
>
> -Chuck
>

[Chuck Harris]

Ian Iveson wrote:
> The generally accepted code for emphasis is enclosure within
> asterisks.

Says you!

Do what you like, but I will interpret capitals as
> shouting coz I can't be bothered to remember you as an exception.
>
> As for the rest, I have lost you somewhere it seems.

If what you are really saying is you cannot be bothered to
accurately read my posts, then I believe you.

>
> The problem with Casino's transformer is the windings, not the core.
> You explain it how you like, you are just wriggling. What's the
> difference between that transformer and one identical except for
> double the voltage? Windings. That is all.

No, "wriggling" implies I am attempting deception. I am just
telling it like it is. My explanation has remained the same
each time I have made it. It has remained correct each time
I have made it. My statement was that the core will saturate.
It will. Casino asked what would happen, I told him. You're
the one who has muddled things up here.

You started out very wrong, and then started to read up on the
subject. After you found out you were wrong, you started to
WRIGGLE your way towards a more correct explanation. And in
your usual fashion, you start to bluster and blow, and twist
and turn the statements of others to serve your own flaccid
arguments.

>
> You have by no means demonstrated that William is wrong. You, OTOH,
> most definitely *are* wrong.

I have made statements, not demonstrations. This medium does not lend
itself well to demonstrations. William was wrong, period!
Your siding up with his advice makes you wrong too.

>
> It will not go "poof".
>
> Nowhere near "poof".

Prove it!

-Chuck

[Jim Adney]

On Thu, 10 Jul 2003 19:10:23 +0100 "Ian Iveson"
> wrote:

>No, John. The original question demonstrates that
>input voltage is not a fixed quantity. That is what the pertinent
>parts of the thread are about.
>
>The core size has nothing to do with voltage for a given quality
>transformer. I'm sure you know this really. You're just trying to
>wind me up.

You really should listen to John here. At least in this case he's
right.

Once a transformer is built, if you take its design voltage and divide
by its lowest design frequency (ie 50 in the case of a 50-60Hz
transformer) you'll get a number which you should be careful not to
exceed if you should want to use this transformer in a different
application. Do this caluclation using any winding you like on this
transformer, but use the same winding each time you do it.

So if you have a 50-60Hz transformer with a 120V primary that you want
to use on 220V this will only be possible if you have 100Hz power
available.

Really. Honest. No one's trying to wind you up.

-
-----------------------------------------------
Jim Adney
Madison, WI 53711 USA
-----------------------------------------------

[Jim Adney]

On Fri, 11 Jul 2003 20:23:19 +0100 "Ian Iveson"
> wrote:

>The generally accepted code for emphasis is enclosure within
>asterisks. Do what you like, but I will interpret capitals as
>shouting coz I can't be bothered to remember you as an exception.

I've got to admit that I don't recall ever seeing this used, while I
find that simply capitolizing a word or 2 for emphasis is quite
common.

As for shouting (writing out the whole message in caps,) I find that
just as distasteful as you do.

-
-----------------------------------------------
Jim Adney
Madison, WI 53711 USA
-----------------------------------------------

[Patrick Turner]

Ian Iveson wrote:

> The generally accepted code for emphasis is enclosure within
> asterisks. Do what you like, but I will interpret capitals as
> shouting coz I can't be bothered to remember you as an exception.
>
> As for the rest, I have lost you somewhere it seems.
>
> The problem with Casino's transformer is the windings, not the core.
> You explain it how you like, you are just wriggling. What's the
> difference between that transformer and one identical except for
> double the voltage? Windings. That is all.
>
> You have by no means demonstrated that William is wrong. You, OTOH,
> most definitely *are* wrong.
>
> It will not go "poof".
>
> Nowhere near "poof".
>
> As for squirming, which you denied, that is what you are doing now.
> Anything other than proof of "poof" is squirming.
>
> So put up or shut up and accept you are wrong. What's your bet?
>
> cheers, Ian

I heard the weather is very hot in the UK right now....

Patrick Turner.

>
>
> "Chuck Harris" > wrote in message
> ...
> > Hi Ian,
> >
> > When one writes on an ascii only medium, one must do something to
> > show emphasis. The only way I know to do that is write a word
> > or two in capital letters.
> >
> > In the culture of email and news that I adopted, when I started
> > messing with DARPA back in the '70s, that is the way it was done.
> >
> > For example:
> >
> > I HATE YOU AND YOUR MOTHER'S NOSE! <- that would be shouting.
> >
> > I think it would best if we REALLY REALLY try and concentrate....
> >
> > ^^^^^^^^^^^^^
> > Emphasis.
> >
> > I am not shouting at you, I don't think I have ever done.
> However,
> > lacking a method of underlining or italicizing a word, I
> capitalize.
> > Please don't take that as shouting. You will find that a large
> > number of people do the same as I. They aren't shouting either.
> >
> > Shouting is when you write whole angry sounding sentences in
> > capital letters.
> >
> >
> >
> > Ian Iveson wrote:
> > > Now my ISP has switched me back on...
> > >
> > > "Chuck Harris" > wrote:
> > >
> > >
> > >>>[see below]
> > >
> > >
> > > I take capitals as shouting. That is usual, I believe. You
> shout
> > > sometimes.
> > >
> > > It has been said that saturation depends on voltage. It has
> been
> > > said that in order to avoid saturation the core should be made
> > > bigger. You began yourself by explaining that the problem would
> be
> > > related to skimpy margins of iron and copper, neither of which
> is
> > > relevant and seems a bit like axe-grinding to me. For the same
> > > quality transformer at twice the voltage Casino requires the
> same
> > > quantity of both.
> > >
> > > The impression of emphasis, and I did say emphasis, is because
> much
> > > has been said about size of core, and virtually nothing has been
> > > said about number of turns.
> >
> > It really doesn't matter to which you attribute the saturation.
> The
> > OP only had the transformer he had. He wasn't planning on
> changing
> > it so that it would work. That being the case, one can explain
> the
> > reason for its failure to work properly at 2x its rated primary
> voltage
> > as:
> >
> > 1) the core is too small for the flux...
> > 2) there aren't enough turns...
> > 3) the frequency is too low...
> >
> > [note, ... means I left out the boring part of the sentence]
> >
> > Pick one. Whichever makes you happy. They are all true, and they
> all
> > explain the deficiencies of the OP's transformer when run at twice
> > its design voltage.
> >
> > There is no implication in any of the above statements that the
> only
> > way to make a transformer that would handle a higher primary
> voltage
> > is to make one with a larger core... not at all! If you kept the
> > turns the same, that would be one solution. But not necessarily
> the
> > best solution.
> >
> > For instance, I can make a quite dandy 110v to 240v transformer
> with
> > one turn of wire on the primary, and two turns on the secondary.
> I
> > can make this transformer, and it will run cool as a cucumber, all
> > day long at 25Hz! How would I do this???
> >
> > Simple, I would need a core the size of a volkswagon for a silly
> > couple hundred VA transformer. (No, I didn't calculate this...
> > I am purposefully exagerating.)
> > >
> > > Yes, what you said about economy of materials is true, and yes
> you
> > > can argue as has John that you merely meant that the symptom of
> > > saturation appears in the core, rather than the cause.
> > > Nevertheless, by not mentioning number of turns, which is the
> crux
> > > of the matter, the impression was created. I was not the only
> one
> > > who had difficulty following a thread that drifted towards the
> > > moans of small producers in the face of the bean-counting
> > > Juggernauts, blah.
> >
> > I am one of those bean counting Juggernauts! When I design a
> > transformer, or anything else, I have to make it as cheaply as
> will
> > meet the design requirements. That is what it means to be an
> > engineer. I have certain preferences when it comes to
> transformers
> > that I use for myself. I don't want to ever have to redo one,
> > so I would prefer it run at a low flux so that it will run cool
> > and have a minimum amount of vibration. Why? Because the
> insulation
> > will last longer that way. Because I don't like burning myself on
> > transformers that are operating "normally".
> >
> > >
> > > The maths bits are fair enough but rather generalised with
> respect
> > > to the original question of what would be the problem with that
> > > particular transformer at that voltage. That left a gap,
> between
> > > the reminiscences, the business interests, and the maths, where
> I
> > > felt a simple answer might have been.
> >
> > The answer was quite simple and was stated over and over again:
> >
> > This transformer will saturate if run at 2x its rated voltage...
> > Don't do it!
> >
> > The message was needlessly complicated when both you and William
> > came in and said ignore the above message, and go ahead. It'll
> > work just fine with reduced load current.
> >
> > The message was further complicated by guys that tried to judge
> > the "book" by its cover, rather than its contents, and said the
> > "saturation" message is wrong, go ahead run the transformer
> backwards,
> > I do it all the time. Of course they meant they put 6V into a 6V
> > filament winding, or 110V into a 110V secondary of a 220->110V
> step
> > down transformer.
> > >
> > > As far as "boning up" is concerned, I had failed previously to
> grasp
> > > the point that currents due to load cancel, more or less, with
> > > respect to saturation. Something I had not noticed in the fog
> of
> > > knowledge
> > > I already have. There is a reason for this that I have just been
> > > wondering about. It explains why my spice transformer model
> still
> > > has a linear core. I must have known everything else that is in
> the
> > > model otherwise it wouldn't work, I suppose. Specified
> transformers
> > > seem to work OK, so I guess I must have known something
> then...can't
> > > really remember. Must have though because some of my
> transformers
> > > are quite complicated and have quite a lot of wires coming out
> of
> > > them.
> >
> > Spice models for transformers are very simple if they don't take
> into
> > account core characteristics, leakage inductances, and
> capacitances.
> >
> > Most of the time, an ideal transformer will adequately model the
> real
> > thing. It is cases like where you increase the primary voltage,
> or
> > drop the frequency so as to cause the core to head for the
> nonlinear
> > region that are not handled well by an ideal transformer model.
> > >
> > > And for audio-intended transformers you don't really need to
> know.
> > > Saturation will be outside the full-power bandwidth. Er...I
> think
> > > that follows anyway...yes it must.
> >
> > When you take the typical audio output transformer, and head for
> the
> > lower end, 20Hz, or so, the transformer is quite likely to be
> starting
> > to saturate... That is, with the typical garden variety
> transformer.
> > Some of the transformers that top off in the 12 lbs range will do
> much
> > better.
> > >
> > > I am fundamentally uneasy about saturation. Coming from a
> > > background with no practical electronic content, I still am
> > > conscious of mapping mechanical dynamics onto electrical things.
> > > But I have never needed to consider the saturation velocity of
> mass.
> > > Or even its equivalent of the BH curve. Never been into rockets
> and
> > > stuff, or really big things, or really small things moving very
> > > quickly. Ballistics is about my limit.
> >
> > Saturation is all about sudden nonlinearities. Some places you
> see
> > it in mechanical models are springs when they run out of range,
> (eg. a
> > compression spring that has its coils touching, a tension spring
> that
> > has its coils pulled all the way out... Or elastic deformation, a
> > spring that has been bent so much that it won't return.) There
> are
> > plenty of other examples, a shaft on a pot can be thought of as
> being
> > in mechanical saturation when it is turned against its stop.
> >
> > >
> > > What I say is related to my purpose here, which is not to
> > > demonstrate what I know. I have never to my
> > > knowledge been responsible for electrocution, and am careful to
> > > frequently point out that I know virtually nothing about
> anything.
> > >
> > > Your original comment to William:
> > >
> > >
> > >>> Please don't give advice about things you don't understand.
> > >
> > >
> > > Is pathetic and silly. This is a discussion group. I give
> opinions
> > > on lots of things, and got sick of putting "IMO" in every
> sentence.
> > > If there were only Chucks here, and no Williams, there would be
> no
> > > point in coming. But William employs a similar strategy to me,
> it
> > > seems, and it was his perseverance, despite your bullying, that
> > > sorted the wheat from your chaff.
> >
> > It puts my hackles up when someone like William tells the group to
> > ignore a person who provided a correct answer, and go with his
> > uneducated, and incorrect solution. That is where the bullying
> > occurred, not with me.
> >
> > If William came in and stated that he didn't think this was right,
> > and gave his reasons, I would have entered this thread in a more
> > friendly manner. But he didn't. He flat out said the answer was
> > wrong! ... and he couldn't have been more wrong himself.
> >
> >
> > >
> > > The rest of your first contribution was pure hyperbole, intended
> no
> > > doubt to rub in your bullying with a bit of guilt-tripping.
> > >
> > >
> > >>> If you try applying 120V, 60Hz to this 60V, 60Hz winding, you
> > >>> will drive the transformer well into saturation. This will
> > >
> > > cause
> > >
> > >>> the primary current to approach that of the DC wire
> resistance.
> > >
> > >
> > >>> In a word, You will let all the smoke out of your
> > >>> transformer.
> >
> > All of which is completely and verifiably true. Why do you call
> it
> > hyperbole? Just because you don't want it to be true isn't a good
> > reason.
> >
> > As to your assertion that I believe current is measured in
> resistance,
> > do you really think I am that stupid? I left a couple of words
> > out in my haste to get the message done. Tell me you have never
> > done that.
> >
> > "This will cause the primary current to approach that caused by
> the
> > DC resistance alone."
> >
> > Does that sound better to you?
> > >
> > > Only if he gets an arc. No poof. Much slower than that. Care
> to
> > > bet? A current can't approach a resistance, BTW. Pedantry,
> yes,
> > > but OTOH that sentence doesn't make sense anyway. Approach
> usually
> > > implies that it gets close. We don't actually know how close it
> > > would get, because we don't know for what proportion of the time
> the
> > > core would saturate.
> >
> > That is why I said "approach", rather than "will be". I have
> taken
> > the "I" laminations off of transformers and used them as
> demagnetizers,
> > so I do have a fair idea how long it would take to ruin the
> primary.
> > Even with a 50W light bulb in series, they get hot very fast.
> > It is not in my nature to needlessly destroy things. I would
> never
> > make a very good UL inspector.
> >
> > The part that you are not aware of is that if you do this
> experiment
> > at 240V, the winding temperature will skyrocket immediately, and
> the
> > insulation will be damaged very quickly. No, a short won't
> > necessesarily happen right away, but the insulation will become
> > so brittle that it won't withstand any bending or vibration
> without
> > flaking off.
> >
> > At saturation, a typical 1 ohm dc resistance 120V primary will be
> > sucking V*V/R watts at Vpeak. So, if you are sticking 240V into
> > the 120V primary, that will result in 240V*240V*2 = 115200Wpeak!
> >
> > Do you think that might heat things up a bit?
> >
> > Do you think it matters if it is only saturating 1/4 the time
> > at the peaks? (that drops the power to somewhere around a mere
> > 28000W. I would have to integrate the waveforms to give you a
> > more exact answer.)
> >
> > About the only thing that will keep these numbers from being
> > realized is the fact that the primary has some inductive reactance
> > even without a core, and the power cord you connect up with has
> > some resistance.
> >
> > There will be a lot of margin to play with, in
> > > terms of power rating of the windings, so I would expect
> minutes,
> > > not seconds, before destruction. You'd notice, wouldn't you?
> The
> > > buzzing, the noise, etc. You wouldn't just walk away and let it
> > > burn.
> >
> > Walk away from it? No, not me, I already know what will happen.
> > But some newbie might if folks such as you, and William,
> > convinced him that it would be ok as long as he just reduced
> > the load current.
> >
> > -Chuck
> >

[William Sommerwerck]

>> If there were only Chucks here, and no Williams, there would be
>> no point in coming. But William employs a similar strategy to me,
>> it seems, and it was his perseverance, despite your bullying, that
>> sorted the wheat from your chaff.

> It puts my hackles up when someone like William tells the group to
> ignore a person who provided a correct answer, and go with his
> uneducated, and incorrect solution. That is where the bullying
> occurred, not with me.

> If William came in and stated that he didn't think this was right,
> and gave his reasons, I would have entered this thread in a more
> friendly manner. But he didn't. He flat out said the answer was
> wrong! ... and he couldn't have been more wrong himself.

You're both right.

The original post wasn't "correct," simply because it gave an answer without an
explanation.

But I was guilty of doing what I so vehemently despise in others -- gainsaying
someone's remark without giving an explanation of my own. I owe the group an
apology, and offer it, forthwith.

[Gerald Stombaugh]

I happened to be designing a transformer when I took a break and
found this discussion. Maybe if I describe this transformer and the
simple math used, it might clear the air.

120v-60Hz Primary input.
Secondary: 12.16V @ 6.61 amps rms for 80Watt lamp load.
Max. temperature rise over 25C = 70C.

Now the most important calculation for any requirement: How
big does the transformer have to be to meet these requirements?

Well, I use: Square root of the watts (va) out divided by 7 equals
the core area. This is an EI transformer so the answer of 1.27 would
equate to a 1-1/8" center leg lam. and a stack buildup of 1-1/8"
This unit would be wound on a three flange glass filled, nylon bobbin.

My curves tell me that in order to meet the 70C heat rise that I can
tolerate 15.20watts combined core and copper losses.

Second calculation: How many primary turns must I use in order to
STAY AWAY FROM SATURATION?? I choose a flux density
of 100,000 lines per square inch density. The formula is :

Ex10^8/4.44 x f x B x A x K.

120 x 10^8/4.44 x 60 x100,000 x .95 = 418 turns on the primary
winding.

4.44 is the factor for sine wave (4.00) for square wave, B is the
flux density, A is the centerleg x the stack in inches and K is the
stacking factor (.95) of the lamination ala that the effective amount
of steel is 95% because of burrs etc.

I can determine the primay current of the transformer by denoting
80 va out plus 15.2 va losses/120v = .793 amps. We determine that
80va/12.16v = 6.61 amperes. SO--
I choose 418 turns of 23 awg wire for the primary and 45 turns 2#17
awg for the secondary.
Rp = 4.25 ohms---Rs = .059 ohms. Pri cu. wgt. is .325# and Sec. cu.
wgt. is .293#

Pri voltage drop is .993a x 4.25 ohms for 3.37 volts. Sec. voltage
drop is 6.61 amps x .059 ohms or .39 volts.

Primary copper losses = .993 x .993 x 4.25 ohms = 2.67 watts:

Secondary copper losses= 6.61 x 6.61 x .059 ohms + 2.58 watts.

Core loss per Tempel steel curves denote 3.5 watts per pound
which is 2.12 lbs for a loss of 7.40 watts.
Total losses of 12.65 gives me a temperature rise of 68C.

The insulation system is rated for 130C total temp. which would allow
me a temp rise of about 100C but the lamp is very close to the
transformer so I design more conservatively.

The transformer, made in Mexico will sell for $6.90 in 2500
quantities.

NOW: If the customer mistakenly applied 140V to the primary
of this transformer you would have HEAP BIG SMOKE AND DEATH
in about 4 min--32 seconds by my curves. We utilize a thermal device
in the primary winding which will immediately open the moment the
copper temp. achives 130C which is usually just over a minute.

That is your lesson for today.

Gerald Stombaugh in hotter than hell Tucson, Az.







On Fri, 11 Jul 2003 09:06:56 -0500, (John Byrns) wrote:

>In article >, "Ian
>Iveson" > wrote:
>
>> "John Byrns" > wrote in
>> > >
>> > > For mains trannies they require the voltage to be specified, but
>> it
>> > > has no effect on core size.
>> >
>> > The input voltage and frequency for a mains transformer are fixed
>> > quantities, so it is only natural to rate the windings in terms of
>> these.
>>
>> No, John. The original question demonstrates that
>> input voltage is not a fixed quantity. That is what the pertinent
>> parts of the thread are about.
>
>Yes, Ian, the input voltage is a fixed quantity, here in the US the
>domestic supply voltage provided to residences is 120 volts at 60 Hz, in
>Europe it is something different, but still fixed. The original posters
>question was, can I hook the 60 volt winding of a power transformer to
>that 120 volt supply without damaging the transformer? It is not
>necessary to get the data on the core and calculate current to provide a
>reasonably accurate to that question. My point is that I am not a
>transformer designer, I am only a user. Sure the transformer designer is
>going to do all the current calculations you talk about, but why should I
>do that as a user, when I can much more easily determine approximately
>what voltage a transformer winding can handle at a given frequency, based
>on the original voltage and frequency ratings of the transformer, which in
>commercial grade power transformers bear a close relationship to the
>saturation point?
>
>> The core size has nothing to do with voltage for a given quality
>> transformer. I'm sure you know this really. You're just trying to
>> wind me up.
>
>Although I may be mistaken, I don't think I have made any claims in this
>thread about core size vs. voltage, I don't think I have even mentioned
>core size. I think you are confusing my comments with those of someone
>else. You are the one that insists on injecting core size into the
>discussion.
>
>I'm not trying to wind you up, I'm simply saying that from the perspective
>of a user of power transformers, the voltage spec. of a winding is the
>most relevant for determining if a given voltage will saturate the core,
>the current is only of interest to the transformer user in calculating the
>power the transformer can pass without overheating. As a user,
>calculating the saturation point based on current is a much more complex
>undertaking than simply looking at the rated voltage and frequency.
>Calculating saturation based on current requires me to have knowledge of
>several characteristics of the core, and do relatively involved math, why
>go to all that trouble when I know approximately how commercial power
>transformers are designed with respect to saturation vs. voltage, and
>frequency?
>
>
>Regards,
>
>John Byrns
>
>
>Surf my web pages at, http://users.rcn.com/jbyrns/



-----= Posted via Newsfeeds.Com, Uncensored Usenet News =-----
http://www.newsfeeds.com - The #1 Newsgroup Service in the World!
-----== Over 80,000 Newsgroups - 16 Different Servers! =-----

[David Stinson]

Wow...
There's a big fight going on and I'm not getting blamed for it...

I feel so lonely....

[Choky]

"Gerald Stombaugh" > wrote in message
...
> I happened to be designing a transformer when I took a break and
> found this discussion. Maybe if I describe this transformer and the
> simple math used, it might clear the air.
>
> 120v-60Hz Primary input.
> Secondary: 12.16V @ 6.61 amps rms for 80Watt lamp load.
> Max. temperature rise over 25C = 70C.
>
> Now the most important calculation for any requirement: How
> big does the transformer have to be to meet these requirements?
>
> Well, I use: Square root of the watts (va) out divided by 7 equals
> the core area. This is an EI transformer so the answer of 1.27 would
> equate to a 1-1/8" center leg lam. and a stack buildup of 1-1/8"
> This unit would be wound on a three flange glass filled, nylon bobbin.
>
> My curves tell me that in order to meet the 70C heat rise that I can
> tolerate 15.20watts combined core and copper losses.
>
> Second calculation: How many primary turns must I use in order to
> STAY AWAY FROM SATURATION?? I choose a flux density
> of 100,000 lines per square inch density. The formula is :
>
> Ex10^8/4.44 x f x B x A x K.
>
> 120 x 10^8/4.44 x 60 x100,000 x .95 = 418 turns on the primary
> winding.
>
> 4.44 is the factor for sine wave (4.00) for square wave, B is the
> flux density, A is the centerleg x the stack in inches and K is the
> stacking factor (.95) of the lamination ala that the effective amount
> of steel is 95% because of burrs etc.
>
> I can determine the primay current of the transformer by denoting
> 80 va out plus 15.2 va losses/120v = .793 amps. We determine that
> 80va/12.16v = 6.61 amperes. SO--
> I choose 418 turns of 23 awg wire for the primary and 45 turns 2#17
> awg for the secondary.
> Rp = 4.25 ohms---Rs = .059 ohms. Pri cu. wgt. is .325# and Sec. cu.
> wgt. is .293#
>
> Pri voltage drop is .993a x 4.25 ohms for 3.37 volts. Sec. voltage
> drop is 6.61 amps x .059 ohms or .39 volts.
>
> Primary copper losses = .993 x .993 x 4.25 ohms = 2.67 watts:
>
> Secondary copper losses= 6.61 x 6.61 x .059 ohms + 2.58 watts.
>
> Core loss per Tempel steel curves denote 3.5 watts per pound
> which is 2.12 lbs for a loss of 7.40 watts.
> Total losses of 12.65 gives me a temperature rise of 68C.
>
> The insulation system is rated for 130C total temp. which would allow
> me a temp rise of about 100C but the lamp is very close to the
> transformer so I design more conservatively.
>
> The transformer, made in Mexico will sell for $6.90 in 2500
> quantities.
>
> NOW: If the customer mistakenly applied 140V to the primary
> of this transformer you would have HEAP BIG SMOKE AND DEATH
> in about 4 min--32 seconds by my curves. We utilize a thermal device
> in the primary winding which will immediately open the moment the
> copper temp. achives 130C which is usually just over a minute.
>
> That is your lesson for today.
>
> Gerald Stombaugh in hotter than hell Tucson, Az.
>


****ing period!!

good engineering is not bigger than life.
all we need is few calculations from olde books,few graphs and conditio sine
qua non is also that we don't have short circuit between two earphones.
one of biggest threads lately is about flux or not flux........
o tempora ,o mores!

hehe
cheers to all Rodents!

--
Choky
Prodanovic Aleksandar
YU

[Choky]

"David Stinson" > wrote in message
...
> Wow...
> There's a big fight going on and I'm not getting blamed for it...
>
> I feel so lonely....
why?
that is fun on RAT ;))
flux or not flux...........

bzzzzzzzzpoooooooffffffffzzzzzzzzzzsssssssssskkkc.


--
Choky
Prodanovic Aleksandar
YU

[Ian Iveson]

Jim, you are quite right. But I don't think John and I disagree
about that.

Since the mains is not 100Hz, your's is not a practical
approach to transformer design.

If Casino simply wants a transformer for twice the mains voltage
whilst not changing his other requirements or his mains frequency,
then he wants one with more turns of thinner wire. That is all.

I don't think you have read my posts very well. Or John's. Or we
are not making ourselves clear. Not coming from an electronics
background, I sometimes don't fit in with received wisdom. Perhaps
I need a translator?

Current saturates. Not voltage. The relationship between the
current and the voltage depends on frequency. Hence the voltage at
which a transformer saturates depends on frequency. The *current*
at which the transformer saturates does *not* depend on frequency.

John got very cloudy, and increasingly muddled about saturation, the
more he got into fencing for the sake of it. He knows about it if
he considers the issue in cooler light.

But listen...Adrian and I seem to be the only ones who have learned
anything here. Since we have done learning now, and you apparently
know already, we might as well stop, eh?

Unless you want to bet on whether this 120V tranny will go "POOF" or
"BzzzzzzZzzzZzzZzZzZZZZ......pft, or whatever, at 240V.

cheers, Ian


"Jim Adney" > wrote in message
...
> On Thu, 10 Jul 2003 19:10:23 +0100 "Ian Iveson"
> > wrote:
>
> >No, John. The original question demonstrates that
> >input voltage is not a fixed quantity. That is what the
pertinent
> >parts of the thread are about.
> >
> >The core size has nothing to do with voltage for a given quality
> >transformer. I'm sure you know this really. You're just trying
to
> >wind me up.
>
> You really should listen to John here. At least in this case he's
> right.
>
> Once a transformer is built, if you take its design voltage and
divide
> by its lowest design frequency (ie 50 in the case of a 50-60Hz
> transformer) you'll get a number which you should be careful not
to
> exceed if you should want to use this transformer in a different
> application. Do this caluclation using any winding you like on
this
> transformer, but use the same winding each time you do it.
>
> So if you have a 50-60Hz transformer with a 120V primary that you
want
> to use on 220V this will only be possible if you have 100Hz power
> available.
>
> Really. Honest. No one's trying to wind you up.
>
> -
> -----------------------------------------------
> Jim Adney
> Madison, WI 53711 USA
> -----------------------------------------------

[Ian Iveson]

"Gerald Stombaugh" > wrote

>>> ...[excellent worked example]

Thanks very much Gerald. You are *very* convincing!

Could you describe the noise, and perhaps in the later stages the
smell, of this process of destruction?

cheers, Ian

[Ian Iveson]

Your entire post [below] demonstrates exactly what I mean. If you
transform equations willy-nilly you can end up very confused about
the direction of causality in a given practical context.

See Gerald's excellent illustrative example.

cheers, Ian

"Jim Adney" > wrote in message
...
> On Sun, 6 Jul 2003 16:33:27 +0100 "Ian Iveson"
> > wrote:
>
> >Equations can be rearranged. You can just as easily say that
> >voltage heats resistors but it would be a highly misleading
> >confusion of dependent and independent variables in most
contexts.
>
> You seem to be arguing that
>
> P = Isquared * R
>
> is a more correct form and that
>
> P = Vsquared/R is less correct.
>
> How about P = I * V?
>
> >So it is with cores. They are power rated, not voltage rated.
>
> Yes, cores themselves are often sold with a power rating, but this
has
> everything to do with the space available for copper around the
core
> (the winding window) and nothing to do with the flux density in
the
> core.
>
> For example, if I want to design a transformer using standard E-I
> cores I would first pick a size that would allow a large enough
> winding window. Then I have to use a thick enough stack of these
> pieces to support the flux that I will get from my voltage, number
of
> turns and frequency. I may have to adjust N, or the stack
thickness to
> get what I need, and I may even have to pick a different size E-I
part
> if things get out of hand. It's an iterative process if you want
to
> get the most economical finished part. In practice this isn't done
> very well unless the production volume is large.
>
> In fact, if you could wind your transformer with superconducting
wire,
> you could put ANY amount of power thru it, as long as the windings
and
> the core were properly designed with respect to the voltage and
> frequency.
>
> The core is there to handle the flux. The copper is there to
handle
> the current.
>
> What most people don't get, is that in a real transformer the flux
> peaks are the highest when there is NO LOAD on the transformer.
>
> -
> -----------------------------------------------
> Jim Adney
> Madison, WI 53711 USA
> -----------------------------------------------

[Ian Iveson]

"Jim Adney" > wrote

> Well, it's important to keep in mind that the V and I approaches
to
> flux use completely different formulas. They are both RIGHT, but
one
> (the V method) is a LOT easier to get close answers with.

But it is indirect. It depends on frequency. Current does not.

But now we have Gerald, who seems to know what he is on about.

cheers, Ian

[Ian Iveson]

"Patrick Turner" > wrote

> I heard the weather is very hot in the UK right now....

You might learn something from Gerald if I am too hot for you,
Patrick.

Looks like he has successfully made some transformers that work
properly.

cheers, Ian

PS I see you manage to guzzle heaps of bandwidth even when you have
nothing to say.

[Ian Iveson]

We all know you're behind this, Stinson.

It wasn't a fight until you injected your poisonous cynicism into a
friendly and productive discussion. So it's is all your fault.

Feeling better?

cheers, Ian

"David Stinson" > wrote in message
...
> Wow...
> There's a big fight going on and I'm not getting blamed for it...
>
> I feel so lonely....

[Chuck Harris]

Why Ian, in your effort to elect Gerald your new best
friend, you completely failed to notice how he
calculated the number of turns in his primary!

-----gerald's method----

Ex108/4.44 x f x B x A x K.

120 x 108/4.44 x 60 x100,000 x .95 = 418 turns on the primary

-----gerald's method----

Now, I bet I know why you missed it; because, being as you are
a self professed nonengineer, you probably didn't know that
engineer speak for voltage is "E".

But just so you won't miss out, the "120" above is 120 Volts RMS,
and the "60" is 60Hz, and the "100,000" is a number of lines of
flux that Gerald picked based on the core loss he was hoping
to achieve.

And surprise, out of the equation (that has no mention of current)
comes 418 turns of primary.

Another silly little thing you will notice about the above
equation, is it is completely independent of the current in
the transformer winding! Gerald would get the same number
of turns for the same flux in the core if the primary drew
1A fully loaded, or 101A... or 1001A.

The funny thing is it is the same equation I use when I design
transformers. I sometimes look it up in the "ITT Reference Data
for Radio Engineers", sometimes in RDH4, sometimes in "Electronics
Engineers' Handbook"... It is always the same??? It never uses
current??? It MUST be wrong!

If you re read Gerald's article, you will notice that the only
use he had for current was in calculating the wire size used
in his windings. He also used the current to calculate the
Trise of his windings. You know, I^2 R and all that.

Gerald, didn't willy nilly transform equations, like you have
suggested, he left them the way they are most useful....the
way he found them.

-Chuck


Ian Iveson wrote:
> Your entire post [below] demonstrates exactly what I mean. If you
> transform equations willy-nilly you can end up very confused about
> the direction of causality in a given practical context.
>
> See Gerald's excellent illustrative example.
>
> cheers, Ian
>
> "Jim Adney" > wrote in message
> ...
>
>>On Sun, 6 Jul 2003 16:33:27 +0100 "Ian Iveson"
> wrote:
>>
>>
>>>Equations can be rearranged. You can just as easily say that
>>>voltage heats resistors but it would be a highly misleading
>>>confusion of dependent and independent variables in most
>
> contexts.
>
>>You seem to be arguing that
>>
>>P = Isquared * R
>>
>>is a more correct form and that
>>
>>P = Vsquared/R is less correct.
>>
>>How about P = I * V?

[John Byrns]

In article >,
(John Byrns) wrote:

> In article >, "Ian
> Iveson" > wrote:
>
> > Are you completely daft or what?
> >
> > That is what I have been saying. Double the voltage, and you need
> > double the turns. But you only need half the cross-sectional area,
> > so it should all fit together with the same core, give or take a
> > little to take account of scale effects of geometry and losses.
>
> That has nothing to do with the original question, but going with it, if
> we need twice as many turns, and each has half the cross sectional area,
> doesn't that mean the copper losses for the winding, at a given power,
> will double?

oops, forget it, that was wrong, that's what I get for trying to do simple
math in my head.


Regards,

John Byrns


Surf my web pages at, http://users.rcn.com/jbyrns/

[Ian Iveson]

Actually I wasn't making myself perfectly clear anyway.
"Cross-sectional area" could be ambiguous...I meant of the wire.

And I don't think it would work out quite so tidily in practice,
which is why I hedged round the scaling of losses.

> oops, forget it, that was wrong, that's what I get for trying to
do simple
> math in my head.

That's why I want an active core for my preferred simulator. Too
strung out for mental arithmetic. Forget the question by the time I
get half way to the answer.

cheers, Ian

[David Stinson]

Ian Iveson wrote:
>
> We all know you're behind this, Stinson.
>
> It wasn't a fight until you injected your poisonous cynicism into a
> friendly and productive discussion. So it's is all your fault.
>
> Feeling better?

Ahhhhhhhh.... Thanks, Ian.
Like relaxing in an old, familiar, comfortable chair....
Dave S.

[Jim Adney]

On Sun, 13 Jul 2003 00:24:40 +0100 "Ian Iveson"
> wrote:

>One thing I have just begun to wonder is how EI-core folk deal with
>the directional property of GOSS. A wrapped core like a torroid or
>a C can complete the magnetic circuit with all the grain aligned
>with the field. But an EI transformer has the grain in the wrong
>direction down the I and down the side of the E.

EI cores are punched from strips of sheet steel slit to exactly the
right width. The dies are arranged such that the 2 Es face one another
leaving slots which become the Is. This means that the field can be
parallel to the grain in all the areas of the EI core except for the
main body of the E. But if you look at it, you will see that the flux
density in this area is about half that of the center leg.

The advantages of GOSS only make a difference where the flux density
approaches saturation, so this approach takes care of us.

-
-----------------------------------------------
Jim Adney
Madison, WI 53711 USA
-----------------------------------------------

[Jim Adney]

On Sat, 12 Jul 2003 20:58:19 +0100 "Ian Iveson"
> wrote:

>Your entire post [below] demonstrates exactly what I mean. If you
>transform equations willy-nilly you can end up very confused about
>the direction of causality in a given practical context.

I think you're on extremely thin ice in claiming, if I understand you
corrrectly, that there is causality implied in the way an equation is
stated. This gets philosophical very quickly, but physicists like
Steven Hawking work really hard to find a time direction preference in
the equations that we find that govern our world. In the end I'm not
sure that he, or anyone else, has found any examples, and certainly
there are none in the areas that we're discussing.

All the equations that I'm familiar with are time INvariant, meaning
that they work equally well in both directions in time.

An equality is nothing more than the statement that 2 quantities are
exactly the same. You're trying to read too much into this. While your
misunderstanding of this is common, I'm afraid it's still wrong.

Once you realize that equations are relationships which are simply
true (equal) under all circumstances, you will find that the realm of
problems which are solvable has suddenly gotten much larger.

I will try to find the exact title and author for the book that opened
my eyes on the current topic (transformers.) It certainly improved my
understanding, which started out where you are now.

-
-----------------------------------------------
Jim Adney
Madison, WI 53711 USA
-----------------------------------------------

[Jim Adney]

On Sat, 12 Jul 2003 21:05:31 +0100 "Ian Iveson"
> wrote:

>"Jim Adney" > wrote
>
>> Well, it's important to keep in mind that the V and I approaches
>to
>> flux use completely different formulas. They are both RIGHT, but
>one
>> (the V method) is a LOT easier to get close answers with.
>
>But it is indirect. It depends on frequency. Current does not.

But you only know the current if you also know the inductance. Then
you can add in f and find the current.

This only works if you're willing to "design" by trial and error: That
is, make up a prototype, measure the L and whatever else you're
interested in and then make another guess to refine the "design."

If you work from the V perspective you can just work thru everything
before making your first prototype. That's what I do, and that's what
the people I buy custom transformers do. I know they get quotes back
to me in less time than it would take to actually build up a
prototype, and the calculations can all be computerized these days so
they only take a few moments to actually figure out what is needed.

-
-----------------------------------------------
Jim Adney
Madison, WI 53711 USA
-----------------------------------------------

[Jim Menning]

"Tim Williams" > wrote in message
...

>
> > Jim Adney
> > Madison, WI 53711 USA
>
> Hey whaddya know, I'm in Beloit!
>
> Tim
>


Then I expect both of you to be at the Wisconsin Antique Radio Club
swap meet in Milwaukee next weekend.

Bring some stuff to sell or just to show off!

Sunday, July 20

5917 S. Howell Street
Milwaukee, WI
(Immediately West of the Airport)
Hours: 8am-3pm
(But if you are smart, you'll be there earlier!)

jim menning

[Jim Adney]

On Sun, 13 Jul 2003 00:24:40 +0100 "Ian Iveson"
> wrote:

>This is quite an interesting article, Jim.
>
>http://www.metglas.com/downloads/lit/amor_elec_pow_dist_appl.pdf
>
>It has commercial interests attached of course, and compares GOSS
>with amorphous iron, but is informative nonetheless.

Okay, I looked it up and read it. It's mostly a historical summary of
METGLAS with a bit of technical data in there. It seems mainly to be
aimed at selling METGLAS core power transformers to the power
industry. Some of the data looks interesting, but most of it is given
without explanation, such as the data on telephone interference....

One thing that seemed really odd was the section claiming to show that
a METGLAS core would have a lifetime of at least 30 years. They really
didn't come right out and say that METGLAS cores deteriorate over
time, but the implication is there, and this would certainly bother me
if I was considering investing in a bunch of them. Transformer cores
should be durable over centuries or longer. If this stuff breaks down
in decades it's got an uphill battle ahead of it.

There's also nothing in there about the comparative economics of
making high efficiency transformers by simply using a bit more iron
and copper. This has become popular in transformers and motors in the
past couple of decades and may be a better investment than a METGLAS
cored transformer.

Nothing that I could see on transformer design proceedures other than
specs on saturation flux density and the very narrow hysterisis curve,
which shows low losses.

Certainly nothing about running a given transformer at twice the
design voltage.

SO who's got that transformer? Can we actually do the test and settle
the debate?

-
-----------------------------------------------
Jim Adney
Madison, WI 53711 USA
-----------------------------------------------

[Jim Adney]

On Sun, 13 Jul 2003 02:29:47 -0500 "Tim Williams"
> wrote:

>"Jim Adney" > wrote in message
...
>> Sure, that's a no brainer. If you have a commercially built
>> transformer with a winding rated at 120V/60 Hz I would be ready and
>> willing to bet that it will be in instant trouble if you put 240V/60Hz
>> across it.
>>
>> Let's try it!
>
>Uh oh...
>
>*Wraps arms tightly around stock of transformers*

Don't worry. For peace of mind I'd include a fuse anyway. Fuses are a
lot cheaper than transformers. ;-)

-
-----------------------------------------------
Jim Adney
Madison, WI 53711 USA
-----------------------------------------------

[Jim Menning]

"Jim Adney" > wrote in message
...

> >
> >Bring some stuff to sell or just to show off!
>
> Maybe we should just bring a stock of antique transformers and blow
> them up by running them on 240V. You know, sort of like the engine
> blows they do at drag races... ;-p
>
>
Little late for the 4th of July, isn't it?

jim menning

[--Bill--]

Jim Adney wrote:
> On Sun, 13 Jul 2003 15:49:18 GMT "Jim Menning" >
> wrote:
>
>
>>"Tim Williams" > wrote in message
...
>>
>>
>>>> Jim Adney
>>>> Madison, WI 53711 USA
>>>
>>>Hey whaddya know, I'm in Beloit!
>
>
>>Then I expect both of you to be at the Wisconsin Antique Radio Club
>>swap meet in Milwaukee next weekend.
>>
>>Bring some stuff to sell or just to show off!
>
>
> Maybe we should just bring a stock of antique transformers and blow
> them up by running them on 240V. You know, sort of like the engine
> blows they do at drag races... ;-p

Considering that so many Airline and Zenith (yeah, blame it all on 6X5s)
are barely capable of 125 volts, this could be a sleeper. They'll burn
up before the crowd gets seated.
-Bill

[Tim Williams]

"Jim Adney" > wrote in message
...
> Don't worry. For peace of mind I'd include a fuse anyway. Fuses are a
> lot cheaper than transformers. ;-)

Not to mention the $0.05 fuses being protected by $15 three-legged fuses ;)

Tim

--
In the immortal words of Ned Flanders: "No foot longs!"
Website @ http://webpages.charter.net/dawill/tmoranwms

[Michael A. Terrell]

Jim Adney wrote:
>
> On Sun, 06 Jul 2003 08:34:29 -0400 Chuck Harris >
> wrote:
>
> >Ian,
> >
> >Please point out the article that "concentrates only on primary voltage
> >and core size".
>
> That would be my posts. I've tried to lay out the theory, but I
> realize I'm swimming against the current here, even though I'm
> probably one of the few here who's actually made a living doing this.
>
> There are others here who also clearly understand, and I appreciate
> their posts, but we are in the minority.
>
> Most people here are just going on what they have gleaned from a few
> web sites and don't really understand the theory. Once you understand
> the theory you will quickly understand why any transformer designer
> will want to know voltage and frequency first, and power (or current)
> right after that.
> Jim Adney

Jim, some of the comment others have made remind me of a ham I went to
high school with. He was building a linear with a 304TH and he needed
to wind a plate transformer. He found a surplus core and unwound it,
then called me to see if I could tell him where to #2 AWG magnet wire to
re-wind the primary of his plate transformer.

He told me he had calculated that he was going to wind a 10 turn
primary, put a single layer of wax paper and wind a secondary, and only
fill about one third of the core with copper. It took a lot of
explaining that it wouldn't work, and why. He just couldn't wrap his
mind around the relationship between (turns per volt, core size, and
frequency), even with the math in front of him.

He believed the only important numbers were the turns ratio, and he
planned to use 500 Circular mils per amp. He was convinced his 24 volt
per turn would work as well as the proper numbers, and the only reason
he finally wound it with the proper wire was he couldn't find the absurd
size he wanted.
--


Michael A. Terrell
Central Florida

[t.hoehler]

>
> I realize that most people here are not comfortable with a lot of
> equations, but you won't find a LOT of equations here, or a lot of
> difficult math. The truly amazing thing about this stuff is that it
> really is so EASY once someone lays it all out for you.
>
> -
> -----------------------------------------------
> Jim Adney
> Madison, WI 53711 USA
> -----------------------------------------------

I have been lurking, just reading the posts about reversing power
transformers. A step down xformer from 120 to 60 volts has a totally
different kind of winding for the primary and the secondary. Before you
shout "DUH", think about the turns versus the voltage in the primary. That
choice of wire size and number of turns in the primary is not accidental. It
is based on current capacity, flux saturation in the core, inductive
reactance, many factors. I promise you, with today's price point
engineering, there is not one extra turn on that primary that is not
absolutely needed. Having said that, you just flat don't have enought turns
on the secondary to use it as a primary. Additionally, if you have torn
apart enough power transformers, you know well that the primary is almost
always wound next to the core. That is not accidental either. The reason
that winding is on first is to ensure adequate coupling of the winding turns
to the core structure. These are simply observations on my part, but they
sure make sense to me.
Regards,
Tom