I am well aware of the controversy surrounding the purported transformer
test that gives little information as to which transformers were actually
tested and nothing about how they were tested and what the actual technical
specifications are for each transformer.

I thought it more appropriate for RAT listers to provide real world
assessments of their own empirical use of any of the readily available
transformers. To that end, I have had in my SE amp over the last 6 years
two well regarded output transformers, the Magnequest DS025 and the Tamura
5002.

I had the Magnequest DS025 in a SE amp for about 5 years. Over the course
of that time, I had several variations on the circuit topology shown he

http://home.comcast.net/~vhbrown/misc/2A3SEamp.gif

The circuit shown is the one that I finally settled upon except that the
AVVT 2A3 was the tube of choice after all the listening to different tubes
was completed (note a slight error in the schematic as the two 2.5V
secondaries were in parallel for the 2A3 application, the series connection
is for the 300B). Also tested in this amp were engraved base WE 300B, AVVT
300B, and the chinese and RCA 2A3. The B+ was adjusted by inserting a cap
before the first inductor to increase the plate voltage for the 300B so that
it would operate at the correct operating conditions. Let me say right off
that, with the exception of the chinese 2A3, I enjoyed every minute of
listening to this amp. In the end, I thought that the 2A3 was a much
better match for the DS025 than the 300B of any variant. I don't really
want to speculate, but it is possibly because of the higher inductance at
the lower current of the 2A3 over the 300B or it could be just the matching
of the 2A3/transformer/speakers was better. The speakers over the 6 years
have been modified Altec A7s. I don't really think it is necessary to go
into audiospeak about glorious midrange, crystalline highs and veils being
lifted etc. I think it more important to affirm that I enjoyed listening to
this setup and that I think any listener on a budget who wants to make a SE
amp would be satisfied with the Magnequest DS025.

Now on to the Tamura 5002. The sale of the WE 300Bs to a gentleman in Hong
Kong afforded me the opportunity to purchase any transformers I wanted and I
chose the amorphous core Tamura 5002 based solely on their reputation. I
reconfigured the amp to this topology:

http://home.comcast.net/~vhbrown/misc/300BSEamp.gif

Note again there is an error, this time it the 351V going to the 5002 is
closer to 385V At any rate, these transformers cost almost $1000 more than
the Magnequest! Do they sound $1000 better? Well, that is one of those
questions that is impossible to answer. I think the fairest answer is that
I am very happy that I bought them. These transformers work wonderfully
with the AVVT 300Bs where the DS025s just did not seem to mate as well with
them. I think it is safe to say that the AVVT 2A3 and 300Bs have
substantially the same sound and that these two transformers allowed each
tube to show their respective strengths. The difference between the
2A3/DS025 and the 300B/5002 is not subtle. With respect to frequency
response, the 300B/5002 has considerably more authoritative bass. The
midrange is more liquid (I really hate audiospeak but the metaphor is so
accurate) and the soundstage is really breathtaking. The entire audio
presentation is just more three dimensional and reed instruments, strings,
and female voices just sound more realistic. Well recorded female voice is
stunning and a well recorded string bass has all the woody resonance and
punch that it should while horns have all the appropriate overtones and bite
that they should.

I suppose it is easy to summarize by saying that it is impossible to get any
kind of a clear understanding of the sound or quality of an output
transformer by reading a silly revue like the one that is periodically
dredged up as an attempt to discredit certain winders. I would guess that
any of the low to medium priced transformers by respected manufacturers (and
there are many!) will provide hours of satisfying musical experience and a
wonderful way to get into this great hobby on a budget. It is now clear to
me that there are some extraordinary and elegant transformers available to
the hobbyist, but the price of admittance is high.

Regards,
Sandy Brown

Centifolia wrote:

I am well aware of the controversy surrounding the purported transformer
test that gives little information as to which transformers were actually
tested and nothing about how they were tested and what the actual technical
specifications are for each transformer.

Indeed. one would need to take a grain of salt with those tests.



I thought it more appropriate for RAT listers to provide real world
assessments of their own empirical use of any of the readily available
transformers. To that end, I have had in my SE amp over the last 6 years
two well regarded output transformers, the Magnequest DS025 and the Tamura
5002.

I had the Magnequest DS025 in a SE amp for about 5 years. Over the course
of that time, I had several variations on the circuit topology shown he

http://home.comcast.net/~vhbrown/misc/2A3SEamp.gif

The circuit shown is the one that I finally settled upon except that the
AVVT 2A3 was the tube of choice after all the listening to different tubes
was completed (note a slight error in the schematic as the two 2.5V
secondaries were in parallel for the 2A3 application, the series connection
is for the 300B). Also tested in this amp were engraved base WE 300B, AVVT
300B, and the chinese and RCA 2A3. The B+ was adjusted by inserting a cap
before the first inductor to increase the plate voltage for the 300B so that
it would operate at the correct operating conditions. Let me say right off
that, with the exception of the chinese 2A3, I enjoyed every minute of
listening to this amp. In the end, I thought that the 2A3 was a much
better match for the DS025 than the 300B of any variant.

It seems you are not needing a high level of power, maybe a
watt, so a 2A3 should do all that you'd want , and having
an amp capable of 8 watts, or 80 watts, is not needed.

I don't really
want to speculate, but it is possibly because of the higher inductance at
the lower current of the 2A3 over the 300B or it could be just the matching
of the 2A3/transformer/speakers was better.

The inductance, whatever it measures at 70 mA, won't be much less
at 35mA. Air gapped transformers have a more constant inductance
regardless of DC flow, within their non saturated region.
The inductance of an SE OPT is like that of a DC choke.

The speakers over the 6 years
have been modified Altec A7s. I don't really think it is necessary to go
into audiospeak about glorious midrange, crystalline highs and veils being
lifted etc. I think it more important to affirm that I enjoyed listening to
this setup and that I think any listener on a budget who wants to make a SE
amp would be satisfied with the Magnequest DS025.

Now on to the Tamura 5002. The sale of the WE 300Bs to a gentleman in Hong
Kong afforded me the opportunity to purchase any transformers I wanted and I
chose the amorphous core Tamura 5002 based solely on their reputation. I
reconfigured the amp to this topology:

http://home.comcast.net/~vhbrown/misc/300BSEamp.gif

Note again there is an error, this time it the 351V going to the 5002 is
closer to 385V At any rate, these transformers cost almost $1000 more than
the Magnequest! Do they sound $1000 better? Well, that is one of those
questions that is impossible to answer. I think the fairest answer is that
I am very happy that I bought them. These transformers work wonderfully
with the AVVT 300Bs where the DS025s just did not seem to mate as well with
them. I think it is safe to say that the AVVT 2A3 and 300Bs have
substantially the same sound and that these two transformers allowed each
tube to show their respective strengths. The difference between the
2A3/DS025 and the 300B/5002 is not subtle. With respect to frequency
response, the 300B/5002 has considerably more authoritative bass. The
midrange is more liquid (I really hate audiospeak but the metaphor is so
accurate) and the soundstage is really breathtaking. The entire audio
presentation is just more three dimensional and reed instruments, strings,
and female voices just sound more realistic. Well recorded female voice is
stunning and a well recorded string bass has all the woody resonance and
punch that it should while horns have all the appropriate overtones and bite
that they should.

I suppose it is easy to summarize by saying that it is impossible to get any
kind of a clear understanding of the sound or quality of an output
transformer by reading a silly revue like the one that is periodically
dredged up as an attempt to discredit certain winders. I would guess that
any of the low to medium priced transformers by respected manufacturers (and
there are many!) will provide hours of satisfying musical experience and a
wonderful way to get into this great hobby on a budget. It is now clear to
me that there are some extraordinary and elegant transformers available to
the hobbyist, but the price of admittance is high.

Well it seems the Magnequest DSO25 is their cheapest SE tranny,
and weighs 6 lbs, at $180, but their FS030 is specified as having nearly twice
the primary inductance, and is 11 lbs, and costs $360.
I am left wondering whether the 11 lb tranny may or may not have given you the
rapture than the
amorphous cored Tamura at $1,000 gave you.

It also seems to me that with the larger tranny, using the same amount of output
power,
the iron generated distortions would be less, and even if you are only using a
1/2 watt.
I looked for details on the Tamura 5002, but couldn't find any.
I do know amorphous cores are good for audio, but how much better than
grain oriented silicon steel, if at all, is a moot point.
If one built two SE trannys of equal weight, one with amorphous cores, the other
with
GOSS, I'd put my money on the amourphous being best,
if the turn numbers and interleaving were identical.

But what if we used twice the weight of GOSS, and thicker wire,
same turns? Unfortunately, one could spend a lifetime fiddling around with
such concerns, and never conclude anything.
There are rather a lot of "what ifs" about audio trannies.

My own experience is that one should spend on weight, and inductance
for SE trannies, and better that you have a tranny larger than you may
think is really needed at first.
The bandwidth of the tranny should not reduce the bandwidth of the rest of the
circuit, and the thd caused by the presence of the OPT should be minimal,
and a good deal less than the tubes, even at full throttle at 30 Hz.

I think some of the glory you hear with the 300B and Tamura is due to
the low distortion imposed on the music by the iron, which allows the
300B to get on with what its good at.
There are many folks who say the 300B has great midrange, but
no bass or treble.
Methinks perhaps the 300B reputation is smeared by inadequate transformers,
because nobody listens to 300B without a tranny connected, eh.
The other thing about 300B sound is that its so often heard without
an NFB applied.
If you use a 2.5k to 8 ohm OPT ratio, then the impedance ratio is 312.5 : 1,
so the Ra of the tube is transformed to 800/312 = 2.56 ohms,
so Ro = about 3 ohms including winding resistances.
Such a high Ro might lead one to make the wrong conclusions about
bass sounds. Any speaker would want to have a more ot less constant
impedance to prevent troughs and peaks appearing in the response.

Now what if the 300B is set up with a 5k to 4 ohm tranny?
The ratio is 1,250 :1, so the Ro would fall to around 1.2 ohms,
if winding resistances were 0.6 ohms.
The thd/imd at all levels of operation would be a lot lower.

Its hard to think that such re-arrangements of the circuit wouldn't cause
some changes to the sound.

The use of the higher value RL allows a slightly higher plate voltage and lower
plate current,
for a wider undistorted voltage swing.
Since the OPT has less DC, its thd contribution is a bit less.

Just how much difference the brand of transformers would make is a moot point.
It is what is in the transformer, not the brandname that counts.

Patrick Turner.


Regards,
Sandy Brown

"Patrick Turner" wrote:
Just how much difference the brand of transformers would make is a moot
point.
It is what is in the transformer, not the brandname that counts.

That really was the point of my post. I enjoy both transformers because
they are competent at what they do. There are many brands and types of SE
output transformers and it is most important to choose the one that fits
one's budget and technical requirements. This is an important point as I
really agonized for a while over whether I should go for a 3K, 3.5k, or 5k
primary impedance. Just as you illustrate in your post below, primary
impedance can have a meaningful effect on distortion and power. I opted for
the "standard" 3k and I could spend endless hours worrying about the "what
ifs" or I can just listen to the transformers and enjoy them.

Well it seems the Magnequest DSO25 is their cheapest SE tranny,
and weighs 6 lbs, at $180, but their FS030 is specified as having nearly
twice
the primary inductance, and is 11 lbs, and costs $360.
I am left wondering whether the 11 lb tranny may or may not have given you
the
rapture than the amorphous cored Tamura at $1,000 gave you.

Wonder no more. A friend of mine bought a couple of Wavelength Cardinals
with the FS030s just after I bought the Tamuras and he wanted to see how
they would sound with my speakers. We both agreed that they sounded superb
and he was thrilled with his purchase. They were only in my system for an
evening so it would be impossible to give a full characterization of the
differences with respect to the Tamuras, particularly as the circuit
topology is different. What is safe to say is that he enjoys the
Magnequests and I enjoy the Tamuras and I am sure that we would both be
satisfied with many other transformers in the same technical class.

I looked for details on the Tamura 5002, but couldn't find any.

Here is the catalog. It still doesn't give much info:
http://www.tamura-ss.co.jp/jp/catalo.../C-1005-10.pdf

I think some of the glory you hear with the 300B and Tamura is due to
the low distortion imposed on the music by the iron, which allows the
300B to get on with what its good at.

That is really what I understood the strengths of the amorphous core to be:
it responds to the most subtle and slight changes in voltage, or in other
words, low level detail.

Again, I think you summed it up perfectly:

Just how much difference the brand of transformers would make is a moot
point.
It is what is in the transformer, not the brandname that counts.

Thanks for your thoughtful response.

Regards,
Sandy Brown


Centifolia wrote:

I am well aware of the controversy surrounding the purported transformer
test that gives little information as to which transformers were
actually
tested and nothing about how they were tested and what the actual
technical
specifications are for each transformer.

Indeed. one would need to take a grain of salt with those tests.



I thought it more appropriate for RAT listers to provide real world
assessments of their own empirical use of any of the readily available
transformers. To that end, I have had in my SE amp over the last 6
years
two well regarded output transformers, the Magnequest DS025 and the
Tamura
5002.

I had the Magnequest DS025 in a SE amp for about 5 years. Over the
course
of that time, I had several variations on the circuit topology shown
he

http://home.comcast.net/~vhbrown/misc/2A3SEamp.gif

The circuit shown is the one that I finally settled upon except that the
AVVT 2A3 was the tube of choice after all the listening to different
tubes
was completed (note a slight error in the schematic as the two 2.5V
secondaries were in parallel for the 2A3 application, the series
connection
is for the 300B). Also tested in this amp were engraved base WE 300B,
AVVT
300B, and the chinese and RCA 2A3. The B+ was adjusted by inserting a
cap
before the first inductor to increase the plate voltage for the 300B so
that
it would operate at the correct operating conditions. Let me say right
off
that, with the exception of the chinese 2A3, I enjoyed every minute of
listening to this amp. In the end, I thought that the 2A3 was a much
better match for the DS025 than the 300B of any variant.

It seems you are not needing a high level of power, maybe a
watt, so a 2A3 should do all that you'd want , and having
an amp capable of 8 watts, or 80 watts, is not needed.

I don't really
want to speculate, but it is possibly because of the higher inductance
at
the lower current of the 2A3 over the 300B or it could be just the
matching
of the 2A3/transformer/speakers was better.

The inductance, whatever it measures at 70 mA, won't be much less
at 35mA. Air gapped transformers have a more constant inductance
regardless of DC flow, within their non saturated region.
The inductance of an SE OPT is like that of a DC choke.

The speakers over the 6 years
have been modified Altec A7s. I don't really think it is necessary to
go
into audiospeak about glorious midrange, crystalline highs and veils
being
lifted etc. I think it more important to affirm that I enjoyed
listening to
this setup and that I think any listener on a budget who wants to make a
SE
amp would be satisfied with the Magnequest DS025.

Now on to the Tamura 5002. The sale of the WE 300Bs to a gentleman in
Hong
Kong afforded me the opportunity to purchase any transformers I wanted
and I
chose the amorphous core Tamura 5002 based solely on their reputation.
I
reconfigured the amp to this topology:

http://home.comcast.net/~vhbrown/misc/300BSEamp.gif

Note again there is an error, this time it the 351V going to the 5002 is
closer to 385V At any rate, these transformers cost almost $1000 more
than
the Magnequest! Do they sound $1000 better? Well, that is one of those

questions that is impossible to answer. I think the fairest answer is
that
I am very happy that I bought them. These transformers work wonderfully
with the AVVT 300Bs where the DS025s just did not seem to mate as well
with
them. I think it is safe to say that the AVVT 2A3 and 300Bs have
substantially the same sound and that these two transformers allowed
each
tube to show their respective strengths. The difference between the
2A3/DS025 and the 300B/5002 is not subtle. With respect to frequency
response, the 300B/5002 has considerably more authoritative bass. The
midrange is more liquid (I really hate audiospeak but the metaphor is so
accurate) and the soundstage is really breathtaking. The entire audio
presentation is just more three dimensional and reed instruments,
strings,
and female voices just sound more realistic. Well recorded female voice
is
stunning and a well recorded string bass has all the woody resonance and
punch that it should while horns have all the appropriate overtones and
bite
that they should.

I suppose it is easy to summarize by saying that it is impossible to get
any
kind of a clear understanding of the sound or quality of an output
transformer by reading a silly revue like the one that is periodically
dredged up as an attempt to discredit certain winders. I would guess
that
any of the low to medium priced transformers by respected manufacturers
(and
there are many!) will provide hours of satisfying musical experience and
a
wonderful way to get into this great hobby on a budget. It is now clear
to
me that there are some extraordinary and elegant transformers available
to
the hobbyist, but the price of admittance is high.

Well it seems the Magnequest DSO25 is their cheapest SE tranny,
and weighs 6 lbs, at $180, but their FS030 is specified as having nearly
twice
the primary inductance, and is 11 lbs, and costs $360.
I am left wondering whether the 11 lb tranny may or may not have given you
the
rapture than the
amorphous cored Tamura at $1,000 gave you.

It also seems to me that with the larger tranny, using the same amount of
output
power,
the iron generated distortions would be less, and even if you are only
using a
1/2 watt.
I looked for details on the Tamura 5002, but couldn't find any.
I do know amorphous cores are good for audio, but how much better than
grain oriented silicon steel, if at all, is a moot point.
If one built two SE trannys of equal weight, one with amorphous cores, the
other
with
GOSS, I'd put my money on the amourphous being best,
if the turn numbers and interleaving were identical.

But what if we used twice the weight of GOSS, and thicker wire,
same turns? Unfortunately, one could spend a lifetime fiddling around with
such concerns, and never conclude anything.
There are rather a lot of "what ifs" about audio trannies.

My own experience is that one should spend on weight, and inductance
for SE trannies, and better that you have a tranny larger than you may
think is really needed at first.
The bandwidth of the tranny should not reduce the bandwidth of the rest of
the
circuit, and the thd caused by the presence of the OPT should be minimal,
and a good deal less than the tubes, even at full throttle at 30 Hz.

I think some of the glory you hear with the 300B and Tamura is due to
the low distortion imposed on the music by the iron, which allows the
300B to get on with what its good at.
There are many folks who say the 300B has great midrange, but
no bass or treble.
Methinks perhaps the 300B reputation is smeared by inadequate
transformers,
because nobody listens to 300B without a tranny connected, eh.
The other thing about 300B sound is that its so often heard without
an NFB applied.
If you use a 2.5k to 8 ohm OPT ratio, then the impedance ratio is 312.5 :
1,
so the Ra of the tube is transformed to 800/312 = 2.56 ohms,
so Ro = about 3 ohms including winding resistances.
Such a high Ro might lead one to make the wrong conclusions about
bass sounds. Any speaker would want to have a more ot less constant
impedance to prevent troughs and peaks appearing in the response.

Now what if the 300B is set up with a 5k to 4 ohm tranny?
The ratio is 1,250 :1, so the Ro would fall to around 1.2 ohms,
if winding resistances were 0.6 ohms.
The thd/imd at all levels of operation would be a lot lower.

Its hard to think that such re-arrangements of the circuit wouldn't cause
some changes to the sound.

The use of the higher value RL allows a slightly higher plate voltage and
lower
plate current,
for a wider undistorted voltage swing.
Since the OPT has less DC, its thd contribution is a bit less.

Just how much difference the brand of transformers would make is a moot
point.
It is what is in the transformer, not the brandname that counts.

Patrick Turner.


Regards,
Sandy Brown

Centifolia wrote:

"Patrick Turner" wrote:
Just how much difference the brand of transformers would make is a moot
point.
It is what is in the transformer, not the brandname that counts.

That really was the point of my post. I enjoy both transformers because
they are competent at what they do. There are many brands and types of SE
output transformers and it is most important to choose the one that fits
one's budget and technical requirements. This is an important point as I
really agonized for a while over whether I should go for a 3K, 3.5k, or 5k
primary impedance. Just as you illustrate in your post below, primary
impedance can have a meaningful effect on distortion and power. I opted for
the "standard" 3k and I could spend endless hours worrying about the "what
ifs" or I can just listen to the transformers and enjoy them.

I wind a few transformers for the amps I sell, so
I worry about the what-ifs all the time.

I aim to please that part of the market who would like something SE,
and all the magic that it seems to convey, but with more grunt
than a 2A3 or 300B. So 11lbs for an OPT is a bit lightweight for what
I like to build.

The power that you get is all class A from SE, and even if the load is not
perfectly matched to the tubes, the first few watts are very listenable.
Ideally, my SEUL should be used with 8 ohm speakers, but I have used 4 ohm
Duntechs, in a demo, and ppl enjoyed the sound.
300B would have been quite limited imho.

But for sound at home, you don't need too much power, and
the last time I listened to a stereo Valve Mark amp with two 300B,
it sounded better than a 100 w/pch solid state ME amp, and had *more* bass,
although I think that was a function of the highish tube amp Ro,
and the peaking impedance curve of the Vaff speakers used.
At between say 20 and 100Hz, speakers will have a peak in Z,
and the acoustic output, so if Ro is high, its
easy to get a couple of dB bass boost, but I think
if Ro is too high, the bass then sounds wooly with some speakers.



Well it seems the Magnequest DSO25 is their cheapest SE tranny,
and weighs 6 lbs, at $180, but their FS030 is specified as having nearly
twice
the primary inductance, and is 11 lbs, and costs $360.
I am left wondering whether the 11 lb tranny may or may not have given you
the
rapture than the amorphous cored Tamura at $1,000 gave you.

Wonder no more. A friend of mine bought a couple of Wavelength Cardinals
with the FS030s just after I bought the Tamuras and he wanted to see how
they would sound with my speakers. We both agreed that they sounded superb
and he was thrilled with his purchase. They were only in my system for an
evening so it would be impossible to give a full characterization of the
differences with respect to the Tamuras, particularly as the circuit
topology is different. What is safe to say is that he enjoys the
Magnequests and I enjoy the Tamuras and I am sure that we would both be
satisfied with many other transformers in the same technical class.

I looked for details on the Tamura 5002, but couldn't find any.

Here is the catalog. It still doesn't give much info:
http://www.tamura-ss.co.jp/jp/catalo.../C-1005-10.pdf

I had a look at the F response quoted, and it starts to sag at each end
of its range over 4 decades, but my pdf reader couldn't read what I think
are japanese characters, so I assume the 3 decades are from 100Hz to 100kHz,
with -3 dB? at 30 Hz and 200 kHz.
The test conditions are given with the source impedance, but the pdf
didn't show a driving impedance value, but perhaps it is 1 kohm,
and I assume the load on the sec reflects the 3k primary load on the tube.

Therefore the RA, ie the 1k Ra of a tube and 3k RL in parallel is 750 ohms,
and its this is in shunt with the Lp, so
if the -3 dB point is 30 Hz, then Lp = 750 / ( 6.28 x 30 ) = 4 Henrys.

This seems far too low a value of Lp imho, and perhaps I have made a mistake
reading
the Tamura data sheets, and the response may cover 10Hz to 10 kHz, ( flat ),
with -3 at 3 Hz to 20 kHz, and Lp at 40H, which sounds pretty good to me.

Have you got the details on response for this Tamura?

The LF response from a transformer varies with the driving impedance.
If the load was 3kohms, and the Ra was say 15k, like a KT88,
then RA would be 2.5kohms, so if the Lp was 4H,
the -3 dB point would be at 100 Hz, but if the Lp
was 40H, it would be at a respectable 10 Hz for the high impedance tube.
But the speaker loads rise at LF, and seldom is there a load of 3k
on the amp at say 50 Hz, more like 30k, so with a pentode/tetrode
SE tube, there is nothing to damp the circuit, so the response can be anywhere.
Triodes avoid the trouble, and with pentodes, loop FB must be used somewhere
to reduce Ra to get triode like performance.
Triodes have their own internal loop of electrostatic feedback,
placed into the tube by the God of Triodes, when every triode is born.
Such feedback, when presented to us by a god, is sacred and entirely
permissable, ihmo.
An SE KT88 in beam pentode mode can be set up easily with FB so
it has effective Ra' = 800 ohms approximates what a 300B does.
Does it sound as good? I leave that for others to find out.

I could say the last SE EL34 amp with 12 dB of FB I heard which made 9 watts
max,
and with 12 dB of FB sounded pretty dreadful with some cheap Thai made OPTs,
and with 450v Ea, and low bias current, and 2.5 k load.
The thd at normal listening levels was ~20 dB above a well done 300B amp,
and the thd has all those odd order harmonics.
The guy who bought it reckoned it was no improvement on his SS amp.
But this cheap Thai made stereo kit was only us $ 750.
To me this amp was totally ill-concieved by a designer or production
engineer who had no understanding of tube craft, and just anted to make a fast
buck.
To anyone in Thailand, $750 is a fortune. I was given the task of trying to make
it sound better,
and all I could do was to lower the screen voltage a bit, and
filter the HT better, since no other alteration was feasible.
And no, I won't divulge the brandame.
The owner has progressed to thinking about 300B amps.

So for me, there is good and bad in the world of SE amps.

I think some of the glory you hear with the 300B and Tamura is due to
the low distortion imposed on the music by the iron, which allows the
300B to get on with what its good at.

That is really what I understood the strengths of the amorphous core to be:
it responds to the most subtle and slight changes in voltage, or in other
words, low level detail.

Again, I think you summed it up perfectly:

Just how much difference the brand of transformers would make is a moot
point.
It is what is in the transformer, not the brandname that counts.

Thanks for your thoughtful response.

I get good results with designing for low thd, and low F at which
saturation starts to occur in transformers, rather than
worrying too much about primary inductance.

I am building a pair of 45 watt SE amps now with a
load of only 1.2k, since there are 4 x 6CA7.
The prototype amp can be taken to -1dB below clipping at 1 kHz,
and then the F rolled down to 20 Hz, where the wave form starts to go
a bit wobbly, then quite a bit wobbly at 14 Hz, and fairly well distorted
at 10 Hz, but still not fully saturated, or allowing huge
sudden current peaks in the output tube due to the
complete saturation of the core from AC and DC.
I could get this performance at half the frequencies quoted,
but I would need to double the size of the core cross section
or double the turns, ( roughly speaking ) and the transformer would need a
forklift to lift it.
As it is there is *plenty* of inductance as a result, and in my case, about 12H,

so there is this ratio of 12H to 1.2kohms between Lp and RL.
If the Tamura has 40H to 3k, then its a good ratio, imho,
but then when the DC is present, which is *not shown*
at the Tamura pdf detail diagram, then the Lp/Rl ratio may not be so hot.

I don't know how to corelate the use of amorphous cores to
fine sound, because to me thay are just
iron cores, although prepared fro use in a hi-falooting way,
and with optimal magnetic qualities for HF.

I would always be ready to expose my amps to any competition
with any others with fancy trannies from OS.
There is little opportunity, and I get buy without
trapsing around the world with amps tucked under the arm,
to impress the so called cognescenti.
I would have to do that if I wanted to make a big splash in the world of amps,
and get more sales.
The amorphous cores would be somewhat expensive for me to purchase in sizes
I would want them to be, so I soldier on with fine GOSS E&I lams,
which is made in Oz, and which has a U max higher than anything I have tested,
at 17,000 when in a core used for an AC tranny.
The use of such core material in SE trannies gives superlative measured results,

according to the standards laid out in books like RDH4, and others.
The GOSS mentioned in RDH4 has a U max of around 5,000,
and it seems that after 50 years of R&D, the industry standard GOSS
is now a lot higher.
This allows far less losses in mains transformers, which then can be designed
more
critically close to saturation.
But the side benefit is that the tranny steel made now happens to also suit
our audio needs. And especially those of the SE tranny winder.

The higher the U of the GOSS, the higher the effective U can be for
a given number of DC ampturns, or DC flux, or if you like,
for a given weight of iron, turns, DC, and well adjusted gap the
inductance will be higher than the same amount of iron
with a lower U.
The GOSS displays far less thd than plain old low U non oriented silicon steel
But both GOSS and plainer old NOSS will still saturate at around the same F.
Its still only iron which is the main component in the core, even with
amorphous.
To significantly reduce the size of the cores to get lower Fsat,
you'd have to go to cobalt cores.
Then your'e talkin real money, for the real thing, which is rare.

You may yourself not be all that interested in the salient issues I ahve raised,

but others may be, and the group is what I have in mind with ALL my postings.

Happy sunday listenings,

Patrick Turner.






Regards,
Sandy Brown

Centifolia wrote:

I am well aware of the controversy surrounding the purported transformer
test that gives little information as to which transformers were
actually
tested and nothing about how they were tested and what the actual
technical
specifications are for each transformer.

Indeed. one would need to take a grain of salt with those tests.



I thought it more appropriate for RAT listers to provide real world
assessments of their own empirical use of any of the readily available
transformers. To that end, I have had in my SE amp over the last 6
years
two well regarded output transformers, the Magnequest DS025 and the
Tamura
5002.

I had the Magnequest DS025 in a SE amp for about 5 years. Over the
course
of that time, I had several variations on the circuit topology shown
he

http://home.comcast.net/~vhbrown/misc/2A3SEamp.gif

The circuit shown is the one that I finally settled upon except that the
AVVT 2A3 was the tube of choice after all the listening to different
tubes
was completed (note a slight error in the schematic as the two 2.5V
secondaries were in parallel for the 2A3 application, the series
connection
is for the 300B). Also tested in this amp were engraved base WE 300B,
AVVT
300B, and the chinese and RCA 2A3. The B+ was adjusted by inserting a
cap
before the first inductor to increase the plate voltage for the 300B so
that
it would operate at the correct operating conditions. Let me say right
off
that, with the exception of the chinese 2A3, I enjoyed every minute of
listening to this amp. In the end, I thought that the 2A3 was a much
better match for the DS025 than the 300B of any variant.

It seems you are not needing a high level of power, maybe a
watt, so a 2A3 should do all that you'd want , and having
an amp capable of 8 watts, or 80 watts, is not needed.

I don't really
want to speculate, but it is possibly because of the higher inductance
at
the lower current of the 2A3 over the 300B or it could be just the
matching
of the 2A3/transformer/speakers was better.

The inductance, whatever it measures at 70 mA, won't be much less
at 35mA. Air gapped transformers have a more constant inductance
regardless of DC flow, within their non saturated region.
The inductance of an SE OPT is like that of a DC choke.

The speakers over the 6 years
have been modified Altec A7s. I don't really think it is necessary to
go
into audiospeak about glorious midrange, crystalline highs and veils
being
lifted etc. I think it more important to affirm that I enjoyed
listening to
this setup and that I think any listener on a budget who wants to make a
SE
amp would be satisfied with the Magnequest DS025.

Now on to the Tamura 5002. The sale of the WE 300Bs to a gentleman in
Hong
Kong afforded me the opportunity to purchase any transformers I wanted
and I
chose the amorphous core Tamura 5002 based solely on their reputation.
I
reconfigured the amp to this topology:

http://home.comcast.net/~vhbrown/misc/300BSEamp.gif

Note again there is an error, this time it the 351V going to the 5002 is
closer to 385V At any rate, these transformers cost almost $1000 more
than
the Magnequest! Do they sound $1000 better? Well, that is one of those

questions that is impossible to answer. I think the fairest answer is
that
I am very happy that I bought them. These transformers work wonderfully
with the AVVT 300Bs where the DS025s just did not seem to mate as well
with
them. I think it is safe to say that the AVVT 2A3 and 300Bs have
substantially the same sound and that these two transformers allowed
each
tube to show their respective strengths. The difference between the
2A3/DS025 and the 300B/5002 is not subtle. With respect to frequency
response, the 300B/5002 has considerably more authoritative bass. The
midrange is more liquid (I really hate audiospeak but the metaphor is so
accurate) and the soundstage is really breathtaking. The entire audio
presentation is just more three dimensional and reed instruments,
strings,
and female voices just sound more realistic. Well recorded female voice
is
stunning and a well recorded string bass has all the woody resonance and
punch that it should while horns have all the appropriate overtones and
bite
that they should.

I suppose it is easy to summarize by saying that it is impossible to get
any
kind of a clear understanding of the sound or quality of an output
transformer by reading a silly revue like the one that is periodically
dredged up as an attempt to discredit certain winders. I would guess
that
any of the low to medium priced transformers by respected manufacturers
(and
there are many!) will provide hours of satisfying musical experience and
a
wonderful way to get into this great hobby on a budget. It is now clear
to
me that there are some extraordinary and elegant transformers available
to
the hobbyist, but the price of admittance is high.

Well it seems the Magnequest DSO25 is their cheapest SE tranny,
and weighs 6 lbs, at $180, but their FS030 is specified as having nearly
twice
the primary inductance, and is 11 lbs, and costs $360.
I am left wondering whether the 11 lb tranny may or may not have given you
the
rapture than the
amorphous cored Tamura at $1,000 gave you.

It also seems to me that with the larger tranny, using the same amount of
output
power,
the iron generated distortions would be less, and even if you are only
using a
1/2 watt.
I looked for details on the Tamura 5002, but couldn't find any.
I do know amorphous cores are good for audio, but how much better than
grain oriented silicon steel, if at all, is a moot point.
If one built two SE trannys of equal weight, one with amorphous cores, the
other
with
GOSS, I'd put my money on the amourphous being best,
if the turn numbers and interleaving were identical.

But what if we used twice the weight of GOSS, and thicker wire,
same turns? Unfortunately, one could spend a lifetime fiddling around with
such concerns, and never conclude anything.
There are rather a lot of "what ifs" about audio trannies.

My own experience is that one should spend on weight, and inductance
for SE trannies, and better that you have a tranny larger than you may
think is really needed at first.
The bandwidth of the tranny should not reduce the bandwidth of the rest of
the
circuit, and the thd caused by the presence of the OPT should be minimal,
and a good deal less than the tubes, even at full throttle at 30 Hz.

I think some of the glory you hear with the 300B and Tamura is due to
the low distortion imposed on the music by the iron, which allows the
300B to get on with what its good at.
There are many folks who say the 300B has great midrange, but
no bass or treble.
Methinks perhaps the 300B reputation is smeared by inadequate
transformers,
because nobody listens to 300B without a tranny connected, eh.
The other thing about 300B sound is that its so often heard without
an NFB applied.
If you use a 2.5k to 8 ohm OPT ratio, then the impedance ratio is 312.5 :
1,
so the Ra of the tube is transformed to 800/312 = 2.56 ohms,
so Ro = about 3 ohms including winding resistances.
Such a high Ro might lead one to make the wrong conclusions about
bass sounds. Any speaker would want to have a more ot less constant
impedance to prevent troughs and peaks appearing in the response.

Now what if the 300B is set up with a 5k to 4 ohm tranny?
The ratio is 1,250 :1, so the Ro would fall to around 1.2 ohms,
if winding resistances were 0.6 ohms.
The thd/imd at all levels of operation would be a lot lower.

Its hard to think that such re-arrangements of the circuit wouldn't cause
some changes to the sound.

The use of the higher value RL allows a slightly higher plate voltage and
lower
plate current,
for a wider undistorted voltage swing.
Since the OPT has less DC, its thd contribution is a bit less.

Just how much difference the brand of transformers would make is a moot
point.
It is what is in the transformer, not the brandname that counts.

Patrick Turner.


Regards,
Sandy Brown