Overvoltage protection

[Alex Pogossov]

If you are experimenting on a running tube amp, you might be at some stage
pulling a hot output tube from its socket while the amp is running.
Alternatively you might accidentally send a high negative pilse to a grid of
the tube while working on the circuit.

In any case, an abrupt interruption of the plate current will cause a
voltage spike on the OPT primary. The stored magnetisation energy of
Lp*(Ia^2)/2 will have to be dissipated. Lp can be quite high in a decent
amp. If a load is connected, then the energy will be dissipated in the load
giving you a loud crack in the speaker. A residual smaller energy stored in
the leakage inductance Ls*(Ia^2)/2 will most likely be safely dissipated in
a snubber of say 2200pF+5K usually connected in parallel to the OPT primary.

But what is the load is not connected? Then the huge magnetisation energy
will either cause:
- arcing in the tube;
- breakdown in the OPT winding insulation;
- breakdown of the subber capacitor (say 2200pF);
- arcing elsewhere in the wiring.or in the tube socket.

Would it be a good idea to placa a varistor rated slightly above the +B
across the primary? Or a spark gap of some sort, or gas discharge surge
arrester tube (the later have miniscule capacitance and very reliable). I
remember in vertical deflection stages of old TVs such varistors were a
must, since the abrupt cutoff of the current (during flyback) in EL84 often
used for this purpose was the mode of operation.

However I have never seen any discussions on the varistor protection issue
on this site.
What is your opinion?

Regards,
Alex

[Phil Allison[_3_]]

"Alex Pogossov"

>
> In any case, an abrupt interruption of the plate current will cause a
> voltage spike on the OPT primary. The stored magnetisation energy of
> Lp*(Ia^2)/2 will have to be dissipated. Lp can be quite high in a decent
> amp. If a load is connected, then the energy will be dissipated in the
> load giving you a loud crack in the speaker. A residual smaller energy
> stored in the leakage inductance Ls*(Ia^2)/2 will most likely be safely
> dissipated in a snubber of say 2200pF+5K usually connected in parallel to
> the OPT primary.
>
> But what is the load is not connected? Then the huge magnetisation energy


** Why huge ???

If Lp is say 100H and Ia = 0.1A then how many Joules is that ?




..... Phil

[Alex Pogossov]

"Phil Allison" > wrote in message
...
>
> "Alex Pogossov"
>
>>
>> In any case, an abrupt interruption of the plate current will cause a
>> voltage spike on the OPT primary. The stored magnetisation energy of
>> Lp*(Ia^2)/2 will have to be dissipated. Lp can be quite high in a decent
>> amp. If a load is connected, then the energy will be dissipated in the
>> load giving you a loud crack in the speaker. A residual smaller energy
>> stored in the leakage inductance Ls*(Ia^2)/2 will most likely be safely
>> dissipated in a snubber of say 2200pF+5K usually connected in parallel to
>> the OPT primary.
>>
>> But what is the load is not connected? Then the huge magnetisation energy
>
>
> ** Why huge ???
>
> If Lp is say 100H and Ia = 0.1A then how many Joules is that ?

Apparently 0.5J.

It is the same as charging a 10uF to 315V or 47uF to 160V. If you are in
Europe or Australia, try charging 10uF from the grid through a diode and
then discharging it with a screwdriver... The closer you keep the cap to
your face when discharging -- the better... Then you will see if it is huge
or not... It is all subjective after all... If you are in the US, use 47uF
for a similar experiment.

Enjoy

Alex

[Phil Allison[_3_]]

"Alex Pogossov"
> "Phil Allison"
>> "Alex Pogossov"
>>
>>>
>>> In any case, an abrupt interruption of the plate current will cause a
>>> voltage spike on the OPT primary. The stored magnetisation energy of
>>> Lp*(Ia^2)/2 will have to be dissipated. Lp can be quite high in a decent
>>> amp. If a load is connected, then the energy will be dissipated in the
>>> load giving you a loud crack in the speaker. A residual smaller energy
>>> stored in the leakage inductance Ls*(Ia^2)/2 will most likely be safely
>>> dissipated in a snubber of say 2200pF+5K usually connected in parallel
>>> to the OPT primary.
>>>
>>> But what is the load is not connected? Then the huge magnetisation
>>> energy
>>
>>
>> ** Why huge ???
>>
>> If Lp is say 100H and Ia = 0.1A then how many Joules is that ?
>
> Apparently 0.5J.


** IOW - **** all.

Even the smallest Varistor you can buy will absorb it with ease.

Like this one for $ 1.58 + gst

http://au.element14.com/epcos/b72205s0301k101/varistor-9-6j-300vac/dp/1004399?Ntt=100+4399


..... Phil

[Alex Pogossov]

"Phil Allison" > wrote in message
...
>
> "Alex Pogossov"
>> "Phil Allison"
>>> "Alex Pogossov"
>>>
>>>>
>>>> In any case, an abrupt interruption of the plate current will cause a
>>>> voltage spike on the OPT primary. The stored magnetisation energy of
>>>> Lp*(Ia^2)/2 will have to be dissipated. Lp can be quite high in a
>>>> decent amp. If a load is connected, then the energy will be dissipated
>>>> in the load giving you a loud crack in the speaker. A residual smaller
>>>> energy stored in the leakage inductance Ls*(Ia^2)/2 will most likely be
>>>> safely dissipated in a snubber of say 2200pF+5K usually connected in
>>>> parallel to the OPT primary.
>>>>
>>>> But what is the load is not connected? Then the huge magnetisation
>>>> energy
>>>
>>>
>>> ** Why huge ???
>>>
>>> If Lp is say 100H and Ia = 0.1A then how many Joules is that ?
>>
>> Apparently 0.5J.
>
>
> ** IOW - **** all.
>
> Even the smallest Varistor you can buy will absorb it with ease.
>
> Like this one for $ 1.58 + gst
>
> http://au.element14.com/epcos/b72205s0301k101/varistor-9-6j-300vac/dp/1004399?Ntt=100+4399
>
Agree, it is no problem to clamp the surges, but the point is: why I have
not seen varistors across the OPT primaries in various schematics? Is this
sort of protection not needed? Or is the assumption that no one will be
changing tubes on a working amp? Or occasional arcing and sparking is not a
problem?

[[email protected]]

On Jul 8, 11:16*pm, "Alex Pogossov" > wrote:
> "Phil Allison" > wrote in message
>
> ...
>
>
>
>
>
>
>
> > "Alex Pogossov"
>
> >> In any case, an abrupt interruption of the plate current will cause a
> >> voltage spike on the OPT primary. The stored magnetisation energy of
> >> Lp*(Ia^2)/2 will have to be dissipated. Lp can be quite high in a decent
> >> amp. If a load is connected, then the energy will be dissipated in the
> >> load giving you a loud crack in the speaker. A residual smaller energy
> >> stored in the leakage inductance Ls*(Ia^2)/2 will most likely be safely
> >> dissipated in a snubber of say 2200pF+5K usually connected in parallel to
> >> the OPT primary.
>
> >> But what is the load is not connected? Then the huge magnetisation energy
>
> > ** Why huge ???
>
> > If Lp is say 100H and *Ia = 0.1A *then how many Joules is that ?
>
> Apparently 0.5J.
>
> It is the same as charging a 10uF to 315V or 47uF to 160V. If you are in
> Europe or Australia, try charging 10uF from the grid through a diode and
> then discharging it with a screwdriver... The closer you keep the cap to
> your face when discharging -- the better... Then you will see if it is huge
> or not... It is all subjective after all... If you are in the US, use 47uF
> for a similar experiment.
>
> Enjoy
>
> Alex- Hide quoted text -
>
> - Show quoted text -

We used to use 100 uF@ 500VDC in my highschool to play jokes in the
electronics shop. Gives a very nice wake-up jolt , especially to those
with sweaty palms . . .

[Phil Allison[_3_]]

"Alex Pogossov"
"Phil Allison"
>>
>>>>> But what is the load is not connected? Then the huge magnetisation
>>>>> energy
>>>>
>>>>
>>>> ** Why huge ???
>>>>
>>>> If Lp is say 100H and Ia = 0.1A then how many Joules is that ?
>>>
>>> Apparently 0.5J.
>>
>>
>> ** IOW - **** all.
>>
>> Even the smallest Varistor you can buy will absorb it with ease.
>>
>> Like this one for $ 1.58 + gst
>>
>> http://au.element14.com/epcos/b72205s0301k101/varistor-9-6j-300vac/dp/1004399?Ntt=100+4399
>>
> Agree, it is no problem to clamp the surges, but the point is: why I have
> not seen varistors across the OPT primaries in various schematics?


** Cos the usual RC network suppresses the spike sufficiently, along with
the load.


> Or is the assumption that no one will be changing tubes on a working amp?

** Well, not with no load connected.


> Or occasional arcing and sparking is not a problem?


** The Aussie made 6V6GT used in my first valve amp liked to spark
internally with no load.

Never did it, the A&R OP tranny or the socket any permanent harm.

But if you wanna go belt and braces - fine with me....

BTW - how much HT have you got ??


..... Phil

[Alex Pogossov]

"Phil Allison" > wrote in message
...

> ** Cos the usual RC network suppresses the spike sufficiently, along with
> the load.

If a load is connected -- correct. If no load, then as I have mentioned you
need microfarades + several kiloohms to absorb it, not a few nanofarades
usually connected to dump HF ringing of the leakage inductance.
>
>
>> Or is the assumption that no one will be changing tubes on a working amp?
>
> ** Well, not with no load connected.

What about an amp being fool proof then?

>> Or occasional arcing and sparking is not a problem?

> ** The Aussie made 6V6GT used in my first valve amp liked to spark
> internally with no load. Never did it, the A&R OP tranny or the socket any
> permanent harm. But if you wanna go belt and braces - fine with me....
>
> BTW - how much HT have you got ??

Only 250V. But I do not want to risk any possibility of insulation breakdown
in an old lousy OPT.
Probably a small varistor rated at 350...400Vdc and with less than 500pF of
capacitance will be OK.

The tread is also of general nature -- why do not we see varistors in the
tube amp circuits. Probably you are right -- no one suffered serios damage
of this particular stress mode.

[mick]

On Sat, 09 Jul 2011 21:25:39 +1000, Alex Pogossov wrote:

> "Phil Allison" > wrote in message
> ...
>
>> ** Cos the usual RC network suppresses the spike sufficiently, along
>> with the load.
>
> If a load is connected -- correct. If no load, then as I have mentioned
> you need microfarades + several kiloohms to absorb it, not a few
> nanofarades usually connected to dump HF ringing of the leakage
> inductance.
>>
>>
>>> Or is the assumption that no one will be changing tubes on a working
>>> amp?
>>
>> ** Well, not with no load connected.
>
> What about an amp being fool proof then?
>
>>> Or occasional arcing and sparking is not a problem?
>
>> ** The Aussie made 6V6GT used in my first valve amp liked to spark
>> internally with no load. Never did it, the A&R OP tranny or the socket
>> any permanent harm. But if you wanna go belt and braces - fine with
>> me....
>>
>> BTW - how much HT have you got ??
>
> Only 250V. But I do not want to risk any possibility of insulation
> breakdown in an old lousy OPT.
> Probably a small varistor rated at 350...400Vdc and with less than 500pF
> of capacitance will be OK.
>
> The tread is also of general nature -- why do not we see varistors in
> the tube amp circuits. Probably you are right -- no one suffered serios
> damage of this particular stress mode.


It's probably traditional. :-) Small, effective varistors are a newer
invention than valves. The two would never have met when the circuits
were being developed. That, and engineers having it hammered into them
that you *never* run a valve amp without a load.

--
Mick (Working in a M$-free zone!)
Web: http://www.nascom.info
Filtering everything posted from googlegroups to kill spam.

[John Byrns]

In article >,
"Alex Pogossov" > wrote:

> The tread is also of general nature -- why do not we see varistors in the
> tube amp circuits. Probably you are right -- no one suffered serios damage
> of this particular stress mode.

It's interesting to note that 1950s, 1960s, and 1970s Telephones used to be full
of varistors, both for protective purposes, and for regulating and controlling
the speech level.

--
Regards,

John Byrns

Surf my web pages at, http://fmamradios.com/

[Patrick Turner]

On Jul 9, 9:25*pm, "Alex Pogossov" > wrote:
> "Phil Allison" > wrote in message
>
> ...
>
> > ** Cos the usual RC network suppresses the spike sufficiently, along with
> > the load.
>
> If a load is connected -- correct. If no load, then as I have mentioned you
> need microfarades + several kiloohms to absorb it, not a few nanofarades
> usually connected to dump HF ringing of the leakage inductance.
>
>
>
> >> Or is the assumption that no one will be changing tubes on a working amp?
>
> > ** Well, not with no load connected.
>
> What about an amp being fool proof then?
>
> >> Or occasional arcing and sparking is not a problem?
> > ** The Aussie made *6V6GT *used in my first valve amp liked to spark
> > internally with no load. Never did it, the A&R OP tranny or the socket any
> > permanent harm. But if you wanna go belt and braces - *fine with me.....
>
> > BTW *- *how much HT have you got ??
>
> Only 250V. But I do not want to risk any possibility of insulation breakdown
> in an old lousy OPT.
> Probably a small varistor rated at 350...400Vdc and with less than 500pF of
> capacitance will be OK.
>
> The tread is also of *general nature -- why do not we see varistors in the
> tube amp circuits. Probably you are right -- no one suffered serios damage
> of this particular stress mode.

This issue of back emfs in unloaded tube amps was never worried about
in the thousands of old radios with 1 x 6V6 used without any NFB and
connected to OPT and a speaker with no Zobel damping network.
The reactance of speaker inductance rises so that by 3kHz there is
virtually no load on the 6V6 and the shunt C is low and leakage L also
high. RDH4 does not mention that arcing in the tube would be a
problem, and I've seen no evidence of it. Stored magnetic energy in
bean couter designed radio OPTs is low; it uses a tiny core which is
air gapped. B+ voltages in radios tend to be less than +280V, and even
if anode goes to +560V, the insulation will survive.

But in PP amps the B+ can be +500V and as we all know the anode swing
can be to +/-1,500V, and thus threaten to arc. There's a simple answer
which is well know and been around ever since Si diodes began to get
good reverse voltage ratimgs, and that's to connect diodes from each
end of the OPT to 0V with diode anode connected to tube anodes. I like
to use 3 x 1N7007 in series with 1M strapped across each diode to
equalise reverese voltages, and when one end of the OPT tries to rise
to more than 2 x +B, the other end tries to go below 0V so diodes
conduct, and thus the OPT voltage swing is CLAMPED. Voltage clamping
with diodes has been around for ages.
There is a schematic example at http://www.turneraudio.com.au/100w-monobloc2-2004.htm
The amp also has Zobel networks across each 1/2 primary of the OPT to
load the OPT at HF.

The book I have with 17 designs of amps from 5W to 1,100 W written in
anout 1960 has no provisions for dealing with excessive unloaded OPT
voltages except on the high power types where an adjustable spark gap
is used with a 10k0 load resistance in series between each end of the
OPT.

But arcing in OPTs is not confined to amps which have B+ of over 1kV
or more. I once rescued a Geloso PA amp made in 1950s moments before a
trundling bulldozer went over it after someone dumped it when people
used to be able to take stuff to a tip and then raid each other's tip
off's. This kind of amp was used at Italian weddings for an acordian
player and singer. 2 x EL84 for about 15W max. The amp arced just
sitting there doing nothing while turned on; there was a fault in
insulation where B+ wires went into the winding and it arced between B
+ and the core. Some careful observation showed a blackened area and
it had been doing is for awhile and probably because it had got DAMP
at some time. I scraped out the carbon and wax in the area and dressed
wire positions, heated the tranny to dry it, re-waxed it, and it never
ever arced again. I then rewired it totally to give better guitar amp
usability and sold it for $350, about 12 years ago, to a guy here who
calls himself Dr Zot, and who dabbles in vintage electric guitars and
amps.

Patrick Turner.

[Phil Allison[_3_]]

"John Byrns" "
>
> It's interesting to note that 1950s, 1960s, and 1970s Telephones used to
> be full
> of varistors, both for protective purposes, and for regulating and
> controlling
> the speech level.


** More likely called " VDRs" ( Voltage Dependant Resistors ) and not
varistors or MOVs.

Although in the same general category, they are not identical.

Most modern MOVs are made from Zinc oxide, which gives them very low leakage
currents and sharper turn on characteristics compared to other types like
the older Silicon carbide VDRs.




..... Phil

[John Byrns]

In article >,
Patrick Turner > wrote:

> But in PP amps the B+ can be +500V and as we all know the anode swing
> can be to +/-1,500V, and thus threaten to arc. There's a simple answer
> which is well know and been around ever since Si diodes began to get
> good reverse voltage ratimgs, and that's to connect diodes from each
> end of the OPT to 0V with diode anode connected to tube anodes. I like
> to use 3 x 1N7007 in series with 1M strapped across each diode to
> equalise reverese voltages, and when one end of the OPT tries to rise
> to more than 2 x +B, the other end tries to go below 0V so diodes
> conduct, and thus the OPT voltage swing is CLAMPED. Voltage clamping
> with diodes has been around for ages.
> There is a schematic example at
> http://www.turneraudio.com.au/100w-monobloc2-2004.htm
> The amp also has Zobel networks across each 1/2 primary of the OPT to
> load the OPT at HF.

There seems to be an inconsistency here, between what you said above, "connect
diodes from each end of the OPT to 0V with diode anode connected to tube
anodes", and the way the "schematic example" shows the diodes wired?

--
Regards,

John Byrns

Surf my web pages at, http://fmamradios.com/

[John L Stewart]

This issue of back emfs in unloaded tube amps was never worried about
in the thousands of old radios with 1 x 6V6 used without any NFB and
connected to OPT and a speaker with no Zobel damping network.
The reactance of speaker inductance rises so that by 3kHz there is
virtually no load on the 6V6 and the shunt C is low and leakage L also
high. RDH4 does not mention that arcing in the tube would be a
problem, and I've seen no evidence of it. Stored magnetic energy in
bean couter designed radio OPTs is low; it uses a tiny core which is
air gapped. B+ voltages in radios tend to be less than +280V, and even
if anode goes to +560V, the insulation will survive.

But in PP amps the B+ can be +500V and as we all know the anode swing
can be to +/-1,500V, and thus threaten to arc. There's a simple answer
which is well know and been around ever since Si diodes began to get
good reverse voltage ratimgs, and that's to connect diodes from each
end of the OPT to 0V with diode anode connected to tube anodes. I like
to use 3 x 1N7007 in series with 1M strapped across each diode to
equalise reverese voltages, and when one end of the OPT tries to rise
to more than 2 x +B, the other end tries to go below 0V so diodes
conduct, and thus the OPT voltage swing is CLAMPED. Voltage clamping
with diodes has been around for ages.
There is a schematic example at http://www.turneraudio.com.au/100w-monobloc2-2004.htm
The amp also has Zobel networks across each 1/2 primary of the OPT to
load the OPT at HF.

The book I have with 17 designs of amps from 5W to 1,100 W written in
anout 1960 has no provisions for dealing with excessive unloaded OPT
voltages except on the high power types where an adjustable spark gap
is used with a 10k0 load resistance in series between each end of the
OPT.

But arcing in OPTs is not confined to amps which have B+ of over 1kV
or more. I once rescued a Geloso PA amp made in 1950s moments before a
trundling bulldozer went over it after someone dumped it when people
used to be able to take stuff to a tip and then raid each other's tip
off's. This kind of amp was used at Italian weddings for an acordian
player and singer. 2 x EL84 for about 15W max. The amp arced just
sitting there doing nothing while turned on; there was a fault in
insulation where B+ wires went into the winding and it arced between B
+ and the core. Some careful observation showed a blackened area and
it had been doing is for awhile and probably because it had got DAMP
at some time. I scraped out the carbon and wax in the area and dressed
wire positions, heated the tranny to dry it, re-waxed it, and it never
ever arced again. I then rewired it totally to give better guitar amp
usability and sold it for $350, about 12 years ago, to a guy here who
calls himself Dr Zot, and who dabbles in vintage electric guitars and
amps.

Patrick Turner.[/QUOTE]

Before 1960 the Selenium Thyrector Diode by GE & others was available to circuit designers. These were meant for power applications & came in both single ended & back to back stacks in several voltage ratings. Reverse leakage current specs were low enough so as not to adversley effect the questionable fidelity of those Class AB & B PA amplifiers. In a guitar amp they might add something.

Cheers, John

If the designer chose to he could have selected a pair of back to back devices attached across the OPT primary. But caps & resistors were less money.

[Patrick Turner]

On Jul 11, 1:00*am, John Byrns > wrote:
> In article >,
> *Patrick Turner > wrote:
>
> > But in PP amps the B+ can be +500V and as we all know the anode swing
> > can be to +/-1,500V, and thus threaten to arc. There's a simple answer
> > which is well know and been around ever since Si diodes began to get
> > good reverse voltage ratimgs, and that's to connect diodes from each
> > end of the OPT to 0V with diode anode connected to tube anodes. I like
> > to use 3 x 1N7007 in series with 1M strapped across each diode to
> > equalise reverese voltages, and when one end of the OPT tries to rise
> > to more than 2 x +B, the other end tries to go below 0V so diodes
> > conduct, and thus the OPT voltage swing is CLAMPED. Voltage clamping
> > with diodes has been around for ages.
> > There is a schematic example at *
> >http://www.turneraudio.com.au/100w-monobloc2-2004.htm
> > The amp also has Zobel networks across each 1/2 primary of the OPT to
> > load the OPT at HF.
>
> There seems to be an inconsistency here, between what you said above, "connect
> diodes from each end of the OPT to 0V with diode anode connected to tube
> anodes", and the way the "schematic example" shows the diodes wired?

Gee, I thought you slept through most days. But you are correct and
the schematic IS RIGHT, and what I said here in the post WAS WRONG.
Readers take note !

But the diodes work excellently to limit or clamp the Va swings to no
more than +/- B+ at each anode.

Now the same scheme would be impossible to implement on an SE amp
because there's only one phase of voltage and nothing much to stop Ea
rising enormously when no load is present and the tube cuts off when
in pentode or beam tetrode mode where grid current usually has no
limiting effect on Ea swings and voltage gain approaches gm x Ra, or
perhaps 130 for an EL34 or 170 for KT88.

But one could rig a voltage doubler supply to create a rail at 2 x B+,
and run SI diodes with their cathodes taken to tube anodes, and diode
anodes taken to 2 x B+, say +800V, so that when and not if Ea rises
that far the diodes conduct, and Ea is clipped. There will be a
negligible rise in the B+ rail as a a result.
Diodes may be also placed between tube anode and 0V to prevent Ea ever
going negative.

I find that SE amps don't tend to exhibit the excessive Ea problems as
much as PP pentode/tetrode amps, and despite building and selling
quite a few SE amps I've never seen a problem with arcing tubes or OPT
insulations.

The other way to clamp output voltage and thus clamp the anode swing
voltages in a tube amp with OPT is to use two networks of 5W rated
Zener diodes in series with a normal 1N5404 3 amp Si diode between the
OPT secondary outlet and 0V so that if the Vo exceeds the rated Vo for
maximum rated power by say +3dB, then the zeners+diodes conduct to
clamp the voltage. Such clamping of outpur voltage is done routinely
on many SS amps where a diode such as 1N5404 is taken to each + and -
rail so that if the Vo ever exceeds the rail voltages then the diodes
conduct. Back emfs generated in speakers and inductiove loads are thus
prevented from flow *backwards* through output SS devices which will
destroy themselves instantly is the Ve-b back flow V exceeds about
7Vpk. The same sort of diode
clamping is used routinely in SS amps with mosfet OP devices to limit
the Vg-s, and thus not exceed a rated Vg-s which would punctuture gate
insulation which is very thin indeed.

Voltage clamping is thus very easy for anyone to achieve.

Patrick Turner.
>
> --
> Regards,
>
> John Byrns
>
> Surf my web pages at, *http://fmamradios.com/

[Patrick Turner]

>
> Two Zener diodes connected facing in opposite directions to each other
> in parallel ? And a voltage dropping resistor in series . . .- Hide quoted text -

But not in parallel, because zeners act like ordinary diodes in one
direction of current, so the zeners must be in series, and then it
clamps. Using a series R might be OK to give a higher load once
clamping limiting begins to avoid the low Z when zeners turn on. It
should not matter in a tube amp though because the zener voltage
should just exceed the likely peak Vo max when normally loaded. The
zeners have some capacitance and are accused of being slow to act, but
they'll act quite fast enough, and the capacitance is orders of
magnitude below what might cause HF instability because of capacitive
loading of the amp.

I have used about ten 68V x 5W zener diodes in parallel as diode vari-
caps to generate +/- 40kHz of FM signal deviation of a 455kHz RF
signal. The modulation applied to the diodes can be a saw tooth, and
this gives a "wobbulated" IF signal which allows the selectivity curve
of a radio set to be seen on a CRO, so that when one aligns an AM
radio IFTs, or RF sections, or experiments with an IFT teriary
winding, the effects on pass bandwidth and skirt selectivity can be
seen easily. The amount of capacitance with each zener is only tens of
pF maximum. This form of FM modulation would be most difficult to
otherwise achieve unless one had a tuning gang which was driven at say
40Hz rotation by a motor. The electronic method of wobbulation has no
mechanical parts which tend to wear out.

Patrick Turner.

[[email protected]]

On Jul 12, 4:47*am, Patrick Turner > wrote:
> > Two Zener diodes connected facing in opposite directions to each other
> > in parallel ? And a voltage dropping resistor in series . . .- Hide quoted text -
>
> But not in parallel, because zeners act like ordinary diodes in one
> direction of current, so the zeners must be in series, and then it
> clamps. Using a series R might be OK to give a higher load once
> clamping limiting begins to avoid the low Z when zeners turn on. It
> should not matter in a tube amp though because the zener voltage
> should just exceed the likely peak Vo max when normally loaded. The
> zeners have some capacitance and are accused of being slow to act, but
> they'll act quite fast enough, and the capacitance is orders of
> magnitude below what might cause HF instability because of capacitive
> loading of the amp.
>
> I have used about ten 68V x 5W zener diodes in parallel as diode vari-
> caps to generate +/- 40kHz of FM signal deviation of a 455kHz RF
> signal. The modulation applied to the diodes can be a saw tooth, and
> this gives a "wobbulated" IF signal which allows the selectivity curve
> of a radio set to be seen on a CRO, so that when one aligns an AM
> radio IFTs, or RF sections, or experiments with an IFT teriary
> winding, the effects on pass bandwidth and skirt selectivity can be
> seen easily. The amount of capacitance with each zener is only tens of
> pF maximum. *This form of FM modulation would be most difficult to
> otherwise achieve unless one had a tuning gang which was driven at say
> 40Hz rotation by a motor. The electronic method of wobbulation has no
> mechanical parts which tend to wear out.
>
> Patrick Turner.

Ooh, yes, the Zeners have to be connected in series . Absolutely and
positively .

[NT]

On Jul 9, 3:08*am, "Alex Pogossov" > wrote:
> If you are experimenting on a running tube amp, you might be at some stage
> pulling a hot output tube from its socket while the amp is running.
> Alternatively you might accidentally send a high negative pilse to a grid of
> the tube while working on the circuit.
>
> In any case, an abrupt interruption of the plate current will cause a
> voltage spike on the OPT primary. The stored magnetisation energy of
> Lp*(Ia^2)/2 will have to be dissipated. Lp can be quite high in a decent
> amp. If a load is connected, then the energy will be dissipated in the load
> giving you a loud crack in the speaker. A residual smaller energy stored in
> the leakage inductance Ls*(Ia^2)/2 will most likely be safely dissipated in
> a snubber of say 2200pF+5K usually connected in parallel to the OPT primary.
>
> But what is the load is not connected? Then the huge magnetisation energy
> will either cause:
> - arcing in the tube;
> - breakdown in the OPT winding insulation;
> - breakdown of the subber capacitor (say 2200pF);
> - arcing elsewhere in the wiring.or in the tube socket.
>
> Would it be a good idea to placa a varistor rated slightly above the +B
> across the primary? Or a spark gap of some sort, or gas discharge surge
> arrester tube (the later have miniscule capacitance and very reliable). I
> remember in vertical deflection stages of old TVs such varistors were a
> must, since the abrupt cutoff of the current (during flyback) in EL84 often
> used for this purpose was the mode of operation.
>
> However I have never seen any discussions on the varistor protection issue
> on this site.
> What is your opinion?
>
> Regards,
> Alex

I've seen spark gaps on the primary on amps occasionally. But not
often.


NT