Help with Traynor YSR-1 B+ issues?

[Jim]

I just spent a couple of hours trying to get a Traynor Custom Reverb
back up to snuff. Schematic:
http://www.prowessamplifiers.com/schematics/misc/traynor_customreverb_ysr1b.pdf

Symptom: Low output. Non functioning reverb and tremolo (I think, but
didn't try tremolo again after getting the bias under better control).

I'm seeing much higher than expected preamp plate supplies. I don't
have the notes in front of me, but I think some preamps had about 400V
on the plates!

OT and screen supplies are smack on (a couple percent high, but with
122V line), and the voltage dropping resistors are in spec (and filter
caps are new).

I ran out of time at this point, and haven't been back on it.

This was done before firing it up: Grounded AC cable, replaced all
electrolytics, and brought some minor changes back to the stock circuit.
New Mullard re-issue EL34 idling at about 17W, and strong preamp tubes.

To the more experienced: How would you proceed?

I tend to learn on a "need to know basis." So all help is appreciated.

Thanks in advance,
Jim

[RS]

I'll take a look at the schematic later, but if the preamp tubes have
400v on the plates, they aren't conducting. Any current through the
string of power supply dropping resistors will decrease the voltage,
and current through a pre tube's plate resistor will drop it further.
That's why I spec electrolytics for preamp section at full B+ voltage;
If the amp is turned on without the pre tubes plugged in, those caps
will get full B+.

Sounds like there's no current, and hence maybe no heater voltage or
something? Measure the voltage across each preamp tube's plate
resistor to determine how much current is being drawn by the tube. If
they're all drawing next to nothing (as would probably be the case if
400v is correct), then the problem is global, like a filament supply.

Anyway, that's just a guess until I take a closer look.

[Jim]

I'll measure the voltage drops across the plate resistors. But the
tubes are doing something, because I get some output. Less than what
I'd expect for a 1W amp, and it distorts when volumes are 3/4's up or so.

[RS]

If it's not distorting til 3/4, then the plate voltages must be closer
to normal. See if you can confirm voltages on the supply and on preamp
stage plates.

1W amp? BTW, you might want to get an o'scope. Sounds like you'd make
good use of it.

[Jim]

On 10/31/2011 12:16 AM, RS wrote:
> On Sun, 30 Oct 2011 21:20:25 -0700, Jim<ask@first> wrote:
>
>> On 10/30/2011 9:07 PM, RS wrote:
>>>
>>> I'll take a look at the schematic later, but if the preamp tubes have
>>> 400v on the plates, they aren't conducting. Any current through the
>>> string of power supply dropping resistors will decrease the voltage,
>>> and current through a pre tube's plate resistor will drop it further.
>>> That's why I spec electrolytics for preamp section at full B+ voltage;
>>> If the amp is turned on without the pre tubes plugged in, those caps
>>> will get full B+.
>>>
>>> Sounds like there's no current, and hence maybe no heater voltage or
>>> something? Measure the voltage across each preamp tube's plate
>>> resistor to determine how much current is being drawn by the tube. If
>>> they're all drawing next to nothing (as would probably be the case if
>>> 400v is correct), then the problem is global, like a filament supply.
>>>
>>> Anyway, that's just a guess until I take a closer look.
>>
>> I'll measure the voltage drops across the plate resistors. But th

>> tubes are doing something, because I get some output. Less than what
>> I'd expect for a 1W amp, and it distorts when volumes are 3/4's up or so.
>
> If it's not distorting til 3/4, then the plate voltages must be closer
> to normal. See if you can confirm voltages on the supply and on preamp
> stage plates.
>
> 1W amp? BTW, you might want to get an o'scope. Sounds like you'd make
> good use of it.

I have an old but functional 'scope. I just didn't have time to do much
else. I basically powered down as soon as I verified high preamp plate
supplies because I had some nice tubes in it. I'll feed it with less
valuable tubes and get some voltages ASAP.

[RS]

On Mon, 31 Oct 2011 11:14:03 -0700, Jim <ask@first> wrote:

>I have an old but functional 'scope. I just didn't have time to do much
>else. I basically powered down as soon as I verified high preamp plate
>supplies because I had some nice tubes in it. I'll feed it with less
>valuable tubes and get some voltages ASAP.

Did you measure just one tube plate, or did you read 400v on multiple
plates? Max plate v on a 12ax7 is 330v, so yeah, you'd want to be
careful, but I doubt that any harm was done. Given your comments on
operation, I'm guessing that something was wrong when you got the 400v
reading. If you are getting signal, and aren't getting severe
distortion, the plates are probably in operational region. That's a
simple circuit, after all.

I can't read the exact numbers on the schematic, but I believe the
plate -supply- is only around 230+v there. A couple different things
would have to go wrong in order to measure 400v on the plates
themselves. If any of the input tubes is pulling current through R50
in the supply (that's what...33k?), then there will be a voltage drop.
So again, likely that there was an error in measurement somewhere.

A couple other comments on the circuit:

They use a shared screen resistor. That sorta works, but I wouldn't do
that. You may want to replace that with separate 1K resistors for each
output tube.

The trem works by varying the EL34s' bias. The trem depth control has
the bias voltage attached to one leg, and the output to the control
grids comes from the wiper. If that pot gets noisy or the wiper loses
contact, the control grids will float and you're liable to lose two
EL34's. You can tie a high value resistor between the two mentioned
terminals of that pot to make sure there's a path for bias voltage if
the pot fails.

I'm not a fan of shrieking treble. If you are, disregard this: The
lower preamp channel on the schematic has a .001 cap (C6) bridging the
volume control. I never understand why some designers think that
wide-open treble is a great idea, but that's like a Fender Bright
switch x5. You can probably just lift one end of the cap, tape it, and
leave it in place.

Same for C14, which is a permanently wired Presence turned to 10.
Boosts only extreme highs, which I find annoying. Same remedy.

Odd but sort of interesting tone control topology. High insertion
loss. You may want to leave them as is for more novel effect, or
change to Fender type controls if you want a more 'direct' sound.
Since you have 500k pots in there already, you could look at the old
Magnetone amps that have Fender controls scaled up to double impedance
(500k pots).


I could make a couple other comments on reverb, etc, but best to just
address the larger problems. I wouldn't even count the tone controls
among those, but personally, I would ditch the extreme treble boost
circuits. I never liked that about some Traynor designs.

[Jim]

On 10/31/2011 10:49 PM, RS wrote:
> On Mon, 31 Oct 2011 11:14:03 -0700, Jim<ask@first> wrote:
>
>> I have an old but functional 'scope. I just didn't have time to do much
>> else. I basically powered down as soon as I verified high preamp plate
>> supplies because I had some nice tubes in it. I'll feed it with less
>> valuable tubes and get some voltages ASAP.
>
> Did you measure just one tube plate, or did you read 400v on multiple
> plates?

Both, quickly. Didn't even write down the numbers.

Max plate v on a 12ax7 is 330v, so yeah, you'd want to be
> careful, but I doubt that any harm was done. Given your comments on
> operation, I'm guessing that something was wrong when you got the 400v
> reading. If you are getting signal, and aren't getting severe
> distortion, the plates are probably in operational region. That's a
> simple circuit, after all.
>
> I can't read the exact numbers on the schematic, but I believe the
> plate -supply- is only around 230+v there. A couple different things
> would have to go wrong in order to measure 400v on the plates
> themselves. If any of the input tubes is pulling current through R50
> in the supply (that's what...33k?), then there will be a voltage drop.
> So again, likely that there was an error in measurement somewhere.
>
> A couple other comments on the circuit:
>
> They use a shared screen resistor. That sorta works, but I wouldn't do
> that. You may want to replace that with separate 1K resistors for each
> output tube.

I can do that. I'll just bypass the 10W on the board, and put 1k 5W on
each tube.

>
> The trem works by varying the EL34s' bias. The trem depth control has
> the bias voltage attached to one leg, and the output to the control
> grids comes from the wiper. If that pot gets noisy or the wiper loses
> contact, the control grids will float and you're liable to lose two
> EL34's. You can tie a high value resistor between the two mentioned
> terminals of that pot to make sure there's a path for bias voltage if
> the pot fails.

I just saw the same design in the pre-CBS Princeton that is now back in
shape. That tremolo didn't work until I put tubes in that drew the
right current range.

What value bypass resistors?

>
> I'm not a fan of shrieking treble. If you are, disregard this: The
> lower preamp channel on the schematic has a .001 cap (C6) bridging the
> volume control. I never understand why some designers think that
> wide-open treble is a great idea, but that's like a Fender Bright
> switch x5. You can probably just lift one end of the cap, tape it, and
> leave it in place.

I also noticed that. I believe that's the only difference between the
two input stages. I was going to see how it runs by bridging the
"channels" at the inputs. Then decide if I want to change that or do
more of a Marshall type change.

>
> Same for C14, which is a permanently wired Presence turned to 10.
> Boosts only extreme highs, which I find annoying. Same remedy.

I didn't pay attention to that, so thanks. I did notice the high value
on the NFB loop, 100k.

Mine has push/pull on the bass that looks to be either original, or
early mode. It puts in a smaller coupling cap to the PI stage, thinning
the tone. Not sure I'll find that useful. If not, I suppose I could
make use of the SPST on C14 for two choices.

>
> Odd but sort of interesting tone control topology. High insertion
> loss. You may want to leave them as is for more novel effect, or
> change to Fender type controls if you want a more 'direct' sound.

I'll see how much drive I get as-is.

> Since you have 500k pots in there already, you could look at the old
> Magnetone amps that have Fender controls scaled up to double impedance
> (500k pots).

I also have the Duncan tone stack calculator.

>
>
> I could make a couple other comments on reverb, etc,

Reverb had some odd tweaks that I removed, but noted on my schematic.
So it may have been a fix for poor operation. There was a parallel 470k
plate resistor tacked piggyback on the stock 470k for V8 pin 6, and the
pin 3 cathode resistor had another 1.5k and a diode in parallel. So it
looks like they were after more reverb gain.

but best to just
> address the larger problems. I wouldn't even count the tone controls
> among those, but personally, I would ditch the extreme treble boost
> circuits. I never liked that about some Traynor designs.
>

Maybe I'll get back to it tonight.

[RS]

On Tue, 01 Nov 2011 11:33:14 -0700, Jim <ask@first> wrote:

>On 10/31/2011 10:49 PM, RS wrote:

>> The trem works by varying the EL34s' bias. The trem depth control has
>> the bias voltage attached to one leg, and the output to the control
>> grids comes from the wiper. If that pot gets noisy or the wiper loses
>> contact, the control grids will float and you're liable to lose two
>> EL34's. You can tie a high value resistor between the two mentioned
>> terminals of that pot to make sure there's a path for bias voltage if
>> the pot fails.
>
>I just saw the same design in the pre-CBS Princeton that is now back in
>shape. That tremolo didn't work until I put tubes in that drew the
>right current range.
>
>What value bypass resistors?

Control grids are very high impedance, so a high value will do fine,
maybe 470k. It will exceed the EL34 spec for "DC resistance to
ground", but the idea is to keep the output tubes alive until the pot
can be fixed.

>> I'm not a fan of shrieking treble. If you are, disregard this: The
>> lower preamp channel on the schematic has a .001 cap (C6) bridging the
>> volume control. I never understand why some designers think that
>> wide-open treble is a great idea, but that's like a Fender Bright
>> switch x5. You can probably just lift one end of the cap, tape it, and
>> leave it in place.
>
>I also noticed that. I believe that's the only difference between the
>two input stages. I was going to see how it runs by bridging the
>"channels" at the inputs. Then decide if I want to change that or do
>more of a Marshall type change.

Marshall?

>> Same for C14, which is a permanently wired Presence turned to 10.
>> Boosts only extreme highs, which I find annoying. Same remedy.
>
>I didn't pay attention to that, so thanks. I did notice the high value
>on the NFB loop, 100k.

You can't judge amount of NFB purely from the value of that feedback
resistor. It's ratiometric, with that resistor working against R14,
the 4.7k (and that's how the 'presence' cap works, as well). So the
overall ratio of 100k:4.7k is about 20:1. Fender's 820 ohm to 47 ohm
is roughly in the same range. Those values are often scaled up
intentionally in circuits that use a presence cap so the value of the
cap can be kept relatively small. The equivalent cap in the Fender
circuit would have to be around 10 uf...impractical.

>Mine has push/pull on the bass that looks to be either original, or
>early mode. It puts in a smaller coupling cap to the PI stage, thinning
>the tone. Not sure I'll find that useful. If not, I suppose I could
>make use of the SPST on C14 for two choices.

That bass cutoff cap will prove useful when you crank the amp. Bass
frequencies will more easily overwhelm the output stage and force the
tubes into grid conduction. A subject for another post if you're
interested. The switch will be useful when playing clean at lower
levels vs dimed.

>> Odd but sort of interesting tone control topology. High insertion
>> loss. You may want to leave them as is for more novel effect, or
>> change to Fender type controls if you want a more 'direct' sound.
>
>I'll see how much drive I get as-is.

Not as much a question of drive, as possible phase cancellations,
though I haven't done any math on the existing tone stack. The Fender
circuit is very simple, with frequencies strategically placed for
their sound, and it does have the overall effect of 'directness',
especially in upper mids.

>> Since you have 500k pots in there already, you could look at the old
>> Magnetone amps that have Fender controls scaled up to double impedance
>> (500k pots).
>
>I also have the Duncan tone stack calculator.

Useful. But the Mag circuit is already there, and sounds nearly
identical to Fender's. You'll see that if you model it on Duncan's

calculator.

>> I could make a couple other comments on reverb, etc,
>
>Reverb had some odd tweaks that I removed, but noted on my schematic.
>So it may have been a fix for poor operation. There was a parallel 470k
>plate resistor tacked piggyback on the stock 470k for V8 pin 6, and the
>pin 3 cathode resistor had another 1.5k and a diode in parallel. So it
>looks like they were after more reverb gain.

Actually, the higher the plate resistor value, the higher the gain.
But there are diminishing returns. Not a lot of sense in going much
past 220k for a guitar amp circuit. Hence the paralleled 470k. Do you
think that was a factory mod?

> but best to just
>> address the larger problems. I wouldn't even count the tone controls
>> among those, but personally, I would ditch the extreme treble boost
>> circuits. I never liked that about some Traynor designs.
>>
>
>Maybe I'll get back to it tonight.

OK.

[Jim]

I was thinking about the 1959 / 1987 preamps. Including the .68 uF
cathode resistor bypass cap. Then change that .001 to something
smaller. But first things first...

>
>>> Same for C14, which is a permanently wired Presence turned to 10.
>>> Boosts only extreme highs, which I find annoying. Same remedy.
>>
>> I didn't pay attention to that, so thanks. I did notice the high value
>> on the NFB loop, 100k.
>
> You can't judge amount of NFB purely from the value of that feedback
> resistor. It's ratiometric, with that resistor working against R14,
> the 4.7k (and that's how the 'presence' cap works, as well). So the
> overall ratio of 100k:4.7k is about 20:1. Fender's 820 ohm to 47 ohm
> is roughly in the same range. Those values are often scaled up
> intentionally in circuits that use a presence cap so the value of the
> cap can be kept relatively small. The equivalent cap in the Fender
> circuit would have to be around 10 uf...impractical.

Thanks, that makes sense, NOW.

No. Different type resistor, even.

So what about that big old silicon diode on the other triode. Is it
something with biasing it with the forward voltage?

>
>> but best to just
>>> address the larger problems. I wouldn't even count the tone controls
>>> among those, but personally, I would ditch the extreme treble boost
>>> circuits. I never liked that about some Traynor designs.
>>>
>>
>> Maybe I'll get back to it tonight.
>
> OK.

Thanks for the help!