JCM800 2204 clone red plating (please, help)

[IvanMan]

Hello everybody.
I bought this ten years old clone and I'm working on it by many weeks. Replaced one filter caps, some coupling caps, cleaned everything, replaced output tubes.
I find also that the turret board was conductive (with the help of guys in another forum) and replaced it.

Now the last issue is red plating of power tubes, that start (really a few at crossing of plates) with Gain and Master at 12 o'clock.
Sounding guitar (my final scope) it seems that I cannot reach full power and the power itself growth only a little and then nothing after 12 o'clock of the Master.

In the following some measures I take. For each set of measures I take also the picture of scope images across 1 Ohm bias measurement resistor I've between cathode and ground (trace below is V4 and the above is V5).

0) In idle conditions
B+ before choke +459Vdc
V4 Bias current 29.5mA, GR2 voltage -35,9Vdc
V5 Bias current 27,8mA, GR2 voltage -35,8Vdc

In the other conditions:
1) Pre & Master 12 o'clock (image1)
B+ 414Vdc
very few red plating in the middle, where plate crossing
Grid voltage V4 -37.5Vdc, V5 -36Vdc

2) Pre 12 o'clock - Master fully clockwise (image2)
B+ 396Vdc
more consistent red plating, quite more on V4
Grid voltage V4 -52Vdc, V5 -62Vdc

3) Pre fully clockwise - Master 12 o'clock (image3)
B+ 403Vdc
red plating more than 1) but less than 2), more on V5
Grid voltage V4 -44Vdc, V5 -42Vdc

4) Pre & Master fully clockwise (image4)
B+ 392Vdc
red plating more or less like 2), more on V4
Grid voltage V4 -56Vdc, V5 -61Vdc

Have you suggestion of what may be the cause of red plating? Or what can I check to help troubleshooting?

[IvanMan]

May be that the redplating is a consequence of ripple voltage? Results of measures on CT of OT (between low and peak):
1) With no input signal and gain/master fully CCW (x10 probe, 0,1Vdc/division, 4 division on screen) --> 4Vdc
2) with triangular wave input, 400Hz, and gain/master fully CW (x10 probe, 0,5Vdc/division, 4 division on screen) --> 20Vdc
Frequency (1msec/division, 10 divisions) --> 100Hz
I tried to replace C19/C20 (bias caps), diode D1 and the 220K resistor R30, but with no results.

[shooter]

Results of measures on CT of OT

4vac is probably normal, probably 20vac isn't. 
that ripple is from bridge (D2-5) n C23 (50+50)

I'd put in 1 ohm 1-3W R's on the cathode to ground pins of the 34's that way you get a good idea of I flow at idle.
what is the bias range with NO 34's in?
your 1st readings 28-ish mA at 460vdc is like 13W nowhere near redplate range.

with bias more neg you should have less current, less red plate.  once you can accurately measure idle current, might be easier, also verify you have a good verified load on the OT secondary, if that's sketchy the primary gets pretty cranky

[IvanMan]

Hello Shooter, and thanks for your suggestions.

Results of measures on CT of OT
4vac is probably normal, probably 20vac isn't. 
that ripple is from bridge (D2-5) n C23 (50+50)

In the mean time I try to replace the D2-5 rectifier diodes but no changes.
Do you think the ripple can generate redplating?

I'd put in 1 ohm 1-3W R's on the cathode to ground pins of the 34's that way you get a good idea of I flow at idle.

I still have a 1 Ohm resistor between cathode to ground of the power tubes, and the first 4 picture I posted (from image1 to image4) are taken from scope with probes at two end of that resistors, for V4 & V5, so they represent the flowing of bias current at the different position of Gain and Master, as I indicated.
At idle there isn't current flowing (the minimum scale of the scope is 2mV and both the lines are flat).
In the other 4 cases, with waveform sometimes quite different, the maximum value of bias current I see on the scope is 400mA (scale 200mA, 2 division height).
The bias current seems quite symmetric on the 2 tubes, with the exception of Gain=12 o'clock Master=12o'clock --> in this case I've V4=400mVpp and V5=300mVpp
Red plating starts a little onV4 at plate crossing in this condition (may be the 400mA are a level that start generating the issue?)

what is the bias range with NO 34's in?

At pin 5 of the sockets I've V4 -36.6Vdc and V5 -36.5Vdc

your 1st readings 28-ish mA at 460vdc is like 13W nowhere near redplate range.

Yes. I confirm no redplating. That readings ware at idle condition (Gain and Master fully CCW)

also verify you have a good verified load on the OT secondary, if that's sketchy the primary gets pretty cranky

At OT secondaries I've a pure resistive dummy load, a little higher than 16 Ohm (16.8 Ohm).

[VMS]

Does it redplate when playing thru speakers and how does the amp sound?

[shooter]

after you evaluate reply#4
plug in a 800hz(sounds better than 1k:), 100mV rms where the guitar goes, all knobs @5, and give me a scoped  pic @ speaker, or DL, extra credit; meter reading at cathodes while the sig is playing, VDC range, I like the min/max feature if you have one
WATCH for RP
got new tubes?

[IvanMan]

Does it redplate when playing thru speakers and how does the amp sound?

Yes, more or less when the Master is around 5 (twelve o'clock) redplating starts at the plate crossing and arise moving the master CW.
The sound is quite good up to that level (Master on 5), but not moving in the zone between 5 to 10: the volume seems not growing more, the amp not able to reach the full power, the sound become confused.

plug in a 800hz(sounds better than 1k:), 100mV rms where the guitar goes, all knobs @5, and give me a scoped  pic @ speaker, or DL, extra credit; meter reading at cathodes while the sig is playing, VDC range, I like the min/max feature if you have one
WATCH for RP

Test at 800Hz, 100mV RMS (means around 280Vpp I supposed, quite higher than previous test I made were I generally used 100Vpp), all knobs @5).
The results are in the picture attached. The line below is the BIAS current (voltage at the end of 1 Ohm resistor) the scale is 200mV=200mA. The line above is at speaker output jack and the scale is 10V.
1) Picture 1 is: Gain&Master at 5 and tone controls at 0 (Bias for V4)
2) Picture 2 is: all knobs at 5 (Bias for V4)
3) Picture 3 is: all knobs at 5 (Bias for V5)
My comments on redplating:
- with higher signal in input (vs. previous test) the redplating is quite higher with knobs at 5
- Bias values are around 400mA, a little higher in picture 3 for V5. V5 is redplating before and more than V4

One question, sorry but I search all around the forum but didn't find the answer. What is DL?

got new tubes?

The tube are quite new. I replaced them one month ago because one of the previous set blows the HT fuse.

[shooter]

You're images appear to be over-driving the tubes, without a meter reading your scope is showing peak-peak values and if I did the math right you're biased way hot;

Bias values are around 400mA,

.4 * .3535 = .1414A * ballpark 400vdc = 56.6W

[IvanMan]

You're images appear to be over-driving the tubes, without a meter reading your scope is showing peak-peak values and if I did the math right you're biased way hot;

Bias values are around 400mA,

.4 * .3535 = .1414A * ballpark 400vdc = 56.6W

I just recheck the Bias values at idle but they seem correct: Bias current is 27.4mA for V4 and 29mA for V5, Bias voltage (at pin 5) is -36Vdc for both.
When I say Bias is 400mA is bacause I see on the scope (from peak to peak of the grid voltage trace --> with tips at the ends of 1 ohm resistor) 2 division with a scale of 0.2V/division = 0.4V
What other readings can help your evaluation? In case I can also send you a short video of the scope trace?
What is your opinion about the 20V ripple still present at max power? May be influencing?

[shooter]

When I say Bias is 400mA

Bias current is 27.4mA

I never use a scope to measure my DC characteristics, just for the AC signal path.
IF you use a meter at the top of a 1ohm Resistor the mVDC = Tube current in mA, no guess, no mess.

27.4mA is good, .4A isn't

[IvanMan]

When I say Bias is 400mA

Bias current is 27.4mA

I never use a scope to measure my DC characteristics, just for the AC signal path.
IF you use a meter at the top of a 1ohm Resistor the mVDC = Tube current in mA, no guess, no mess.

27.4mA is good, .4A isn't

You are completely right and I'm sorry. At school I'd learn that measures, to be comparable, have to be taken with the same method, but problably with the age I forgot.
Now I taken them (BC Bias Current) with the meter at the end of 1 Ohm resistor, BV Bias Voltage at pin 5, with all knobs on 5 and varying the Master only:
V4 (redplating less the V5, starting a little with Master on 5)
idle - master 0: BC 26mA (varying a bit from 27.7 to 26.4) - BV -35.6Vdc
master 5: BC 165mA - BV -57Vdc
master 10: BC 190mA - BV -63Vdc
V5 (redplating very soon with master on 5)
idle - master 0: BC 28.6mA (same variations around) - BV -35.6Vdc
master 5: BC 230mA - BV -49Vdc
master 10: BC 227mA - BV -58Vdc

Hope this can help.
Ivan

[shooter]

master 5: BC 165mA - BV -57Vdc

are you injecting a signal at the input?
If there is no signal and it's doing this, my gut says you have DC getting through a cap.  If you are injecting a signal my gut says you're overdriving the  outta the PA causing the grid to conduct, BUT, I must be missing something since my understanding of fixed bias; The bigger the negative # the closer you are to cutoff, no current, the smaller the  negative number the closer you are to saturation - max current.

[shooter]

wanted to make sure I wasn't out in the tall weeds;

http://www.valvewizard.co.uk/bias.html

good read on biasing.
I'm guessing your screens are ~~~~450VDC, EL34's/10 ~~~ bias at idle -45vdc

[IvanMan]

master 5: BC 165mA - BV -57Vdc

are you injecting a signal at the input?
If there is no signal and it's doing this, my gut says you have DC getting through a cap.  If you are injecting a signal my gut says you're overdriving the  outta the PA causing the grid to conduct, BUT, I must be missing something since my understanding of fixed bias; The bigger the negative # the closer you are to cutoff, no current, the smaller the  negative number the closer you are to saturation - max current.

Yes, I injecting a signal as you sugegst previously: sinewave, around 300mVpp, 800Hz. The strange things I cannot understand is that both current and negative voltage both increase. What can generate this? Again, may be something generated by the voltage ripple (I measured around 20V B+ ripple at maximum power)?
May be a probleam with the sockets? I replaced the full turret board because it bacame conductive with HV present (nobody understud why). May something similar happened to sockets?

[sluckey]

Disconnect the NFB wire from the OT secondary. Any better?

Post some hi-rez pics of the amp guts.

[IvanMan]

Disconnect the NFB wire from the OT secondary. Any better?

No changes

Post some hi-rez pics of the amp guts.

Here the first two: a general overview and the preamp section. Next post contains the other ones with power part and tubes

[IvanMan]

Post some hi-rez pics of the amp guts.

Sorry for the two cable you see. I unsoledered the bias trimmer and connected a new one with cable. But no changes again.

[shooter]

   I use either all red, or all yellow for tube hook-ups

If your big filter cap is above V1's parts, I'd un-bolt it, GENTLY hang it outside the chassis, see if that changes anything, very doubtful, but easy.

[zendragon63]

Hi Ivanman - these kinds of projects can be really, really frustrating.  And I have done this more than once...

Because there are things that just don't seem to make sense--what might make the most sense is to go back to the fundamentals of wire routing and checking very closely ALL the soldering that has been done. For example, just two things that look suspicious are solder joints like the one on the 1K 5W screen resistor on pin 4 of V4 and the wire routing to the control grids of V4 and V5. Also having the main filter cap hanging over the input components/V1--probably the most sensitive and vulnerable area of the amp in terms of signal--can contribute to extra noise that could be avoided. You might shorten the leads on the bright cap as it kind of acts as a antenna for noise. With all due respect, you might find that a couple of hours invested just re-routing wires and reflowing all solder joints will eliminate a lot of gremlins--as well as ones laying in wait. (Check out dummyload's Boilermaker as it looks similar in layout)

I could be wrong (and according to my bride, I often am ) but I have found that when it comes to amps, good layout beats good luck nearly every time.  Regards

dennis

[pdf64]

...injecting a signal as you sugegst previously: sinewave, around 300mVpp, 800Hz. The strange things I cannot understand is that both current and negative voltage both increase. What can generate this?

The control grid and cathode of a tube form a diode. When Vg1=>Vk, that diode will conduct. So if the positive peak of the signal applied to the grid is greater than the negative bias voltage, the g1-k diode chops off the top of the signal.
That is tube amp overdrive.
And, due to the capacitive coupling to the grid, it will charge up the coupling cap so as to make the effective bias voltage greater; this is known as bias shift, and, in extreme, can cause blocking distortion, see http://www.aikenamps.com/index.php/what-is-biasing

[Ed_Chambley]

The amp may have problems you cannot change without a new layout.  Appears the pots are on the tube side, which can be done and I have done it, but it is not good practice.  I have never taken high voltage near the input.


The filter cap over the bypass section of V1 I would not do.  I would get a Marshall Layout and make this the same.

[sluckey]

That dual cap can on the preamp end of the chassis feeds V1 and V2. It's fine where it is.

[IvanMan]

  I use either all red, or all yellow for tube hook-ups

If your big filter cap is above V1's parts, I'd un-bolt it, GENTLY hang it outside the chassis, see if that changes anything, very doubtful, but easy.

Done but no result. I try again to trace the input signal in all the preamp stages, but it seems good. At the output of PI the two signals are symmetric (good sinewave, same peak values).
In any case, when the power tubes are not redplating, the sound is quite good, very "Marshallish" and with no particular noises.

Hi Ivanman - these kinds of projects can be really, really frustrating.  And I have done this more than once...

Because there are things that just don't seem to make sense--what might make the most sense is to go back to the fundamentals of wire routing and checking very closely ALL the soldering that has been done. For example, just two things that look suspicious are solder joints like the one on the 1K 5W screen resistor on pin 4 of V4 and the wire routing to the control grids of V4 and V5. Also having the main filter cap hanging over the input components/V1--probably the most sensitive and vulnerable area of the amp in terms of signal--can contribute to extra noise that could be avoided. You might shorten the leads on the bright cap as it kind of acts as a antenna for noise. With all due respect, you might find that a couple of hours invested just re-routing wires and reflowing all solder joints will eliminate a lot of gremlins--as well as ones laying in wait. (Check out dummyload's Boilermaker as it looks similar in layout)
I could be wrong (and according to my bride, I often am ) but I have found that when it comes to amps, good layout beats good luck nearly every time.  Regards
dennis

Today I checked again all the solders, replaced R24-R25-R31-R32 (with this I replaced the whole bias circuit and the grid resistors) and cable to the tubes (with AWG20, solid core) but no results.
I replaced also C23 (the first filter cap) because I had a quite high ripple (4V at idle, 20V at max power). Ripple is eliminated but no changes in redplating.

Measuring at ends of the 1 Ohm resistor I use for biasing, with a signal applied at input (I know is not a real measure, but I use it as an indicator) the redplating starts when the value is 120mA and growth if I raise the input signal (no matter if this is obtained raising the input signal at High input jack, or rotating CW the gain or the master volume).
When the value goes beyond 120mA the tube starts redplating a little at cross of plates, and much more if raise the input up to 200-220mA (max value I measured).

In the mean time I searching on the web about redplating and find this very interesting thread on Metropoulos forum: http://forum.metropoulos.net/viewtopic.php?t=30036
There are a lot of things I haven't the knowledge to understand exactly , but there a couple of comments of Randall Aiken about the too high voltage we apply to the power tubes and the fact that this situation force the tube to work in a critical area where redplating could happen more easily. In fact observe that all my voltages on B+ are quite higher than the original Marshall ones (see attachment).
Thare are also a couple of suggestions in that thread:
- try to test with a load of 16 Ohm but with the selector on 8 Ohm --> no results, redplating still the same
- try to reduce to around 400Vdc the B+ value (at the moment I've around 460Vdc).
What do you think about? Can you suggest a simply way to make this test (it's enough a resistor in series)? Thanks to everybody for your comments and helps.



   

[shooter]

up to 200-220mA

in fixed bias you set a "nominal" value, no signal, using ohm, that can be set ~~60-70% of max plate dissipation.  The signal is what drives it to those limits you measured, the closer you get G1 to 0, or past, the crazier things get.  So you need to compromise on drive and bias, and ultimately sound/tone.

lowering the B+/plate, changes the amount of signal you can "swing" before clipping.  You can probably experiment with a ~~~ 470ohm 20W between the bridge diodes and the 1st filter cap C23.  This WILL effect all downstream volts, which will effect all the downstream signal "swing".  Who knows, you might like it 

you can also limit the drive to the PA tubes so no red-plate happens and jam for awhile, who knows, you might like it 

[zendragon63]

Hey IvanMan, how are you fairing with this project? Regards

dennis

[IvanMan]

Hey IvanMan, how are you fairing with this project? Regards

dennis

Hi Zendragon63, thanks for your question.
After some other check I find that the OT primary windings are defective:
- The two primaries have different resistance (when measured between CT and pin 3 of the PT socket): 24.7 and 45.9 Ohm
- I try to apply a 100Vac to each winding of the primary and measure the secondary: 12.4Vac in one case and 9.2Vac in the other
-  Applying 9Vac to the secondary I've 25.3 on one primary (CT to pin 3) and 14.8 on the other primary (CT to other pin 3)

Probably this also cause redplating?
Now I'm waiting for a new OT and then I can go ahead with the test.

When I start to working on this clone my scope was to reproduce the original JCM800 2204. One og the first things I found was that one B+ resistor was eliminated and the preamp part is feeded with a quite  higher voltage than expected. May be this generate an overdriving of the PT and of the OT and this damage both (I have also replaced the two EL34 bacause one of them blow the HT fuse).


Do you know this amp? What do you think about reducing B+ that now is around 470Vdc? I read that the first series the B+ was much lower.
 

[IvanMan]

Hey IvanMan, how are you fairing with this project? Regards

dennis

Hello. Installed the new OT = no more redplating  and full output power again
I'm in doubt, now, about trying however to reduce B+ (switchable?) to complete the transformation to a old style ('80) JCM800 2204. Your opinion?

In any case thanks again for your patience and your suggestions: I learnt a lot!

[sluckey]

I think 470V B+ is fine for that amp. I would not worry about that being too high.

[zendragon63]

Glad you found the culprit Ivanman. That it was the OT is a bit of a surprise but then these trouble shooting expeditions can just as often turn up the unexpected. I will readily admit that I have seen virtual rats nests--and your build didn't drop to that level--that apparently worked despite itself. But that said, good layout and neat wiring is still pretty good insurance against many a troublesome parasitic.


Coincidentally, I still happen to have my 1979 JCM800 MarkII 2204 half stack (even at 50W, way louder and troublesome to cart around than I care to deal with) and it runs about 397V on the El34 plates. Played that set up for almost 20 years and it always sounded great. Apparently some versions ran at a bit lower voltage than you would expect but like sluckey said, 470V shouldn't be an issue.


And don't forget your hearing protection  Regards


dennis