Dual 50 Plexi/JCM 800

[purpletele]

It does happen when the meter is hooked up.

IS this what doesn't happen;

feels real thin and tinny

in other words; it works fine without meter?????

No, it performs the same with or without the meter hooked up.  It warms up fine, plays fine for a minute and then gets weak.


BV

[shooter]

ok, I'm back one the same page, nut I still got your meter in the 70% chance bad bucket

what kinda AC on the cap?

[purpletele]

ok, I'm back one the same page, nut I still got your meter in the 70% chance bad bucket

what kinda AC on the cap?

When I am hooked up as described my Alternating Current is  going from 210 V to 77 V then to 28 V.

When I switch over to DC it starts at 497 V  and bottomed out at 341 V

[pdf64]

I think that connecting a meter to the power tube plate is a bad idea, as it effectively puts a massive transmission antennae there, invoking feedback loops to earlier stages.
It's no surprise then that the amp oscillates, resulting in crazy readings (and potentially damaging levels of Vac on to the meter).
Just measure the HT at its feed to the OT, the Vdc at the plate/s is hardly going to be that much lower.
I'll hazard a guess that the crazy readings stop if the same test is repeated with LTP tube removed (thereby breaking the most problematic loops)?

[purpletele]

I think that connecting a meter to the power tube plate is a bad idea, as it effectively puts a massive transmission antennae there, invoking feedback loops to earlier stages.
It's no surprise then that the amp oscillates, resulting in crazy readings (and potentially damaging levels of Vac on to the meter).
Just measure the HT at its feed to the OT, the Vdc at the plate/s is hardly going to be that much lower.
I'll hazard a guess that the crazy readings stop if the same test is repeated with LTP tube removed (thereby breaking the most problematic loops)?

Thanks PDF,

I had performed the test as you suggested previously, but I did it again to see what occurs.

I am getting the exact same result when I read the voltage from the OT feed at the fuse.

The voltage starts in the high 470's and then just about 45 seconds it starts ticking away volts.  Then it will drop quickly to 345 V and hold.

When I play the amp I can hear and feel the amp lose power and volume which is consistent with the visual on the meter.

Thanks for the input

BV

[MFowler]

Is the filter cap charging up at turn on and then discharging somehow?


You get the same decreased voltage reading at the screens?


Mark

[Willabe]

Is the filter cap charging up at turn on and then discharging somehow?

MFowler might be on to something?

I think at this point, I'd change all the B+ caps.

You could try disconnecting the B+ line after the screen grid B+ filter node, and pull all the tubes except the power tubes. That would eliminate everything down stream from there. 

Maybe it's a new/bad cap that's leaking?

[st]

If it's one of the filter caps, there's two ways to know. 1. High voltage capacitor tester, which most people don't have. 2. Take caps out of circuit, tack in new ones (different values are okay for now, as long as they can stand the voltage).

Note, however, that it has been observed that the problem only happens with the power tubes installed. This might indicate the following: 1. It has to do with the power tubes themselves but if I remember correctly different pairs have been tried. 2. It has to do with circuitry associated with the power tubes. A. This could something like oscillation happening in the power amp, but i guess removing preamp tubes or disconnecting nfb should affect this, which apparently it doesn't. Can you conform this? B. It could also be a physical problem. Look carefully at the rear of the sockets when you insert and remove the tubes (Amp off). Anything touching anything? getting really close? Any dirt/gunk/deposit on the sockets? burn marks? Take closup pics. C. It could also have to do with  the power tubes' current draw and a possibly associated thermal issue (although you say there are no signs of heat). When it happens, do all voltages go down, e.g. preamp? And have you by any chance measured the ac at the pt secondary (at input of recitifier)? Does it also go low?

[purpletele]

Is the filter cap charging up at turn on and then discharging somehow?


You get the same decreased voltage reading at the screens?


Mark

That seems logical, I had ordered a new 50/50 Cap to test, I'll grab some of the on-board caps as well.

BV

[purpletele]

Is the filter cap charging up at turn on and then discharging somehow?

MFowler might be on to something?

I think at this point, I'd change all the B+ caps.

You could try disconnecting the B+ line after the screen grid B+ filter node, and pull all the tubes except the power tubes. That would eliminate everything down stream from there. 

Maybe it's a new/bad cap that's leaking?

It is worth investigating.  That is where I will be focusing on the next few days.

[purpletele]

If it's one of the filter caps, there's two ways to know. 1. High voltage capacitor tester, which most people don't have. 2. Take caps out of circuit, tack in new ones (different values are okay for now, as long as they can stand the voltage).

Note, however, that it has been observed that the problem only happens with the power tubes installed. This might indicate the following: 1. It has to do with the power tubes themselves but if I remember correctly different pairs have been tried. 2. It has to do with circuitry associated with the power tubes. A. This could something like oscillation happening in the power amp, but i guess removing preamp tubes or disconnecting nfb should affect this, which apparently it doesn't. Can you conform this? B. It could also be a physical problem. Look carefully at the rear of the sockets when you insert and remove the tubes (Amp off). Anything touching anything? getting really close? Any dirt/gunk/deposit on the sockets? burn marks? Take closup pics. C. It could also have to do with  the power tubes' current draw and a possibly associated thermal issue (although you say there are no signs of heat). When it happens, do all voltages go down, e.g. preamp? And have you by any chance measured the ac at the pt secondary (at input of recitifier)? Does it also go low?

ST,

All good points and test options, thank you.  I'll respond after I work through some of the suggestions. 

BTW, I did not try another set of tubes, but I will today.

BV

[Willabe]

BTW, I did not try another set of tubes, but I will today.

Try swapping the power tubes 1st.

You have EL34's in there? They have 3 grids, that when they heat up can move and that could cause them to short.

[purpletele]

BTW, I did not try another set of tubes, but I will today.

Try swapping the power tubes 1st.

You have EL34's in there? They have 3 grids, that when they heat up can move and that could cause them to short.

Willabe,

I just put in new tubes and it performed just as the others.  The voltage built up to the high 400's and then started dropping.

BV

[shooter]

this will be a ugly hail-Mary suggestion
pull all the pre an PI tubes
If you don't have a spare 32uf+32uf PS cap, use any collection of caps that will get you to >500vdc, >20uF. make that your  "new" PA tube PS tap. (It don't have to be pretty).  BUT you NEED to be safe!
disconnect NFB and tape off.
disconnect the secondary wires on the OT.  Use gator clips to connect them directly to the speaker, which is isolated from amp, chassis, etc.
ONLY measure at your new "tap", not plates.   

[sluckey]

Don't forget... Once he disconnects the NFB wire, everything is fine.

[shooter]

ya, my thought, get the cap outta the way cuz it ain't a prob, but it's proven now.
then add the NFB back with gator clip once the speaker jacks & wiring are gone from the equation.
then it's wonky meter time 

can the coupler caps from PI be safely grounded on the left side?  my "uneasiness" there is DC creep that could effect bias 

[pdf64]

...when I read the voltage from the OT feed at the fuse.

The voltage starts in the high 470's and then just about 45 seconds it starts ticking away volts.  Then it will drop quickly to 345 V and hold...

So the above happens only if the NFB loop is connected?
Is that true even with no tube in the LTP phase splitter?

While it is happening, what is the mV across the 1ohm power tube cathode resistors?

[purpletele]

...when I read the voltage from the OT feed at the fuse.

The voltage starts in the high 470's and then just about 45 seconds it starts ticking away volts.  Then it will drop quickly to 345 V and hold...

So the above happens only if the NFB loop is connected?  No, it happens either connected or not connected


Is that true even with no tube in the LTP phase splitter?  It is true

While it is happening, what is the mV across the 1ohm power tube cathode resistors?

The bias setting is stable at 52 mV, however as the amp was in the peak voltage stage (Assuming the same pattern) the bias was around 75mV.  I heard a mild pop through the speakers and the bias voltage dropped and stabilized at 52 mV.

[st]

Don't forget... Once he disconnects the NFB wire, everything is fine.

I somehow missed this...

[purpletele]

Don't forget... Once he disconnects the NFB wire, everything is fine.

I somehow missed this...

That is not true anymore.  Originally there was a big difference with the NFB connected vs. not connected.  I am not sure what happened but the NFB connected or not no longer has a significant difference. 

I have a new cap can coming tomorrow and some more 47 UF 450V caps by Saturday, I think that will be the next test.

Thanks for checking it out.

BV

[shooter]

I think that will be the next test.

while you're waiting you can still eliminate the secondary of the OT.   gator clip the OT wires directly to a "free-standing" speaker, NO nfb.

IF PA tap is stable, gator clip the spkr - (neg) to your cap neg, recheck.
IF it's still stable, add the PI tube, recheck.
IF it's still stable, use a 2nd gator clip from speaker + (pos) to the NFB point, recheck.

[pdf64]

...The bias setting is stable at 52 mV, however as the amp was in the peak voltage stage (Assuming the same pattern) the bias was around 75mV.  I heard a mild pop through the speakers and the bias voltage dropped and stabilized at 52 mV.

It would be good to see what is happening to the bias supply output voltage over the course of the above pattern, ie the feed to the 220k pair of power tube grid leak resistors.

If that is stable, what is happening to the bias voltage at the power tube control grids, socket terminals #5?

[purpletele]

I think that will be the next test.

while you're waiting you can still eliminate the secondary of the OT.   gator clip the OT wires directly to a "free-standing" speaker, NO nfb.

IF PA tap is stable, gator clip the spkr - (neg) to your cap neg, recheck.
IF it's still stable, add the PI tube, recheck.
IF it's still stable, use a 2nd gator clip from speaker + (pos) to the NFB point, recheck.

I had installed a new fresh OT last week and that obviously didn't change things, but it eliminated the possibility that I damaged it early on.

I'll still follow the instructions and run the test, I want to see the reactions.

Thanks,

BV

[purpletele]

this will be a ugly hail-Mary suggestion
pull all the pre an PI tubes
If you don't have a spare 32uf+32uf PS cap, use any collection of caps that will get you to >500vdc, >20uF. make that your  "new" PA tube PS tap. (It don't have to be pretty).  BUT you NEED to be safe!
disconnect NFB and tape off.
disconnect the secondary wires on the OT.  Use gator clips to connect them directly to the speaker, which is isolated from amp, chassis, etc.
ONLY measure at your new "tap", not plates.

I missed this note.

[shooter]

Ya, got that you swapped OT, I'm still in the HF osc camp or meter gone wild, since the only thing left in the PA is cap and secondary wiring/jacks......

I do have a nagging little helper whispering something about the bias circuit, or DC creep, but I'm ignoring him for now 

[purpletele]

Ya, got that you swapped OT, I'm still in the HF osc camp or meter gone wild, since the only thing left in the PA is cap and secondary wiring/jacks......

I do have a nagging little helper whispering something about the bias circuit, or DC creep, but I'm ignoring him for now 

The meter gone wild seems to be a potential culprit, but the amp acts the same whether I measure the voltage or plug in and play the amp. 

I am looking at putting $300 into a nice meter if anyone has suggestions I would certainly appreciate it.

I placed a close up of the bias circuit and trim pot.  I had the trim pot set up incorrectly originally, then I added turrets to set it the way it is now, which I think is correct, but maybe not.

I found a 32 /32 500V cap that I can run tests with, so I'll start with that.

Thanks for the help

[DummyLoad]

does the bias supply drift with the B+? e.g. does bias voltage at pin 5 fall? you may simply have an issue with the bias supply. check the pot for drift or temporarily tack in a fixed resistor to validate.

[Willabe]

does the bias supply drift with the B+? e.g. does bias voltage at pin 5 fall? you may simply have an issue with the bias supply. check the pot for drift or temporarily tack in a fixed resistor to validate.

I had a small bias pot, like you have, in my BF Princeton reverb that went bad. I would turn on the amp and after a 30 seconds or so, Iwould hear a small pop and the bias would shift.

You said you hear a small pop.

Tack in a R in place of the -bias pot like DL wrote.   

-Bias cap could be bad too.   

[purpletele]

Wow, this is getting interesting.

1. I swapped the 'A' Cap with a 32/32 500V Cap without change.  I changed it to the original and tested with same results.
2. I swapped the 'B' Cap with the 32/32 500 V Cap without change.
3. I changed out the 25 K bias pot with a new one and the same result.

4. I have a speaker here that I am going to run the isolation test that Shooter recommended. 
5. I hooked up the secondary to the speaker +.  I clipped the ground to the speaker ground.

6. I pulled the pre amp and PI tubes.
7. Flipped the switch and the circuit loaded to 490 V and started dropping quickly.

8. I shut it down and poured a glass of wine.  Time to play a working amp.

Thanks for the help

BV

[sluckey]

Has this amp been plugged into that lamp limiter this whole time?

[purpletele]

Has this amp been plugged into that lamp limiter this whole time?

No, not the whole time.  I plugged it in for the speaker isolation test.

[Willabe]

1. I swapped the 'A' Cap with a 32/32 500V Cap without change.  I changed it to the original and tested with same results.
2. I swapped the 'B' Cap with the 32/32 500 V Cap without change.

When you did this, 1 & 2, did you have the rest of the down stream B+ disconnected and the PI and preamp tubes out?

If you didn't disconnect the down stream B+ filter caps, then you don't know if it's 1 of them that's bad and bleeding the B+ off. Trying to isolate, divide/conquer, just the power tubes/PT B+ OT CT node and screen node/OT.

If that stuff is OK, then go through the -bias circuit test's/voltage readings the guys suggested.

[purpletele]

1. I swapped the 'A' Cap with a 32/32 500V Cap without change.  I changed it to the original and tested with same results.
2. I swapped the 'B' Cap with the 32/32 500 V Cap without change.

When you did this, 1 & 2, did you have the rest of the down stream B+ disconnected and the PI and preamp tubes out?

If you didn't disconnect the down stream B+ filter caps, then you don't know if it's 1 of them that's bad and bleeding the B+ off. Trying to isolate, divide/conquer, just the power tubes/PT B+ OT CT node and screen node/OT.

If that stuff is OK, then go through the -bias circuit test's/voltage readings the guys suggested.

Thank you I'll be back on it later today.

[purpletele]

I am studying the Hoffman Bias section and I think I have this particular Piher 25KL Trim Pot installed incorrectly.

1. I have the middle leg of the trim pot tied one of the Pair of legs, and that connects to the 47K resistor.

2. I haven't been able to determine the designation of the legs on this vertical trim pot, but I am thinking the middle leg is really the conductor that should be connected to the two 220K resistors.

I attached a photo and a cut sheet.  Please let me know if I botched the trim pot installation.

Thanks BV

[sluckey]

You have a choice. Do you connect the center leg to this outer leg or the other outer leg? Doesn't matter.

Then you have another choice. Do you connect the outer leg and center leg (that are connected together) to the 47K or to the 220K grid resistors? Again, it doesn't matter.

The bottom line... you want the negative voltage on pin 5 of the output tubes to change when you turn the pot. If it does change then don't worry. Be happy.
 

[purpletele]

You have a choice. Do you connect the center leg to this outer leg or the other outer leg? Doesn't matter.

Then you have another choice. Do you connect the outer leg and center leg (that are connected together) to the 47K or to the 220K grid resistors? Again, it doesn't matter.

The bottom line... you want the negative voltage on pin 5 of the output tubes to change when you turn the pot. If it does change then don't worry. Be happy.

Thanks, digging deep looking for clues.

[sluckey]

I know. And I feel for you. I keep thinking we are overlooking something simple. It seems to be a power amp problem. I would pull all the preamp tubes, leaving only the PI and PA tubes in and troubleshoot that. Maybe open up that other 50W amp you just built. Put them side by side and make some voltage comparisons. I suspect the power amps in both have nearly identical circuits.

I want to see some voltage readings for ***EVERY*** pin on those EL34s while the amp is misbehaving.

[purpletele]

I know. And I feel for you. I keep thinking we are overlooking something simple. It seems to be a power amp problem. I would pull all the preamp tubes, leaving only the PI and PA tubes in and troubleshoot that. Maybe open up that other 50W amp you just built. Put them side by side and make some voltage comparisons. I suspect the power amps in both have nearly identical circuits.

I want to see some voltage readings for ***EVERY*** pin on those EL34s while the amp is misbehaving.

I'll run some more tests today,

Thanks, it is all in good fun.

[shooter]

it is all in good fun.

 
that goes miles in tough troubleshooting!
along with your favorite "adult distraction" and a good amp 

[purpletele]

The first test I ran today was the stability of the filament voltage.

Check this video out as I start up the amp cold.  half way through the video you hear a series of knocks, that is the amp making the noise as it is dropping voltage.

[purpletele]

Positive Update

1. I installed the Classictone 40-18023 PT after I ran the Filament test today.

2. The amp is now running stable but it is running hot.

3. The plate voltage is 478 and I went to set the bias to 34 and I could only go down to 66 mV.

4. I replaced the 220K V Bias resistor with a 100K Bias resistor and couldn't get a reading.  I need a 180K Resistor.

The amp plays really stiff but the voltage is stable.

5. How can I reduce the bias voltage to balance the circuit? 

6.  Getting closer, long day of quiet amp work, with multiple bursts of loud guitar throughout the day.

7. My pre-amp voltages are a little hot as well.

8. Current voltage chart is attached for review

How is that for perseverance   

[DummyLoad]

>I replaced the 220K V Bias resistor with a 100K Bias resistor and couldn't get a reading.  I need a 180K Resistor.

use 2 x 330K 1W in parallel. should be close enough to trim.   

--pete

[purpletele]

>I replaced the 220K V Bias resistor with a 100K Bias resistor and couldn't get a reading.  I need a 180K Resistor.

use 2 x 330K 1W in parallel. should be close enough to trim.   

--pete

Thanks Pete!

[sluckey]

This is sounding encouraging. 

[purpletele]

This is sounding encouraging. 

  Still losing juice.  The dual 330K resistors worked great to get me in range for the bias.  It was looking very good for a few fleeting moments, then I started leaking again..

I'll try rolling through the 47/450 caps.  I have a 47 uF/500 V on hand to use.

[MFowler]

On the video as the power tubes warmed up it popped and reduced the voltage, has to be bad power tube or tubes.

[purpletele]

On the video as the power tubes warmed up it popped and reduced the voltage, has to be bad power tube or tubes.

Mark,

I just found it, or at least I think I have.  The 47 uF/450 V cap in the Plexi Pre-Amp section must have been leaking.  The replacement held the voltage as I rolled the Cap through the 3 on-board caps.

The voltages all look much more in line , plate Voltage of 478, etc.

I also found I have a bad connection somewhere that is affected when I push down on the board.  So I'll be looking for that, but I am encouraged about the steady voltage!

[purpletele]

Success

I got it running smooth and stable.  I have had it on and off for quite a few hours today.

I have to assume that the 47 uF 450V cap was definitely one of the culprits. 

I like the Plexi side best.

I remember why I didn't have a JCM 800, too harsh!!  I had used 500 pF silver micas on the JCM and plexi circuit instead of 470 pF as designed.

Any suggestions to take the iciness away from the JCM, it's a bit sterile and will make your eyeballs bleed.  I want to soften it and smooth it out if possible.

Thanks, that was quite a ride.

I have a couple of head shells that I can use but I thought Red looked cool.

[sluckey]

Glad you hung in there! You can probably cut some of the harsh ice pick by removing those two 470pF caps on the voltage dividers in the V2 circuit. Maybe remove the .001 from the volume pot also. The red head looks bold and daring! 

[shooter]

what He said;
if you're feeling froggy, un-do the select switch and replace with a blend pot for testing, it might be worth it