Bias & Volume troubleshooting on my NAD '88 JCM800 2205

[Spanngitter]

C37 is a Snubber for the Diodes, normally not really required and if underrated a possible root cause for a catastrophic fail.
for a last time: Issue are problems with your resistors in the Bias Supply so consider replace R51/52/53/55/56/60/61 with high quality MO types, also include the bias caps (get good 105°C rated Version 10uF/100V or 22uF/100V) and it should be good.

 

[Silverburst]

C37 is a Snubber for the Diodes, normally not really required and if underrated a possible root cause for a catastrophic fail.
for a last time: Issue are problems with your resistors in the Bias Supply so consider replace R51/52/53/55/56/60/61 with high quality MO types, also include the bias caps (get good 105°C rated Version 10uF/100V or 22uF/100V) and it should be good.

Thx for chiming in again. jup I heard you guys. just waiting for my order or to arrive that contains all these things. But it is just not 100% clear what the stock amp is supposed to be seeing these different boards that differ from the schematic. I'll just have to experiment. and i'm learning on the fly here. so thx for your patience

 

[Silverburst]

Ok, I got a small step further. Apparently both tube sets I was testing with were defect!

Before I went in to replace all suspicious bias parts, I tested one last time with other tubes and in the meantime a 1ohm resistors installed at the pin 8k cathode to rule out any strange things with my simple bias probe. Now everything reads normal afaik:
- plate voltage 428v
- 0,035-0,040mV drop
-> about 20watts going for 6CA7's with still some room @ my bias trim pot.

But now I get barely any volume on either channel. I really have to dime channel volumes to hear a very little bit (tried master at 9). Clean channel does sound clean, Boost channel does sound distorted.

So I replaced ALL pre amp tubes at once with working tubes I have. But still no change. Fuses are fine btw. tried different speaker Ohm settings as well to rule out bad contacts over there. Tried different guitars, cables, speakers out->no change.

 

[XTRXTR]

Plug directly into the return and see if it is loud there, check and clean the return jack, might as well do all jacks while you are at it. If still not loud then; Check your screens, re-flow the wire that feeds them, pin 4 connection, pin 6 connection, and the post from the board. Also the picture you posted looks like the grid stopper on pin 5 may need to be re-flowed. Also verify the cathode on power tubes are grounded very well.

Keep in mind this an old-ish amp and may have some corrosion and use issues here and there, not a bad idea to reflow the whole board eliminate any cold solder or worn joints.

 

[Silverburst]

Plug directly into the return and see if it is loud there, check and clean the return jack, might as well do all jacks while you are at it. If still not loud then; Check your screens, re-flow the wire that feeds them, pin 4 connection, pin 6 connection, and the post from the board. Also the picture you posted looks like the grid stopper on pin 5 may need to be re-flowed. Also verify the cathode on power tubes are grounded very well.

Keep in mind this an old-ish amp and may have some corrosion and use issues here and there, not a bad idea to reflow the whole board eliminate any cold solder or worn joints.

Thanks! Tested the return and I get zero volume over there, even with master on 10.

Then I went back to the front input and also there, the master has no effect whatsoever. All the volume seems to be coming from the channel volumes. Even with the master at zero, I hear the same slight volume as with the master on 10. (but I have to dime the channel volumes).

I'll clean jacks and reflow what you suggest next.

 

[Silverburst]

All contacts and tube sockets cleaned with kontakt60, reflowed suggested pins -> no effect

pin 4 : 430v on both tubes
pin 6 : 436v on both tubes

Initially, before power tubes went bad and fried HT fuse, I did not have any volume problems btw.

 

[Chris-in-LA]

Make sure that your output selector is in the right position. My 2210 was in between 16 and 8 once and sounded very faint. Or try a jumper to bypass that switch just to rule it out as the problem.

 

[Silverburst]

Ok, think I got it! went in to replace the 2 ugly resisitors because that wouldn't hurt anyways, and noticed one was actually a bit loose on one legg. when I aplied some pull with my pliers it came right out. Looks cleaner now.

Thanks so much for all your help. Going to check the other tubes again now.

I asume the original tubes got maybe fried due to this resistor failing...

 

[Silverburst]

hmmm... volume gone again... there must be some bad contact elsewhere...

 

[Silverburst]

captain's log: When I fiddle with a chop stick around the PI socket connections, I had a few times volumes swells down over a few seconds.... then I fiddle a bit further over there for a few minutes and the volume swells back up again. can't pinpoint the culprit yet but it must be over there. That is probably why volume went and came back after some soldering because I was moving the wires. quite an invisible enemy grrr. In the meantime all bias caps and resistors on the pcb have been replaced as suggested by @Spanngitter

This socket has been replaced as well and solder joints are not shiny, so I'll reflow them as well.

When volume is gone, the tube it still glowing so I presume it is not the 6.3v going of? (I already replaced the tube).

So as long as I don't touch it, I can play and it sounds great!

 

[Pete Farrington]

volumes swells down over a few seconds.... then I fiddle a bit further over there for a few minutes and the volume swells back up again.

Smooth fades up and down indicates a bad connection in the heater circuit. eg a failed solder joint at a valve socket terminal. So one of the triodes somewhere won’t be conducting. A voltage survey whilst it’s in the fault mode will reveal whereabouts it may be.

 

[Silverburst]

Smooth fades up and down indicates a bad connection in the heater circuit. eg a failed solder joint at a valve socket terminal. So one of the triodes somewhere won’t be conducting. A voltage survey whilst it’s in the fault mode will reveal whereabouts it may be.

Thx Pete, I was thinking in that direction, but if a tube isn't "heated", doesn't it stop glowing? It kept glowing in fault mode...

 

[Silverburst]

Something else I wondered. these are the 2 remaining components I didn't replace yet in the bias circuit. these 22nF caps look kinda 'smokey' compared to all other white Wima caps on the board. I that a sign of wear/heat over time?

 

[Pete Farrington]

Bear in mind there's 2 triodes inside the ECC83 bulb, one might be getting heater power, the other not, yet there's still a glow and the bulb still warms up.
And a bad joint at V5 phase splitter may only affect V1-4.
Hence I suggest a voltage survey of at least V1-5 pins 1,3,6 & 8 whilst the amp is in fault mode.

Yes, C34 looks kinda smokey, I suspect R51 may have burnt and dumped on it. I'd clean the area with pure IPA, definitely don't want conductive carbon deposits around there!
https://en.wikipedia.org/wiki/Isopropyl_alcohol

 

[tschrama]

Something else I wondered. these are the 2 remaining components I didn't replace yet in the bias circuit. these 22nF caps look kinda 'smokey' compared to all other white Wima caps on the board. I that a sign of wear/heat over time?

If you suspect the bias circuit ti be faulty, measure the bias voltage. Dont start randomly replacing components.

If the bias voltage is fine, look else where for your problem.

 

[Spanngitter]

Something else I wondered. these are the 2 remaining components I didn't replace yet in the bias circuit. these 22nF caps look kinda 'smokey' compared to all other white Wima caps on the board. I that a sign of wear/heat over time?

Those are the main coupling caps to the power stage. They tend to get "smokey"/dusty due to some charging caused by HV. They normally are quite robust and I only had a very few of these fail but if in doubt you can replace them, preferrably with a 630V Version (Marshall is a bit on the limit with a 400V Rating on those). You can use current production WIMAs or (like I do) try to find some ERO MKT1822 which do fit. However, normally a failed coupling cap from PI to Power Stage does usually reveal itself by the Bias Voltage being pushed off into a more positive direction (e.g. good side -40Vdc, bad side +10Vdc) usually contributing to a tube redpalting at that point.

 

[Silverburst]

Bear in mind there's 2 triodes inside the ECC83 bulb, one might be getting heater power, the other not, yet there's still a glow and the bulb still warms up.
And a bad joint at V5 phase splitter may only affect V1-4.
Hence I suggest a voltage survey of at least V1-5 pins 1,3,6 & 8 whilst the amp is in fault mode.

Yes, C34 looks kinda smokey, I suspect R51 may have burnt and dumped on it. I'd clean the area with pure IPA, definitely don't want conductive carbon deposits around there!
https://en.wikipedia.org/wiki/Isopropyl_alcohol

@Pete Farrington I did full measurement in fault (grey notes at top) and working mode (blue notes at bottom). This time fiddling around v4 triggered the problem a few times, but it is quite hard to replicate. Now I can do what I want and the amp keeps working. can't get it into fault mode anymore. Nudging tube v4 makes the amp crackle now. But anyways, 2 things I notice:

- there is a big difference at V4A plate anode going from 306v in fault mode to 188v in working mode. Otherwise eveywhere else there is only a few volts difference.

- fillament/heater measueres 3v AC on all tubes in either situations. shouldn't this be 6.3v? (missnoted in grid collumn below)

 

[Tatzmann]

Heaters should read about 3.15 theyre
centertapped.

Some later 80s Marshalls can have wonky
valvesockets, very loose and wobbly.

 

[Silverburst]

Heaters should read about 3.15 theyre
centertapped.

Because the feeding wire indeed feeds 2 pins/filaments?

Some later 80s Marshalls can have wonky
valvesockets, very loose and wobbly.

Yeah they are. if you move the wires or tubes, the whole sockets wiggles. PI has already been replaced.

 

[Tatzmann]

If i remember correctly the platevoltage
on my v3b is fairly higher because it is
connected before the first dropping resistor,
more like 480 to 490v.