Bias & Volume troubleshooting on my NAD '88 JCM800 2205

[Silverburst]

Hi all,

Went to get this 2205 for a good price because it had some issues.

While testing the amp at the seller's house, it had some hum from the start but within a minute the hum became very loud with a lot of rrrr-sound, which reminded me of the sound I once had with a power tube that was failing. I found the EL34B's tubes were glowing quite a bit, so we shut down the amp and I decided to take it home for a good price.

At home a few days later the amp played again quite fine with a bit of hum but again after a few minutes the hum suddenly became a lot worse with some seriously loud squeaks, so I shut it down immediately. Afterwards when restarting it, no sound at all. I noticed the HT fuse had blown. So I assumed a power tube had died on me during that heart attack sound a few minutes ago.

Took it to my bench and opened it up. I noticed the Bias resistors had been tempered with, but the guy claimed it had run fine for years. Pics and values in post below (Tech note on the inside mentioned new tubes in 2015 and a preamp socket replaced, but I don't think the amp was played that much)

I replaced the HT fuse, put another set of EL34B's I happen to have in there, turned the Bias pot all the way up to get maximum negative voltage and to start I did some measurements with a 1ohm tube probe socket:

• EDIT SOLVED WITH NEW TUBES +/- 300 mA's across the cathode, like 10x too much! I didn't dare to let it settle to avoid damaging the new tubes so shut it down immediately. Neither did I dare to test the other socket in this condition. UPDATE: I tested the other socket and 2nd tube. Both sockets same behavior. 2nd tubes goes up to about 220 mV's thus 220 mA's.

After reading up, I removed the Power tubes and did some extra measurements:

• pin 5 voltage on both sockets, ranging between -36v to -44v when dialing Bias pot from min to max.
• Pin 3 voltage on both sockets = 462v after a few minutes.
• Resistors Pin4-6 on both socket still 1000 Ohm
• Filter cap 1: leg 1 328v and leg 2 378v Filter cap 2: 455v both legs

Thanks for reading until here!

EDIT: NOW HAVING VOLUME TROUBLES

 

[Silverburst]

So about the Bias resistors. I my case:

• R51 is 56K original + added 15k
• R52 & R53 are 220K
• R55 is 2x470k in parallel = 235k

EL34 Schematic below says the following:

• R51 is 56K
• R52 & R53 are 220K
• R55 is 220k

So not sure why the added the 15k to R51 and inserted the R55 is 2x470k (I've seen some pictures of 2205 boards where this position is blank).

The owner used it for years like that, I assume the bias resistors have been like that for years.

 

[Pete Farrington]

Are the 1ohm current sensing resistors in the socket probes definitely good?

 

[Silverburst]

Are the 1ohm current sensing resistors in the socket probes definitely good?

I measured them and they read fine. How can I tell otherwise?

 

[Sapient]

Might wanna put the resistors back to original configuration and see what the amp does.
If it appears to work normal you could then maybe try to figure out what the heck he
was trying to do.

"Normal" may be a good default start state.

 

[Pete Farrington]

I measured them and they read fine. How can I tell otherwise?

That’s all you can do really.
The amp may be oscillating; does it behave the same if the phase splitter valve is removed temporarily?

 

[tschrama]

or temporally desolder the PI coupling caps, C32,C34. That way you break the loopgain, and also break current leak if any of these caps is leaking...

 

[Spanngitter]

I would rebuild the Bias Circuit with high quality metal oxide resistors to exclude a possible issue coming from the modifications.
It would not be the first time that resistors with a weak temperature stability / negative temp coefficient cause Bias to run away, Marshall has a history there...

 

[Silverburst]

Might wanna put the resistors back to original configuration and see what the amp does.
If it appears to work normal you could then maybe try to figure out what the heck he
was trying to do.

"Normal" may be a good default start state.

I would rebuild the Bias Circuit with high quality metal oxide resistors to exclude a possible issue coming from the modifications.
It would not be the first time that resistors with a weak temperature stability / negative temp coefficient cause Bias to run away, Marshall has a history there...

Might be a good idea yes. But do my measurements actually indicate that these resistors would be off?

 

[Spanngitter]

You would need to measure them and heat them up during measurement to see any change. Since this would require removal from circuit it’s faster to rebuild with known good hardware…

 

[Sapient]

Might be a good idea yes. But do my measurements actually indicate that these resistors would be off?

That's what I'm saying ...put things back and see if "normalcy" develops. Reducing variables is a wonderful thing.

 

[neikeel]

If you have the correct grid voltages (and your measurements sound like they are) then you have to wait until it goes funky and remeasure. I would be surprised if the pi caps were leaky (given normal grid voltages). It could be oscillating (presume you were checking bias cathode voltages at idle with no signal and the screens resistors are not shorted.
I agree with the others rebuild with stock values and new bias caps and new screen resistors. I would not be worried as to type as long as not cc and all at least 1w modern (5w screens).

 

[PentodeLicious]

I'm biased towards rebuilding the bias circuit.
I know my jokes become more lame as i get older.
It will happen to you too!!!

Also hum and sequels are usually in the preamp department YMMV

 

[Silverburst]

btw, is it normal both filter caps measure differently? ->

• Filter cap 1: leg 1 328v and leg 2 378v Filter cap 2: 455v both legs

 

[tschrama]

Please specify “filter cap 1” etc..
Or put your Vdc values in the schematic and post it in your message. Much easier for everyone than guessing what you mean by “filter cap 1”

 

[Silverburst]

Please specify “filter cap 1” etc..
Or put your Vdc values in the schematic and post it in your message. Much easier for everyone than guessing what you mean by “filter cap 1”

Does this indicate cap 1 is fruitcake? visually no leaking nor bulging. The amp has some base hum, but no sure this is related to the tube that was failing or maybe this cap.

 

[Silverburst]

That’s all you can do really.
The amp may be oscillating; does it behave the same if the phase splitter valve is removed temporarily?

Yes, with the PI removed, it also rises up to 300-400 mV but slower (takes a about 5-7 secs whereas with the PI inserted it jump to these high voltages in 1 sec).

I also probed the 2nd tube socket and it has the same behavior.
I also probed the 2nd tube and it has likewise too high voltage, but only +/-220mV, not as much as the 1st tube which goes above 300mV.
To be sure the probe with the 1ohm resistor was ok, I used the probe with the same DMM in my 1959 and there it reads nicely and stable 37mV.

 

[Silverburst]

I also measured these resistors (installed) with amp not running. (I wasn't sure this would give exact readings while the resistor is installed because I thought you don't really know if you are measuring anything else along. Can you actually measure resistors correctly while they are installed in the circuit? or do you always need one leg out of the circuit?)

• R51 is 56K original + added 15k. -> There the original only measures 41Kohm, so it makes sense they added the 15km in series. (too lazy to entirely replace with 1x 56k or that value not in stock at the time of repair.)
• R52 & R53 are 220K -> they both measure 75k ???
• R55 is 2x470k in parallel = 235k -> measures 240K

I'm about to order new resistors to replace these, but not sure it will resolve anything since pin 5 voltage on both sockets ranges from -36v to -44v, which seems quite correct, no?

 

[Silverburst]

Another thread I also found https://www.marshallforum.com/threads/jcm-800-bias-problem.51809/

mentiones:
The resistor in front of the rectifier diode was defective (1.2m instead of 220k), after replacing that everything looks fine. It's exactly located beside the defective rectifier diode of the HV power supply. I've put in a 1/2 Watt resistor, now. Is this sufficient or do I need a higher wattage?

-> is this R55 and D7 ? D7 measures .633 in correct direction and open in the other direction.

 

[Silverburst]

Hmm, while waiting for parts, these pictures of later IC versions make it even more confusing:
On some boards, R55 on the scheme is:

• stock there (1986)
• stock not there (1987)
• stock not there and clearly added later

One board even has c37 which I can't find on the scheme. (1987)