• We are looking to make improvements to the Classifieds! Help us determine what improvements we can make by filling out this classifieds survey. Your feedback is very appreciated and helpful!

    Take survey

JCM900 Help - fault

  • Thread starter fastdave
  • Start date
  • This site may earn a commission from merchant affiliate links like Ebay, Amazon, and others.

fastdave

Member
Joined
Jan 17, 2011
Messages
91
Reaction score
19
Location
Fife,Scotland
I became aware of a hum, more of a buzz, coming from the JCM900 4100, and being intent on playing, politely ignored it.
A few nights later, I could smell burning, and ignored this also, until I realized I was some way away from the amp, and it wasn’t the normal ‘valve heat I was smelling, but something more sinister. I checked the ‘fuse fail’ indicators on the back, to discover one fuse o/c. searched everywhere, and in my stock, found 2 x 500maT fuses – replaced that one and switched on, holding a mirror round the back (amp was against the wall, pedals on floor, cables trailing, etc., and to my horror, witnessed the plates of V4 glowing red and becoming white.
I switched off, cleared the area, and pulled out the chassis. A 1.5K perhaps 1W resistor, strung across 5 and 6 of the 5881 furthest from the mains switch was completely fried and o/c (open circuit – I don’t know the nomenclature for the American, if there is any).
I managed to source a 3W wirewound 1.5 k from my stock, suspecting the wattage was perhaps higher than 1/2W, duly replaced the resistor, and switched on, having replaced the valve with an old EL34 which I knew worked, this having been one of the ones I replaced with nice new matched quad 5881’s.
The amp started, but seemed to burp intermittently until it warmed up, and it played ok, but when I tapped the cabinet or valves (tubes), it appeared to be microphonic (this may have been the reverb coil, but I don’t think so), and the noise was not from dirty pots.
I put it back together, and played for 10 mins or so, (one gets involved, you know the feeling), and then checked the back with the mirror again.
This time the valve closest to the mains switch seemed to be getting redder and hotter.
Throughout this I was still getting a buzz – better described as a buzz, because it wasn’t smooth like 50 Hz ( British mains frequency ) through the speaker.
I haven’t checked any voltages thus far, since my valve theory is more than rusty, but I had set the bias when I got the new 5881’s and had played successfully every other night for the past year or so – always set on high output.
One transformer seemed different from the other, so it may have been replaced, and I am reading that Marshall stock trannies aren’t very good.
Any ideas? Can anybody give me a quick solution or speed diagnostic action?
Dave.
If I was near you, Mr Wilder, I would be on your doorstep!!
 

fastdave

Member
Joined
Jan 17, 2011
Messages
91
Reaction score
19
Location
Fife,Scotland
Sorry, mate, appreciate where you're going, but do you mean that the resistors should be within their specified limits, or if they are in the amp spec?
The answer to both questions any way, is yes, and the good news is that I went back last night, checked the solder joints on the power supply resistors and diodes to get rid of the buzz, then tried replacing V3 with a previous EL34, and the buzzing went - prognosis - valve duff.
Will carry out soak test and renew 5881's anyway.
Thank you.
DAve.
 

fastdave

Member
Joined
Jan 17, 2011
Messages
91
Reaction score
19
Location
Fife,Scotland
I took Marshall's advice on this one and set the bias voltage to -49V.
When I blew the first valve, it took out the grid 1.5K.
After blowing another, I took the Cathode currents by inserting a mA meter - (only recommended if you know what you're doing), to find the bias current up around 60 odd mA on all valves (tubes).
I discovered the amp had been modded from 5881's to EL34's! (which is what it came with - because they run ice cold, and they were a mix match of makes, and one was an e34L, I changed them)
I changed the voltage divider circuit to achieve -65V bias voltage - it is now running very happily at 37 mA - sounds great - cost me hundreds on tubes - always follow your hunch.
Measure the stuff accurately - and don't always believe the manufacturer's tech.:applause:
Dave.
 

littlewyan

Member
Joined
Jul 15, 2012
Messages
33
Reaction score
1
Location
Wiltshire
-65V sounds a bit high on the negative bias voltage, on my 6100LM head its set to -49V which is what Marshall recommends
 

saranden

Member
Joined
Feb 4, 2009
Messages
43
Reaction score
0
Location
UK
I bought a faulty JCM900 4100. It has EL34's. One tube was broken so I bought 4 new ones. I replaced the four 2K2 4w resistors with 1K 5w and also installed R35 which was an absent 100R 5w. The pico fuse was dead and cut! I replaced C15 with a 47n 1000v instead of 47n 250v. Seemed good to go. I plugged in to read my plate volt before installing new tubes. The volts were reading 598V. I thought this was a bit high as from nearly all of the forums I've read indicated that something near 470V is the norm. I worked out the bias to be just over 29. I put the new tubes, warmed them up, still got the same plate volt when they were installed. The tubes glow, I could only get 2.9mv reading max on the octal bias reading. When I turned the amp to low on the back, there was quite a bit of a hum and the tubes heated up rapidly starting to smell. I switched off. Put the amp back on high, and turned it back on. Tubes back to normal. I am getting no sound, and I have cleaned all of the sockets with cleaner and the jack sockets. Zero output.
Why am I getting too much voltage on the plate? Why do you think I am getting no output? Appreciative.
 

fastdave

Member
Joined
Jan 17, 2011
Messages
91
Reaction score
19
Location
Fife,Scotland
-65V sounds a bit high on the negative bias voltage, on my 6100LM head its set to -49V which is what Marshall recommends

Just got to see your reply there, mate - the problem was actually that - Marshall recommended setting the bias voltage at -49 - but remember the bias current is the product of the voltage through the resistance, that is to say V/I, but if the voltage is too low, one cannot generate enough current, so I actually had to modify the voltage divider circuit to get 37mA. remember -49V is higher than -60V.

Quote Jon Wylder-
Assuming around a 425-450V plate voltage you'll wanna see roughly about 40mA on pin 8 of each socket. Some of these amps have been known to be biased cold from the factory. People have also installed EL34s in the 4100s that are set up for 5881s and EL34s will run ice cold if you install them in an amp set up for 5881s without modding the bias supply.
It's actually running beautifully.
Cheers, Dave.
 

saranden

Member
Joined
Feb 4, 2009
Messages
43
Reaction score
0
Location
UK
Hi all. I hope you don't mind if I jump in here to carry on with this thread.
I bought a Marshall JCM900 4100 for repair. The guy I bought it off, told me that it just faded out.. I thought great.. New tubes and I'm away!
One of the pins was missing off one of the EL34A's so I bought four new tubes.
HOWEVER.....The pico fuse had been snipped on one end off of the pcb, I replaced the poco as it was open circuit, I replaced R31-R434 1K 5W resistors, and R35 had been removed, so I replaced it with 100R 5w all wire wound.

I then turned the amp on let it warm up then measured the plate voltage.
It was 598V. After reading loads of threads about plate voltage, I assumed it would be about 460V - 480V. I decided to install the new tubes and measure the plate voltage again also assuming it would drop a little.. It didn't.

When I read the bias measurement, I could not get it above 2.9mv when according to the plate voltage my bias should be around 29.2mv. I expected to start setting at above 37mv and a 470V plate voltage.

Anyway...there's my problem, I am reading nigh on 600V from the transformer and at the plate voltage, also still no audio output! Could anyone help me with this repair. Should I consider that the transformer or caps are faulty?
 
Last edited:

Jonathan Wilder

Well-Known Member
Joined
Oct 12, 2009
Messages
6,163
Reaction score
747
Location
Greenville, SC
Just got to see your reply there, mate - the problem was actually that - Marshall recommended setting the bias voltage at -49 - but remember the bias current is the product of the voltage through the resistance, that is to say V/I, but if the voltage is too low, one cannot generate enough current, so I actually had to modify the voltage divider circuit to get 37mA. remember -49V is higher than -60V.

Actually, voltage does not flow. Voltage is a measurement of static potential (i.e. charge difference between two points) and is the product of current flowing through a resistance.

-49V is "less negative" than -60V.

Resistance to current flow generates voltage...just as resistance to water flow generates pressure.

Just thought I'd throw that one out. ;)
 

Ken

Well-Known Member
Joined
Feb 24, 2009
Messages
9,944
Reaction score
3,172
Location
Arlington Heights, IL
Actually, voltage does not flow. Voltage is a measurement of static potential (i.e. charge difference between two points) and is the product of current flowing through a resistance.

-49V is "less negative" than -60V.

Resistance to current flow generates voltage...just as resistance to water flow generates pressure.

Just thought I'd throw that one out. ;)

Jon! We miss you!!!!!!!!!!!!

Ken
 

fastdave

Member
Joined
Jan 17, 2011
Messages
91
Reaction score
19
Location
Fife,Scotland
uhh!
Just found your reply Jon.
I didn't state that voltage flows.
Is less negative not 'higher' with respect to earth, that is - more positive?
To say that the bias current was the product, I was using the mathematical term 'product' as the produce of the equation V/I as you well know.
Dinnae cross swords wi' me big man!
 

fastdave

Member
Joined
Jan 17, 2011
Messages
91
Reaction score
19
Location
Fife,Scotland
LBK,
Appreciate your point, and it is certainly a possibility that there could be a leakage somewhere, but all the other voltages are nominal.
One thing intrigues me though, being a verbose geek, interested in nomenclature and etymology, how did you mean 'modality'? - I'm getting - 'of mode or form not of substance' - 'used to express the mood of another verb' or 'style of classical music.
Can I infer that this is an americanism, or is it very correct, in having a suffix denoting quality or condition?
I'm not being smartassed - I am genuinely interested, for I would have used the word 'tendency' in this instance - so increase my vocabulary if you will.
Dave.
 

Latest posts



Top