JCM900 4500 Extra 2.2k resistor soldered to pin 5 of V5 (EL34)

[milfordcubicle]

Hi, this is my first post on the forum, and probably not my last! I have a 1990 JCM900 4500 DR that we've owned since ~1994, purchased used. This model has to JMP50B power stage board and came stock with EL34s (no reason to think this was modded from 5881 given the confirmed year of the amp, screen resistor values and pins 1-8 jumper). I recently changed the tubes and went to adjust the bias, realizing it was between 7-8ma on each tube. I maxed out the bias pot and could get ~20ma best, so I swapped out the 56k bias range resistor (R30 on this board) to a 47k and that allowed me to adjust the bias to ~36ma on one tube and ~37ma on the other. Plate voltage on both is reading ~450 volts, so these are biased at a safe 65% or so.

Amp sounds great but I noticed that pin 5 on V5 (control grid) has a 1/4 watt 2.2k resistor soldered to the pin. That pin then runs to an orange wire to R27, a 1.5k resistor (the "swamp" resistor, I believe). V4 does not have any resistor on pin 5 and has a green lead running to R24, also a 1.5k resistor. The schematic properly identifies these 1.5k resistors on the board and the green and orange leads are correct, but one of the tubes has this extra 2.2k resistor in series with the 1.5k. The schematic does not show an additional resistor in series. The solder job doesn't look original on the V5 pin 5 2.2k resistor but I wouldn't actually know.

I'm still new to this and I have read that it can be beneficial to run up to 5.6k swamp resistors on pin 5 of EL34s, but I would expect the resistance on both control grids to be symmetrical (currently, it is 1.5k on pin 5, V4 and 3.7k [1.5k + 2.2k] on pin 5, V5). I have no reason to believe this is affecting performance, but I don't know why this extra resistor is there. Any suggestions?

 

[cozmacozmy]

Just opened up my 4500 JCM900 to check bias settings and I have the same resistor on that orange wire pin #5 of v5. I was wondering why it is not in V4?

Did you ever find out what it is for? I don't see it anywhere on the schematic. I have the same JMP50B power board.

 

[milfordcubicle]

It's still a mystery, but the voltages on pin 5 for both tubes are the same, so I have no idea. The fact yours has it too suggests Marshall did this from the factory, so there's that.

 

[cozmacozmy]

Yeah mine looks like the day it came off the assembly floor and doesn't look like anything has been molested internally. Even being over 30 years old I was surprised how clean it was inside there.

 

[milfordcubicle]

Yeah mine looks like the day it came off the assembly floor and doesn't look like anything has been molested internally. Even being over 30 years old I was surprised how clean it was inside there.

I thought the same thing. Super clean inside! I see yours is the same model from the same year; does yours also have a QC sticker inside signed by a "Denise" and/or "Karen"?

 

[cozmacozmy]

How did you go about measuring your bias?

Yes it does have the Karen signed QC sticker! Trying to upload a photo for the sticker but it says the file is too large.

 

[cozmacozmy]

The date code from the sticker on the side of the chassis says 26/6/90 for mine.

 

[milfordcubicle]

Next time my chassis is out I will take a look at the build date. Transformers are Drakes and have 3/90 and 5/90 on them, so probably assembled around the same time. My serial number indicates it is a 1990.

For bias I just use a single probe and DMM and adjust the hotter tube to ~70% dissipation, letting the other tube run a bit cold. I had issues with the bias pot range so had to change the 56k R30 (which you should also have) to a 47k. That worked well for my Ruby EL34s, but I recently swapped the KT77s in and had to drop the resistor value lower to get enough current. I had to piggyback a 220k on the 47k to get it down to 38.5k.

 

[cozmacozmy]

I have a plate current probe that goes between the tube and socket on order and my Fluke 88. But I've been reading up on how to go about checking these and seeing so many ways it made my head spin. So that's why I came lookin here and saw your post and remembered seeing the same extra resistor on the V5 of mine.

Mine has Sovtex EL34WXT's in them. Those are what was in there when I got this years ago but I have never checked the bias before. Just was checking for curiosity sake.

 

[milfordcubicle]

Check it for yourself but you might be surprised to see that they are biased super cold. I don't know why Marshall used a 56k bias pot resistor but if you have to increase the current and max out the pot before you hit your target, try swapping a 47k in R30. Just be careful when de-soldering because the contacts on those boards are very soft and you can easily chip away the surface. I found that avoiding solder wick helped and would just re-melt the existing solder and add a bit more when installing a new component. Even if you scratch off the conductive material, you can just use the leg of the component you are working with to solder on to the next component on the board, following the path.

 

[cozmacozmy]

I'll make sure to keep an eye on that, thanks!

 

[cozmacozmy]

Right after talking to you about that 2.2k resistor on pin 5 that you were asking about. I ended up having red plating on my power tubes in my 4500. I finally got the replacement in for the blown C15 safety cap in the bias circuit and fixed it a few hours ago.

https://www.marshallforum.com/threads/jcm900-4500-el34-red-plating-run-away-bias-current.122500/

 

[Spanngitter]

Normally in a JCM900/4500 there should be
1. Screen Grid Resistor 2.2k 4W (Ceramic block) on PCB (R33/34/35/36) with cable to Pin4
2. 1.5k GridStopper on PCB (R24/25/26/27) with cable to Pin 6 and an additional 2.2k from Pin 6 to 5 which makes it a total of 3.7k for the Grid Stoppers

the above redplating has nothing to do with the GridStoppers, it shows that bias supply failed and tube did run away but then normally the GridStopper blows and acts as a fuse to protect the valve from screen grid / core melt down.

 

[cozmacozmy]

Correct, yes I know the red plating wasn't the cause of the gridstoppers.

 

[cozmacozmy]

Revisiting this post once again on my 4500 900 DR.

After fixing the red plating issue due to the bias cap failing on me 2 years ago. The amp was sounding great and working really good... Especially after getting ahold of a GT-7025 for V1 that I have posted about before here. Until recently that is... I started getting this not quite a "squeal" noise... but more like the distortion getting out of control noise any time it was on & plugged in. The more gain you add (pedal or volume/gain), the worse it would get. But it never sounded like a squeal, just uncontrolled distortion? I first thought I had a microphonic pickup, but all my guitars do this now.

So, being all the caps in the power section except the failed bias control one I had replaced were original, I thought I would replace all the rest of them and see what this uncontrolled noise is afterwards (hoping the total cap job would fix it). I ordered all the power section caps, the solid state caps along with the 2 other of the bias caps that were still original all from amprepairparts.com.

I had completely forgot about this extra 2k2 1/4 watt resistor that is on pin 5 that the above OP also had on his model 4500 JCM 900. I have this 2k2 resistor going from the R27 1k5 resistor & orange wire to pin 5 but not on the R24 green wire going to pin 5 of the other El34. So while I was in here, I figured I would fix this. I do have the 2k2 4watt brick resistors at R33 & R36


So do I add a 2k2 1/4watt to the green wire pin 5 of the other EL34? When I went to rebias this 2 years ago, I ran out of adjustment but it got it pretty close to within 62% dissipation if I remember right so I left it alone for the time being. It sounded great so I wasn't worried about it.

The only wiring diagram I can find that somewhat matches what I have, is this 100watt diagram but take away the 2 extra tubes and it is the closest match I can find to mine. Why did Marshall leave this out? I measured all the resistors related to this here and they all are pretty spot on.
I guess I am more curious as to why it is missing on one of the EL34's? I also need to look at why the reverb isn't working anymore either, but that is very low priority.

 

[Rudy v]

Here is a schematic and its different.

 

[cozmacozmy]

Here is a schematic and its different.

Thank you Rudy, why are 2 of the tubes at 22k and the other 2 at 4k7?

The one I linked is still the closest (other than being a 100 watt and mine being a 50 watt) to what I have actual in my amp.

 

[booradley]

did this get resolved? I have a 100 watt JCM900 4100 EL34's with a JMP50B PCB and V4 and V5 have the 1/4 watt 2.2K's but not on V6 or V7. The schematic has the 2.2K's on all pin 5 power tubes. They are not the same 2.2K screen grid 4 watt resistors. Wondering why they are on 2 tubes and not all 4 like the schematic?

 

[cozmacozmy]

Yes, I replaced all the power caps and while in there I straightened out the missing resistors on the power tube in my 50watt.

Guitar CPR talks about this issue in this YouTube video of his and on the other JCM 900 videos he has on his channel. Seems they all have this problem.

 

[dtier]

Those tube mounted resistors may have been added to deal with a board artwork induced parasitic oscillation that occurred only on the one side of the power amp. Just guessing. Sometimes production time/cost call for minimal solutions. The You Tube authors solution of adding 5k6 grid stopper resistors at the tube bases and bypassing the board mounted stoppers is probably the best arrangement although not a necessary one in all cases.