JCM900 conversion with 550vDC on the B+. how to reduce WITHOUT CT

[Curranm]

I am doing a JCM900 4100 100 watt conversion for a friend of mine. the 900's where god awful with IC chip clipping and LED clipping. the tubes where basically for buffering between IC ships(evil of the highest order). so I have cobbled together a merlin preamp design, with a standard marshall PI and output coupling section. The bias circuit is a merlin design due to the PT not having a CT. I am getting 550vDC off of the bridge rectifier (1000V/3A) going into the 1st filter cap(F+T 100uF-100uF 500V -10%+30%). After the 1st stage filter I have a 2K5/5watt resistor(*choke*) going to the board and to the screen grids(1K5/5watt). my plate to cathode is 547vDC. my control grid voltage is set to -50vDC, and I have 1R/5watt resistors between cathode and ground of the power tubes (JJ E34L's). I would normally reduce the B+ with zener's between the HT CT and ground BUT this is not possible since the PT HAS no CT. so how do I reduce this B+ after the rectifier but before I hit the 1st B+ filter? would a resistor (5 watt) be 'ok'? OR could I zener off of the negative lug of the rectifier? and if so what type of zerner would be appropriate? well thanks for reading and thanks in advance for nay help you are able or willing to offer! Love and Hugs!!!

[Curranm]

So I am doing some VVR, and VCB research for this marshall JCM900 4100 with the 550vDC at the rectifier B+. it looks like most of the guys doing VVR are using them on 18-30 watt amps. makes me think that the 100 watt marshall is not really suited for something like that. I have found 600V mosfets on mouser. BUT working out a VVR with a 1M pot in there really looks and feels like a NO NO.........Any feedback is good

[kagliostro]

Most of the 100W amps are fixed bias and VVR is more complicated for fixed bias amps because they require that also the bias is changed according with B+

However VVR for fixed bias exist and perform well if well implemented

--

If you want only to tame B+ at a fixed level, you don't require a "special" VVR, a simple VVR for cathode bias circuit is what you need and you can use a pair of resistor (voltage divider) instead of a pot to "set & discard" the level of B+

--

Merlin published in one of his books a VVR circuit (also for fixed bias amps) that is planned to solve the problem of high voltage on the pot, but a friend tried it without good results because it requires a low impedance raw negative voltage source (a small separated transformer may solve the problem)

--

At London Power you can find explanations about VVR (they call it Power Scaler, don't confuse it with Super Scaler that is one other thing) and they sell also Power Scaler kit and low impedance bias power supply kit

Franco

[PRR]

550V on a 500V cap is NOT a happy situation. Keep your face away and the windows open.

550V sounds just right for a 100W Marshall. You are only tampering the preamp? Then why question Marshall's wisdom?

Links to schematics would be far more helpful than long text descriptions. Quicker too.

[92Volts]

You can put a Zener in series with the HV supply, after the positive end of the recto for example.

I'm not sure why it's often on the CT/ground side. Maybe: when bolting SS components to heatsinks, the shell/case is often electrically connected to part of the component. Cooling is safer and easier near ground. This may or may not apply to any particular diode...

Regarding power, let's say the amp pulls 400ma at full volume. You need to drop 50v, that's 20W.

That would be a HUGE zener and likely need a proper heatsink to reach its potential. Or you could use a string of several zeners.

A resistor is cheap, but it will drop more voltage when you least need it to drop voltage-- at high power output when the supply likely sags anyways.

On that note... are these voltages with or without tubes? Also, you can use caps in series (with balance/bleed resistors) for higher voltages and I thought that would have been included in this Marshall's original design.

[alerich]

On that note... are these voltages with or without tubes? Also, you can use caps in series (with balance/bleed resistors) for higher voltages and I thought that would have been included in this Marshall's original design.

The schematics I have seen have 2x50uf 500VDC caps in parallel for the first filter node for the B+. 550VDC on the plates with the tubes in is gonna be a problem. I never saw a JCM900 preamp schematic until today. Yikes, that's a lot of sand in the preamp. You have to hunt to actually find the preamp tubes on the prints.

[92Volts]

The schematics I have seen have 2x50uf 500VDC caps in parallel for the first filter node for the B+. 550VDC on the plates with the tubes in is gonna be a problem. I never saw a JCM900 preamp schematic until today. Yikes, that's a lot of sand in the preamp. You have to hunt to actually find the preamp tubes on the prints.

You're right, and that's unfortunate!

I googled JCM900 plate voltage. I don't think it's consistent over the years or between EL34/5881 models. At least one JCM900 owner reports unloaded voltage over 500v... but not this high.


Curranm, can you check your wall voltage and/or check accuracy of your measuring equipment? These amps are designed for 117v mains. If it was designed for 500v at 117v input (not smart, but possible) you'd see ~550v with ~128v from the wall. Yours "shouldn't" be that high, but it's not safe to assume all houses get 120v exactly. Always worth checking, and if you find yours is crazy-high you can usually have the power company address that since the official range is narrow.

[Bantam]

When we had a store, a lot of those amps came in for trade (broke) and repair. Don't think it's a very robust design. I wasn't working/building on amps at the time, so can't say exactly what, or if any percentage of similar probs though. Guessing power supply caps....or Satan!?