Marshall bias problem

[Johns62312]

I am looking at 1972 50 watt marshall that the bias goes up to 144 milliamps with no adjustment to it. With the tubes pulled it only gets -23 volts. I have checked every part in the bias circuit and everything is good. Also disconnected the splitter resistors feeding the power tubes. Added a 220K accross the existing 220K and it made no difference. Did the same with the 47K with no results. From what I have read it should range in the -34 to -40 volt range. Has anybody ever run into  a problem like this? With out the power tubes in it the B+ is 470 volts. I can't find any reason for it to act like this. The two 10uf caps check good with a meter and a esr meter so I assume they are good. The 47K, 15K, and the 220K all check good, and the 25K pot checks good all thru it's sweep. Any suggestions appreciated.

[Johns62312]

I guess I should add that the guy that brought it to me said he got it out and it quit on him. He went to guitar center and bought new tubes (I know it's a no no). He installed the tubes with no adjustment and it worked for a minute or two before it burned out a tube. It's blowing the mains fuse when it does this and not the ht fuse.

[plexi50]

With the power tubes pulled are you getting 22MVDC on pin 5 of each tube socket? Does this change when you adjust the bias trimmer pot? Is there a bias trimmer pot?

[Colas LeGrippa]

output tubes good ?

[Johns62312]

I am assuming that you mean volts instead of mv. Yes it gets -24.5 bias pot turned one way to -22.2 going the other way. This is with the tubes pulled. I tried known good tubes in it and the bias just keeps on going to .144. I don't know about this year of marshall but the latter ones all show -32 to -40 on bias voltage.

[Johns62312]

I just tried the tubes again and it goes from 114mv one way to 88mv with the pot turned the other way and it looks like it slowly rises. Afraid to leave it turned on to long. I'm assuming you can use the bias probe with 1 ohm resistor. I've never had a problem using it in anything else. It still checks good.

[sluckey]

Ther problem is most likely in the bias supply. Not too much to go wrong. Check everything including the AC voltage feeding the bias rectifier. Does this amp use a bridge recto or conventional rectifier? What's the exact model? Got a schematic?

[Johns62312]

I don't know what model number it is, there is no letters or numbers on it. Date code inside in 1972, turret board construction. I've checked everything in the bias supply, has 200 volts ac feeding it through the 220K resistor. I figured this would be simple when it was dropped off. I can't see anything obvious that I have missed.

[phsyconoodler]

Well to start with,you need about -35v or more to get close before you put tubes in.Locate the resistor that is in the bias supply that is 47k or close to that.Remove it and install a 70k.That should get you in the ballpark for negative voltage.Measure the bias voltage BEFORE you install the power tubes.
  If you get a steady -35 to -45v then put the tube back in and see what the current is.Make sure you measure the plate voltage again too.

[Johns62312]

Didn't have a 70k so I put a 33k in series with the 47k. Voltage was -27.24 the bias pot one way and -26.01 all the way the other direction. Right off the diode is -34.2. This of course is without the power tubes installed. I don't see anyway to get it higher. Disconnected the wire going to the splitter resistors to grids and it still don't come up. Bias pot checks good all thru is's sweep.

[sluckey]

has 200 volts ac feeding it through the 220K resistor.

Put the other resistors back to stock and replace the 220K with a 100K and see if the voltage increases.

I'm curious about the 200VAC input. If this is a conventional rectifier then 200vac would be low for one of the HT leads to center tap. If this is a bridge rectifier then there would have to be a cap connected between the HT lead and the diode. Look at page 5 of this pdf.

http://home.comcast.net/~seluckey/amps/misc/Amp_Scrapbook.pdf

Does your bias circuit look like the "Typical Marshall" (top right)? If so, your PT may be bad. Check the ac voltage on both HT leads. I'd expect 300 to 350VAC on both leads, 200VAC.

[Johns62312]

Yes it's identical to the diagram. 200v is between the 220k resistor and diode. It has a bridge rectifier but it don't use neg. connection. Each leg has 346 volts on it. I tried a 220k resistor in pallalel with the existing one and that only made a few volts difference.

[phsyconoodler]

Replace all the bias supply components including the diode and see what happens.
   You have something amiss and it may be one of those caps or the diode.You should be able to get more than -35v.
It would not surprise me if one of the caps is bad.

[Johns62312]

I tried a different diode, that made no difference. Lifted the caps one at a time and checked them with a capacitor meter and esr meter and both passed. Tried it when each one was pulled up and the voltage still didn't came back up. I will have to do a search and see if I have any caps I can try in there place. Thanks for all the suggestions everybody.

[Johns62312]

Does anybody know by chance if that 200v between the 220k and diode sound right on these?

[phsyconoodler]

What about the bias pot? You should have no more than 50-60v from the bias tap.Maybe the problem is BEFORE it gets to the bias supply?

[Johns62312]

I pulled up the 47K resistor and the bias pot check good all thru it's sweep and the voltage still did not come up.

[Johns62312]

I pulled the 15k up from the diode leaving the 10uf hooked up and it does not go up but a couple volts. Man, this is a puzzle.

[phsyconoodler]

Follow the bias tap to it's origin.Does it come from one side of the rectifier? Maybe one of the rectifier diodes is shot?
 On a 50 watt the bias is from one side of the rectifier and the 100 watt has a 100v bias tap.
There is also a couple of coupling caps in that circuit connected to the two 220k resistors.If one is bad you will get wonky stuff going on.
 

[Johns62312]

The B+ runs at 470 volts without the tubes in, ac measures the same on both sides of the rectifier. The bias connection comes off one side of the standby switch, I moved it to the other side, that did not make any difference. It's a double pole single throw switch that also goes straight to both sides of the bridge rectifier.

[sluckey]

Lifted the caps one at a time and checked them with a capacitor meter and esr meter and both passed. Tried it when each one was pulled up and the voltage still didn't came back up.

Substitution is the best check on those caps. Just replace them.

You say you have 200VAC on one side of the diode but only about -34v on the other side.... It's almost got to be the diode or caps (or a wiring problem). Since you've already replaced the diode, now replace the caps.

[Johns62312]

Disconnected the ac going to the bridge-no change in the bias voltage.

[Johns62312]

Many thanks to Sluckey and Phsyconoodler for you help. I pulled both 10uf up on one end and tacked in a 22uf and the voltage went up -41. I guess I have to learn not to trust meters that much. In the past I never relied on one metet to do it, always used a capacitance meter and then checked with an esr meter. Never had a case where they checked good on both meters and were bad. You guys are the greatest, many thanks. I guess this guy just wants it working to sell it. It needs a recap job for sure. This is one of the old ones with a polarity switch on the front.

[PRR]

> a capacitance meter and then checked with an esr meter

Such meters usually test with a very low voltage, 3V or so.

A cap may have correct uFd and ESR at 3V, yet break-down at 23V even though it says 50V on the label.

That's a very common failure mechanism in some older e-caps.

In the MAIN power supply, a "450V" cap decayed to 300V break-down will reveal itself by sucking HUGE current, overheating, and bursting.

In bias supplies like these, the maximum current available is small. Perhaps 1mA. If all of this is dissipated in a 23V breakdown, that is 0.023 Watts. Any cap big enough to hold can dissipate this 0.023W forever.

The "Typical Marshall" circuit has two caps. You can remove either one with only "slight" change of output voltage. (Removing both will give wonky readings.) If snipping one end of one cap gives a "big" change, such as 23V to 45V, then that cap is very dubious.

[Johns62312]

I guess both of these are bad cause I lifted them one at a time and it really didn't make any difference. Thanks for the info.