Lo Plate Voltage: Opinions needed

[paully]

Working on an Ampeg V4 guitar amp's reverb circuit, and at power up I get
the right supply voltage, but after the plate resistors, the (recovery) 12AX7 plate reads way low, less than half the voltage at plate 1 of the 6CG7. It should be 3 volts more! The cathodes all have proper readings, it's that one plate. I used the voltages from the original Ampeg schemo. Wondering if it could be a misprint there. Otherwise, I'm guessing a bad 12AX7(?). But that correct cathode voltage has me scratching my noggin. Circuit diagram moved to post below. Opinions much appreciated.Thanks.

Best, Paul

[themagicmanmdt]

does it sound good, though?

[nclayton]

If the cathode voltage is right but the plate voltage is wrong, that means there's the correct current running through the cathode resistor, but more current running though the plate resistor. If the extra current running through the plate resistor isn't coming out the cathode, then it must be going somewhere else. Possibly the blocking cap going out to the reverb level pot is leaking some DC, and that's where the extra current is going. Check to see if there's any DC voltage on the reverb level pot. Another possibility is that the tube has a plate to grid leak and some current is flowing out the grid. Check the DC voltage on the grid.

Both of those 2 voltages (the grid and the reverb level pot) should be 0 volts, but probably you'll find one of them is positive, and that will tell you your problem.

Ned

[O_ellinas]

Hi

in old tube equipment, carbon resistors in plate circuits tend to go high as they age. That 270k resistor could be out of specs. Measure it with an ohmmeter and see if it is above its nominal value + tolerance (10 or 20% was common back then, look at the resistor, silver band=10%, no fourth band=20%). If it is, then change it.

Leaky coupling caps give horrible distortion, motorboating or no sound at all depending on the amount of leak.

[paully]

Thanks guys.

Major fopah. Same problem but different tube(6CG7). I was reading the wrong pins. The 12AX7 reads to spec. The low voltage is after the 120K plate resistor at pin 1 of the 6CG7. I'm also having a problem with the 10K/5W plate resistor. Slightly low voltage(re-calculated to match lower variac AC line voltage) and very hot, even at 1/2 line AC power with no signal input.

When I beta checked the rev module with a bench power supply and new tank, it sounded great and the voltages were near spec. After inserting into the amp(which is a stick-built hybrid borrowed from the V4's reverb design) using it's PS, that's when I noticed the problems. Haven't checked how it actually sounds in the amp cause I'm kinda afraid to bring the line voltage all the way up till I find the problem. BTW, all new caps(rated double the associated voltage) and resistors. Don't remember if I checked the grid for DC leakage, but I'll do it this afternoon. Beyond that, I'm stumped!

That 6CG7 is a real old tube with some mileage. Might just be time for an upgrade . Thanks again!

Best, Paul

[nclayton]

Hi. Well, if the resistors are the correct values, then it doesn't matter how old or bad the tube is, it can't just make the voltage go to low without bringing down the proportional amount of current. If the cathode voltage is the correct voltage and the plate voltage is wrong, the point still holds that the extra current in the plate resistor has to be going somewhere. The tube can't just magically evaportate that current, it has to run out either the cathode, the grid, or else it's not getting into the tube in the first place, ie, it's going through the plate resistor but then getting diverted through a blocking cap or something else attached to the plate.

But the question is, did you recheck your 6CG7 cathode voltage and verify that it's correct? Maybe for that tube the cathode voltage actually is too high, in which case you know the extra current's going all the way through the tube and down through the cathode resistor.

Usually old tubes pull too little current rather than too much, but it is possible the tube is leaking between elements.

Another possibility is that the quiescent operating point is actually setting up right, but your meter is reading a funny DC voltage because there might be a big AC signal voltage on the plate throwing it off. Maybe the new power supply and connections are making the tube go unstable and oscillate. If you have an oscilloscope you might want to check whether you see oscillation at the plate. But it would be a little unexpected for that tube to oscillate since it's not working at super high gain, has an unbypassed cathode resistor and isn't directly inside a feedback loop. Still, anything is possible (other than current disappearing).

If there's a 5 watt resistor it's supposed to be hot. The reason high wattage resistors are big is not so that they'll run cool (5 watts of electricity = 5 watts of heat which is always hot to the touch) but so that they can take the heat without catching fire or cracking. So it being hot isn't NECCESSARILY a sign of a fault.



Ned

[nclayton]

It looks like there's something wrong with your schematic by the way.

At the top of the picture there's a circled voltage that says 120. There's no way that's right. It should probably be more like 300-350 volts. If you've arranged for that voltage to actually be 120, then that explains why your plate voltage is too low.

The first 6CG7 has a 2.2K cathode resistor, and it says the cathode voltage should be 4.1 volts. That means that tube is expected to pull just under 2mA. Since the plate resistor is 120K, there should be a difference of about 220 volts between the plate and the B+ voltage. But the schematic shows the B+ as 120, the plate as 112, a difference of 8 volts. Same problem with the 12AX7 that's looks like it's developing 5 volts across a 270K resistor.

Ned

[paully]

Thanks Ned. Couple things to explain.

I checked the PS rails with a scope, and there's no ripple at all in the lower stages. No ripple or anything strange at the plate in question either.

The 5w resistor is unusually hot at quiescent state(no signal applied), and about 5% below spec at all line AC steps. I swapped it out for an 8.5K 10 watter, and the votage is dead on. Still gets hot even with that resistor. I can live with that.

The cathode(pin 3) on the low plate side of the 6CG7 is also low, but in step with the low plate voltage. I disconnected the grid and got no change. Checked the blocking cap at the plate. No DC leakage there.

You're right to question the 120 VDC on the schemo. The DC for those plates in a V4 started at a 354 VDC rail, but there's a lot of other things sharing that rail in the V4. This amp's power supply design is from an old Magnatone MP5 amp. I used that supply because the preamp, driver/inverter and output stages are from the same amp, and I had the original transformers. Choice of DC rails for alien circuit designs(V4 reverb and Fender Trem) is a matter of trial and error and resistor matching. The 120 is derived from the lowest rail in the supply(285 VVDC), which has been stepped down to where the plate voltages and related resistors are correct(or should be) as per the V4 rev schematic. Had to do the same at the pre and the driver sections to get them on target. No problems there. They're working fine.

I just gave it a full line power test with signal, and everything is actually working, including the reverb. I've been building these things for years, but this one has me stumped. I'm seriously thinking it may be a misprint in the V4 schemo. AAR, I've got a replacement 6CA7 on the way. Should tell a lot.

Thanks again for the suggestions. If you spot anything else, please holler!

Best, Paul

[nclayton]

The cathode(pin 3) on the low plate side of the 6CG7 is also low, but in step with the low plate voltage.

If the plate voltage is reading too low AND the cathode voltage is also reading too low it can only mean one thing: your B+ voltage is DEFINITELY too low. If your B+ voltage was right and the plate voltage was too low, then you'd actually find the cathode voltage was proportionally too high, not too low as you've stated. They're inverse...when the plate voltage goes down the cathode voltage goes up and vice versa.

Since you know the values of the plate and cathode resistors and you know the voltages you're shooting for on the plate and cathode pins (112 volts and 4.1 volts) it's totally simple to figure out what the B+ voltage should be. In fact, assuming the tube is running with no grid current, you can calculate it exactly (assuming the resistances are exact). In this case it should be exactly 335 volts. Of course, if you make the B+ 335 volts, you might not get those exact plate and cathode voltages, but the ONLY B+ voltage that will make that exact pair of plate and cathode voltages both simultaneously possible according the laws of physics is 335 volts.

I doubt if the plate and cathode voltages on the schematic are typos. You'd expect a 6CG7 and 12AX7 in a tube amp to run on a B+ of something between 250 and 350 volts, and the numbers (other than the 120) on your schematic jibe with that. You definitely wouldn't expect a 6CG7 and 12AX7 to run off of a 120 volt B+. It's possible to run them that low, but they won't run optimally at that low of voltage.

If you do the same calculation as above using the 12ax7 voltages and resistances you get a B+ of 320 volts. Didn't come out exactly the same...probably cause the voltages on the original schematic were measured with an analog low res. meter rather than calculated. Still...it bears out the point: you won't get the correct plate and cathode voltages unless your B+ is a good bit over 300 volts.

I hope this is helpful. It really sounds to me like you should be working on the power supply to try to get the B+ up to where it ought to be.

Ned

[paully]

Thanks Ned. Got the theory, but not te application. BTW, disregard that 120 supply voltage. That was before I started testing different supply legs. I think it's around 190 or better as I write, which did get the 12AX7 to the proper plate/cathode readings.

Now the question is, how do I take 335 VDC(I have a 340 VDC option on the PS) and get it to hit the right plate voltages on both plates without changing the plate resistors or messing with the cathode side? How would you do it? Hope I'm not being a pest, but I'm a little lost here.

Best, Paul

[nclayton]

I don't mind replying, I just hope I'm being helpful, not strictly lame.

Probably the 12AX7's plate voltage measures about right cause it's pretty hard to get a 12AX7 plate to go down below a particular voltage. It will tend to run dry on current rather than go much below 100 volts. Probably that 12AX7 is running at a current way lower than the .76 mA it's supposed to.

190 volts is better than 120, but of course still lower than would be ideal, and also I'm sure it's a lot lower than the other preamp tubes would like to have (I think you said that point used to be 285 volts?)

What's the highest voltage in your amp? Does the power supply have a choke?

Probably the easiest way to solve the problem for the new tubes as well as the other tubes in the amp is to break off a new power supply chain for the 6CG7 and reverb 12AX7 immediately after your amp's choke (if it has one). This would mean you'd have to add 2 power supply caps and 2 dropping resistors, but if you can measure the voltage after the choke, then figuring the values of the resistors to use will be pretty straightforward. Also, the current that the reverb tubes draw won't mess with the other preamp voltage hardly at all.

The other option is more complicated. Basically you'll have to temporarily cut the new reverb tubes' power supply connections then go through the amp with just the original tubes running, measuring all the supply voltages from the power tube screen supply, to the phase inverter, to the other preamp tubes, all down the line. Make note of all the voltages, but especially the two that are closest to 400 volts and something like 320-335 volts (the 340 volt point would probably work, but if you've got a point like 320 or so it might be better). I guess you probably don't actually have to do this first step since it sounds like you already know the voltage available in your amp prior to adding the reverb.

Now go back in and attach the reverb tube plate resistors to those points. Once you turn the amp on, you'll have to measure the supply voltages again. Obviously they'll have dropped a lot because of the extra current. Now you can start reducing the values of the appropriate power supply dropping resistors until the voltages come back up to their original values. Basically you'll have to adjust the resistor feeding in to the 400 volt-ish point first...once you've got it to its original value you can adjust the resistor feeding into the 320-340 point. When you get those two voltages right PROBABLY all the other voltages in the amp will be pretty close and you won't have to mess with anything else.

On the other hand if you play the amp for a while and it sounds good to you, maybe there's no reason to do anything. The preamp voltages might be unconventionally low, but if it sounds good, it might not cause any problems at all.

Ned

[paully]

Hey Ned,

That's what I was thinking.. separate the 2 tubes from the common 190(?) rail, and go from there. That's what I was gonna do tonight. Yes, I added a huge choke(from an old Leslie 122) after the main b+ leg. I really appreciate the input.

Best, Paul

[mu_amps]

The first 22k resistor in the PSU looks like it should be 2.2k.

The screen supply is also omited from the PSU. The schematic suggests that the screens are not in series with the rest of the rails in your supply. If that's the case I would recommend arranging all resistive PSU elements in series.

[paully]

Thanks for the reply and suggestions, Mu.

The first 22k resistor in the PSU looks like it should be 2.2k...

>>>>>>>>>>>>>>>>>
Except for the addition of a choke and rearranging the negative voltage supply, I followed the original PS circuit. Don't really know which 22K you're referring to. They're all 22K as they should be. The rail voltages reflect a fully loaded amp with all circuits connected as per the original schemo. I've deviated by adding a couple different circuits(reverb and trem) and eliminating a few originals, so those PS voltages could be way off from originals at 117 VAC.

The screen supply is also omited from the PSU. The schematic suggests that the screens are not in series with the rest of the rails in your supply. If that's the case I would recommend arranging all resistive PSU elements in series.

>>>>>>>>>>>>>>>>>
Again I'm using the original MP5 schemo, and the only screens are in the the EL34s. They're fed from the B+ rail(400 VDC), and stepped down with a resistor. Not exactly sure what you mean. Except for the -30 tap, all other PS rails are in series as far as I can tell.

Best, Paul

[nclayton]

Yeah, I think that 22K resistor is what it should be, it's just that the original magantone amp dropped the preamp voltages down quite a bit lower than the most common tube amps like fenders and marshalls.

I haven't sent any replies, because I figure if the amp is sounding good, you probably don't really need to fuss over the exact voltages. It's probably enough just to know that the tubes you were worried about are running low just because they're starting off with a low voltage. So you don't need to worry that there's something really wrong like a bad tube or cap or something like that.

If you do still want to get the voltages to the level originally spec'ed by ampeg you could do it, but you'll also have to decide whether you want the OTHER voltages to be the same as in the original magnatone amp, which I believe is not the case right now.

I think you might have already tried supplying the 6CG7 and 12AX7 from the magnatone's original 340 node, is that right? If so, since your preamp draws less current than the magna's, I think that is probably a good bit more than 340 volts, right? But it should be OK.

The one other thing I'd suggest is just supplying the plates and screens before the choke and then supplying all the preamp stages after the choke. You could do this using only the caps and resistors you already have in place like this:




The voltages written there are what I would about guess you might actually measure if you did it this way (and assuming you really have about 400 volts to begin with).