B+ Plummets When Power Tubes Warm Up

[The New Steve H]

I have discovered why the volume was so low, but it would be too embarrassing to go into further detail.

The amp is now at a B+ of about 292. It sags to around 280 when the amp is played hard. The screens are a couple of volts lower. The cathodes are at 28.5, which is 5 volts higher than expected.

My only complaints are that it seems to break up a lot earlier than I expected, and it's still bearable to play it full blast while standing near the speaker. The clean headroom is so low, I'm not sure if I'll be able to get anywhere with pedals.

The amp seems to do better with a 12AX7 in V1. Usually I like a 12AY7.

[diagrammatiks]

at least nothing blew up!

you should have about 8-10 watts of volume.

the breakup volume shouldn't be too loud.

actually what value of cathode resistor are you using?

[The New Steve H]

It's either 420 or 430.

The plate voltages on V1-V3 (and the voltage on the cathode of the cathode follower) were kind of low, so I bumped them up. I changed some resistors and removed the resistor between the last two filter caps to jack the voltage a little bit on V1 & V2. Seems to have worked, but now I have an oscillation at high volume. With all these wires hanging out, I don't see any point in chasing it down until the amp is cleaned up.

I'm going to fool with it tomorrow.

Why would I get 8-10 watts when Da Geezer got 7 for half as many tubes?

[The New Steve H]

When I say it's a mess, I'm not kidding. I could build a new amp with the wire I'm going to take out of this one.

[diagrammatiks]

It's either 420 or 430.

The plate voltages on V1-V3 (and the voltage on the cathode of the cathode follower) were kind of low, so I bumped them up. I changed some resistors and removed the resistor between the last two filter caps to jack the voltage a little bit on V1 & V2. Seems to have worked, but now I have an oscillation at high volume. With all these wires hanging out, I don't see any point in chasing it down until the amp is cleaned up.

I'm going to fool with it tomorrow.

Why would I get 8-10 watts when Da Geezer got 7 for half as many tubes?

I still think you'd be current limited by the pt. but maybe not depending on how you were biased.

at any rate you need to half the value of the resistor if you are sharing it with 4 tubes rather then 2.

so you gotta use 220 or so. If it's a single resistor for the 4 tubes you'll need like a 10-15 watt resistor.

[The New Steve H]

I originally bought 2 resistors (one per cathode pair), but when I looked at the circuit, I kept coming to different conclusions as to whether I needed both. When the bias voltage originally settled right where it was supposed to be, I decided not to install the second resistor. Now it has increased.

The math says the resistor in the Little Wing puts out something like 1.2 watts of heat. I have huge 10W resistors, but I am hoping I can replace them with a couple of paralleled 2W jobs, to save space.

I also have some resistors missing on the grids. The design calls for one 220K per grid, but they were omitted from the PPIMV, and when I was trying to get the amp running, and the PPIMV had to be removed, I jury-rigged the grids with 2 resistors. I wasn't looking for perfection; just a functioning amp. I have to fix that. I guess I'm going to have to order a new pot, since the ones I got were no good.

I suppose I should fix everything that can be fixed, before considering a new PT. I may have a spare 250K pot somewhere. If so, I can put it in the treble spot and use two identical pots for the PPIMV.

I'm amazed how little this thing hums. It's all over the place, and one OT is right next to the PT, with the plates in the same plane.

[diagrammatiks]

well you actually have two choices

you can use 1 resistor for all 4 tubes and use 210-220 at 10watts

or you can use 1 resistor per tube at 860-900 and have them be like 2-3 watts.

you'll need one capacitor per resistor as well in that case.

you should get the grid leak resistors, it's bad to run your power tubes without a ground reference.

people make a bigger issue out of pt placement then it needs to be.

there's no problem with the ot being right now to the pt.

[Firestorm]

I changed some resistors and removed the resistor between the last two filter caps to jack the voltage a little bit on V1 & V2. Seems to have worked, but now I have an oscillation at high volume.

That means you've paralleled those last two caps and now you're running two preamp stages and your phase inverter off the same supply node. That should oscillate.

[The New Steve H]

Here's something weird. It only oscillates on one channel.

Is there any way I can fix it without losing the higher voltage?

Interesting side note: the Classictone people say their Trainwreck-style PT is capable of 300 mA.

[Firestorm]

I don't think you should worry so much about your preamp voltages. How much did you pick up by removing the resistor anyway? How big was the resistor? The dropping resistors don't just drop voltages, they also decouple the different stages from each other. You could use a smaller value resistor instead of taking it out.

But why are you so concerned about preamp voltages (unless they are hugely off)? There are two voltages of interest on a common cathode preamp tube: the plate voltage and the supply voltage. The supply voltage sets the peak value of the positive signal swing through the stage (it doesn't actually get to that level, it's sort of a "limit"). The plate voltage is controlled by the plate resistor and the cathode resistor and sets the "midpoint" of the signal swing. You can move it higher or lower by manipulating the cathode resistor and/or the plate resistor. But in the V1 stage, you are almost certainly not clipping (nor even compressing the signal) so you won't likely hear any difference unless you make the voltage dramatically different. Cathode follower stages are more dependent on voltage in order to make them "misbehave" in the good way that they do, but if your voltages are significantly lower from those in a 5F6A for example, you can still make the stage compress positive peaks by making the load resistor (on the cathode) smaller.

Whether lack of decoupling alone is causing the oscillation is hard to says. You need to get rid of the extraneous wire and get the lead dress as good as it can be before you start manipulating component values. Otherwise you won't know why particular effects are happening.

[The New Steve H]

I was getting voltages that were something like 15% off. I thought that was a little excessive. And increasing them did make the amp work better. I no longer remember the figures, but the cathode follower was the thing that alarmed me the most. I probably should have recorded the numbers; all I remember is that they disturbed me.

I think wire dress is the entire explanation for the oscillation. If it oscillated on both channels, I'd think differently. It amazes me that I can hear anything BUT noise from that mess.

Anyway, I have a day's worth of work to do, before I go any further.

[The New Steve H]

I got a lot of it cleaned up. The oscillation is gone. The amp sounds wonderful at low volume, but when you crank it, the bass notes start to sound a little bit like a synthesizer. It's not an unpleasant sound, but it's a little more distorted than I expected in the 10 o'clock position. I may try the bigger transformer and see what happens.

[John_P_WI]

I would guess the "synthesizer" is inter stage modulation or distortion due to the preamp and pi nodes coming from the same filter node.

[The New Steve H]

The grid on the cathode follower is at 154, and the schematic (5F6A) says 180. The Plate is at 254, and the schematic says 325. The cathode is at 154, and the schematic says 180. I was trying to get it closer to the values on the drawing, and this is what I ended up with.

If I want to put a resistor back in the power supply, how small can I go? I want the least voltage drop possible.

I don't think the distortion comes from removing the resistor, because I'm pretty sure it was always there. Also, playing the amp with two tubes removed doesn't make any difference, so I don't know if a bigger PT will accomplish anything.

[John_P_WI]

A 1k resistor with the 16 uF cap will give a RC break point response at 10 hz, if you cut the resistance in half @ 500 ohms, the frequency will be 20 hz, 250 ohm will give 40 hz. Changing the 16 uf to 22 uF will also lower the frequency allowing a smaller valued resistor to be used.