Allied/Knight problem

[dmartn149]

I've converted this Allied/Knight PA to a guitar amp. All I've done is change the obsolete in/out jacks to 1/4", change the power chord to a 3 prong, and replace the e caps.
It sounds pretty good turned all the way up, but at partial volume it has a really ugly distortion and the volume fluctuates. I noticed that the first power tube (the one connected to pin 5 of the PI) doesn't get as hot as the other.
I compared voltages on the two tubes and found that pin 5 had only 10v on the first tube and 35v on the second. I checked pins 2 & 5 at the PI and got 60 and 30v respectivley. I thought maybe I had a bad PI tube, so I switched it with one of the other tubes, and it sounded the same, first power tube still running cold.
I don't know what I should check next.
Advice anyone?Thanks

[loudthud]

I compared voltages on the two tubes and found that pin 5 had only 10v on the first tube and 35v on the second. I checked pins 2 & 5 at the PI and got 60 and 30v respectivley. I thought maybe I had a bad PI tube, so I switched it with one of the other tubes, and it sounded the same, first power tube still running cold.
I don't know what I should check next.

Swap the power tubes to see if you have a bad tube or the problem is with something inside the amp connected to that socket.

AC or DC voltages? The DC voltage on pin 5 of a (good) 6L6 in this amp should be 0V. It should be zero volts iven with the tube removed. If it's not, replace the 0.05uF cap(s). That's probably what is pulling down the voltages on the PI pins 2 and 5 and making the 6L6s hot.

[dmartn149]

Thanks again Loudthud for your help.
I tried other power tubes before I started measuring. They all sounded the same, one hot the other just warm.
I was measuring DC on pin 5. It went to about 100v when I first turned on the power then dropped to 10 and 35 on the two tubes.
I'll change those caps. I guess I should change all of them. I think there are 7.

[dmartn149]

How about these?
0.05uF @ 630V Tubular Axial Film Capacitors:QTY=30 (.05uF for Tube Electronics) | eBay
or these from Hoffman amps
047 mallory 473
Mallory 150 series capacitor - 630 volt
Price each = $1.08

[loudthud]

Any of those caps should help you get the amp going. Physical size is the only thing to watch out for. Mallory 150's are small and kind of mellow sounding. "Orange Drop" types are somewhat brighter sounding but bigger. Just get the amp working and add the mojo later

[dmartn149]

Thanks, I ordered the mallorys. My ears are pretty old. I don't think they can hear mojo anymore

[dmartn149]

WOAH! It's loud now. LOL I literally laughed out loud when I heard this amp after replacing all the caps. No shortage of gain either. In fact I think it needs to be reined in some. With my guitar turned all the way up it starts overdriving the preamp right from the start. If I turn my guitar down to 4 or 5 It stays clean all the way up (still plenty loud) So do I change that first 10 meg resistor? Would a 5 meg be appropriate? I was thinking maybe I could rewire the "phono control" pot to trim the input on one channel, and have the other permanently trimmed with a resistor.
Any advice would be appreciated

[loudthud]

The input stages of this amp have what is called "grid leak bias". The 10 meg resistors are part of this circuit. Very old Tweed Fender amps like the 5C3 used this type bias. Later Fender amps and virtually all amps made today use cathode bias. The original intent was that grid leak bias was ok if the signal was very small, like a microphone. Gibson guitars make a much bigger signal so cathode bias will give a cleaner signal that won't distort much before you can turn it down with the volume control.

I suggest you change one input to cathode bias. The Phone control could be re-wired to be a Master Volume. It would turn down both channels.

[dmartn149]

I think I have to have both channels biased the same as they each use 1/2 of the same 6sc7 with a shared cathode.
Is the second gain stage in this amp an example of cathode bias? Would I use the same value resistor in the first stage 1.8k? and would I keep the .05 cap and 10meg resistor just after the input?
Thanks

[dmartn149]

Okay, I've been looking at all the fenders that use 6sc7 preamp tubes. It looks like whenever they used cathode bias on a 6sc7 (not often) it was a 2.5k resistor and a 25mfd cap. So I guess that's what I should do. Should I plan on doing all 3 6sc7s or just the first one and see how it sounds? The other two are already cathode biased, but w/o the cap. (there is a schem earlier in this thread)
Thanks

[loudthud]

I suggest you change one input to cathode bias. The Phone control could be re-wired to be a Master Volume. It would turn down both channels.

Ooops, I didn't notice that the 6SC7 has the cathodes tied together internally so different bias on each side is not possible in the normal sense.

You could add a cap across the 1.8K of the second 6SC7 and that will increase the gain a little. In the phase inverter, the unbypassed cathode resistor helps couple the signal to the bottom side, but you could try bypassing the 1.5K too see if it sounds good.

[dmartn149]

Lots of options. I think I should go one at a time, starting with cathode bias on the first tube. The nice thing about working on this amp is it's easy to see where everything goes, being "point to point"

[dmartn149]

I just realized, I still need a grid leak resistor. Right? 1 meg?

[loudthud]

1 meg is pretty much standard.

[dmartn149]

Okay, It sounds pretty good now, maybe still just a little bit over cooked with my guitar all the way up. I haven't decided yet. If I was going to bring the gain down a little more, would I need MORE resistance between the cathode and ground?

[loudthud]

Okay, It sounds pretty good now, maybe still just a little bit over cooked with my guitar all the way up. I haven't decided yet. If I was going to bring the gain down a little more, would I need MORE resistance between the cathode and ground?

It's a complex problem that has no simple answer. Because of the common cathode, options are limited.

I need to know the cathode, plate and B+ voltages to see if the biasing is optimal. Simply changing the current won't change the gain that much, only about 20%. I'm thinking the problem is limited headroom because of poor biasing, not too much gain. The best option might be to chage to a different tube type.

[dmartn149]

Okay, so I'd measure cathode v at pin 6 and plate v at pin 5. Would I measure B+ at the other end of that 220k resistor? (the end away from the tube)

Here is part of the schem showing the changes so far. The grid stoppers are 47k and the cathode resistor and cap are 2.2k and 22uf . Not exactly the values I was looking for, but I was trying to make due with what they had at my local Cel Phone Shack

[loudthud]

Okay, so I'd measure cathode v at pin 6 and plate v at pin 5. Would I measure B+ at the other end of that 220k resistor? (the end away from the tube)

Yes, it says 295V there now but that voltage may have changed.

[dmartn149]

I'm measuring 2v at the cathode, 165v at the plate and B+ is 287v. Are these okay?

[loudthud]

I drew a load line on the curves in an old RCA data book. The bias looks about as good as it could be. Lowering the plate resistor will lower the gain but also reduce the headroom on the plate. That must be one hot guitar. Since the cathodes are tied internally, you can't add cathode degeneration because that will inject an out of phase signal into the other channel. So to improve headroom on the input you would have to re-wire for a different tube or add an attenuator on the input, in effect another volume control between the guitar and the tube.

[dmartn149]

Well the amp isn't really distorted, just not super clean, and over all it sounds pretty darned good, and really loud too.
I sure do appreciate all your help and advice. It's very unlikely that I would have ever gotten this amp working right on my own.
Thank you very much Loudthud