Debugging some home-made amp problems

[Theashe]

Hey guys,

I've been slowly tinkering with designing a tube amp over the last year and a bit, and around (Canadian) Thanksgiving I finally had some time to put it together. It's quite a complicated amp, with a lot put into a small space, but it's got a lot of features that I wanted:
- Tube reverb driver and recovery
- Effects loop and parallel mixer
- Dumble-style tone controls with jazz/rock switch
- Presence and resonance controls, adjustable feedback loop
- PI cathode scaling pot and Dumble trimmer

I've attached a gutshot for your viewing pleasure (or displeasure).

Anyway, I plugged it in for the first time, and it actually sounds pretty good at low volumes. Maybe not the incredible tone-monster that I was dreaming of, but I can work with it. I uploaded a couple of videos for you to see. I'm in the process of fine-tuning it and I have a few problems and have noticed some odd behavior. I was wondering if you guys could maybe shed some light on these issues:

1. Firstly, despite my simulations in PSUDII, the transformer I bought is about 60V too hot. Rather than drop that with a resistor somewhere in the amp, I'm gonna see if I can go get a 300-0-300 instead of the 330-0-330 I'm using. Some of my other problems might be related to this, but bear with me for now.

2. The screens are drawing a different amount of current at quiescent. Right now, my measurements on my 6L6GCs show a plate voltage of 442, screen voltage of 432, and cathode current of 56mA per tube (35V drop across Rk). One of the 1k screen dropper resistors drops 2V, while the other drops 3V.

3. This one's odd. With no signal applied to the amp, if I turn the pre-PI master volume up all the way, the 1k screen droppers then drop 16V and 30V, respectively. This puts one tube's screen at 416V, and the others' at 402V. The tubes have identical B+ and share the same Rk and Ck. The tube with the greater voltage drop glows blue when first turned on, but this disappears over time.

4. When the amp is first turned on, it has a hum, and adjusting the humdinger in the amp apparently has no effect. If left to warm up for 5 minutes, said hum dies down and disappears.

5. The reverb is far too quiet. I believe this was due to me underestimating the amount of signal attenuation in the amp, so I took the signal for the reverb driver off of the cathode of the stage before it instead of off the plate, hoping to get a quieter signal. Since banging on the reverb tank makes a very loud and echo-ey BOOM with the volume turned up, I think the recovery circuit is just fine.

6. When playing guitar through the amp, if the pre-PI master volume is turned up too loud, the amp makes a horrible "squeaky" sounding distortion, the volume dies, and then it recovers. Possibly blocking distortion? I've got 0.22uF coupling caps into 220k grid leak/1.5k grid stop resistors on the power tubes.

Videos as promised. (Sorry that my chops aren't up to snuff.)
A little overdriven: https://www.youtube.com/watch?v=tZDex4V6AXM
Clean: https://www.youtube.com/watch?v=7tE9rt85E0k

Thanks in advance for your help, guys! I won't have time to work on this until the Christmas holidays because of final exams approaching, but I can't wait to get back to it and I'll keep you posted.

[Theashe]

Update: problem 3 above was actually my fault. I had the reverb volume turned up - so when I turned up the master volume, the power tubes were receiving signal from the reverb tank. Still not sure why the two power tubes are drawing different screen currents, though.

[Leo_Gnardo]

Still not sure why the two power tubes are drawing different screen currents, though.

Not unusual even in "matched" sets. Lots of vendors match by emission at some - unpublished - test setting. Fewer match also by gain, we like that! I know of none that go the extra dimension, screen grid current.

[Theashe]

Still not sure why the two power tubes are drawing different screen currents, though.

Not unusual even in "matched" sets. Lots of vendors match by emission at some - unpublished - test setting. Fewer match also by gain, we like that! I know of none that go the extra dimension, screen grid current.

That's some kind of bad good news. Lol.

As for the "squeaking" distortion when the master volume is turned up... any ideas?

[sunnydaze]

Have you tried swapping the power tubes to see if the screen current diff follows the tubes?

That will at least eliminate whether it's a design or wiring issue.

Best regards,
Mike

[Theashe]

Good idea. I'll give that a shot and get back to you.

[Structo]

Power tubes rarely match all characteristics between them, so the differing measurements.
Even matched sets.

Many feel this is a blessing in disguise because an imbalance in certain characteristics can cause a richer harmonic tone from the amplifier.
As long as there isn't a huge difference in current draw.

Hi-Fi guys obsess over slight mis-matches because they are trying to
reproduce audio as close to the original as possible without coloration.

I use dual bias pots on a couple D style amps.

I know it doesn't make up for mis-matched tubes, but it gives me confidence in what the power tubes are doing between each other.

First of course is to try the power tubes in different sockets in the push/ pull amp, listen and measure between each change.

[jjman]

Near the top of the pic is a red wire that is black-taped to the chassis. Above the tape it appears to have burnt insulation. No?

[Structo]

Also, using that black tape to hold wires down is not good.

Tube amps get hot and the adhesive will lose grip and probably fall off.

This could be a source of future problems.

[Theashe]

Near the top of the pic is a red wire that is black-taped to the chassis. Above the tape it appears to have burnt insulation. No?

Luckily, that's just a light reflection. I bent that wire to keep it far away from the heater wires. Here's a picture from another angle, while I was putting it together. The insulation on that wire is still fine after running.

Also, using that black tape to hold wires down is not good.

Tube amps get hot and the adhesive will lose grip and probably fall off.

This could be a source of future problems.

Good advice. I custom-bent every wire for flat runs against the chassis, and used the tape to keep it all from moving while I was putting it together. It shouldn't move even if the tape falls off, but 'tis better to future-proof now.