Could this be a bad OT??

[crbowman]

OK guys, I’ve got a weird one……
Just got through reworking an old Spitfire clone I built a few years ago into a sort of stripped down AC15 with the EF86 preamp like the AC30/4. Basically just grafted the EF86 pre onto the Spitfire.
As a Spitfire, it was working fine but I wanted to try the EF86.
Now the amp sounds beautiful clean but as soon as the volume goes over about 9:00 I get popping noises and really horrid distortion. Sounds kinda like a broken fuzz box running parallel with the straight signal. As the signal decays, it reaches a point where this distortion stops before the signal completely dies off. All voltages are within 5-10% of where they should be. My DVM squeals at me when I take voltage on power tube plates, but the voltage is still around 340 volts.
On the scope I’ve got a decent sine wave all the way through the amp up to the grids on the power tubes (EL84’s), but when I put the scope on the output jack I’ve got a seriously squashed square wave with very distinct crossover notch. The fact that this amp is really dead simple is kinda insult to injury.

So far I have:
Swapped out all the tubes twice over.

Replaced the 120 Ohm cathode resistor and bypass cap with known good ones.

Verified that I have a good signal coming out of the preamp buy taking a line out after the Vol contol and run into separate amp (this sounds great!).

Verified that all solder joints are good and clean.

I have completely rebuilt the output section and power supply and replaced all components except the output transformer.

On the OT, I have roughly between 75 – 100 ohms from the center tap to either plate, and open to ground. Fuse doesn’t blow. I have another suitable OT that I could try swapping out, but it’s a pull I got from somebody who wasn’t exactly sure why it was pulled so that could be a whole other can of worms.

Has anyone experienced this behavior from a bad OT? If not I’m stumped.

Any guidance is greatly appreciated.
Thanks!
CB

PS: Yes, I have tried different speakers, cords, and guitars.

[sliberty]

Blocking distortion maybe???

Check out this article:

http://www.aikenamps.com/BlockingDistortion.html

and try some large value grid stoppers as described at the end.

[Firestorm]

[quote="crbowman"]when I put the scope on the output jack I’ve got a seriously squashed square wave with very distinct crossover notch./quote]

Squashed on both sides? How much current is going through the tubes?

[crbowman]

Blocking distortion maybe???

Check out this article:

http://www.aikenamps.com/BlockingDistortion.html

and try some large value grid stoppers as described at the end.

Thanks sliberty. This does sound similar to what I'm experiencing.
Definitely worth a shot.

when I put the scope on the output jack I’ve got a seriously squashed square wave with very distinct crossover notch.

Squashed on both sides? How much current is going through the tubes?

Squashed on both top and bottom but more so on the top. IIRC, current is pretty high. Don't remember the exact figure but I'll check again when I get home. Tubes aren't red-plating though.
Thanks guys! I'll report back.

[crbowman]

OK,
This is where my fundamental lack of electrical knowledge shows up.
I've got 343 volts on each of the EL84 plates, 341 on the screens, and 11.2 from cathode to ground on a 2XEL84 cathode biased amp with a 120 ohm resistor, cap bypassed (25/50).
So using Ohm's law I'm coming up with what seems like a crazy number here.

[Structo]

I had an amp act like this before when it had a bad preamp cathode ground.
But it wouldn't explain why it sounds good at low volume.
Probably is blocking distortion.

Here's how I figure cathode bias.

1. Measure the voltage drop on the cathode resistor to ground. Write the value down.

2. Divide this voltage by the value of the cathode resistor. This gives you the amount of current being drawn by both power tubes in milliamps. Write this value down.

3. Measure the voltage on the plates of the power tubes to ground. Write this down.

4. Now, subtract the voltage from the cathode resistor in step 1 from the voltage measured on the plates. Write this value down. Take this value, and multiply it by the current (milliamps) from step 2. This will give you the dissipated power (in watts) of both power tubes. Write this figure down.

5. Take the figure from step 4 and divide by 2. Write this figure down. This is the power dissipation (in watts) of each tube. For EL84s, if it is over 12 watts, then you need to install a higher value cathode resistor. If it's 10.5 watts or less, you need to install a lower value cathode resistor.

6. After installing the new cathode resistor, do ALL of the steps again to see what you now have. You may have to repeat this process several times to get it dialed in, but it is worth it, and your ears will thank you.

[Phil_S]

OK,
This is where my fundamental lack of electrical knowledge shows up.
I've got 343 volts on each of the EL84 plates, 341 on the screens, and 11.2 from cathode to ground on a 2XEL84 cathode biased amp with a 120 ohm resistor, cap bypassed (25/50).
So using Ohm's law I'm coming up with what seems like a crazy number here.

Check bias first.

343 on the plates and 341 on screens is too much voltage for an EL84. At the very least, increase the value of the dropping resistor to lower the screen voltage. You'll cook those tubes PDQ with the info you stated.

Here's worksheet to calc your bias. I inserted a second scenario, but all the voltages will change if you put in a 150 ohm cathode resistor, so it only suggests that would be an improvement in terms of watts. With cathode bias, your target is around 100% of max.

[crbowman]

Thanks Guys!

OK,
This is where my fundamental lack of electrical knowledge shows up.
I've got 343 volts on each of the EL84 plates, 341 on the screens, and 11.2 from cathode to ground on a 2XEL84 cathode biased amp with a 120 ohm resistor, cap bypassed (25/50).
So using Ohm's law I'm coming up with what seems like a crazy number here.

Check bias first.

343 on the plates and 341 on screens is too much voltage for an EL84. At the very least, increase the value of the dropping resistor to lower the screen voltage. You'll cook those tubes PDQ with the info you stated.

Here's worksheet to calc your bias. I inserted a second scenario, but all the voltages will change if you put in a 150 ohm cathode resistor, so it only suggests that would be an improvement in terms of watts. With cathode bias, your target is around 100% of max.

Phil, did you mean to post a link to the worksheet? I'm not seeing it.

[Phil_S]

duplicate post

[crbowman]

Got the worksheet. Thanks Phil!

[crbowman]

OK put in a 165 ohm cathode resistor which according to Phil's worksheet puts me at 97.75 % of max. Didn't have the right resistors on hand to adjust the dropping string.
Amp is a little quieter, however, did not solve the problem. Popping seems to have subsided, broken fuzz sound was unaffected by this change.
Voltages went up slightly:
Plates and screens at 348V (which is higher than Phil indicated they should be but is still slightly under what is specified on the Spitfire schematic), and cathode is at 12V.

[crbowman]

Got it figured out.
I had to move the OT primary wires to the back wall of the chassis and get them away from everything else. It seems that with the extra gain of the EF86, lead dress issues which hadn't been a problem before with the 12AX7,
become much more significant. Amp sounds beautiful now at all volumes.
Thanks to everyone who responded.