The sound of a cathodyne PI dying inside

[jordan86]

My new Princeton is great. Stays mostly clean up to about 6.5-7 on volume dial. As I pull in to the neighborhood around 10 though, things get real interesting. Audio and video below.  It's like this crazy gated fuzz. While experimenting, it did this notably less with a 5v4, and even less so with a different 12ax7 in V1.  In this video I am running a preferred series 7025 from the tube store in V1, as well as their preferred series (chinese) 5ar4.

Would love to know what's dying here, and if this is bad/harmful to my amp. Does this sound like a cathodyne PI distorting? Would a different grid stopper help? Or would raising the PI plate voltage (stokes mod) get me to "10" on the volume dial before this happens?

If helpful to know, I have done most of the Rob Robinette mods in the attached schematic. My PI grip stopper measured 433K FWIW.

Plate voltage is 435v if I recall correctly, biased to about 21ma of plate current, roughly 9 watts of plate dissipation.

Side note: I get no sound until the volume knob hits about 2-2.5...is that common?

[sluckey]

Ever solder those resistors???

[jordan86]

Caught red handed. I did not yet :(

[PRR]

You don't have to solder resistors. Bogen missed one G2 resistor on one of their biggest amps. Crimped, not soldered. Sold it to a film-lab on the 13th floor of a NYC hi-rise. Worked great several years. Caught FIRE. Unlike my house, they could not kick the burning corpse out in the yard. Controlled panic.

[jordan86]

Ha! Sluckey is referring to my bypass caps. I paralleled all the resistors around the bypass cap leads by twisting them several wraps around, but forgot to go back and solder. Will be done tonight?

Is that really the culprit for the fuzz tones?

[shooter]

fwiw;
a grid stopper helped my distortion ALOT same PI
I also bumped the cathode/plate R's to 68k
where are you biased?

[jordan86]

My 6v6 Plate voltage is 435v, Biased to about 21ma of plate current, roughly 9 watts of plate dissipation.

The voltage on pin 6 of my PI is 207v.

[shooter]

have you read this;
http://www.valvewizard.co.uk/cathodyne.html

got a scope?
here's a before/after from my cathodyne

[jordan86]

I have read that. But no scope.

I was more so curious if more trained ears could confirm this was indeed the PI.

I currently have a 470k in there. I assume I would need to try a higher value? Between 500k and 1 meg. Easy solution.

This would potentially create some additional high end loss? The Princeton is a tad warmer than I would like. Considering a bright cap anyhow.

[SILVERGUN]

Ha! Sluckey is referring to my bypass caps. I paralleled all the resistors around the bypass cap leads by twisting them several wraps around, but forgot to go back and solder. Will be done tonight?

Is that really the culprit for the fuzz tones?

Those resistors matter more than the caps do so if I was using that technique I would make the mistake the other way around. Either way I wouldn't come to any other engineering decisions until they were soldered in.

[tubeswell]

I vote for blocking distortion.

[pdf64]

...Would love to know what's dying here, and if this is bad/harmful to my amp...
Side note: I get no sound until the volume knob hits about 2-2.5...is that common?

So what happens to the background hum / hiss when the vol control is turned from 9 to 10?
With the vol on 10, do the instrument’s vol / tone control settings affect the horrid sound?
With the vol on 10, does the amp’s treble control setting affect it?

I strongly suspect ultrasonic oscillation, to me the sound and way it manifests is characteristic of that.
ie intermodulation sum and difference frequency byproducts of the guitar and oscillation signals seem identifiable to me.
High resolution photos would be beneficial.

Apart from sounding bad, it's not harmful to the amp, though I guess that the power tubes may wear out faster if used like that.

Bearing in mind that full counter clockwise on the knob should be '1', getting signal first appearing between 2 and 3 seems normal.

[Willabe]

I vote for blocking distortion.

At the PI?

Or which other tube(s)?

[jordan86]

Thank everyone. I will test some more today and report my findings. I soldered the bypass cap/resistor leads. No change. So I tried the stokes mod since I was in there and the iron was hot. What a jump!

Plate voltage in PI went from 207 to 260v. Seemed slightly less noticeable in my quick test. But again, it is worse with the 5ar4 and this particular 7025 at V1.

Will do some more research with tone pot interactions and let you all know. Thanks.

[shooter]

I strongly suspect ultrasonic oscillation

fwiw;
if it's an oscillation you can try "walking" a 500pF to .001uF on plate R's for your pre amp tubes so;
gatorclip a cap in parallel with a plate R, evaluate, move to next, repeat, STOP prior to Pi

[pdf64]

As I see it, the main unknown is whether the loop includes the reverb driver, or power tubes, or both 

[jordan86]

The fuzz seems to have gone away with stokes mod, a 5v4 and a lower gain V1 tube. 

BUT.....

Made another discovery today...I have found a new sort of crackly distortion that goes away when I either disconnect the the reverb tank, or pull the 12AT7 altogether. Stupid me. I never jumpered pins 3 and 8 on the 12AT7, or pins 2 and 7 for that matter. They're practically invisible on the Mojotone schematic. Like 100% would miss it unless you knew it was supposed to be there. Ugh. Don't get me started on their "instruction manual".

There was a comment from Sluckey in an old thread that brought it to my attention. Man that guy is good. :)

Will jumper those connections in the morning and hope for some good juju.

Thanks again all.

[PRR]

Put 100k in series with the 12AT7 grids. And don't make the grid coupling cap (500pFd?) too big.

Distorting in the reverb driver is no big deal. But distorting at the driver *grid* can muck-up signal in the dry path unless isolated some.

[jordan86]

The grid coupling cap is indeed 500pf. I don’t have any 100k’s. Would a 68k work? In series, meaning it would go between the 500pf cap and pin 7 or do you mean across pin 7 and pin 2. Like a typical grid stop on a power tube?

[tubeswell]

https://m.youtube.com/watch?v=bCN4LH4urns

[pdf64]

The grid coupling cap is indeed 500pf. I don’t have any 100k’s. Would a 68k work? In series, meaning it would go between the 500pf cap and pin 7 or do you mean across pin 7 and pin 2. Like a typical grid stop on a power tube?

Between 7 and 2 would be no good, as only one grid would get the stopper. It needs to be in the series feed to the grids.

[jordan86]

Update: Connecting both sides of the grids and cathodes on the 12at7 seemed to have solved all problems. No need for a grid stopper at this point. I do have one on the PI fwiw. 

Operator error is the worst. Sorry guys. Appreciate the help. The Princeton is sounding great.