6V6 Hiwatt issues

[Smokebreak]

After watching Isle of Wight for the 50th time the other night I figured I needed a Hiwatt right now, and that's been my project over the past few days. I worked off the M.Huss CP103 preamp :http://hiwatt.org/Schematics/CP103Pre.pdf and referred to the Trinity Triwatt :http://www.trinityamps.com/ForumGallery ... out_v1.pdf
My issues are oscillation and blocking distortion.
I admittedly half-cocked the layout, and used an 18w TMB(I think) prepunched chassis, tagboard, and generally rushed through things. I used some beautiful iron from RJ, his "15w" Edcor set, 540CT, which is way low voltage for a Hiwatt I understand. I have 360V plates. I made it a single input amp and my split after the 1st stage gain pot is 470K/150K, with 2K2/47n cathodes for the 3 gain stages. No extra OD stage here as in the Triwatt. Grid bias resistors are 100K then 22k stoppers for 6V6. All of my voltages are consistently 10-20% lower than Triwatt voltage chart, so nothing totally out of whack, as my B+ is equally as low.
On fire up the cleans were quite stunning and bell-like, way different than Fender sparkle or Vox chime. Love it! When everything is close to dimed, in comes the oscillations and what I think is blocking distortion. I haven't experienced this kind of oscillation, as it's a strange whirring that I can kill by touching chassis, guitar strings, tailpiece, metal on guitar cable, etc. I've got shielded cable to input, from board to 2nd stage, and from MV to 3rd stage. This helped some. I've got everything in preamp grounded via buss by input chassis, and everything from PI on down next to PT . Chopsticking doesn't change anything, really, but it's kind of hard to tell because getting my hand close to the chassis brings on the noise.
The blocking distortion isn't that noticeable with chords, but single notes(only at near dimed) have a nasty quality that sounds like a good note and a crackly farty note together. I've just found Aiken's page on the matter, so I'll go through that, but wanted to get the opinions from the experienced minds here, as well.
Also of note is the 6V6 starts pulling 35-40mA when the trouble starts, I've got it biased around 22mA. I've swapped tubes and that definitely has had an effect, for better or worse, but the problems remain. I keep staring at the fact that the last B+ node feeds the first stage, then the 3rd stage and CF, and some grounding issues there, possibly...

[Reeltarded]

Inducing... Inducting... MARTIN!!! help! lol

I think it had a questionable ground, and/or a preamp tube. Only one way to figure that out.

Pull the first preamp tube and see if it still reacts to your floating above the circuit. If yes, pull the next tube. If the PI is all that's left and you have swapped in the other tubes in it's place it is probably a looped ground or maybe one that isn't quite good enough. Cleans can get away with stuff distortion never could. Is it doing any idle buzzing noise?

Ahh.. another thing. Are all the stages and outputs snubbed with grid swampers?

Interesting and I think I have experienced this same thing before and fixed it (if circuit based) by even so much changing a pot that tested ok... ghosts in the machine?

[martin manning]

This thing has an unusual NFB loop, with phasing opposite to a normal PI-tail FB. Have you tried swapping the OT primary leads?

[Smokebreak]

Inducing... Inducting... MARTIN!!! help! lol

I think it had a questionable ground, and/or a preamp tube. Only one way to figure that out.

Pull the first preamp tube and see if it still reacts to your floating above the circuit. If yes, pull the next tube. If the PI is all that's left and you have swapped in the other tubes in it's place it is probably a looped ground or maybe one that isn't quite good enough. Cleans can get away with stuff distortion never could. Is it doing any idle buzzing noise?

Ahh.. another thing. Are all the stages and outputs snubbed with grid swampers?

Interesting and I think I have experienced this same thing before and fixed it (if circuit based) by even so much changing a pot that tested ok... ghosts in the machine?

Pulling the first preamp tube kills the wackiness. Is that proof my problem is before 2nd stage? Btw since this is single input amp, v1a and b are stage 1 and 2.
At idle, maxed out controls, it definitely hisses and buzzes with guitar plugged in, and it quiets down considerably with no input.
I've only got a 22k stopper between MV and 3rd stage, and 22k on the outputs. I'd go higher, but 6v6 datasheets show 100k max at grids and bias resistors are 100k so I'm already at 122k..Hogwash?
I found a v1 tube that sounds great, and I can't really hear the nasty distortion, but it squeals in addition to the other noises. Other tubes sound awful maxed, but aren't as problematic. Uggh. I think I've been through 10 12ax7s. I actually have a Chinese at7 in PI, swapped in a JAN Philips at7 and it was waay worse.
Martin I had to swap the OT leads on first startup because it immediately had low frequency overwhelming rumble. I switched them again just now to check, and the rumble returned.
What about my assessment of simulating the 4 hole input using 1 w/ 470k/150k divider after pot? I know Pete used 1 input with the rest turned down...in Marshall's when simulating this (470k/470k mixers), a 470k/470k split works, so I figure on this, 3 470k in parallel puts the leg to ground around 150k?

[Firestorm]

I'm not sure pulling V1 tells you anything in this case because it kills two cascaded stages; the problem signal won't show up. From a lead dress perspective, you might try labeling the phase relationships of signal leads to make sure you haven't got same phase leads too close. The "blocking distortion" sounds like it could be the coupling caps out of the PI (per Aiken) or overdrive characteristics of the LDR (per Blencowe). You could go with higher grid stops on the 6V6s; Fender ran 220k grid loads at 400v and got away with it. But maybe the PI grid would be the place to play. Granted, the PI cathode resistance isn't huge.

[Smokebreak]

I'm not sure pulling V1 tells you anything in this case because it kills two cascaded stages; the problem signal won't show up. From a lead dress perspective, you might try labeling the phase relationships of signal leads to make sure you haven't got same phase leads too close. The "blocking distortion" sounds like it could be the coupling caps out of the PI (per Aiken) or overdrive characteristics of the LDR (per Blencowe). You could go with higher grid stops on the 6V6s; Fender ran 220k grid loads at 400v and got away with it. But maybe the PI grid would be the place to play. Granted, the PI cathode resistance isn't huge.

Ok so do I have it correct that signal from first stage plate to second stage grid is inphase with signal fom 3rd stage plate to CF grid?
I've now got stoppers on every pre stage, and more shieled wire than I've ever used, and it's maybe helped a bit. Removing cathode bypass on stage 3 helped a good bit, but lost gain...now the cleans aren't quite as beautiful, but the OD is much better and focused, maybe.
This is drivin me nuts. At high control settings., even with the guitar volume turned down, a played note is lower in volume and a bit overtaken by noise. I can touch the chassis and all the noise disappears, and the volume and tone gets to where I can tell it should be.

[Smokebreak]

More strange behavior!! I didn't notice until now that my 6V pilot light is flickering, and when I touch the metal speed selector on the ceiling fan above me, the amp loudly crackles and pops . Also, when I touch the chassis, the 6V6s exhibit a bit of blue.
I have turned everything off in the apartment, only now to be sadly sitting in the dark watching the same flicker and oscillations. Lol I gotta put this amp down for a day or so..

[martin manning]

Give me $200 on Parasitic Oscillation to win.

[Reeltarded]

And it's developed in those first two stages.



Umm...

Place the grid resistor right on the socket (68k?) both 1st and 2nd stages. 2nd can be lower but you wouldn't notice a difference.

I still didn't look over the scheme. What is the interstage mixer arrangement between 1-2? Holy shit! 220k plates everywhere. My god that thing must me smoking. No, I mean I bet it smells like a new toaster everytime you fire it up.. no, I mean really fire.

Throw a 500v 250p cap over that first plate load resistor. Buh byes PO BS.

[Smokebreak]

I appreciate the help guys. I think I've tried every band aid in the book. I'm thinking the stages are too close together. It's rebuild time.

[Cliff Schecht]

Without seeing a schematic I can also guess that it's oscillating somewhere. Pictures are worth a 1000 words here, we might be able to point you at something that you aren't seeing.

[Smokebreak]

Well I ended up moving v2 and v3 up on the chassis to isolate any phasing issues but that didn't help like I thought it would. I took out the bus ground for a star ground , and that did the trick.
Miles, my split at the getgo is 470k/150k. I'm actually thinking of going down a bit on the lower leg.
I can't help but think the fact that the way the last B+ node feeds the first stage as well as the CF stage has something to do with the way grounds go there.