Source Of Squeal On 5C8 Clone?

[Cliff Schecht]

Something else just popped in my head. The second amp I ever built used a 6SJ7 for the first stage. Squealed to the moon and back no matter what I did (other than shunting highs with caps in parallel with plate resistors, a band-aid fix in my book) until I connected the 6SJ7 shield pin to either the cathode or ground (don't remember which worked). The glass 6SC7 doesn't have internal shielding like the glass 6SJ7's do but the metal can ones do allow you to connect pin 1 (shell) to something. If you just leave the metal can floating, however, you might cause it to squeal. Same for the 6J5, the metal ones are prone to squeal if the shield is floating. From your pictures I actually can't really tell what you are doing with the shields, one of them looks grounded maybe so try moving the shield to cathode potential.

FWIW I stay away from metal can tubes when I can, I've had more bad luck than good luck and they tend to just not sound as good IME. The glass encapsulated versions just sound better. Maybe IMO somewhat, but the general online public seems to agree with this notion.

Also make sure that ALL of your hardware is cranked down well. I've had amps that had problems for months before I figured out on one build that loose hardware (pots, grounding chassis screws, inputs/speaker ins/outs, etc) can cause all sorts of weird problems, including untamable squealing!

[Cliff Schecht]

One other thing I'm staring at now is the 250uF bypass cap on the 6J5 cathode. It doesn't benefit that stage to be that large and IME, trying to get too much gain like they were doing here can cause instability. 22uF-47uF is good enough for here for full cathode bypass.

What is the capacitance you used on the cathode biased input stage? I think in my 5C3 build I left it unbypassed as I found that there was TOO MUCH GAIN with any additional bypass capacitance. IME the 6SC7/6SL7 tube types have even more gain that some AX7's and really get after it when driven hard. But too much gain and they will most definitely let you know - luckily they oscillate in the audible range so we can at least hear there is a problem!

I also agree above that the 75k resistors can be made 220k-270k without an audible consequence (and this might also help cut down on squealing).

[martin manning]

rp, this wire going from the 6J5's plate to the treble nw has signal in-phase with V2's input, and it passes near the V2 socket; maybe something to check out. You could disconnect it from V3 (disabling the treble) and see if the squeal goes away.

Cliff, earlier rp says he's grounded the shields...

[rp]

Fixed, 100% I think. Sure glad the amp doesn't have to live with a cocktail shaker forever on it. Still, I have no idea why I was able to stop it from the top that way.

Looks like Milkman called it, or maybe Martin, they'll have to split the winnings. I ran shielded for both those wires. The teflon I used is the bad stuff too, 50pf a foot, but if I lost any high end with 8" of wire I sure can't hear it. I was suspicious of those runs from the start but best to always try and avoid shield wire first IMO, Leo pulled it off, and with reverb and trem too.

Mark, changing the voltage divider to 68k/220k neuters V2, also it causes too much of a gain difference btw the two channels which bugged me. For a low gain amp this thing has a lot of kick, it's very dynamic and immediate, I want to keep it right were it is.

I thought maybe the lower value had to do with the 6J5s modest mu but I checked the pictures I got off ebay of the 1953 5C8 with the 6SC7 for V3 and Fender used 75k/75K there too. So even with my mod the balance btwn the channels sounds perfect as is, which is good cause you can really hear the specific difference going from the grid leak side to the cathode biased side - and both are great. Very cool.

Cliff - of course I grounded the sockets. I'd be terrified of having four metal cased tubes that were ungrounded! Yikes! The glass tubes now sound great - same hum but are no more microphonic than novals - but I want to spend more time with the metal ones. They are brighter (sharper) and more 2D but they have an edgy gritty grind I like. Don't write them off. Glass for all 7 tubes in the hifi chassis is pretty with the lights low.

Soon to come the high volume test.

Thanks all.

[Structo]

I'd love to hear that amp if you get a chance to record a clip.

Very old school.

When I was about 15 I had a tube amp that was point to point, not sure what it was out of, probably a console or something.
But it had a couple metal tubes on it.

It really had a cool tone when it was cranked!

Wish I still had it.

[martin manning]

Good to hear you've got it sorted. 33pF to ground? pFFFt! that can't do much... The low mu on the 6J5 actually would let you use larger mix resistors without fear of HF roll-off, so I'm not sure why they are only 75k in the 5C8. Soon after Fender went to 270k into a 12AX7 with 5x the mu (5F6), and then settled on 220k. Most guitar amps are arranged in a linear fashion input to output, so you're pushing your luck a bit here.

[rp]

Squeal happened on channel two with any input jack open or closed. And no difference in the squeal either way. I discovered I can make it a bit worse by bringing up V1 while V2 and the treb are maxed.

That's 'cause you're decreasing the mixer's attenuation of signal out of V2 by bringinging up the V1 channal volume control. That's about the same as you'll get miss-matching the mix resistors. I guess I would try increasing both 75k's to say 220k. The larger they are the more HF roll-off you will get from the 6J5's Miller capacitance, which might actually help.

update: depending on my preamp tube rolling I can still get a touch of the squeal happening when I do the above knob twiddling. As I'm not jumping channels and they're not interactive like a 5E3 and I could just leave it. But before I try upping the 75K mixers (to the 6J5) to 220K, does anyone think I could benefit from stoppers on the power tube control grids? I had already wired the coupling caps through pin 6 (NC) to pin five just in case, so I could just clip the wire and solder in 1.5K stoppers easy. Question: do these only prevent oscillations developing at the power tube stage or cure what has come before?

Or should I not bother and just concentrate on the preamp stages? I like the 75K combo going into the 6J5. It's very perky as is, I'd like to leave it.

I read this and learned something http://www.aikenamps.com/InputRes.htm

[martin manning]

But before I try upping the 75K mixers (to the 6J5) to 220K, does anyone think I could benefit from stoppers on the power tube control grids? I had already wired the coupling caps through pin 6 (NC) to pin five just in case, so I could just clip the wire and solder in 1.5K stoppers easy. Question: do these only prevent oscillations developing at the power tube stage or cure what has come before?

The grid stopper makes a low-pass filter, which reduces gain at high frequencies across the stage. If that stage is the source of an oscillation, or it is in the feedback loop, it might reduce the HF gain enough to stop it. In your case the oscillation seems to be up front, possibly involving the input stage(s) and the 6J5, so I don't think grid stoppers on the power tubes will help. It's a bit of a shot in the dark, but I would try increasing the mix resistors to 220k or 270k.

[Cliff Schecht]

Just some food for thought RP.. I recently finished a little handheld amp (5Y3, 6AU6, 6AQ5) and for the first day when I flipped it on, it squealed like a mofo. I tried all sorts of patches until I got frustrated and started chopsticking. The two leads for B+ and plate from the power transformer got moved maybe a cm and the oscillations stopped. I pulled all of the band-aids out of that amp and it's still squeal free. Maybe more chopsticking before you pull parts? Move around the power/transformer wiring and see if that helps..