Single-ended Class A Another Dilemma

[Crispy]

This is still about my 5f2a I built using Classic Tone Iron with multiple taps. So far I've tried a single 6v6S (great sound, but not quite giggable volume) and a single 6L6GC (Lots more headroom and overall volume, but I think I liked the tone better with the 6v6). So the next thing to try (well, unless I get a KT-66 or a better 6L6) is either 2 6v6's in parallel, or even 2 el-84's. What I want to know is, is having a matched pair even a concern in a true Class A amp? I don't see why I can't just get another JJ 6v6S and not worry about it. But I'm starting to figure out sometimes it's better to ask somebody before diving in. I can't believe I'm asking this, when I haven't even finished reconfiguring it for the 6L6 yet. ( I had it up and running for a minute or 2 but the big old 6L6 was hitting the speaker magnet, so I had to move all the tubes over one hole to the left... I had anextra hole because I thought I might try the parallel output some day...and 6v6's or el-84's will clear the speaker, no problem)

[shooter]

having a matched pair even a concern

All but 1 of my Class A's have been either PSE or QuaDSE.  I use a Cathode R for each, not mandatory, I just like to get them as close as possible.  I liken it to a tandem bike with the guy in the back just sorta there for the ride, you're gonna wear out faster.

[PRR]

Matched or not-matched will make a small difference in the overload sound.

I say just do it, and experiment with various pairs in your free time.

[Crispy]

Thanks, guys. I didn't see any reason not to try it, but I don't see lots of things until it's too late.

[Crispy]

Got it back together, hooked it up to the light bulb thingy, and as the rectifier warmed up and the bulb started to glow again, had a hum that was never there before. Shut it down. Still looking.....

[John]

If you moved stuff around, maybe you got a signal wire close to the heater wires, or something like that. With a plastic or wooden stick, and 1 hand behind your back, try moving any loose wires around to see if that helps the hum. Sometimes just a half inch does the trick.

[Crispy]

That's what I thought, too. A half hour with a flashlight and a magnifying glass, and I found a couple strands of heater wire bridging from pin 2 to pin 3. Got it going. Had to add a 1.2K resistor in parallel with the 700 ohm cathode resistor, measured out 432 ohms, biased right around 63% with the JJ. Couldn't find a data sheet for the Sovtek I had that wasn't in Russian, so I just plugged it in. Sounds fantastic! Much louder and cleaner than the JJ, which is what I wanted. I've got lots of pedals to make a clean amp dirty, but I don't have any to make a dirty amp clean! I've got enough clean headroom for the small clubs I play in now! Think I'll leave it as is for a while. Might not need the 2x output tubes, after all! Of course, that doesn't mean I won't be trying it anyway. Next time I order a board from Hoffman, I think I'll have him add about 4 extra lugs at each end, just for tinkering. For instance, although the Princeton sounds great, there's a tiny bit of "warble", for lack of a better word. Kinda like a very slight tremolo... Needs more filtering! So, now I'm thinking about adding a choke (even though the "warble" is very slight, and kind of cool....)

[shooter]

biased right around 63%

You said it's loud enough but fwiw I normally run my Class A's at about 90%, fixed bias AB at about 65%

run it trough a 2X12 or 4X for even more scare the neighbors full sound

[Crispy]

Don't know what it's biased at with the Russian 6L6GC, can't read the data sheets in cyrillic... But it's a lot louder and cleaner than it was with the JJ, so I'm thinking this particular Sovtek has a lower max plate dissipation, therefore biased hotter, maybe? Sounds fantastic, so why sweat it? Oh, and that "warble" sound? Turns out the guitar pickups were too close to the strings. Stepped on the tuner and still heard it, with the amp silenced. A few turns of the screwdriver and it was gone. That's how sensitive this amp is to string dynamics! Never heard that warble thru any other amp! Takes pedals real well, too!

[Crispy]

Worth noting in regards to my post #6 on this thread, that my light bulb current limiter saved my *ss on this one. Yes, there was a hum, but what really got my attention was when the light bulb got to its "normal" brightness, and kept getting brighter. I shut the amp down immediately. "Light gets bright, something ain't right!". So I kept looking until I found it. This sucker sounds fan-friggin'-tastic now! Glad I didn't let the smoke out!

[shooter]

fan-friggin'-tastic now

it's good when the planets align

some of the best guitar player responses came from my cloned PSE 6L6 GA-8.

[Crispy]

Well, I couldn't leave well enough alone... thought it sounded "fan-friggin'- tastic", until I got a chance to really crank it. Decided it really WAS still biased a little cold. Swapped the Cathode resistor for a 350 ohm. Calculator says I'm at 80% now, 24.9W dissipation. But it seems like every time I fix one thing, I screw up something else! Now, when I get either the Volume or Tone knob near "10", output drops to next to nothing, and I get a nasty hum and not much else (I've only bumped it there momentarily a few times, just to see if it still does it ). Amp sounds fine, if you don't dime either knob. Tried it with the limiter in place, and instead of the bulb getting brighter when this happens,it actually dims? First thought was a bad 12ax7, but that's not it. Any thoughts what to check next? Chopsticking found nothing. All voltages in spec. I'm stumped, for now. Update.. It occurs to me that perhaps when you change the value of the cathode resistor, it could necessitate a change in the value of the bypass capacitor, as well? Trying to get my head around that one, now.

[shooter]

output drops to next to nothing

disconnect the NFB as a test, get better or worse?

Where did your bias and plate B+ land when you swapped the cathode R?

guessing you *dimed" the amp before the change and it was ok?

[Crispy]

When I said I used a 350 ohm resistor, that was only half the truth. I actually used 2 700 ohm resistors, because I didn't have a 350. Turns out, this was the source of the problem. Parasitic oscillation. I still don't have a 350 ohm, but separating the 2 by a 1/4" ALMOST eliminated the problem. Now I can dime both knobs at once without causing it, unless I plug in a guitar, then I can ALMOST dime either, but not quite. But at least I know where the trouble lies. I'll try later to see if I can get another 1/8" between them, to get me by until I get the right resistor.

[AZJimC]

When I said I used a 350 ohm resistor, that was only half the truth. I actually used 2 700 ohm resistors, because I didn't have a 350. Turns out, this was the source of the problem. Parasitic oscillation. I still don't have a 350 ohm, but separating the 2 by a 1/4" ALMOST eliminated the problem. Now I can dime both knobs at once without causing it, unless I plug in a guitar, then I can ALMOST dime either, but not quite. But at least I know where the trouble lies. I'll try later to see if I can get another 1/8" between them, to get me by until I get the right resistor.

Do you mean you put a 700 ohm to each tube, or that you actually paralleled the two resistors?

If you placed one to each cathode, then try tying the resistors in parallel. In a Push Pull amp, the two cathodes tied together help maintain both cathodes at the same level. As one pulls voltage higher, the other is going lower, and they offset one another. If they are 700ohms apart electrically, this effect would be less, allowing either tube to shift cathode voltage.

[Crispy]

Never did do the parallel tube thing, just 2 700 ohm resistors in parallel to a single 6L6, which should work fine, but it turns out, if they're touching one another, or even just too close, that can cause problems you might not expect!

[AZJimC]

Sorry,
I'm a dummy, I completely overlooked the whole SINGLE ENDED thing.... my bad.... carry on....

[Crispy]

Sorry,
I'm a dummy, I completely overlooked the whole SINGLE ENDED thing.... my bad.... carry on....

Don't be so hard on yourself. If you've actually built an amp that works right, you've got me beat! Actually, this one worked fine as a 5 watter, with the 6V6, but wasn't quite loud enough for gigging. I've been "Tweaking and repairing" ever since!

[shooter]

I've been "Tweaking and repairing" ever since!

Then you are on the right road, addiction can be a good thing in the right hands

[Crispy]

Moving the resistors further apart made no further improvement. Thinking by making an approximately 12w amp out of a 5w amp may have overcome the NFB loop's limitations, in another stroke of genius brought on by not having the proper parts at hand, I put a 33k resistor in parallel with the 22k NFB resistor, for a measured value of 13.84k in circuit. Didn't solve the problem, but it did improve the tone. Think I'm going back to a slightly cooler bias next, something's just not liking it at 25w plate dissipation.

[Crispy]

Well, further experimentation shows I can set the volume at 0, and the tone at max, without inducing this problem. Bringing the volume up just a little, brings it on quite quickly , say 1 or 2, if there were actually numbers there. Also works the same in reverse, with the tone at 0, I can max the volume. Bringing the tone knob up just a little is exactly the same as the other way around. This leads me to believe that the fault lies in the preamp section, since I am obviously not stressing the power tube with the volume on 0 and no signal applied. Tomorrow, just for the hell of it, I'll calculate the plate dissipation on V1b. I hope to find I've exceeded the 1W max dissipation for a 12ax7, and need a 12at7 instead, with a gain factor of 60 and mad PD of 2.5w, and twice the current rating of a 12ax7. At least, I HOPE that's what it is! Even if that's not the case, the lower gain tube may prevent whatever is oscillating from reaching the threshold where that event takes place.

[John]

If you haven't tried it yet, pop your 6V6 back in. I don't think you'll burn it up too quick with the 300R there. I only say this because I've had problems with the 6P3S tubes. They sound harsh to me, and 1 just stopped making noise and started sucking down voltages everywhere.


If your 6v6 doesn't exhibit the symptoms, try finding a 5881 - I don't think Doug sells them right now - or a 6P3S-E. The -E tubes are excellent.


I don't really see how your power tube can be causing these symptoms, but it's easy enough to try a different one.

[shooter]

volume at 0, and the tone at max

might be a long shot but you might have a leaky/bad coupling cap in the TS/Vol area passing DC volts

[Crispy]

If you haven't tried it yet, pop your 6V6 back in. I don't think you'll burn it up too quick with the 300R there. I only say this because I've had problems with the 6P3S tubes. They sound harsh to me, and 1 just stopped making noise and started sucking down voltages everywhere.


If your 6v6 doesn't exhibit the symptoms, try finding a 5881 - I don't think Doug sells them right now - or a 6P3S-E. The -E tubes are excellent.


I don't really see how your power tube can be causing these symptoms, but it's easy enough to try a different one.

Well, that didn't fix it. It did remind me how much warmer the 6V6 sonds though. I don't know what a 63ps tube is, and I'm pretty sure I'm not running one. It's a 6L6GC, Sovtek. Got a JJ, too, but the JJ sounds like ass. I raised the value on all 3 grid stoppers, which made an improvement in the tone with each one, but didn't solve the problem. It's getting late in the game (got a gig tonight, and I really wanna use this amp) so the short term fix is to undo the last thing I did before it started acting up, and put the last cathode resistor I had back in, going back to a colder bias, which sounded pretty good, probably to most people's ears, but I know it CAN be better. Just not tonight.

[Crispy]

Biasing it back where it was didn,t fix the problem, just made it sound bad, so I put it back where it was (82%,24.7W). All the changes I made didn't affect where on the knobs the problem starts, but did tame it down so it's not a loud, horrendous noise, but it does let you know something's not right. But I have a working solution... don't turn the knobs up that far! Sounds pretty darn good with volume and tone both at about 2 o'clock, and if I use the high gain input, I get back the treble I lost. I'm gonna see if it will make it through 2 gigs in a row without crashing on me, then re-visit this thing next week.

[PRR]

> when I get either the Volume or Tone knob near "10", output drops to next to nothing, and I get a nasty hum and not much else

It is oscillating. Probably super-sonic.

It is a single-ended amp. Leaving it oscillating (probably) won't do it any harm. (Yes, a simple tube oscillator's plate current usually does DROP as it comes into oscillation.)

With it whining its supersonic song, chop-stick and see if you can find the key wire which is too close to some other key wire. My semi-Champ was amazingly critical in the tone-control wiring. I could push one way and increase, the other way and stop. Actually it was so critical that putting a pencil-tip between two wires made it worse (the carbon helps the signal sneak-hop from one wire to the other). Do be careful, pencil-carbon conducts well enough to upset tube circuits if you get right across two leads. (I suppose I coulda put tape on the pencil.)

Alternative is brute-force high-cutting. Since you can't hear it, you don't need the amplifier to be amplifying it. 500pFd from 12AX7 plate to ground or cathode reduces everything above the guitar band. 5K-50K series grid resistors on a few tubes will also shave the supersonic response.

[Crispy]

I used it for 2 gigs this weekend, with the tone set just below where the oscillation starts, and it sounded pretty damn good! I did miss the "presence" I get with the tone knob maxed out. I upped the value of all the grid-stoppers and filtered out 98% of the audible part of the oscillation, trying the amp after each swap, to make sure it didn't hurt my tone any. It actually sounded BETTER with each swap. Just used what I had on hand... 2.7K for the grid of the 6L6,(from 1.5K), changed 470 to 1K on the screen, changed the 68K's on the input to 100K, which made NO difference that I could tell. Before I did all this, I put the cathode resistor back to what it was before the oscillation started, and it still oscillated. I didn't like the colder bias, so I put the 2 700's in parallel back in. Now it occurs to me the oscillation just may be from all the wires I moved around in order to get access to the cathode resistor spot. I'm gonna try moving those around and see what happens, but I've also decided to get a 500K trim pot and wire it in as a cathode resistor, with the test points, so instead of trial and error swapping resistors, I can just dial it in to that magic 90% number!

[John]

but I've also decided to get a 500K trim pot and wire it in as a cathode resistor,

That won't work well, IMO, too high a value. Try finding a 500 - 1K ohm pot, and it should be rated for a couple watts at least, even for a short trial period. Otherwise I'm afraid your 1/2 watt pot will melt. This is only my opinion, others have more experience with that than I do.

[shooter]

+1 John, I would use a 500ohm 5watt rheostat for test.

You'll find if you do enough SE cathode biasing is sorta a trade-off between B+, IpMax and G1 drive.