misbias issue??

[yunger]

Schematic-2TA.pdf
when i add a bypass capacitor to the 12at7 i get some crazy motorboating and the amp sounds like a misbiased fuzz pedal. i feel like the tube may be too hot. should i have a larger rk there? i wasn't sure exactly how to bias a transformer coupled tube properly.
thanks

[Chuck H]

The bypass cap shouldn't be increasing gain much if the output is biased properly. But an increase in gain would be a tell tale precursor to oscillation. Layout could be an issue. As could ground scheme. You seem to be biased a little hot (I figured 3.8 watts at idle for a 2.5 watt tube). I do think a larger value rk is in order. But I'll guess that the problem will persist due to layout or grounding.

[jmaf]

Also, after R6(plate load, triode 2) and before R13 you need a decoupling capacitor.

[Enzo]

C6?

[jmaf]

C6?

I hate them stinkin schematicals made just to trick me.

My OCD says the transformer is upside down and the +B should connect to the top of the drawing, not loop up like that. And that dang C6 should be on the right, right before R13.

I'd add another cap just to satisfy my pride

[yunger]

The bypass cap shouldn't be increasing gain much if the output is biased properly. But an increase in gain would be a tell tale precursor to oscillation. Layout could be an issue. As could ground scheme. You seem to be biased a little hot (I figured 3.8 watts at idle for a 2.5 watt tube). I do think a larger value rk is in order. But I'll guess that the problem will persist due to layout or grounding.

thanks chuck. sounds like you are saying it is oscillating, perhaps inaudibly, all the time....not just with rk bypassed. dangit. i was so happy when i fired it up and everything was quiet.
i grounded two paths to a single ground:
1.-input->preamp->that pesky c6->ground
2.-rk of power tubes->filament 100 ohm resistors->ground<-power diode bridge....(oh, speaker jack is grounded in place)
maybe too much is going on in the 2nd ground scheme.

[yunger]

so i just changed r12 to 820 ohm and i measure 2.24 watts. sounds nice too. perfect amount of volume from this little amp for the apartment.
i tried bypassing it with 25mf and it doesn't motorboat but is way louder and fizzy.
a scope sure would be nice to have to see if weird things are happening in the twilight zone of frequencies.

[Chuck H]

I can't see how or where your OT is mounted, but if my guess is correct it's right under the preamp close to the input. Trying to get full output and gain with a layout like this is nearly always a problem.

[Enzo]

It is probably nothing, but...

I see just a peek of the blue OT wire along the bottom row of eyelets. Looks like it comes from the OT to the eyelet next to the 100k grid resistors, and from there a wire up to the tub. WHy not run the blue wire direct to the tube? That would be less plate wiring right next to grid wires.


Google up "RF Probe." There will be numerous little articles about them. An RF probe is a very simple little circuit made from a diode and a cap and maybe a couple other parts. You can make one simply, then use it with your meter. Your meter will not respond to frequencies above audio that an oscillating amp might produce. But in the absence of a scope, the probe would allow you to detect it.

[Rhodesplyr]

Two good comments: layout is tricky, and a scope would be nice :-)

On one of my first from-scratch builds, I had to move the power transformer and re-do the grounding. I thought I had thought it out in terms of transformer field direction, etc..., but the real world proved me wrong.

Also, my Dean Markley CD-120 amp had an oscillation problem above the range of human hearing that caused fizzy sound. The source turned out to be the lead wires from the PCB to the tube control grids. I replaced two of them with shielded cable, and that cured it. Of course, I diagnosed it as oscillation with a scope. Maybe there's still time to ask Santa for one ;-)

[yunger]

yes chuck the o.t. is right above the speaker jack. unfortunately i'm going to have to live with it. every bit of space is taken up back there and would require starting from scratch. maybe i could build some sort of cage around the output w/in the chassis.

..and yes enzo that blue wire is the o.t. wired to the board. it was, annoyingly, just a nat's hair from reaching the tube socket. i might be able to drill next to the socket to make it work. btw, the "rf probe" sounds like something i'll look into until i come across a scope. thanks.
there is a scope for sale at the variety shop down the street but it is gigantic and i have nowhere to put it now.

with the power tubes' cathodes unbypassed it's impossible to tell that there even is a problem. can negative feedback of an unbypassed rk hide problems like that? i won't bypass anyway...plenty of volume for what i need as is. but being the ocd person i am, i must verify.

on a side note: i bypassed the second triode's rk and turned the 100k/100k voltage divider into a 22k/220k. i needed more signal to drive the power tubes and the 100k was cutting too much highs. sounds great now!

[Chuck H]

Good that it's sounding the way you want it too.

As Enzo mentioned, some leads can be moved around. Also, you could use copper foil on the chassis to help shield the OT. Move that input lead the other way and change it's termination point. There are SOME things that can be done if you decide it matters.

Basically, having the input of an amp near the output is the amp building equivalent of a live mic in front of the speaker. Phase relationships and specific fequency sensitivities can make it better or worse. But you have your preamp going one way, your power amp going the other, they cross over each other and the start shares space with the finish. It's sort of textbook "don't do this".

There have been amps, even high gainers, that have the OT on the input side of the chassis. It can be done. It's not common to see the preamp and power amp crossing over each other though. But if your getting performance from it now, a little re routing and shielding may be all that is needed if you decide to get finicky.

[Rhodesplyr]

with the power tubes' cathodes unbypassed it's impossible to tell that there even is a problem. can negative feedback of an unbypassed rk hide problems like that? i won't bypass anyway...plenty of volume for what i need as is. but being the ocd person i am, i must verify.

The answer is a qualified 'yes,' but I'd take issue with the word "hide" because, based on your observations, the unbypassed cathode isn't masking a problem; rather, the increased gain of a bypassed cathode is an essential part of making the circuit unstable. If you don't need the extra gain, the negative feedback of an unbypassed cathode is a very simple way to keep the circuit operating within the desired parameters. Put in a larger context, there are many amps out there with global feedback loops that would become unstable if those loops were removed, so, in that context, your case is truly not that unusual. During the "no negative feedback" craze of the last decade, I know of several DIY'ers who attempted to eliminate negative feedback from their amps and discovered that their amps had become oscillators.

One other random question: considering that NOS 12AT7s are fairly cheap and plentiful, was there a specific reason you chose a JJ ECC81 as your output tube? You certainly have a lot of choices among old stock types.

[yunger]

The answer is a qualified 'yes,' but I'd take issue with the word "hide" because, based on your observations, the unbypassed cathode isn't masking a problem; rather, the increased gain of a bypassed cathode is an essential part of making the circuit unstable. If you don't need the extra gain, the negative feedback of an unbypassed cathode is a very simple way to keep the circuit operating within the desired parameters. Put in a larger context, there are many amps out there with global feedback loops that would become unstable if those loops were removed, so, in that context, your case is truly not that unusual. During the "no negative feedback" craze of the last decade, I know of several DIY'ers who attempted to eliminate negative feedback from their amps and discovered that their amps had become oscillators.

One other random question: considering that NOS 12AT7s are fairly cheap and plentiful, was there a specific reason you chose a JJ ECC81 as your output tube? You certainly have a lot of choices among old stock types.

very informative answer. thank you. i feel better about stamping this little project a success.

i didn't realize nos 12at7's were so available. i'm gonna jump on that and see if they really do sound "better". maybe they will last longer anyway.

[yunger]

i hate to admit it but i DO have a problem. there's a nice sounding amp in there, i can hear it, but i realized that tweeking the design is pointless with the layout problem.
all i seemed to accomplish by biasing the power tube colder was reduce the output volume enough to mask the issue. bypassing rk and/or biasing hotter just reveals the problem that you guys diagnosed right away.
i hear a fizz (most noticably lower notes) each time i hit a string. cranking the volume makes it so i can just tap the body of the guitar and it fizzles away. with a bypassed rk, it just fizzes itself into motorboating.
chopsticking tube socket wires affects the motorboating rpm so to speak but doesn't eliminate. the most noticable change is touching and moving the power tube grid wire.
so you have all given me advice on how to fix the problem and i will address them all. i just wonder if there is an order to proceed, from less invasive to all out surgery on the amp? maybe the first step is to use enzo's rf probe to find the location and go from there.