Don't Know What I'm Hearing

[boots]

Well, here's a little update for anybody that cares......

I tried most of the things you guys suggested, except for Reel's suggestion of grid stoppers on the PI - I was saving that for last.

I think I have the oscillation minimized to the point that I can live with it now. The thing that seemed to make the biggest improvement was changing the bias feed resistors from 220k to 100k, as suggested by Ken Fischer in his Trainwreck pages. Now the oscillation only amounts to a single "thump" with both bright and normal volumes dimed (before it would take off and thup thup thup until I stopped it).

I tried using Ken's Type 3 PPIMV just for kicks, and the oscillation totally went away, even with the MV pot maxed. But I didn't like the sound of the amp when using the MV. Although it sounded OK with the MV maxed, i.e., essentially out of the circuit. Just having the 1M pot in the circuit seemed to kill the oscillation though. That's what led me to try changing the bias feed resistors.

Along the way I did end up replacing all the 20 uF filter caps with new 22 uF, and the amp is generally quieter now. I also upped the B+3 node to a 47 uF, although I don't think that had any bearing on the original oscillation problem. I do think I may have damaged one of the original caps with too much heat while soldering.

I also played around with different tubes types and found that less overall gain in the system reduces the problem. Using a 12AY7 or 5751 in V1 reduces the problem, and a 12AY7 or 5751 in the PI reduced it even more. Using 6L6's instead of EL34's along with the AY7/5751 virtually eliminated the problem. I was on the verge of putting grid stops on the PI, but I think that would only be masking the symptom, not curing the problem.

I figure there should be no reason the amp shouldn't work right with 12AX7's and EL34's, because there are lots of amps that work OK with those tubes. I still think there may be some sort of lead dress issue that is causing an oscillation to start, but I guess I am happy with the amp at this point.

Anyway, thanks to everyone for their suggestions!

[Littlewyan]

When the motorboating starts can you see on the scope where in the circuit its coming from?

This amp should be able to withstand having all controls set to max. I have a JTM50 which I can dime quite happily, both channels linked and not have an issue and I use a 47K Feedback resistor on the 8Ohm tap.

The thin yellow wire coming off one of the speaker sockets is, I assume for the negative feedback, but its orange when it connects to the bottom of the PI?

Also it looks like the OT Secondaries are running right next to the PI socket, have you tried moving these about whilst the amp is 'motorboating'? I'm guessing the black wire going to one of the speaker sockets is the OT Centre Tap.

Is the 100K Resistor on V2 connected to the middle lug as well? Looks like it goes through it.

[boots]

Hi Littlewyan,

I pulled the output tubes and followed signal through the amp with the scope and never did see any motorboating up through the PI. I guess that doesn't tell me much, since it is likely a feedback/oscillation involving the output stage. The motorboating is tough to see on the scope - I'll try it again.

The yellow wire from the speaker jack is for feedback, and goes to the 27k resistor on the left end of the board. For whatever reason, I used an orange sire from the other end of the resistor, going to the 6.8k/0.1 uF a little further to the right. Maybe I need a bigger value for the feedback resistor?

I did move the OT secondaries away from where they are in the photo - I moved them to the left, and they now cross under where the two boards meet. It never made any difference with the motorboating. But it looks better.

The black wire going to the speaker jacks is a ground, and it connects to the chassis nearby, at the same point as the B+2 & B+3 grounds. I originally had the speaker grounds floating, but grounded them to this point because I had a lot of noise and problems with the feedback.

The OT center tap is actually connected to the upper left hand corner of the small board, next to the sloppy looking red wires.

Yes, the 100k on V2 is routed through the center post of the socket.

I appreciate you looking this over. I am thinking about re-configuring the grounds. Also, I do think I need to take a closer look with the scope and try to figure out exactly where the oscillation is occurring. It will probably be another weekend or two before that happens....

[Littlewyan]

I wonder if it is the feedback loop causing it, although you wouldn't normally get motorboating from that.

Why did you use 6.8K for the PI Tail just out of interest? Also could you make a recording of the motorboating?

[Structo]

Sometimes faulty coupling caps can cause that motorboat sound.

Check your caps in the preamp to see if any are leaky or out of spec.

[boots]

I used a 6.8k tail resistor because I was following the "Modernized 5F6A" layout I found in one of my searches. It was posted on TAG by Martin Manning I believe.

Tom, I have been suspicious of the coupling caps because they are old Micamold Tropicaps I salvaged from an organ. I tested them all for leakage using 300 volts and they tested OK (300 volts was all I had available at the time). I will be replacing them all with Jupiters or something similar eventually because I'm suspicious of them, and I know the Jupiters will sound good.

I will try to get a video of the oscillating. I'm still not sure it is actually "motorboating". I've watched videos of motorboating and I know what classic motorboating sounds like, but this is much slower - more of a helicopter sound, about a 2 Hz oscillation. THUP THUP THUP THUP

[gingertube]

In your photo - are the caps on the left the main B+ supply.
3 off 10uF/150V caps?
If so then they need replacing with 450V rated caps. the bottom one looks a bit like it has a bulged end.
If this is the case it would certainly account for your motorboating problem.

If they are something else then ignore this.

Cheers,
Ian

Edit:
P.S. Just re-read above, I think you can ignore this.

[boots]

Thanks Ian, no those 150 volt caps are filters for the bias supply and the filament lift circuit. Turns out one of them was in backwards (I should know better) and it did blow the end out of it. It has been replaced - actually everything worked OK without that cap in the circuit at all.

[boots]

Well, I actually just made some good progress!

I replaced all those old green Micamold Tropicaps with new Mallory 150's, and the problem is essentially gone. I get just the slightest hint of an oscillation trying to start when both volumes are cranked and the bass control is all the way up, and I hit a big chord. It flutters a couple times and stops on its own. Backing off the bass control just a tad eliminates the oscillation altogether.

I can live with it now. Thanks to everyone for all the suggestions. I've learned a lot from this amp.

And it really does sound incredible! I scored a couple used EL34's - one is a GE that looks to be made by Mullard and the other looks like an Amperex. They are fairly well matched. I have Mullard 12AX7's and a Mullard GZ34. I need to get some proper speakers, but it sounds amazing through the 12" Peavey Black Widow I use as a shop speaker.

Thanks again everyone!

[Structo]

Good!

Glad you got it sorted.

[boots]

I'm dredging up this old thread because I think I finally figured out the problem. At long last. I think.

This build had a number of problems that I worked out one by one, but the one thing I couldn't cure was a parasitic oscillation that occurred when the amp was cranked up really loud.

I had put it on the back burner for a while and it was still bothering me every time I looked at the amp, so I dove back into it recently. Of course, it was something fairly simple.

When I drilled the chassis, I screwed up and forgot to drill a hole for the presence control. So I mounted the presence pot on the back panel. Turned out it was a bad place to mount it. I finally said screw the presence control and hard wired it at "full" presence. Cured the problem. I always ran the presence mostly wide open anyway.

Parasitic oscillations are a real pain. I finally found the HF oscillation with the O'scope and it looked to be about 73 kHz. I don't really understand what was causing it, but I'm glad it's gone!

So that's one thing off the list of things that have been bothering me!

[Stevem]

I never build a amp with more than two gain stages on one filter node , and mod many that do!

[boots]

I have read that the Bassman was considered to have insufficient filtering, but I was trying not to stray too far from the original design. However, at this point, I think I am more concerned with improving the amp to the point where I feel good about it.

Not that the Bassman needs improvement.