Ampeg dog whistle

[Prairie Dawg]

I've got a problem that has stumped me so far. It's an oldish Ampeg B12X, which is a portaflex guitar amp, all octal tubes, and strange reverb setup but nevermind about that. Here's what it does.

B12X dog whistle.AVI - YouTube
I cleaned up a pretty messy board and restored it back to stock, and have tried a number of things including chopstick-ing everything, paralleling all the electrolytic sections with a known good cap, installing a couple of .002 caps from grid to ground on the power tubes and a .0005 uf cap that was shown on the schematic on one side of the PI tube but was missing on the chassis, and tried a .1 cap to ground on all the pins of the preamp tubes, the vibrato, and the PI.

The only luck I have had so far is changing the caps that were on the input jacks to yellow polys, and that shifted the frequency of the #1 channel up beyound the range of audibility but I can still tell it's there.

As I surveyed this I noticed that the volume and tone setup on both channels had been completely redone which suggested to me that this has been a persistent problem with this amp and other people have tried, failed and passed it on. Otherwise, why rip all that stuff out?

Any suggestions are welcome.

[52 Bill]

What about lead dress issues? I once had one where the only way to fix it was to remove the signal from the pc board trace and run a shielded wire between circuit points. I know Ampeg used a lot of shielded cables in those amps. Any unconnected grounds?

[km6xz]

What does the oscillation waveform look like, what frequency and do any controls effect the amplitude and frequency of the oscillation? Those answers will give a clue as to the paths for feedback involved.

[Prairie Dawg]

I wouldn't know what a good or bad oscillation wave form would resemble. Volume controls will affect the amplitude of oscillation-doesnt come on until about 70 per cent volume. The treble controls affect the frequency. I shifted the frequency up beyond audible range by changing out the ceramic disc caps someone had recently placed in the input jacks of one channel. From looking at it I suspect someone had chased this problem and gave up on it judging by all new parts (tone caps) associated with the tone and volume controls. The vid I posted is a fairly close representation of what this amp is prone to do. I can do another one.

What I'm aiming for is something that'll change the parameters of the oscillation because I've found that tracking these things down is helped by getting them dialed up to the very edge of oscillation and then probing with a dowel. In fact I had this amp working like a theramin the other night and could induce oscillation and pitch shifts by waving my hands over it. I haven't had a lot of luck with probing so far but I have found a few things that aren't connected to it. I've also found that the only dumb question is the one you or I didn't ask.

All ideas/suggestions are appreciated.

[Chuck H]

Looking at the schem I see that the circuit is very particular about a couple of issues. Like having a phase correction circuit feeding the PI!?! the "echo" circuit is made to feed into a universal speaker!?! Something I wouldn't have tried. My point is...

This is a circit that was clearly made to work "as it was designed". There are some definite no no's going on with the design that the manufacturer made work. This would need to be done by trial and error. Ground scheme, layout and load values will be critical to avoid oscillation such as you have. If you can't exactly replicate an original (which may be impossible unless you have access to a working unit) you will have a hard time recreating working conditions for this circuit.

JM2C

[Prairie Dawg]

Oh. Lucky me. My karma must have taken a bad hit-wonder who I shat on in a past life?.
Access to another working unit is unlikely as this is the only one of these I've ever seen in the flesh. Maybe Ampeg and Danelectro and Magnatone were in some sort of Nobel prize for weirdness and unnecessary complexity contest.

The echo circuit is like a stand alone, single ended reverb amp with its own separate external speaker. Cute little thing, covered in Ampeg herringbone. Fortunately it has nothing to do with this mess-I don't think.

The only bright spot is that the problem, whatever it is, is located in a relatively small piece of real estate and it's not outside in the snow.

[Steve Conner]

Is the whistle affected by what's plugged into the input jacks? Are they shorting jacks? Should they be?

Dog whistle type noises usually come from an unwanted feedback loop formed by stray capacitances between wires. I'm sure it can be fixed by adding some shielded wire (which reduces the capacitive coupling) or some grid stopper resistors or tiny ceramic caps from grid to plate of a tube or two (which reduce the available high frequency gain)

The question is where to put them. Usually it's the first stage or the reverb recovery: the most sensitive places in the system. Poking the wires around with a chopstick is as good a way as any to track it.

I tend to make any long run of grid wire out of shielded cable.

[Prairie Dawg]

Whether a guitar is plugged in or not makes no difference. It's in both channels so the thought was it was downstream from them. Lunch and then back to the laboratory to do more looking. Judging from the look of things this was one that somebody gave up on. I know all three guys in this area who've owned it.

I've got plenty to think about. I was over at the music shop where I get a lot of work from and the manager says "Come check this out." He takes me upstairs and there was a tweed bassman, a tweed deluxe that looks like it was dredged up from the bottom of a river, and a tweed Vibrolux. All from the same stash, and the negotiations are underway. It looks as if I'll have my nose in some real fun stuff pretty soon.

[Gtr_tech]

You really need a scope to see where the oscillation starts in order to correct the problem. Its pretty extreme....IE *well* within the audio range so there's major lead dress issues here. Start moving wires around while the thing is sitting there howling. You'll find at least one that affects it. Start there.....and keep grid leads away from plate leads.

[Chuck H]

Have you considered that, since this was a "project" amp, the OT or PI leads may have ended up swapped around at some point? Try lifting the NFB loop and see if the howling stops.