"Thuddy" background noise during pick attack

[wizard333]

See original post. I can't find a way to get it to show up on the scope from a signal generator, and trying to read a scope while tremolo picking.......well ...............

[wizard333]

Lets ask it this way:

1) anyone heard this effect before and what was the issue?

2) If not (1), what things might I look for? I've already limited the bass through the preamp which was a quick/easy thing to do, while that made it better, it didn't fix it, so what might cause the effect described?

[jmaf]

Lets ask it this way:

1) anyone heard this effect before and what was the issue?

2) If not (1), what things might I look for? I've already limited the bass through the preamp which was a quick/easy thing to do, while that made it better, it didn't fix it, so what might cause the effect described?

IIRC JJ's had a problem with short pins years ago. They're also fatter (?). So on those shielded Fender sockets they'd get stuck a few mm above proper fit.

It's possible that you bought a whole lot of JJ from tube matcher rejects, it's happened before (with me it was a box of Sovtek) so it's not unseen. But it's less probable than some other cause that may be related to the JJ.

Ghost notes are usually from bad electrolytic caps.

[R.G.]

If it was blocking distortion, it would show up with a constant signal from a signal generator.

Your description seems to say (in translation) that the problem happens when you're feeding it many high amplitude picking transients. OKfine, feed it many picking transients. Modulate your signal gen by turning it off and on at a tremolo-picking rate. Or better, record tremolo picking into a computer mic input, make yourself a loop of that, and arrange to feed the canned signal back into the amp while you look at the scope. I ran into the need to get guitar signal while scoping once too often and designed myself a ringing oscillator setup with a timer to kick it every so often. It does repeated guitar-ish ringing sines, and helped me out of something similar. But a CD of recorded bass signal would do fine, probably better. Probably every amp tech and hacker ought to have such a CD.

Hmmmm... ( I can hear you all thinking about now) that would make a great tech aid product! Yep, it would. I once decided to make clean digital recordings of a number of vintage guitars for just such a creation, then my friend sold out his used-guitar shop before I could setup up the recordings. Sigh. Another fortune down the drain.

But I digress. The description raises thoughts of incipient motorboating. Motorboating is a Nyquist, gain-frequency oscillation, but with phase shifts at the low end cutoff, not at the high end. Otherwise, it acts like a squeal that rises when the gain gets too big or the treble knob's too high. A squeal can be not-quite-constant, and only happen on certain notes that excite it to ring, but it doesn't have enough gain to keep on, so the squeal dies out. Your description is suspiciously similar on the motorboating side.

[wizard333]

Thanks for the reply!

I'll try to get a recording I can re-amp, great idea.

Correct me if I'm wrong, but typical motor-boating is caused by an excess of low frequency content. I've limited that with smaller coupling caps (at least the frequencies that might be causing it due to the excessively large stock caps) but I'm still seeing the issue. If it were motor-boating, where else could I look? Possible things are coupling through the power supply?

[Leo_Gnardo]

1) anyone heard this effect before and what was the issue?

Heard it lots of times, mostly in hi gain amps played by metal heads. They loved the effect: "Sounds like a kick drum hit with every touch of the pick." Wap wap wap, wappitawappitawappitawappitawappitawappita wapwapwappita WAP! Oh well, I guess it's not for everybody.

[wizard333]

I wonder if a bad HT cap may be allowing a positive feedback loop to develop?
Try bypassing each in turn and check for any change.

Going to revisit that as well.

[wizard333]

Heard it lots of times, mostly in hi gain amps played by metal heads. They loved the effect: "Sounds like a kick drum hit with every touch of the pick." Wap wap wap, wappitawappitawappitawappitawappitawappita wapwapwappita WAP! Oh well, I guess it's not for everybody.

The gain doesn't have to be all that high for this to happen, although the effect does seem to increase as gain goes up. Might be just a volume affect there to since they are related. I know what you're describing, typically that occurs where guys use a sh*tload of gain and smiley-face their eq with lots of highs and lows, not really the case here. Boosting low end like a lot of those guys do will definitely accentuate the 'thwap' of pick attack.

[MarkusBass]

There is effects loop in the amp. Have you tried checking whether the problem is with the preamp, or the power amp? You can feed signal from another preamp into the power amp. This would at least tell you in which part of the amp there is problem.
Being frankly I was "fixing" an amp with similar problem lately. It was a strange sound on drive channel. I quickly found out that it was caused the reverb spring. I just asked the customer to turn down the reverb .

Mark

[wizard333]

Since it does not occur on the clean channel when using a drive at similar volumes/gain/tone, I'm going on the assumption it is the channel and not the power amp.

The problem occurs with the amp on the bench with no verb tank within 10', so verb isn't the issue.

[g1]

Have you tried contacting Koch?
When I did it took awhile but they did end up being helpful.

[R.G.]

Motorboating is *gain*-*phase* oscillation. That is, there is a phase shift in an amplifier which has feedback, and the amplifier forward phase shifts add up to enough degrees of shift to make negative feedback become positive; AND there is enough gain at the frequency where feedback shifts to positive to make the gain through the amp and back through the feedback path become one or greater.

What makes this hard to pin down is that (1) there is a phase shift associated with every change in frequency response and (2) with enough gain, everything oscillates.

Each coupling cap in an amp is a high-pass filter. It starts rolling off the signal passed through it at a rate of -6db/octave, but also starts phase shifting the signal. The shift is 45 degrees right at the nominal rolloff frequency, and is asymptotic to 90 degrees going lower. Two single-pole (a single cap is one...) rolloffs never get to 180 degrees of shift, so three poles are needed. But there are three coupling caps between stages in almost every amplifier these days.

Then you need gain at the frequency. Since this is gain at frequencies below the high-pass of a coupling cap, you need some path other than the tiny capacitances of every conductor to every other conductor in the universe. Very often this is either the power supply rail, the grounding setup, or the various decoupling caps. All of these are either DC or very low frequency paths for signal. A bum decoupling cap may do it, as may sharing a power supply wire or a ground wire in the wrong place.

Once you get to there, all you need is gain. I once (mis-)modified a Princeton Reverb and was horrified that I had made it motorboat. Better decoupling fixed it.

I'm not surprised that the uber-high-gain metal amps are at the edge of motorboating. The gain doesn't have to be all the way down to DC, just a total making the gain through the amp and the feedback path be greater than one. You can get to that with many time constants/coupling caps and enough gain.

Since it does not occur on the clean channel when using a drive at similar volumes/gain/tone, I'm going on the assumption it is the channel and not the power amp.

Reasonable guess, but not necessarily correct. A sneaky feature of Nyquist oscillation is that there may not be any one part that makes it happen. It's the sum of all of them. And with a lower gain clean channel, there may not be enough overall gain.

Of course, all of what I'm saying is part of that internet-century of speculation, and worth what it cost you.

[g1]

I'm going to suggest you decide on whether you want to treat this as a design flaw or a repair issue.
Scour the net and see if other people are complaining about this with this model. If not, treat it as a repair of a fault.
Changing coupling cap values etc. seems to be operating under the assumption that this is a design flaw that all these amps have.
I think it would be easier to fix a broken amp than to mod a design shortcoming.
Is this your amp? Did the problem always exist for the owner or is it something that changed?

[jmaf]

It'll be easier to diagnose once you post a recording.

[wizard333]

Most of what you're describing would be due to the design/layout if it were to occur. Having worked on these before, they all share the same PCB, and not all seem to do it, so I wouldn't think it was a design issue per se.

It occurs at well below "metal" gain levels, so it doesn't seem to be a case of "well you have so much gain, gd near anything could happen". "Classic rock" gain levels are plenty to present the problem.

AND there is enough gain at the frequency where feedback shifts to positive to make the gain through the amp and back through the feedback path become one or greater.

That definitely is the issue; there seems to be a positive feedback at a very low frequency in response to a series of fast transients. A single note doesn't really have it, so it's the transient arrival velocity that seems to make the difference here; it seems to 'build up'.

But there are three coupling caps between stages in almost every amplifier these days.

Do you mean 3+ stages each with a coupling cap? If so, on the 3 stage channel we have that. Interesting thought, since on the 4 stage channel with V2A in the picture, the effect seems less prominent.

A bum decoupling cap may do it, as may sharing a power supply wire or a ground wire in the wrong place.

Better decoupling fixed it.

Can you elaborate? If the issue were indeed cause by phase differences causing positive feedback between the coupling caps (though I can't imagine why it would on this amp vs. every other damn amp with the same basic staging), what would the solution be?