Can I get some feedback?

[tomrasdf]

I built a dumble a few months ago. Wasn't happy with it. Buzzy, nasaly, etc. Rebuilt it, (non hrm, bm stack and pi, 100k/100k pre, 180k/120k od, highly recommend it) and after much tweaking and help from some on this board (much thanks, jelle) now have a fantastic amp. The only problem I have with it is that it doesn't like to feedback. At all. It has very nice sustain and tons of harmonics, but IF I can get any feedback going, it's a fight to make it stick around. The most bewildering part is that my first build fed-back like a champ, so I know that guitar, cab, and room have little to do with this. I still love the amp, but I keep hearing all these clips on this board and other places on the internet where players are getting great feedback form their dumble-type amps.

1. How do you de-stabalize an amp in a musical way? Is it possible? I can only imagine that internal capacitance adds to it. Where are the hot spots? Between which stages (v1a input, v2b plate)?

2. Are there any good tech articles on this anywhere on the interweb or in any books? Google feedback, guitar, and amp, and you'll find a lot, but it won't be what you're looking for.

I've done a fair amount of experimenting (including some rediculous layouts), but nothing's had any kind of sizable effect. Anyone have any insights?

[LeftyStrat]

Take everything I say with a grain of salt, because I haven't gotten around to any dumble builds yet. This is more general info I have found:

- High end: Overuse of shielded cable, or capacitive coupling of leads (especially the line to the treble control) can dull an amp and make it less likely to feedback. A lot of the harmonic bloom comes from overdriving the high end and using the natural compression of the tubes to tame it.

- Interstage modulation: Probably not the right term, but how one stage affects the others with current draw, and how the power is fed to each gain stage. Sometimes having a couple of gain stages share a PS node can affect the swirl. That swirl can help feedback (or cause nasty things).

- Compression: Overly stiff power supply or too much negative feedback can diminish compression, and compression is your friend when it comes to acoustic feedback.

The first thing I would try in any other amp would be to play with the negative feedback resistor, but I defer to the guys that have actually spent time tweaking the dumble circuit(s).

[briane]

The first thing I would try in any other amp would be to play with the negative feedback resistor, but I defer to the guys that have actually spent time tweaking the dumble circuit(s).

right on with that one. Whats your NFB resistor, and the PI ground ref resistor (the one from the the 1m pair to the end of the .1 cap in the PI)?

Also what guitar you use has a huge effect. solidbody == not so much feedback. Archtop == tons of feedback.

Then again all of my dumbles have tons of feedback. I cant even imagine how to build one without feedback (dont mistake for NFB).

[tomrasdf]

Whats your NFB resistor, and the PI ground ref resistor (the one from the the 1m pair to the end of the .1 cap in the PI)?

4.75k ground ref with 56k off the 4ohm (50 watt, bassman OT)

Guitar's a PRS cu24, for what it's worth.

Maybe up the NFB to 100K?

Interstage modulation: Probably not the right term, but how one stage affects the others with current draw, and how the power is fed to each gain stage. Sometimes having a couple of gain stages share a PS node can affect the swirl. That swirl can help feedback (or cause nasty things).

"swirl", or the concept there of, is a part of amp building that I'm not all that familear with. Any good articles out there on it either on the web or in a book? I'd love an excuse to learn more about a power supply's effect on tone.

Thanks for the replies.

[LeftyStrat]

"swirl", or the concept there of, is a part of amp building that I'm not all that familear with. Any good articles out there on it either on the web or in a book? I'd love an excuse to learn more about a power supply's effect on tone.

Wish I could find those articles too

My knowledge is limited to my own experiments. When two gain stages share a single filter node, they are going to have a certain amount of interaction, since both are dipping into the same well, so to speak.

This interaction can manifest itself in different ways, some good, some bad. The worst is motor boating. But sometimes it seems a little interaction might be a good thing, in that you get some kind of swirl or animation or livelyness.

For example, if you took any Fender blackface era amp, upped the filtering and added enough filter nodes so that every gain stage had it's own node, you probably wouldn't like the amp. It would be stiff and lifeless.

Of course, tastes vary. Tweed era amps are a little too loose for me, others like them.

But it is always interesting in modern uber amps like Dumbles and Wrecks to see how gain stages share filter nodes. Note how in a Wreck, the first two stages share a node, but the third stage has it's own node. Or how the Dumble pairs up the stages on filter nodes.

[briane]

Maybe up the NFB to 100K?

well that will adjust the negative feed back, but not neccessarily the positive feedback. What you are describing is a total lack of positive feedback (AKA feedback).

I like 100k, use it regularly. Youll get a more aggressive amp with less NFB being dumped in the circuit. NFB tends to promote phase cancelation, and smooth out an amp.

5k ground ref is pretty low, which should inspire more feedback to enter the circuit than more standard vals like 10k or 24k. You could go lower, say try a 2.2k. Personally I am of the camp that 24k is the best it can be.

that PRS is a solid body, right? less feedback solid rather than archtop. What you need is an archtop, methinketh.

[captain_rusty]

My considerably more pragmatic solution (with a Liverpool clone) is to have the amp adjusted to what I like, which means it doesn't always want to burst into feedback, so I step on my wah pedal to get it - depending on where you hold the pedal (and your body/guitar), different harmonics feed back.

Not very technical, but nice and effective

[Ears]

I'm with briane in his line of thinking.
The "feedback" you lack is due to positive FB right back through the pickups and first stage.
This suggests to me it is a gain issue rather than a local NFB issue in the power amp loop (i.e. feedback resistor) - UNLESS there is a major phase issue thoughout the amp.

[jjman]

I read someone who says flipping the speaker polarity can be key on this. The whole "positive absolute phase" thing.

[jelle]

I remember that the amp had a different layout than what I would advise and I think that there is some parasitic coupling going on that is preventing feedback.

My suggestion is to look at the pics in the dumble files section and try to adapt the layout to what you see there. Remember that the distances between components and parts of CKT's are here for a reason.

Also I agree about the NFB remarks. I personally like amps with loose NFB circuits a lot. EDIT: your NFB is loose to me already. I'd keep it like this but 100K is an option too.

I hope this helps.

[tomrasdf]

I remember that the amp had a different layout than what I would advise and I think that there is some parasitic coupling going on that is preventing feedback.

Yeah, parts of the layout were a bit of an experiment. Tonecaps are mounted almost on the pots (gerald weber style). The PI plate resistors are mounted right next to the power tubes, coupling caps go right to the grid resistors (almost no grid wire, but longer plate wires which run from the PI under the ps board to the PI plare resistors), and the bias pot is mounted right between the tubes, 124 style. Other than that, all is pretty standard.

My purpose in starting this thread was to see if anyone knew where the hot sports were that could make/break feedback in an amp via phase issues. If it is my layout, I'm sure in the next year, I'll get inspired and either rebuild this one or build another and I can fix the problem then. If it's something else, I'd love to be able to fix it on this amp.

Thanks for the insights, by the way. Next time I get a chance, I'll throw in a 100K and see where it leaves me.