Delay self-oscillation problem with Dumbleator

[Luthierwnc]

Hi All. I'm doing a bit of soldering after a long layoff. One project has been building a daughter-board for reverb and delay that sits on the amp to use in the onboard D'Lator so I can try a four-cable method without twenty extra feet of cable. The reverb works great. I also got an Empress Buffer Plus in case either end needed it. Can't say yet. I'm not using the Boost at all.

The issue is the delay. I've tried it with a TC Flashback II and an MXR Carbon Copy. Any hint of delay eventually goes into self-oscillation. On the TC unit I tried most of the delay types. Some do it more than others but to be audible is to ring. I've tried it with different send and return levels on the active loop too.

Any ideas on how to tame the brute? Thanks, Skip

[dbharris]

I would start by removing the Empress. Your send side will definitely not need another buffer. On the return, you could always use trails mode for the Hydra, which is a buffered bypass system. But honestly you shouldn't need any buffer with short cables like that.

On the d'lator, set the send for about 1/3rd up from CCW. Does adjusting the feedback knob on the delay pedal have any impact on the problem?

-Dan

[Luthierwnc]

Will do when the household is up for joyous music!

Nearing 11 am: Removing the buffer made some difference but devices at 8:00 on the level knob start feeding back. I threw a NU-X Atlantic on there since I know it pads-down to +4 dB. Same thing. I'll just put it back on the board. Thanks, sh

[rootz]

Almost sounds like feedback through the power supply or ground connections of the Dumbleator. You sure there is no link from D'lator output to input via the jacks or an accidental (solder)jumper? I once had a Witchcraft jack that wouldn't switch. It's very rare I suppose, but not impossible.

I don't think it is in any of the pedals as they'd all exhibit the same fault. That is unlikely.

[Luthierwnc]

Thanks Rootz, I'll have to take a look. Not sure when. I've attached the schematic (homebrew). The overall master is really the D-Lator recovery pot. And the clean master is bypassed in OD mode. A difference in the loop is the series/parallel pot. Honestly, I can't tell a difference. That might not be the most effective parallel loop solution. More research.

At any rate, I'm redoing the pedalboard and will report back with that first. Cheers, sh

[Stephen1966]

Don't know if this will help, but have you isolated the loop jacks from the chassis?

[bcook]

You might be getting feedback through the wet/dry pot. Perhaps try a 250k there.

[Luthierwnc]

Stephen1966; Don't remember. It's been a while since I've been in there. I'll get it downstairs and see what it looks like. You can see the pots and knobs on the left lower corner. My guess is the black wire coming from the second jack to the right is grounded on that barrier strip. FWIW, with nothing in the loop it sounds great -- and dead clean. The reverb pedal is fine too. But the trem on the Hydra and any form of delay immediately sends it into freak-out mode when I turn up the recovery output pot. Turning the drive pot up doesn't cause it. Note; I didn't put a pot on the recovery grid leak. It's just a 220k pot -- so pretty hot.

IIRC the waveforms looked pretty good all the way through on the clean side. With the OD you get what you get.

bcook; As above, when I've got it open I'll have a look. I might disable it to straight series too. In the meanwhile I'll read-up on alternative parallel circuits.

Thanks for the ideas, Skip

[Luthierwnc]

Preliminary report: Think I found the culprit. I tried to push the ground wire on the send pot out of the way for a better look and it moved too easily. Solid core wire broken inside the sleeve. All those ground were connected to the pre-amp bus bar. I'll put it through its paces later today. sh

[Colossal]

Preliminary report: Think I found the culprit. I tried to push the ground wire on the send pot out of the way for a better look and it moved too easily. Solid core wire broken inside the sleeve. All those ground were connected to the pre-amp bus bar. I'll put it through its paces later today. sh

Excellent! I like solid core, but yes, you do have to be somewhat careful with it. Glad you found the problem.

[Luthierwnc]

That was it. No ground on the cathode follower makes effects zesty! I made a couple other tweaks when I was in there. Great sounding amp and now the D'Lator is more than a final gain stage.

I do have another issue: That TC Flashback II is a pretty good little delay but it is overwhelmed by the signal level with the loop at unity. Anything past 9:00 on either the level or feed back controls and it goes into 'weird science' mode. I do have an Empress Tape Delay on the main board but it stays there. For this; a more bucket-brigade unit would probably work better. Is there a delay that comes to mind that can handle the levels a dedicated 12AX7 can throw at it? And it's first in the two-pedal chain. Failing that, would something like an Xblender be a better idea?

Thanks, as always, Skip

[Reeltarded]

Unity is attained by feeding the effect hiw much it wants, and the return gain is the make-up gain.

You are frying the input of the delay. Turn that down until it sounds.. impossibly ok.. because pedal, then make-up with return to equal UNITY.

[Luthierwnc]

Thanks Reeltarded -- been a while, By unity I meant the same sine strength at the end of the loop as going in. That's about noon on both pots. Since the loop recovery output pot is also the overall master, fiddling with that to get the loop levels right detracts from the overall performance.

The TC isn't bad but maxes at about 9 o'clock on either the feedback or level. That little sweep puts a fine line between usable and police siren. The reason I haven't ordered-up the latest new delay is I'm reasonably sure I need a more robust way to make this a parallel loop. Maybe a bigger value on the wet/cry pot. The X-blender, minimix and tru-something would let me adjust that manually. Other solutions seem to be mostly rack gear and I'd like to get this on a medium-sized pedalboard.

Cheers, sh

[dbharris]

Glad you got the grounding issues sorted out. I don't have problems using my JM amp with the TC Nova Repeater and Line 6 M5 in my loop. My send is at about 10 o'clock or maybe a little less. Return is up to about 2 o'clock. And output to taste.

What voltage divider did you use to replace the original master volume pot in the amp with? Maybe it's as simple as hardwiring it for the master to be all the way up?

[Luthierwnc]

Thanks dbharris. I forget the voltage divider but think it was 100k to ground after a 270k. I'm coming to the conclusion that I like the D-Lator better as an effect than as an active loop. The FET I could live without so I may convert that switch to a p/p pot and steal the relay to make the loop switchable. That and I don't play dirty enough to really hurt the minimal delay and reverb at the end of the chain. I'll keep fiddling because that's my hobby.

A good solution might be staring me in the face. My little home amplifier has an effects system I borrowed from leonc (a very good So. Cal amp builder in his own right). https://www.youtube.com/watch?v=NO2fqf8jk0Q

He rigged a low-tech version of Carlton's set-up but using an H&K redbox between the speaker and head rather than a microphone. I use a Radial Cablink which has a handy pad to dial in the right output to a Boss SE-50 FX processor. That splits into stereo into a little Stewart PA-50b amplifier and then into two closed-back 1x10s on either side of the wall. Works great. It never leaves the house, though. The attached drawing is a few years old and I never hooked-up the Midi Xchange. Other than that it's as shown.

Cheers, sh