ad 3208...again! Mark? Scott?.

[swt]

Hi guys...right now i'm at the edge of insanity with this pedal!. It works, but the repetitions goes into some kind of feedback that's really annoying. I'm stucked. How can i solve this?. Where to look at?. Have changed the 571 for a 570, but still the same. Any ideas?. I can make an mp3 to send by mail to show you what it does. It's completely different to my ibanez ad99 or my deluxe memory man. i'll really apreciate your help. Thanks a lot in advance!!!

[mattv]

Does this still happen with the Repeats knob turned less than half-way down?

[swt]

With the repeats pot halfway, i have pretty normal delay but the repetitions feedback as in a distortion feedbacker, although they won't get into self oscillation. In my other analog delays, if the oscillation occurs, it's on the same pitch as the first repetition. In this case it goes thinner and saturated one after another. Clues?

[Mark Hammer]

I'm stumped.  One occasional suggestion is to use a diode clipper to "clamp" the feedback signal below a given level, but in this case that is already taken care of by the compander.

One trouble-shooting suggestion is to verify that the compander is working properly.  If you can isolate and listen to or scope the signal coming off pin 10 of the 570/571, that would be good.  Ideally, it should be squeezed senseless.  If there is anything remotely resembling dynamics in that output, then it may not be compressing properly, which may be responsible for runaway feedback.

[Scott Swartz]

If there was an issue with the post delay filter that was distorting the signal such as a bias issue caused by wrong value resistors, bad transistor, etc that could cause this behavior.

Post the voltages for these trannies.

Obviously even better would be if you could inject a sine wave signal and use a scope to determine where the distortion is occuring.

I am assuming that when you state "saturated" the signal is distorted.

[Jason M.]

How did you set the trimpots?
By ear? Scope?

I think the problem could be that the bias or the clock cancel trimpots are not set properly. With two BBDs in series, proper calibration is very important.
If the bias is on the outside range, you could get a slightly distorted sound, which will become more noticeable with each repeat. If it is clock noise, it would be multiplied with each repeat.

If you don't have a scope, use a signal probe at the input of the second BBD to adjust the first BBD's trimpots. Listen to the signal at the wiper of the second BBD's clock cancel trimpot to calibrate the second BBD.

Is this problem only happening on the longest delay settings?
The 1M trimpot that sets the maximum delay allows for some long delay times at the expense of increasing clock noise.
I am using two MN3205s and I have the max delay set for around 450-500ms. Still sounds quite clean.

Don't give up, you are almost there!

[swt]

After a lot of analysis, i get to the point that what i have is feedback noise from bad pcb, because i did it on perf board. The trannies are bc550b, the transistor going out of the compressor is base at 3.09, emitter at 2.50, the first one from the last delay chip is b=5.83, e=5.24. And the next one, goinf into the expander is b=5.20, e=4.60. As Scott, i also thought the problem could come form the filtering, or maybe one transistor not working, so i disoldered the caps going to the emiters, and the problem seem to dissapear. This is when i think it's noise coming from bad coupling. If i put everything as in the schem, and turn the repeats pot up all the way, it beguins to oscillate by itself, with some noise too.
Do you suggest to make a new pcb as in GGG, or maybe those voltages look suspicious to you?. Thanks a lot for helping me with this one. You are all so kind!!

[Maneco]

one thing that drove me crazy sometime ago was  companding chips...i had ordered some ne571 from jameco,and they sent me a mix of ne571 and sa571..."they are they same"you will say...well...all of the ne571 gave me troubles,but the sa571 were all ok...now i only buy sa571...and the problem was with the feedback...

Maneco

[swt]

mmmmmmm!!! I'm using an ne571 and an ne570, that i ripped from older delays. Maybe you're right maneco, and the feedback comes from these ics. I'll try to get a sa571....but it's difficult around here. Thanks for th sugestion!

[Scott Swartz]

The perfboard construction could cause all sort of hard to trace problems.

PCB layouts can oscillate also of course but the layout at GGG has been thoroughly tested  prior to its release.

For the modest cost of buying one from JD, I think its well worth it, or etch one yourself.