AD-3208 Delay distorts w/humbucker

[WLTerry]

So I built the AD-3208 with the two V3205 Coolaudio chips and it works but the delay signal is heavily distorted. I followed Mark Hammer's suggestion to isolate the 2nd BBD and bias the 1st one alone as explained in this thread: http://www.diystompboxes.com/smfforum/index.php?topic=69615.0 I tweaked the trimpots and the signal became very clean and nice while I play my guitar with single coils (two Duncan Designed generic single pickups).

But when I switch to the bridge humbucker (a Seymour Duncan SH-6 passive high output humbucker) the signal became distorted. The funny thing is that when I struck the higher notes it distorts very little and have a decent amount of repeats, but when I struck the lower notes (5-6th string) the signal distorts heavily and produces much less repeats than the higher notes (say.. high notes: 4 repeats, low notes: 2 repeats).

Is there some way to cure this issue? tweaking the pre-emphasis network? reduce the input gain maybe?... or the compandor chip gone bad and should replace it? (Note that I using one BBD in order to bias it properly)

I'll appreciate any input. Thanks in advance.
Will

[Mark Hammer]

Either drop the value of the 10k feedback resistor in the first op-amp down to 8k2 or 6k8 or raise the value of the 1k resistor to 1k5 or 2k2.  Stock, the device has a gain of 11x in that first stage, and if you start out with a hot signal before that it may induce "misbehaviour" in the envelope tracking.  Dropping the feedback resistor from 10k to 6k8 will drop the gain from 11x to 7.8x, a 30% drop in gain.  Doing that, and raising the 1k up to 1k2 makes it a gain of 6.8x.

Alternatively, you can replace the 10k fixed resistor with a 10k trimpot  and gradually adjust down from 10k until you hit the clean spot.  You will likely need to adjust the value of the 10k feedback resistor in the output stage to bring the overall signal level up again to make the bypass/effect volume equivalent.

[WLTerry]

Thanks for replying Mark. I've tried your suggestion using a 10K pot instead of the 10K resistor and a 5K pot instead of the 1K resistor and I've tweaked both pots, but it seems to behave as a filter (it cuts some treble) instead of lowering the input signal.

Is there another way to reduce the gain without loosing highs? Because I'm totally convinced that the delayed signal distorts because my guitar have a hot signal pickup... I dropped the volume control and raiser the volume in my practice amp and the humbucker signal sounds fairly clean with its caracteristic humbucking tone. Raising the guitar volume (and of course dropping the amp vol) the delayed repeats starts to distorts.

Or I'm wrong and I should buy a new compandor chip?

Cheers
Will.

[Rocket Roll]

...I'm totally convinced that the delayed signal distorts because my guitar have a hot signal pickup...

If it's of any help, here's my AD-3208 (V3205's, NE5532 and NE571, BC550C and WIMAs) with a five-string Ibanez bass (which has an active onboard-preamp):

http://www.box.net/shared/juxxo3wkk8

Would that be the distortion you're describing? If so, then I guess it's just normal - mine EHX Deluxe Electric Mistress does the same when I hit the strings hard, even if it's an 18V model.

At least BBD-based circuits overdrive very gracefully...

[WLTerry]

Hey Rocket... yes, I guess analog delays has a little saturation in its delayed signal, BTW yours sounds good! I've been watching in the youtube for analog delays and it seems like they have a little warm overdriving clipping and yes... it's normal. But my case seems to be diffrent, the distorted delays are really annoying when I play with the humbucker, but when I switch to the single coil pickup now I got this caracteristic analog warm delay sound.

BTW... reading carefully the Scot Swartz' build notes, he refers to the 1M trimpot as a "500K" trimpot, so I switched it to 500K and the sound improved a little bit, but still not as good as I would wanted. Raising the power to 12V does nothing too. I promise tomorrow I'll upload some samples so you guys will hear what I'm talking about.

Cheers and thanks
Will

[WLTerry]

So I got a question... since I forgot some op-amp advanced stuff: What's the function of the 1st op-amp? I know it works as a sort of a non inverter (?) amp without the caps.... but if we add the caps and the 4.5V biasing (the "R" point) this network became into some other thing?

I was making some tests but no samples yet because it's too late and neighbors would kill me if I make a recording micking the amp. But I still convinced the pre-compandor input signal is amplified too much so the compandor misbehave and the BBD ourtput became distorted. Sadly, changing resistor values as Mark suggested didn't dropped the gain... so, maybe knowing what is this op-amp for, I could redesign this input stage in order to get the input signal controlled.

Any clue will be appreciated.
Will.

[Rocket Roll]

All I know is that AD-3208 is based on a Ibanez CD-10 (schematic here: http://www.schematicx.com/schematic/ibanez-cd10-delay-champ-schematic/) and on Boss DM-3 (that's DM-3, not DM-2: http://www.freeinfosociety.com/electronics/schemview.php?id=2291). Both schematics show an opamp before the compander, as the first IC that a guitar signal "sees" - right? So, would an opamp with a higher headroom (such as NE5532 or something from Burr-Brown) be more efficient and more distortion-free in that particular spot?

Btw, here's how one biases a Boss DM-2, hope it helps someone:
http://www.godiksennet.com/images/sch/DM2.jpg
http://www.godiksennet.com/images/sch/DM2PG2.jpg

[WLTerry]

Hey Rocket... thanks for posting that schematic. I totally forgot about that one, I will work on the buffer stage and add it into the ad3208 and see what happens.

Will