[sdiy] despicable delta destroys detective's day (VCF insanity)

[Magnus Danielson]

Dear Bob,

On 07/10/2011 09:10 AM, Bob Weigel wrote:
> This tops 'em all. Ok... I'm using a Delta I repaired months ago the
> other night to give it a test and avoid taking the one out of my studio.
> It crackle-pop-fizzles within a few minutes of use. I opened it and
> wiggled/tapped and it seemed to have some mechanical correlation but I
> couldn't isolate it. One of *those*....
>
> So back at the ranch..er...bench...I look and find that the signal is ok
> up past the first ota stage. Noises happen in the signal at the inv. in
> of second stage..which..means nothing since there's feedback there so we
> can't tell where it's originating from that of course w/out some deeper
> evaluation. I *assumed* it's probably just the 13600's and changed them
> both as I've done on several units with success. This time..uhuh.
>
> I re-soldered everything in the area. Still no result.
>
> Now when it's failing...it sound kinda like a resistor going open type
> effect. And indeed when i finally get it to cut in and out regularly
> enough to see what is happening... the offset voltage at pin 5/7 I think
> it is...before the buffer...is zigging from 1.1V or so offset to near
> zero when the signal gets attenuated to a noisy low level. NOW what
> resistors could affect that? Well the 10K feedback, the 10K input or
> maybe the 15K or the 220ohm ones. Replaced them all....no result! Still
> exactly the same. Usually plays a few seconds when you turn it on as you
> hear the noise gradually coming and then BONK....or it would go into a
> rapid 'switching' between working and not working ... anyway....
>
> I'm like..well..there's only one component (after changing the OTA a
> second time...) that we havent' done ..the .001 uF mylar cap which
> showed NO particular corrleation while squeezing or twisting it.
>
> I pulled it..put the diode checker on thinking it would pick up a leak
> if it was leaking since mine goes up to 3. some volts. Showed no issue.
> Anyway I put a new cap in anyway... and it is working now. I haven't
> heard it crack up once in hours now. Has anyone EVER seen a failure mode
> of this type with what appears to be a mylar cap? -Bob
>

I so recognice the frustration you had.

I've found that open up feedback loops can be an aid in fault analysis. 
That way you get a waterfall analysis which can be simpler. You can also 
observe the input offsets of op-amps easier.

Anyway, as for the fault, it could be that the mylar to pin interconnect 
is flakey... so you have quite a different capacitance and all you 
balances are way off. Consider it flakey-fluke when sitting in the 
negative feedback over an op-amp... the gain of that circuit will jump 
like mad.

Thanks for sharing this story. We all end up with some strange bug every 
now and then. I've felt pretty darn stupid myself at times. Ah well. 
That would be buissness as usual then :)

Cheers,
Magnus