Long standing issues with a V3205 BBD and compander

[Kevin Mitchell]

I've been having issues with a BBD for a while now (almost 2 years!) and have been reluctant to toss it out hoping the problems was elsewhere in the circuit. They were given to me in a trade and out of 3 circuits I've tried them in I have not had any success.

First was the AD3208. No luck, scrapped the board.
Then was a DM-2 clone. It worked intermittent and I could not figure out what was causing the issue. The board was benched and eventually scrapped as well
Now they're in a Aquaboy Deluxe board - first try using a board I didn't homebrew.

One of the two BBDs are biasing and passing a delayed signal but the 2nd BBD isn't passing signal at all. A few hours ago it did though I couldn't tell if it was delayed. But now it's not passing anything.

When it was passing a signal it went to the compander and disappeared again (the signal was very, very low) so I inspected the area, solder joints and part values. When I plugged it back in there wasn't any signal coming out of the 2nd BBD at all!

At this point I feel safe saying it's a bad device and could have been DOA. But why intermittent? Is there any hope for this thing to come to life?
Thoughts? Suggestions?

[Mark Hammer]

BBDs have traditionally been considered static-sensitive.  AFAIK, they're not super-sensitive to static, but they can be damaged by static.  My own experience is that not every retailer treats them as static sensitive.

[Kevin Mitchell]

It could very well be a bad device. But I should get another issue out of the way first. Sadly it's not the first time I've been stumped by this.

I subbed in a V3207 to amuse myself and complete the circuit to pass the delayed signal. Well... I'm having issues with the compander. Tried both v571 and NE571 and almost no delayed signal coming through. I swear I had a similar issue with my DM-2 build and I couldn't get around it. I've already inspected everything and can't seem to find why it would squash the signal so badly.

My MN3005 build works wonderfully. The cheaper build seems to be a nightmare.

What's going on here?

[Mark Hammer]

What's going on?  I suspect the supernatural forces that impede my own successful builds have taken up residence at your bench....for the moment, before they return to mine.

[Kevin Mitchell]

I removed the 3207, lost output from the good 3205, couldn't rebias it, put in the "bad" BBD and got that working.

I'm losing my damn mind. You keep your bad juju for yourself.

[Kevin Mitchell]

I'll have to keep looking for errors. A variable is changing and each BBD has it's own preference to decide to work. I'm at a loss.

Signs are showing it might be a connection at the compander but I'm not sure what would cause this madness. One v3205 will bias and pass signal but not the other (and it varies which one will decide to work!) as well as the compander not passing the delayed signal.

Hopefully this all ties into one problem and not a big mess of them.

[bean]

Some of the early batches of v3205 were definitely problematic (this was a few years ago). Don't know the current quality but I do remember Steve mentioning that smallbear had screened theirs at one point as a measure of quality control. Or, that could be a fabricated memory...but I do remember something about this.

If the expander portion is refusing to pass signal first I would look carefully at the traces and your soldering. Do some continuity checks and check your voltages. Verify that you have used 10k in both R13 and R14. If all that checks out, try removing the 20k resistor (R37), or clip one lead at least and see if it passes signal.

And, also make sure the jumpers are set correctly for the v3205.

[Kevin Mitchell]

All connections and values check out, powered up, both BBDs wouldn't bias, pulled out the v3102, put it back in and now both are passing delayed signal. Please let this be it 
The part I thought was working all along was likely not seated well. Even though I did the paranoid thumb press technique to every IC as I usually do when something's awry.

Onto the compander... voltages didn't look unusual. I'll dig in again and post my findings. I'll try to remove the 20k resistor and see what happens. I know it's not the compander itself. It's something I've yet to find or calibrate. There is signal at the output (pins 10 & 11) but it's very low and doesn't really come through the mixer.

I had this same problem with the AD3208. I started a thread that I couldn't get many repeats out of it and was never resolved. I narrowed it down to the compander section but after replacing it with a new one without success I gave up on it. I'm actually using the ICs from that build.

EDIT:
Removed R37 and nothing has changed.

[Kevin Mitchell]

Compander voltages;
1- 1.1v     16- .95v
2- 1.8v     15- 1.8v
3- 1.8v     14- 1.8v
4- 0v        13- 8.9v
5- 1.8v     12- 1.8v
6- 1.8v     11- 5.13v
7- 2.9v     10- 5.13v
8- 1.8v      9- 1.8v

Comparing voltages around the web pin 7 is lower than it should be and pin 16 might be suspect as well.

R13 & R14 are indeed 10k resistors.

[Kevin Mitchell]

Thinking about it further... I'm not sure why the clock would be temperamental as it was. But hopefully that's settled. I have a feeling next time I plug it in I'll be back at square one 

Has anyone else had similar issues with a compander? As I've said I've had this problem with other builds as well.

I've read that pin 7 should be half the supply voltage. I guess I should work on that and see what happens.

[Scruffie]

I shipped you those, I tested them before shipping but if it's giving you crap, I'll send you another pair

Are you using an SA branded compander by the way?

[Kevin Mitchell]

No worries, Scruffie. I'm having this problem with a few v571 chips and the NE571 you had sent but they seem to work fine in other circuits. Is super strange how this problem has happened with everything I've built with the two v3205 chips. I'd like to figure out why and hope it's not some terrible irony like a solder error.

I should get a few NE570 chips in the mail on Monday. I doubt it will help but figured they'd be worth having.

[Scruffie]

I brought up the SA companders as a lot of people seem to have issues with them biasing, I recently got someone to measure the internal resistors and they were off from the datasheet value of 20k, ranging from 15-25k I think he said which affects the biasing, so something worth checking.

IIRC in the past I've used 22k to bias a compander on the compressor section at 9V up to half supply.

BBD circuits can just be a pain in general, after enough of them you get used to the common problems that crop up. In my experience with a lot of BBD's though they have 3 states, working, working but very noisy or no signal passes through them and anything else is the surrounding circuitry and even then 2 of those problems can be the surrounding circuit.

[StephenGiles]

Have all rogue solder bridges been eliminated using a magnifying glass - that has been an indispensable tool for finding them!

[bean]

Your compander voltages look right for the DM-2 except pins 10 and 11 are a little higher than results I've gotten (usually a little over 3v). A 3v output on pin7 is also normal for the two 10k divider and the DM-2. You can bump those up to 22k to get 4.5v on pin7.

I haven't experienced much of a problem in actual chips but I've also been using the same batch for the last 10 years, haha. Overall, the compander part of analog delays is the most problematic area by far. It just does not do a great job as a compressor for guitar - too quick or dynamic of an attack and you can hear the compressor struggling to keep up. You can improve the response some by increasing the value of the two RECT caps but too high and it gets very flubby and counterproductive. The DMM seems to do the best job of it (from the analog delays I've built). Not sure how much that has to do with different component values or the overall headroom from a -15v supply.

Also, for the v3205 I found that a decent size series resistor on the inputs can make a big difference to reduce some noise. Try 22k up to 100k.

[Kevin Mitchell]

Have all rogue solder bridges been eliminated using a magnifying glass - that has been an indispensable tool for finding them!

The pads aren't that close together. But I guess it's worth digging out the trusty lens.

As I feared the v3205s stopped passing signal again. I've touched up solder joints, shortened wires and fit it into it's enclosure since I'll have to do that eventually. Still no luck.

I can't get the clock under 7.5khz while the controls are at the recommended settings for calibration. I've tried other v3102 chips with the same result. I'll post more voltages shortly.

Also, for the v3205 I found that a decent size series resistor on the inputs can make a big difference to reduce some noise. Try 22k up to 100k.

Do you mean R55 and R27? I followed the build instructions and used 100k and 470k. Should I try lowering these values?

[Kevin Mitchell]

Before I start manipulating things again, here are the voltages. The clock is giving it's minimum of 7.2khz and all other trimpots are at center position. Not sure if it's worth posting voltages from the transistors and oscillator circuit. The LFO seems to work as it should according to the LED indicator - unless the LED in the VTL5C/3 is blown. Idk if that would be a factor in of the known problems.


IC1; TL072
1- 4.4v     8- 8.9v
2- 4.4v     7- 4.4v
3- 4.4v     6- 4.4v
4- 0v        5- 4.4v

IC2; v571
1- 1v        16- 1v
2- 1.8v     15- 1.8v
3- 1.8v     14- 1.8v
4- 0v        13- 8.9v
5- 1.8v     12- 1.8v
6- 1.8v     11- 5v
7- 2.9v     10- 5v
8- 1.8v      9- 1.8v

IC3; v3205
1- 0v        8- 7.7v
2- 4.1v     7- 3.6v
3- 3.7v     6- 4.1v
4- 3.7v     5- 8.2v

IC4; v3205
1- 0v        8- 7.7v
2- 4.1v     7- 3.4v
3- 3.7v     6- 4.1v
4- 3.7v     5- 8.2v

IC5; v3102
1- 8.2v     8- 7.7v
2- 4.1v     7- 4v
3- 0v        6- 4.1v
4- 4.1v     5- 4.1v

[Kevin Mitchell]

Now I'm leaning towards one of the BBDs being bad again... Or both. I have no idea why I'm getting intermittent biasing.

[Kevin Mitchell]

Well I guess I'm SOL with this thing.

I'll keep checking for errors. I'm usually pretty good with debugging and luckily don't have to do it that often but this circuit and everything I've used these chips in have been a big pain in the ass with no reasonable explanation.

Thanks for reviewing my troubles. I hope no one else experiences this long standing issue. I'll chime in if and when I get this squared away.
-KM