Analog Delay V3205 (AD-3208) decays too fast?

[Kevin Mitchell]

Hey guys. Last night I've populated an AD-3208 pcb I've prepared some time ago. I wanted to test out some parts as well as familiarize myself with analog delays a bit more. I was expecting some performance downgrades using the cool audio BBDs so there's not much surprised this is happening.

At first - using the original BOM I was only hearing one repeat of the signal. So I did the couple of resistor swaps for delay level and repeat amount. This did bring out one or so more repeats but they seem to decay so fast that I can't get any traditional delay feedback or any wild noises - which I very much desire. In conclusion - I'm only hearing 2 or so repeats even when the 50k "repeats" pot is cranked.

I'm in doubt of there being an error due to soldering shorts or wiring mistakes. All of the controls seem to work as they should also. I've adjusted the trimmers for the cleanest signal passage with close to no clock noise.

There are a couple mods I'd like to do once I improve the current circuit. I have many 512 BBDs I'd like to store away in a safe place. I've been told that if I select ones with the same production stamp (underbelly code) the biasing should be close to the same - meaning I wouldn't need a heap of trimmers to adjust. I will also replace the clock LFO with one with modulation - like a memory man.

Any tips on improving/lengthening the decay of the repeating signal?
Info on the project is found here AD-3208 Project Page

[armdnrdy]

I'm in doubt of there being an error due to soldering shorts or wiring mistakes. All of the controls seem to work as they should also.

Start doubting! 

The "Repeats" control is not working as it should.

I built this circuit with a couple of Cool Audio 3205s for a friend.
The Repeats control worked just as on any other delay.

Check for solder bridges, cold solder joints, or wrong value components.

[Scruffie]

Post the compander voltages, SA vs NE companders bias up differently, I suspect it's mis-biased.

[Kevin Mitchell]

Ah Larry. The man, myth, the legend. I've checked with an audio probe and the continuity on my multimeter. Though I wont deny it as a possibility! The repeat pot seems to work though not as strong as I'd admire.

Very well, Scruffie.
I actually have the NE571 from you in there  I was meaning to post the dual opamp and compander voltages this morning. Was running late    I'll do that this evening after the daily grind.

What should I fix up to compensate for using a NE571 instead of the SA571 as per the BOM? Didn't read anything concerning this parts alternative needing adjusting in the documentation.

I'll put up a quick clip I did before heading out. Taking a bit to upload.

[Scruffie]

The two 22k resistors from pin 10 & 12 of the compander set the compression side bias (and should both remain equal value when adjusting) at pin 10 and the 22k from pin 5 to ground sets the expansion side bias on pins 6 & 7, obviously depending on if the voltage on the output is high or low will dictate which direction they need adjusting.

There is also the fact that this has no feedback trimmer and just a fixed 100k resistor limiting feedback entering at pin 12 of the compander so depending on the exact nature of the issue, that might just need reducing but the cut off decay to me suggests the signal is getting lopped up by poor bias.

[Kevin Mitchell]

Thank you! To note - using the audio probe I've noticed the signal (wet only) vanishes after the 100k but is present at both sides of the 1uf before it. I will certainly work on that tonight. I'll socket those biasing resistors and see what happens.

Here's a short clip from this morning;

[Kevin Mitchell]

Before I think about swapping more resistor values - here are the voltages on the NE571

1-. 91          16- 1.04
2- 1.81        15- 1.81
3- 1.81        14- 1.81
4- 0             13- 9.22
5- 1.81        12- 1.81
6- 4.7          11- 1.81
7- 4.7          10- 4.49
8- 1.81          9- 1.81

I thought 11 and 12 appeared as a short but it's not. 11 gets input and 12 has no signal as I've mentioned yesterday

How does it look to you, Scruffie? Thanks for taking the time.

[Scruffie]

That looks fine and after listening to your demo the problem I thought you had isn't what you have, I thought the notes were getting cut off but sounds to me like you just need to reduce the feedback resistor (100k) going in to pin 12 of the compander to get more repeats.

[Kevin Mitchell]

I've already done the mod before the video which I reduced the 100k on pin 12 to 22k for more repeats. Before the mod is when I would only get one or two repeats at most. I feel each regeneration of a repeat is decaying too quick - or quicker than I'd favor.

Shall I try I lower value?

Was I wrong to think the AD3208 could continuously loop when the repeat control is cranked?

[Scruffie]

No reason it shouldn't, post your full voltages and a photo of the board so fresh eyes can check your values.

[Kevin Mitchell]

Here's the full image of the board

It is very unlike myself to use incorrect values though I'd never deny the possibility.
Just in case you catch it, I had to use a few scrapped box caps. But those aren't the problem.

Voltages;

V3102
1- 8.22      8- 7.73
2- 4.38      7- 4.36
3- .58        6- 6.38
4- 4.38      5- 4.42

V3205 (1)
1- 0           8- 7.73
2- 4.38      7- 5.66
3- 6.01      6- 4.38
4- 6.02      5- 8.22

V3205 (2)
1- 0           8- 7.73
2- 4.38      7- 5.46
3- 5.56      6- 4.38
4- 5.57      5- 8.22

TL072
1- 4.61      8- 9.22
2- 4.61      7- 4.62
3- 4.61      6- 4.62
4- 0           5- 4.19

No shorts spotted, all pots measure good. Hope you have an idea!

[armdnrdy]

Kevin,
Post a shot of the trace side.

[Kevin Mitchell]

It's no work of art... lol here ya go.
Trace Side Image

Let me know if you see anything suspicious!

Oh, the capacitor is there to substitute a .082 - .033 and .047 in parallel.

[armdnrdy]

Hey Kevin,

I went through the component side of your board...the only thing that I saw was a 10K instead of a 22K for R32. (wet signal mixing resistor)

I looked at my drawing for reference....


I routed my own board and made a few changes...one was adding a feedback trimmer and a lower fixed resistor.
Even if the trimmer was up all of the way...the trimmer and feedback resistor's resistance wouldn't add up to 100K.
As I stated before...I built this for a friend some years back so..I don't have access to it to take a measurement.
One thing of note:
The Boss DM-2 uses a 22K feedback resistor, and the AD80 uses a 10K in series with a 20K trimmer.
The AD3208 is based on both of these delays.
I would try lowering the value to 47K...see what you have.

[armdnrdy]

I just noticed reply #8 where you stated that you lowered the 100K feedback resistor to 22K which didn't solve the problem.

Maybe the compander in kaput.

[Scruffie]

Maybe the compander in kaput.

Shouldn't be, it was one of mine and had been tested in the past.

I'm pretty exhausted but later i'll try and have a look over things.

[armdnrdy]

Kevin,

Insert a sinewave into the input, and with an audio probe...

check the input and output of both BBDs for signal loss.

If there is a substantial loss....that could be causing the weakness of repeats.

The Cool Audio V3205Ds seem to differ from the MN3205s.

People (including myself) have encountered too "hot" of a signal from these "clones"
http://www.diystompboxes.com/smfforum/index.php?topic=100645.msg886212#msg886212

Maybe the batch you have produce a weak signal.

It's worth the time to scratch the BBD signal off of the list of potential problems.

[Kevin Mitchell]

Thanks for the advice, Larry.

It seems the BBDs are fine (measuring between pins 7(in) and 3,4(out 1,2)

I'll continue to explore. Any suggestions are still very welcome.

[Kirby]

Hi everyone!
Just rebuilding the GGG AD-3208 Analog delay on a bread board with V3205 BBD's and and my signal from the seems die as it enters the reconstruction filter..
I'm at a bit loss why this happens as I've built other analog delays and haven't had a issue with the LPF's.
If i route the output of the second BBD straight into the compander it gives nice delays and repeats.
As soon as it passes into the fist stage of the reconstruction filter it kills the signal to the point that it will only give a really poor degraded single repeat.
I have rebuilt this part of the circuit several times in case I've made a mistake and checked all the components and every thing appears correct. If i remove the 2n5088 transistor all together it gives about the same results. Has anyone else experienced a similar issue with this delay. I'm using a SA751 as a compander with 2 x MN3205's with MN3102 clock.
Please help I think i'm starting to go a bit stir crazy trying to understand why it wont pass though the last LPF before the expandor!

[j_flanders]

I'm all for necro bumps if it adds to the original question.
But you post your problem at the end of another unrelated thread.
Nobody checks the date on the initial post, they start reading, look for schematics, think about possible causes, potential solutions etc. etc. and then they reach your post, realizing all their effort was for nothing.
This gives necro bumps a bad name.
Why didn't you start your own thread?