DOD 680 Troubleshoot

[Bassisst90]

Hey guys. I've been posting every now and then about this old DOD 680 delay that I have that I've been trying to fix. I eventually made a SAD4096 retrofit using a v3205 chip instead of a MN3205. I got it to work once using a breadboard, then when I put the components together on a vero board, it seems to have stopped working. I would like to set up a circuit to test the v3205 and make sure that it hasn't burned out or anything, but the application circuit in the datasheet seems way too complicated. I was wondering if anyone knew a quicker way to test it?

The schematics I used were these:
https://www.dropbox.com/sh/2eheu6029f2sv7x/AAD5dK8mAFhqQp86xQnwEieCa?dl=0

[Bassisst90]

Does no one know a quick way to test a v3205?

[Fender3D]

The quickest method is to find another pedal with x3205 or x3208 chip in it, and swap it with your...

BBDs need a dual phase clock, an input buffer, bias circuitry and an output buffer to work....

[armillary]

Bassisst90, I used the same schematic from e-repair and successfully built a daughterboard on veroboard to replace the (working) SAD4096 chip on my DOD 680 with the MN3205 that the schematic specifies. In testing the daughterboard in the DOD 680, it provided a very clean and functional delay/echo, but it wouldn't produce the wonderful self oscillating spaceship sound like the DOD 680 will do with the SAD4096. I lost interest in the project because of that, but now I have more time to work on it. I'll take some pictures and start a new thread.   

I liked Fender3D's suggestion to test your V3205 by swapping it into a pedal that uses the V3205. The Rifftone Manta Analog Delay uses the V3205 chip, in fact it uses two and they're socketed. In auction listings on ebay there's a clear gut shot and a video demonstration. It even does the self-oscillation thing (at the end of the video.) And it's cheap. I don't own one or have any relationship to the seller.           

       

[Fender3D]

...but it wouldn't produce the wonderful self oscillating spaceship sound like the DOD 680 will do with the SAD4096...

auto-oscillation depends on the level you feed back to BBD's input,
if SAD4096 is any similar to other SAD devices, it might have a gain of 1.2dB whereas MN3xxx devices usually show an insertion loss.
Bias and Vgg affect MN3xxx output level too, so check it out

[Bassisst90]

I appreciate the info armillary. Did you happen to have a different schematic from the one I posted or a way to show how you built a daughterboard for the pedal? I've been so busy with school that I actually have time to try and fix a few things here and there now.

[armillary]

Thanks to both of you. I'll get started with more info and study.

[armillary]

The schematic that Bassisst90 and I used was very graciously provided on Mark Whiting's website:
http://www.e-repair.org/restoration3.htm

Here's the layout. I did this on paper. I've since learned DIYLC, but this was prior to that.
https://www.dropbox.com/s/83pk4w2h12m98ku/DSCN5820.JPG?dl=0

I rechecked my layout against the schematic this morning, and it still looks correct to me.   

Here are some pictures of the daughterboard created from the above layout.
https://www.dropbox.com/s/3of6kab8ycikdx4/DSCN5822.JPG?dl=0
https://www.dropbox.com/s/qq3cxikdqk2cxtb/DSCN5823.JPG?dl=0
https://www.dropbox.com/s/s6ydzjzwl2xljc3/DSCN5824.JPG?dl=0
https://www.dropbox.com/s/vy3q1jaskzyn6xe/DSCN5826.JPG?dl=0

It's been almost two years since I tested this in the DOD 680. I'll set it up again on my workbench in the next few days and report back.   

[armillary]

When verifying the daughterboard against the layout last week I realized I hadn't installed the .22 uF cap (the big green one). So I added it and took the pictures I posted last week. My point is that the cap wasn't there the last time I tested. 

This morning I opened the DOD 680, pried out the SAD4096, installed the daughterboard and took a picture:
https://www.dropbox.com/s/684veqta2sno1wu/DSCN5828.JPG?dl=0

I'll have to switch to ribbon cable, as you see there isn't enough room to put the lid back on. But without the lid, I tested what you see there with guitar and amp and it does auto-oscillate now with Remote Mix and Local Mix set at 12 o'clock and Repeat and Delay Time set at 3 o'clock. I also got some clean delay out of it. It would take more extensive testing or a side-by-side comparison to be sure, but at this point I think Mark Whiting's schematic is definitely worth a try. I'll do the layout in DIYLC and post in the next few days.     

[armillary]

Inspired by boga's thread, I measured the Vgg and Vdd on my DOD 680 daughterboard project. The Vdd from the 680 power supply is 15V, which matches the 680 schematic. After the 7808 regulator, the Vdd is 8V. After the 1K8/15K voltage divider, the Vgg going to pin 8 of the MN3205  is 7.14V. This all makes sense, although the 14/15 Vgg spec would have it at 7.46V.

If I do a second version of this, I'm thinking of substituting a 7805 regulator to bring the Vdd down to the datasheet recommended 5V. Or am I overthinking this?  (Sometimes I can't see the forest for the trees.)   

   

[Fender3D]

until you are under 10V you're ok.

Some folks out there claim the best Vdd voltage for MN32xx is 8,2V...
Never had a chance to check...

[armillary]

Thanks! In the next version I'll stay with the 7808 and maybe substitute the 1K8 with a 1K or 1K2 if I have one. Next report in a week.

[armillary]

Last week, I measured the Vdd from the DOD 680 power supply at 15V, which matches the 680 schematic. After the 7808 regulator on the daughterboard, the Vdd is 8V, which will be fine for the MN3205.

This week I looked at the two clock signal voltages supplied by the DOD 680 for the SAD4096. They are nice square waves of minimum 0V and maximum 15V. Of course the MN3205 spec sheet expects the clock voltages to be the same as the Vdd, and in no case above 11V.

The schematic (above) I've been working from assumes an original SAD4096 Vdd of 12V, and most likely clocking voltages the same. It uses the 7808 (and voltage divider) to drop the 12V Vdd to 8V. That works.  But what about the clock voltages? Should I use a voltage divider to try to drop 15V pulses (in the low KHz to low MHz range) to 8V?  I'm doubtful. If there's a better way, what would it be? I'm learning a lot here and appreciate any suggestions, as I don't have an EE background. Thanks.     

I forgot to add that for now, I'll update the voltage dividers in the layout I'm using to take the clock voltages from 15V to 8V. Right now they drop it to over 10V. Then check the square waves on the scope and see what they look like. Maybe I'm just overthinking this again.   

[Fender3D]

The whole "digital" section (VCO4013 + BBD + TL081) should work at the very same voltage supply.
I think it won't change a thing if you just swap the 7815 with your 7808 in main PCB and run everything @ 8Vcc

[armillary]

Here's the latest on the DOD 680 fix. It works. I've cleaned up the daughterboard, made some substitutions, but ended up with the same schematic and layout this thread began with. 

Here's the layout:
http://www.aronnelson.com/DIYFiles/up/DOD680fix.pdf

Here's the board installed in the DOD 680:
https://www.dropbox.com/s/6mx08hpt9qd13si/DSCN5864.JPG?dl=0

Here's the latest pic of the board, front:
https://www.dropbox.com/s/snfhbdyqdpfnpcb/DSCN5870.JPG?dl=0

And back:
https://www.dropbox.com/s/5xwnskt5vhd7skb/DSCN5873.JPG?dl=0

Is there anyone out there with a dead DOD 680?

[armillary]

A brief demo of the DOD 680 with the MN3205 fix:
http://www.aronnelson.com/DIYFiles/up/DOD680_withMN3205.mp3