Dirtbag v3205 Problem- some Distortion and lots of noise in the delay [SOLVED]

Started by Thomas_H, April 01, 2012, 06:51:47 PM

Previous topic - Next topic

Thomas_H

HI,

I just finished a v3205 Version of the Dirtbag Deluxe and I dont think its working right.
Some things I can state after 4 hrs of looking at it with a Scope and reading datasheets of the NE 570 and v3205.

Facts:
47k used for R 11/12.

Observations meassured:
a) NE 570 Output on Pin 7 is Biased 7V
b) BBD Chips run at 8,2V  resp. 8V Vgg

Observations soundwise:
a) Not replacing the 1k/33n filter results in distorted repeats - (we know that)
But the repeated signal also has a high level of noise comming with it. So the delay-circuit is somewhere grabbing noise
b) replacing R21&C14 with 100K both: results in a less distorted signal. The signal is weak and still has the same level of noise. It is also not possible to make the circuit self oscillate with this mod even when the gain pots are up and feedback is full.
The noise starts with the signal coming through and ends with the signal.

Another Observation made:
Why is there a 9V clock signal fed into a 8,2V driven BBD ? Shouldnt be the clock signal max the supply voltage of the chip ? - This seems to be a design flaw as this is how the clock is built.

Can someone please supply me with some reference of a working v3205 Dirtbag.
In Detail I am looking for the correct Pin 7 NE 570 Bias. 7V in my build. Also any Scope reading of BBD in and out signals Pin 3 and 7 would be very helpful as I think the clock signal is not filtered out in the later stages.

Any other idea would be very helpful.
Thank you !


DIY-PCBs and projects:

madbean

Hi Thomas,

Please note the following modifications which have been found to improve further upon problems with distortion when using the v3205: http://www.madbeanpedals.com/forum/index.php?topic=3366.0

Of these the most critical are changing IC6 to a buffer by jumpering R16 and omitting R17, and the 1uF/6k8 combo in place of R10. IIRC, on the Memory Boy I read an output voltage on pin7 of the compander to be about 6.3-6.4v, but it's been a few months since checking. If you are reading high, you can lower the two 47ks -- 39k or 36k will probably do the trick.

These mods take the place of the earlier one suggested (R21.C14) and gave the bests results after a lot of trial an error.

The clock and BBDs do not have to be run at the exact same voltage AFAIK since it is the clock frequency driving the BBDs, not the supply voltage. This is something I have checked out before and I found no difference in terms of distortion in running the two at slightly different voltages.

I'm a bit confused as to what you means by "the clock signal is no filtered out in the later stages"....do you mean on your build, or by design? In the latter, this is exactly what the de-emphasis filtering is doing before getting to the expander portion of the circuit. If you are hearing clock signal, it may be due to the clock frequency running the BBDs running below the actual filtered signal. I think the lowest it can be run before bleeding into the audio was a little below 7kHz.

Also, here is a list of some good voltages you can use as reference in debugging: http://www.madbeanpedals.com/forum/index.php?topic=3380.0


One other thing I have found since the different mods/solutions were talked about---some v3205s are just noisier than others. I've got about a dozen in the workshop and I've found myself digging through them more than once to swap out. The degree to which this is a manufacturing issue, or just because I abuse the hell out of my parts through all the building and prototyping I've done with them I don't know. I'm not saying your particular BBDs are at fault and you should run out and replace them, but rather one thing I would also try is swapping in whatever you've got in IC4, audio probe the outputs while biasing it up and see if you hear any variation between them as far as noise or distortion.

Scruffie

Voltage of the clock shouldn't matter, all chip cares is it sees a nice square wave 'tick' & 'tock' to make it do its thang.

QuoteOne other thing I have found since the different mods/solutions were talked about---some v3205s are just noisier than others. I've got about a dozen in the workshop and I've found myself digging through them more than once to swap out. The degree to which this is a manufacturing issue, or just because I abuse the hell out of my parts through all the building and prototyping I've done with them I don't know. I'm not saying your particular BBDs are at fault and you should run out and replace them, but rather one thing I would also try is swapping in whatever you've got in IC4, audio probe the outputs while biasing it up and see if you hear any variation between them as far as noise or distortion.

Interesting... I wonder if examining a Behringer 'Vintage Time Machine' delay might be beneficial to us.
Works at Lectric-FX

LaceSensor

tbh I still dunno what is up with my build. im using the same chips I used in my first attempt (before all the modifications were posted) and this time round its really nasty sounding.

Scruffie, might be another board (with chips this time) heading your way if you are interested?!

Thomas_H

Thanks to all who are trying to help. I will dig into it when I return from work and see what can be improved.

Interesting thing with the 9V clock. I would have guessed this would inject glitches into the BBD. Good to know it doesnt.

@madbean: The de-emphasis filtering is taking place in IC 3B OR 8A ?  I will also check the clock rate to be above 7kHz. 

I also came across the thread you mentioned but wasnt able to figure out that it superseeds the documents that are uploaded. :-)

DIY-PCBs and projects:

Scruffie

Quote from: LaceSensor on April 02, 2012, 09:33:41 AM
tbh I still dunno what is up with my build. im using the same chips I used in my first attempt (before all the modifications were posted) and this time round its really nasty sounding.

Scruffie, might be another board (with chips this time) heading your way if you are interested?!
Go for it! I still haven't got around to ordering the V3205s myself (waiting for Steve to stock B100ks for my Dirtbag build).

I'll use the old crapped out board for my version of the mods and that'll give me some comparisson.
Works at Lectric-FX

LaceSensor

Quote from: Scruffie on April 02, 2012, 10:19:00 AM
Quote from: LaceSensor on April 02, 2012, 09:33:41 AM
tbh I still dunno what is up with my build. im using the same chips I used in my first attempt (before all the modifications were posted) and this time round its really nasty sounding.

Scruffie, might be another board (with chips this time) heading your way if you are interested?!
Go for it! I still haven't got around to ordering the V3205s myself (waiting for Steve to stock B100ks for my Dirtbag build).

I'll use the old crapped out board for my version of the mods and that'll give me some comparisson.

Lets take it to PMs
I can send you some B100ks if needed.
Im quite serious about you having this build too if you can work any magic on it and Ill pay return shipping.
Im kinda at my wits end. Have now resorted to MN3005s and will do my 3rd (and Final I might add) Dirtbag...
If that doesnt work, I might just fork out for an XO model from EH...

Thomas_H

[SOLVED]

Hi, doing all the mods suggested in the December thread made it. I do have a working delay now  ;D
Even the noise got limited to a bearable degree.
The clock signal I heard was not the clock signal but a rather noisy power supply I was using for my test setup  ;)
Thank you to everyone giving input.
DIY-PCBs and projects: