Skoolie (24) build (and calibration questions/issues)

[benh]

Hi Folks !

So I've finally completeled Skoolie :-) I am not 100% I got it properly calibrated, there are some odd things here, so advice welcome, but otherwise it seems to be doing something not too dissimilar to the E MM demo videos I can see on youtube...

Here's what it looks inside (the wires aren't that neat ... I know). I'll follow up with some of the "issues" I had with calibration (I haven't found how to interleave images and text in the forum)

[benh]

So first is clock calibration. The oscilloscope calibration doc says to aim between 120 and 140us of period. The clock trimmer gives me 120us *all the way* to the left. So I moved it a tiny bit right to get around 130us.

https://photos.app.goo.gl/HMjKGhEZnGePuujPA

However with the delay all the way down, it says to expect between 8 and 10us, I get between 11 and 12us.

https://photos.app.goo.gl/fnAKjLKLhMALEAQMA

Does this hint of a problem or we are just playing games with components tolerance ?

(I couldn't figure out how to insert an image, so I just used public google photo links ...)

[benh]

Then the actual calibration, again trying to follow the oscilloscope calibration guide, I find it ... confusing :-)

There are the parts where you have to probe the compander pins which is hiding underneath the BBD board, I've mostly tried to probe IC3 pin 1 instead.

The BBD BIAS, I get a LOT of noise that seems to be transcients corresponding to the clock edges (so 4x the clock frequency):

https://photos.app.goo.gl/g5wue1egnDt6Bs6w7

This makes calibration rather difficult. Then in "TRIM2 (GAIN1)" it starts with "set for unity". Unity between what and what ? Pin 6 and 7 of IC10_B ? TP1 and TP3 ? I generally tried the latter... (which is a bit tricky with that amount of noise).

Then TRIM3 (BALANCE) "Adjust for minimum clock signal (maximum clock cancellation)" .. that doesn't say much to me. There is no effect on the signal at TP3 at this point that I can see. There is a small effect at TP4. The noise can "spikes" can be tamed to avoid going over the main signal vertically (they still go down but not up as much) with the right setting there, that's what to aim for ?

Looks like this:

https://photos.app.goo.gl/hVamzQLrXfrTJCxi8

(I got it even a bit better than this photo eventually).

Then we go back to BIAS/GAIN on the second BBD, though here too, what needs to actually be done and observed is very very much not clear to me

This docs seems to be written for an audience of experts who fully understand the circuit which is definitely not a category I fall into

Next I'm going to build an audio probe and try the ear calibration instead.

Any advice welcome in the meantime !

[benh]

Hrm... I found I could attach photos when doing "Reply" at the top rather than "Quick reply" in the thread. (silly me)

But I can't edit the posts to do so after the fact. "Modify..." returns an error when trying to submit the modified post and "Quick edit" won't let you attach things. Sorry about that. So here are the photos. First the clock calibration (130us then min delay):

[benh]

And here is the noise at TP3 (and a zoomed version where you can see the transcients)

[benh]

So replacing IC4 with a TL022 significantly improves the clock noise situation (it's still there but with much less amplitude) making calibration easier. I'm not 100% certain I fully understand what's expected with the two gains but overall I've aimed for unity across the board and max amplitude without clipping for bias. It sounds pretty good so I think that will do :-)