[sdiy] Ensoniq keyboard calibration advice?

Peter Pearson electrocontinuo at gmail.com
Thu Jun 30 23:54:46 CEST 2022


Interesting... Thanks!

This ASR10 has a ribbon cable hard wired between the 2 boards and I'm
pretty confident it's good.  After a thorough scrub, I had the whole
keyboard working if I selected to skip keyboard cal after it failed.  I
wanted it to pass cal so I did another check through and tweaked the
oscillator trims out of curiosity.  After that, half the keyboard wouldn't
play and the other half seemed basically fine.

Since I can get stuff to play, I think the digital part is more or less
working (though I'll have to look it over again in detail).  My "coils" all
beep out fine and I don't see any obvious board damage.  I'm going to have
to do another pass with a fine-tooth comb.

Wondering if maybe my oscillators are set too high and that's making it
cranky.  It'd be nice to know what frequency they're supposed to be set to
cuz I'd dial them in as close as I could get.

I've seen people say they had bad caps on their boards but I don't want to
recap anything unless it seems really necessary (or I need a hail mary).
Another post somewhere mentioned the 555 going bad.  So far, I don't think
I'm having those issues.

The pain in the butt with these is that as far as I can tell, you have to
fully strip the keyboard apart and fully put it back together again to
check your progress after each repair pass.  Not fun.  Black magic is
definitely an apt description.  Maybe I need to stock up on some goat blood
and black candles to get this working.

On Thu, Jun 30, 2022 at 5:28 PM Rainer Buchty <rainer at buchty.net> wrote:

> On Thu, 30 Jun 2022, Peter Pearson via Synth-diy wrote:
>
> > Anyone on the list have insight or advice on getting Ensoniq gear to pass
> > the keyboard calibration routine?  I'm looking at an ASR10 at the moment.
> > [...]
> > Ensoniq was very stingy with schematics but the keyboard section is
> posted
> > here:
>
> Yeah, so you know what to check for ;)
>
> All Ensoniq polypressure keyboards are essentially 16 sections per four
> keys (or, more correctly, 2*8 sections per 2*2 keys) muxed to their
> corresponding high-speed VCO. The VCOs' frequencies are driving the KPC
> ASIC's high-speed (well, at least back then) counters.
>
> Bluntly put, each key is a little Theremin: The distance of the key's
> metal plate to the PCB-etched coil determins the frequency. For
> scanning, the ASIC's counters are reset, the muxes are set, then the
> counters are enabled and read out. Those counter values are processed
> through a bunch of hysteresis and threshold black magic in order to
> properly determine idle (key up) from "key traveling" and "key down" and
> finally "pressure".
>
> As such, there are multiple failure scenarios for keyboard calibration
> errors (i.e. "unable to determine idle frequency").
>
> - on VFX and possibly other machines that got the dual-PCB keyboard it's
>    the connector between those two halves; remove it and put soldered-in
>    wires there instead.
> - PCB haircrack -- unfortunately, that's more than often the case
> - pass transistor shot
> - multiplexer dead
> - defective KPC ASIC (though have never seen that one so far)
>
> If it signals "calibration failure", then you can at least rule out the
> CPU, as "failure" is a dedicated code sent back to the main board.
>
> (At least on the SQ80, the machine would just simply hang in calibration
> if neither "ok", KPC error #1/#2, or "calibration failure" are sent
> back. I somewhat doubt that they implemented a time-out on the later
> machines. Ensoniq somehow never liked timeouts -- e.g. if your Mirage
> after booting seemingly dies, just wait some 10 minutes until it finally
> fails filter calibration of all 8 voices and it most likely will spring
> back to life.)
>
> HTH,
>         Rainer
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://synth-diy.org/pipermail/synth-diy/attachments/20220630/92bdb9d4/attachment.htm>


More information about the Synth-diy mailing list