Xpantastic!

no one is saying that moving the boards

displaced the problem.You are misreading

his results. Moving the boards resulted

"slot" that handles voices 7-12, regardless

of which board is connected to this second

"slot" for voices 7-12.

this is simple guys: if the problem always stays

in the same "slot" (not the same board)

then this means there is some "fixed" path

that that the CPU or controller uses to get

to these "slots", regardless of what board

is in place. He already proved this by

swapping the boards and the problem did

not follow the board, it stayed in the same

board "slot"

You need to focus on finding this "fixed" path

"slot" and thoroughly test that entire path from

the CPU to the point where it communicates

to the voiceboard connected to "slot 2" (voices 7-12)

-omar

---

sent from iPhone

On Jul 9, 2012, at 5:00 PM, Christophe JANOT <cjanot@...> wrote:

So you swapped the connectors between the 2 voices boards. We are agree that it was the connectors at the upper right corner, right ? I ask because I don't understand how the swap displaced the problem since there is no tuning specific wires on the connectors. This is a classic CPU bus.

Kris

Envoyé de mon iPad

Le 9 juil. 2012 à 18:19, Terje Winther <terje.winther@...> a écrit :

Very interesting update:

The Matrix-12 is just two Xpander voice board on top of each other, right? So it should be OK to shift which one of the cards who is on top?

Since I was always getting tuning errors on voice 7-12 while voice 1-6 was always right, I thought I could swap position of the voice cards to see what happens.

And guess what: it is still voice 7-12 that gets the error - AFTER the swap. The card that was getting all the errors now works perfectly in tune.

So my conclusion is this:

- Both the voice cards now work as they should.

- There is something wrong on the processor board (or in the connection between the cards: I can check the multicable).

Since I am mostly an analog synth repair guy, I am getting towards my limits if I am about to debug the digital electronics. In my experience though, it is often some logic chips and/or some physical connection (on or off board) that fails. I really hope there is nothing wrong with the processor itself.

It would be nice to get this M-12 completely fixed, as it looks good, and it is now so close, so close to being perfect.

Terje

Den 6. juli. 2012 kl. 20.42 skrev Karl Schmeer:

Hi Terje,

Wow Two M12's :-)

The version of the M12 service manual I have " Bought from Factory long ago" is
pretty thin on details.

It's just schematics and a few engineering change orders.
If you have not done so check out the Xpander Service manual. It's in the files
section of this group.
The calibration procedures for the trimmers you speak of are in there, as well
as a detailed

description of what really goes on during the auto tuning routine.

You probably already checked these things but just to be sure:
Do the 2.2uF caps that were replaced have a 50VDC rating? If not they will not
pass the audio signal correctly.

Actually "the entire board failing tune "points to a more systemic failure.
Both the VCO and VCF are tuned the same way with the exception that the VCF's
resonance
is set to oscillate. A specific voltage is applied to the individual VCO or
VCF
and the output frequency is monitored by the CPU to allow tuning.

I would start at the DAC circuit ( page 19 xpander service) There is a trimmer
in that circuit for

DAC offset calibration. ( The procedure is at the end of the Xpander service
manual page 34.
Maybe thats all that is wrong :-)

If not :
There is also a procedure for checking the board reference voltage ( page 34) If
this voltage is off spec it could
cause errors.
Try swaping DAC's between the "known good board" and your failed one. Or any
other parts in the DAC circuit.

There is a circuit on page 31 (Lower right hand corner in Xpander service)
Which is used in the auto tuning routine. It is built around an LM311 voltage
comparator

which ties to the output of each voice.
Something may be amiss here.

Report back after these checks and I will think about it some more.

Best Luck

Karl


________________________________
From: Terje Winther <terje.winther@...>
To: xpantastic@yahoogroups.com
Sent: Fri, July 6, 2012 12:30:38 PM
Subject: [xpantastic] only one board active


Hi all,
Thanks a lot for all the help I have been given on this list. I have
changed parts in the power supply, some CEM chips, all those el.caps
who blocked tuning and sound, swapped some 74 series logic chips (so
now all encoders works), and got new switches and rubber feet, and
even a new key. And of course cleaned it up since all boards were out
anyway.
The Matrix-12 is almost all right now - it is just one peculiar error
left:
It seems that only one of the voice cards gets the proper commands.
Voice 7 - 12 fails tuning of VCOs and VCFs, but all other tuning
passes and there is sound from all 12 voices (if I turn them on) and
the voices are almost in tune, and VCFs seems to work like they should.
I also have another Matrix-12 in for service (yes, two M-12s in front
of me - pretty sight!), and it too reportedly have an issue with only
one of the voice cards working properly.
I have double-checked the connectors (the paralell connector and the
two 3-4 pin connectors) and they all seems to be fine.
Any tips on what to do if one of the voice cards don´t responds?

On another question: While having all the boards out, I noticed that
there are very few manual trimmers on the M-12, but there are some,
and I can´t seem to find any info on how to calibrate them. I have the
service manual, and I find the trimmers there, so I can pretty much
guess what to do, but it would be nice to know for sure.

Terje Winther
terje.winther@...
http://www.wintherstormer.no/

Terje Winther

terje.winther@...

http://www.wintherstormer.no/