This thing is making me crazy. I checked the D00-D07 bus several times with
a ohmmeter from CN10 to IC 24,26,30,31,33, 28, 29, 34, 35 and can't see a
problem. With the oscilloscope, there is a decent signal on every pin of
every chip, although I couldn't say if it is correct. Unless there is some
crosstalk between data lines somehow, I don't think there's a problem here.
I found some problems with the schematic:
IC24 is mislabelled as IC34,
CN10 has two pins labelled D06 and no D07,
IC23 shows pins 5 and 6 connected. They are not.
I went back to the panel knobs and checked for -5v to +5v from each of them.
That looks okay. Then it goes into the MUX, and comes out really messed up.
As expected, the "Data" line from the MUX chips looks like a set of random
analog voltage steps, but each knob does not seem to control just one step.
The knobs that work change 2 different "steps", the knobs that do nothing
also change nothing, and the knobs that are controlling more than one
parameter are changing more than 2 steps. It seems some knobs are not being
read at all and others are being read more than once. Seems like a timing
problem with the signals on A,B,C,Inh1 and Inh2. These lines also check out
okay back to all ICs they are connected to. My old single-trace Heathkit
oscilloscope (built it myself) is not up to the task of analyzing relative
timings of these signals so I'm looking for a 4 channel scope on Ebay.
Just for grins, I popped a different 8048 CPU from another dead Polysix into
this one. It didn't change any symptoms.
I did some more careful measurements of which knob is changing what, as seen
on the outputs of ICs 7,9,10,11,12,13,14,16.
Effect Speed - does nothing,
VCF Cutoff - does nothing,
EG Int - does nothing,
Res - changes Res and Attack,
Attack - does nothing,
Decay - changes Decay and Sustain,
Sustain - does nothing,
Release - OK,
KB Track - OK,
PW/PWM - OK,
PWM Speed - OK,
MG Freq - OK,
MG Delay - changes MG Delay and EG Int,
MG Level - changes MG Level and VCF Cutoff.
It's easy to see where the D/A conversion is on IC33, but where does the A/D
occur? How do knob settings get converted into something that can be stored
in memory?
Thanks,
Don B.
_____
From:
PolySix@yahoogroups.com [mailto:
PolySix@yahoogroups.com] On Behalf Of
Andrew Jury
Sent: Sunday, January 17, 2010 9:37 AM
To:
PolySix@yahoogroups.comSubject: Re: [PolySix] Wrong functions controlled by knobs?
Use CN10 as a reference point and make sure DO0-7 connects to every
terminating IC which requires access to the data bus. There are plenty of
them! When you measure use the metalwork of the connector and the pins of
the ICs as your test points, not the solder side. I have seen pads connected
before which do not connect to the pins themselves. Also, another problem
which comes up time after time is hidden battery dielectric leaked under the
CPU's holder. I generally remove, clean and replace these as a matter of
course.
Good luck,
Andy
On 17/01/2010 14:19, "backshall1 (dsl)" <backshall1@bellsout
<mailto:backshall1%40bellsouth.net> h.net> wrote:
> On the KLM 367 between some specific chips? I started checking resistance
> from the 14051s on the panel boards to the processor on the 367 board to
> verify A,B,C,INH1,INH2 getting through CN11 okay. It looks good.
> Don B.
>
> _____
>
> From: PolySix@yahoogroups <mailto:PolySix%40yahoogroups.com> .com
[mailto:PolySix@yahoogroups <mailto:PolySix%40yahoogroups.com> .com] On
Behalf Of
> Andrew Jury
> Sent: Sunday, January 17, 2010 8:10 AM
> To: PolySix@yahoogroups <mailto:PolySix%40yahoogroups.com> .com
> Subject: Re: [PolySix] Wrong functions controlled by knobs?
>
>
>
>
> Ah, sorry Don, I misread that in a hurry, I thought you were setting up
the
> a/d converter. If only Leds 1-5 light then pay special attention to bits
DO5
> thru DO7!
>
> Let us know how you get on.
>
> Andy
>
> On 17/01/2010 12:57, "backshall1 (dsl)" <backshall1@bellsout
> <mailto:backshall1%40bellsouth.net> h.net> wrote:
>
>> Hi Andy,
>> I'll check out the data bus and inhibits. No, the calibration did not go
>> well. LEDs 1 and 8 are supposed to light up, and I could only get 1 and
5.
>> Thanks,
>> Don B.
>>
>> _____
>>
>> From: PolySix@yahoogroups <mailto:PolySix%40yahoogroups.com> .com
> [mailto:PolySix@yahoogroups <mailto:PolySix%40yahoogroups.com> .com] On
> Behalf Of
>> Andrew Jury
>> Sent: Sunday, January 17, 2010 7:35 AM
>> To: PolySix@yahoogroups <mailto:PolySix%40yahoogroups.com> .com
>> Subject: Re: [PolySix] Wrong functions controlled by knobs?
>>
>>
>>
>>
>> Hi Don,
>>
>> You almost certainly have track disconnections. Try and focus in on two
>> places of the �367. First the data bus DO0-DO7 and the multiplex address
> bus
>> INH1-3. Using your scope scope ensure that you have at least what look
> like
>> �trains� of data pulse when you operate the controls. I find when dealing
>> with multiple problems like this it pays to select one of the problems
and
>> focus in on it. Usually everything else will start to become apparent as
> you
>> make progress. Before you start, however, are you able to perform the
> reset
>> calibration procedure correctly? If you can verify this then we can
> proceed!
>>
>> Cheers,
>> Andy
>>
>> On 16/01/2010 13:56, "longenough2002" <backshall1@bellsout
>> <mailto:backshall1%40bellsouth.net> h.net> wrote:
>>
>>>
>>>
>>>
>>>
>>>
>>> Just got a Polysix in with some bizarre symptoms. It had an old battery
>>> in it that looked like it had just started to leak. Only a couple of
>>> traces had some black spots and still checked out okay with an ohmmeter.
>>> I put in a cr2032 and diode and it is saving patches okay. It's not
>>> until I twiddle some knobs that things start to get weird. Here is what
>>> works: Output, Bend, Tune, VCO, VCA, Arpeggiator, Key Assign, Programmer
>>> sections all OK. Problem sections: MG, VCF, EG, Effects. By Section: MG
>>> - Frequency, Delay and Mod work, but Level seems to be controlling VCF
>>> Cutoff, not modulation intensity. VCF - Cutoff does nothing, Resonance
>>> works when set between 5 and 10, but seems to control Attack when set
>>> between 0 and 5, EG Intensity does nothing, KBD Track works. EG - Attack
>>> does nothing, Decay works, but in reverse, as if EG Intensity is in the
>>> negative range. Release works between 5 and 10 but seems to do nothing
>>> between 0 and 5. Effects - switch works, but Intensity does nothing.
>>> It would seem the D/A is having serious problems, and I can't get LEDs 1
>>> and 8 to light up during the reset procedure. I get 1 and 5. I wish the
>>> service manual had some voltages to measure, instead of just watching
>>> lights. The +/-15 and +/-5v supplies look good. Anybody have ideas where
>>> to start debugging this? I have a scope and multimeter, but no logic
>>> analyzer. Thanks for any assistance you can provide. Don B.
>>>
>>> [Non-text portions of this message have been removed]
>>>
>>>
>>>
>>>
>>>
>>
>> [Non-text portions of this message have been removed]
>>
>>
>>
>>
>>
>>
>> [Non-text portions of this message have been removed]
>>
>>
>>
>> ------------------------------------
>>
>> PolySix "Digiest" Page: http://www.acc
> <http://www.acc <http://www.acc.umu.se/~amber/Poly6Yahoo>
umu.se/~amber/Poly6Yahoo> umu.se/~amber/Poly6Yahoo! Groups
> Links
>>
>>
>>
>
>
>
>
>
>
> [Non-text portions of this message have been removed]
>
>
>
> ------------------------------------
>
> PolySix "Digiest" Page: http://www.acc
<
http://www.acc.umu.se/~amber/Poly6Yahoo> umu.se/~amber/Poly6Yahoo! Groups
Links
>
>
>
[Non-text portions of this message have been removed]