Discussion about the Korg PolySix synthesizer

So I managed to complete my first replacement PCB.

Things don't look too bad but it has a number (?) of problems that are a
bit difficult to debug for me.

Basically it runs somewhat unstable. Occasionally it will more or less
reset itself or will exhibit some other behaviour which will explain in a
second.

Nearly all chips are new, except processor, LM13600's and RAM. All
functionality is there; calibration was more or less successful. I haven't
yet a battery installed.

THe most common symptom at the moment is that some LED's on the KLM371
start to blink or that a number of LED's light up permanently (and others
get turned off). This has happened with different KLM371's so I suspect the
error to reside on the KLM367. Oddly enough often the sound isn't affected
at all initially and the switches still work fine. I can call up other
programs and other banks.

Sometimes shortly (or several seconds after this) the LFO modulation is
turned off (I have programmed this particular test sound to modulate the
VCF). There might some other parameters change slightly as well; I have not
yet narrowed this down completely. Sliding the LFO switch back and forth
restores the modulation; so the actual LFO depth isn't affected, just the
target.

So, it appears to me (grain of salt...) that the parameters controlled by
IC28/29/34/35 may be affected, but not the ones controlled by IC18/19
(post-DAC).
This and the working switches leads me so far to the conclusion that the
data lines D0 to D7 may not be the culprit here and that the problem could
be related to lines P24 to P27.

Could this be another "bad RAM" issue? My first salvaged RAM was definitely
dead; it wouldn't store anything. The current one seems to work as far as
storing goes.

Interestingly I have board with the very same problem which I constructed
for another forum member. The guy is kindly sending it back me me after a
vast amount of remote debugging. From what we found and saw looks like the
CPU or closely associated components may be at fault. If you can wait for a
week or so I shall have this rogue board back in the lab for analysis. Also,
strangely enough, removing the RAM for this board had no bearing on the
problem whatsoever, but removing the comparator (IC6) which handles the data
in from the control panel mux and the feedback from the D/A convert
completely eradicates the instability! You might want to try this with your
board and report back what happens. The only caveat is that you can use the
control panel!

Let me know what you find.

Cheers,
Andy

On 26/08/2010 14:01, "Malte Rogacki" <gacki@gacki.sax.de> wrote:

>
>
>
>
>
> So I managed to complete my first replacement PCB.
>
> Things don't look too bad but it has a number (?) of problems that are a
> bit difficult to debug for me.
>
> Basically it runs somewhat unstable. Occasionally it will more or less
> reset itself or will exhibit some other behaviour which will explain in a
> second.
>
> Nearly all chips are new, except processor, LM13600's and RAM. All
> functionality is there; calibration was more or less successful. I haven't
> yet a battery installed.
>
> THe most common symptom at the moment is that some LED's on the KLM371
> start to blink or that a number of LED's light up permanently (and others
> get turned off). This has happened with different KLM371's so I suspect the
> error to reside on the KLM367. Oddly enough often the sound isn't affected
> at all initially and the switches still work fine. I can call up other
> programs and other banks.
>
> Sometimes shortly (or several seconds after this) the LFO modulation is
> turned off (I have programmed this particular test sound to modulate the
> VCF). There might some other parameters change slightly as well; I have not
> yet narrowed this down completely. Sliding the LFO switch back and forth
> restores the modulation; so the actual LFO depth isn't affected, just the
> target.
>
> So, it appears to me (grain of salt...) that the parameters controlled by
> IC28/29/34/35 may be affected, but not the ones controlled by IC18/19
> (post-DAC).
> This and the working switches leads me so far to the conclusion that the
> data lines D0 to D7 may not be the culprit here and that the problem could
> be related to lines P24 to P27.
>
> Could this be another "bad RAM" issue? My first salvaged RAM was definitely
> dead; it wouldn't store anything. The current one seems to work as far as
> storing goes.
>
>



[Non-text portions of this message have been removed]

Hi Andrew; I'll investigate this further.

The problem didn't go away when using a different processor.

The problems with the missing LFO modulation and so on could be narrowed
down to the parameters controlled by IC28/29; the parameters controlled by
IC34/35 seem not to be affected. The parameters simply jump to the panel
value.
Interestingly enough sometimes it is not possible to call up the correct
parameters afterwards. For example: I program patch A1 to use 4' and store
it. Afterwards I change it on the panel to 16' and call up patch A1 again.
Many times this works; but sometimes the sound remains at 16'. So I guess
either the RAM is corrupted or somehow the stored value isn't getting
through IC28/29.

The occasional resets still happen with RAM removed; however I haven't
gotten multiple patch LED's so far with this constellation (it seems to
switch back to A1 always). It seems that the time between resets is quite
long (right now it has been running for about 20 minutes in manual mode
without switching back to A1 on its own).

Hi Malte,

Looks like you¹ve made a fairly detailed study, one which I will have to
make myself when the board comes back to me. The patch assigned logic was
hopelessly unstable to the point of completely muting any output from the
P6. It was working perfectly well before it was shipped so I wonder what
happened in transit. Static damage perhaps? I am not so sure. Anyway, I will
let you know what I find and we can perhaps compare results.

Cheers,
Andy




[Non-text portions of this message have been removed]

So things are looking a bit worse by now.

I can no longer successfully make the D/A adjustment. This actually had
happened before but the problem had disappeared for a day or so (after
reflowing all possible solder joints). The symptom is that VR6 does not
shift the offset of the left LED. As an example: If I call up the D/A
adjustment routine LED's 4 and 8 light up. By turning VR6 I can make LED 8
move to 7 or 6, but LED 4 stays on all the same. VR7 seems to do what it
should - however I cannot get lower than LED 4 in any case.

Given the other problems the KLM367 has at the moment I'm a bit reluctant
to make a definitive judgement now (perhaps it is just a display problem
and the adjustment procedure actually works?). I've measured all the
continuity around IC33 and everything looks ok; I also swapped IC27 to no
avail.

Hello Malte,
Hi Andy,

I am the forum member Andrew mentioned yesterday. I'm not sure it could help you a lot but here is a description of my symptoms :

Here is a video taken few days ago. At this time, my P6 had been powered up for 3 hours or so :

http://demo.ovh.com/fr/40fbff223d8eac6917e9a7018648dd47/

Usually the PROGRAM LEDs didn't blink like that. It only happened 3 or 4 times.

But I've very often experienced a switch back to A1, after 2-5 minutes. The alternative was three PROGRAM LEDs lit at the same time.

All the patches stored in RAM had troubles :

- most of them seemed like empty or silent
- some others were working but sounded raw or corrupted

After I had made a new D/A adjustment I had more or less the same proportion of "empty" and "corrupted" patches. But not at the same location. For example, if A3 was previously sounding raw or corrupted and A8 was silent, after the new D/A adjustment A3 was silent and A8 was sounding corrupted or raw.

I also noticed the RELEASE to behave strangely, and the MG too.

One last thing : the ATTENUATOR knob didn't produce the expected result. For instance, switching from +10 to +8 or +6 actually made the output sound louder…

Good luck !

--- In PolySix@yahoogroups.com, Andrew Jury <andy@...> wrote:
>
> Interestingly I have board with the very same problem which I constructed
> for another forum member. The guy is kindly sending it back me me after a
> vast amount of remote debugging. From what we found and saw looks like the
> CPU or closely associated components may be at fault. If you can wait for a
> week or so I shall have this rogue board back in the lab for analysis. Also,
> strangely enough, removing the RAM for this board had no bearing on the
> problem whatsoever, but removing the comparator (IC6) which handles the data
> in from the control panel mux and the feedback from the D/A convert
> completely eradicates the instability! You might want to try this with your
> board and report back what happens. The only caveat is that you can use the
> control panel!
>
> Let me know what you find.
>
> Cheers,
> Andy
>
>
> On 26/08/2010 14:01, "Malte Rogacki" <gacki@...> wrote:
>
> >
> >
> >
> >
> >
> > So I managed to complete my first replacement PCB.
> >
> > Things don't look too bad but it has a number (?) of problems that are a
> > bit difficult to debug for me.
> >
> > Basically it runs somewhat unstable. Occasionally it will more or less
> > reset itself or will exhibit some other behaviour which will explain in a
> > second.
> >
> > Nearly all chips are new, except processor, LM13600's and RAM. All
> > functionality is there; calibration was more or less successful. I haven't
> > yet a battery installed.
> >
> > THe most common symptom at the moment is that some LED's on the KLM371
> > start to blink or that a number of LED's light up permanently (and others
> > get turned off). This has happened with different KLM371's so I suspect the
> > error to reside on the KLM367. Oddly enough often the sound isn't affected
> > at all initially and the switches still work fine. I can call up other
> > programs and other banks.
> >
> > Sometimes shortly (or several seconds after this) the LFO modulation is
> > turned off (I have programmed this particular test sound to modulate the
> > VCF). There might some other parameters change slightly as well; I have not
> > yet narrowed this down completely. Sliding the LFO switch back and forth
> > restores the modulation; so the actual LFO depth isn't affected, just the
> > target.
> >
> > So, it appears to me (grain of salt...) that the parameters controlled by
> > IC28/29/34/35 may be affected, but not the ones controlled by IC18/19
> > (post-DAC).
> > This and the working switches leads me so far to the conclusion that the
> > data lines D0 to D7 may not be the culprit here and that the problem could
> > be related to lines P24 to P27.
> >
> > Could this be another "bad RAM" issue? My first salvaged RAM was definitely
> > dead; it wouldn't store anything. The current one seems to work as far as
> > storing goes.
> >
> >
>
>
>
> [Non-text portions of this message have been removed]
>

Hi Malte,

Think we might have a D/A problem here causing instability. Probably why
taking IC6 out of circuit stops the random behaviour. Just looking for
obvious differences between the original board and the clone and one thing
that springs to mind is the D/A buffer op-amp which is used. The spec for
the new board uses a TL072P whereas the original uses a NJR4560. Seeing as
the rest of the original is covered in generic 4558 and 072 there must have
be a good reason why the designers selected the 4560 in preference. Perhaps
the 072 can¹t switch fast enough of the slew rate is too slow, etc. Time to
get out the datasheets and have careful study I think...

Cheers,
Andy





[Non-text portions of this message have been removed]

> Usually the PROGRAM LEDs didn't blink like that. It only happened 3 or 4
> times.

Yes, this is also similar to the symptoms my machine exhibits.

> But I've very often experienced a switch back to A1, after 2-5 minutes. The
> alternative was three PROGRAM LEDs lit at the same time.

...and that's pretty much what my machine does all the time. Sometimes it
takes only a few seconds, sometimes longer.

I've swapped IC27 with a salvaged 4560 now. D/A calibration seems to work
fine now but I still get the other symptoms.

I've also finished a second board; so far this seems to work stable
(although I probably should do a burn-in test before making final judgement
about this).

There are a few more observations but I want to verify them before posting
them.

Just a quick question: Is the RAM (IC32) only used for patch storage? The
actual "workspace" RAM is in the 8049 itself?

--
Malte Rogacki

> Just a quick question: Is the RAM (IC32) only used for patch storage? The actual "workspace" RAM is in the 8049 itself?

I can't answer this question as my cloned KLM-367 is using an original 8048C. But Andrew will certainly do.

--- In PolySix@yahoogroups.com, Malte Rogacki <gacki@...> wrote:
>
> > Usually the PROGRAM LEDs didn't blink like that. It only happened 3 or 4
> > times.
>
> Yes, this is also similar to the symptoms my machine exhibits.
>
> > But I've very often experienced a switch back to A1, after 2-5 minutes. The
> > alternative was three PROGRAM LEDs lit at the same time.
>
> ...and that's pretty much what my machine does all the time. Sometimes it
> takes only a few seconds, sometimes longer.
>
> I've swapped IC27 with a salvaged 4560 now. D/A calibration seems to work
> fine now but I still get the other symptoms.
>
> I've also finished a second board; so far this seems to work stable
> (although I probably should do a burn-in test before making final judgement
> about this).
>
> There are a few more observations but I want to verify them before posting
> them.
>
> Just a quick question: Is the RAM (IC32) only used for patch storage? The
> actual "workspace" RAM is in the 8049 itself?
>
> --
> Malte Rogacki
>

My bad, It's of course an 8048C here as well (8049 is the key assigner).

--- In PolySix@yahoogroups.com, Malte Rogacki <gacki@...> wrote:

>
> So I managed to complete my first replacement PCB.
>
> Things don't look too bad but it has a number (?) of problems that are a
> bit difficult to debug for me.
>
> Basically it runs somewhat unstable. Occasionally it will more or less
> reset itself or will exhibit some other behaviour which will explain in a
> second.
>
> Nearly all chips are new, except processor, LM13600's and RAM. All
> functionality is there; calibration was more or less successful. I haven't
> yet a battery installed.
>
> THe most common symptom at the moment is that some LED's on the KLM371
> start to blink or that a number of LED's light up permanently (and others
> get turned off). This has happened with different KLM371's so I suspect the
> error to reside on the KLM367. Oddly enough often the sound isn't affected
> at all initially and the switches still work fine. I can call up other
> programs and other banks.
>
> Sometimes shortly (or several seconds after this) the LFO modulation is
> turned off (I have programmed this particular test sound to modulate the
> VCF). There might some other parameters change slightly as well; I have not
> yet narrowed this down completely. Sliding the LFO switch back and forth
> restores the modulation; so the actual LFO depth isn't affected, just the
> target.
>
> So, it appears to me (grain of salt...) that the parameters controlled by
> IC28/29/34/35 may be affected, but not the ones controlled by IC18/19
> (post-DAC).
> This and the working switches leads me so far to the conclusion that the
> data lines D0 to D7 may not be the culprit here and that the problem could
> be related to lines P24 to P27.
>
> Could this be another "bad RAM" issue? My first salvaged RAM was definitely
> dead; it wouldn't store anything. The current one seems to work as far as
> storing goes.
>

Is someone keeping track of all the issues with the KLM367A?

I'll be building 3 and want to get as much of the "crappy" stuff out of the way

Thanks....Scott , Nerd, Vermont

...I suspect there is only a single issue here and that is with the d/a
converter. As soon as I get back my errant board, which exhibits the
problem, and discover what it is I¹ll let everyone know. The previous two
issues, which are missing tracks, are clearly documented on the group buy
site for the board.

Cheers,
Andy

On 30/08/2010 13:26, "Scott" <painintheamp@gmail.com> wrote:

>
>
> Is someone keeping track of all the issues with the KLM367A?
>
> I'll be building 3 and want to get as much of the "crappy" stuff out of the
> way
>
> Thanks....Scott , Nerd, Vermont



[Non-text portions of this message have been removed]

So I've tried to get a bit closer to possible problems and differences.

Luckily I have a few old KLM-367 and 367A here (most in various stages of
disrepair though...).

So, what could make the 8048 reset or hang (which is what I assume happens
when it jumps to A1 out of the blue)?
One possible explanation would be clocking problems. Depending on the
severity the 8048 could act "funny" or lock up.

Touching XTAL1 (Pin 2) with a probe works quite well however touching XTAL2
(Pin 3) will likely lock up the processor (as evidenced by Pin 32 losing
its pulse). Still one can measure the amplitude of the signal on both pins.

With the old boards the amplitude of XTAL1 is around 2.2 to 2.5 volts and
XTAL2 is around 2 volts or slightly below. For both my new boards XTAL1 is
roughly in the same ballpark (2.5 volts) however XTAL2 is much lower (about
1.5 volts).

I've swapped the circuit parts (X1, C22, C23 and even the processor) from
one of the older boards over to the new board; this did not change the
results at all. The value of XTAL2 stays the same on the new board.

Could this be a possible source of trouble?

Update: A change of the value of C23 resulted in a higher amplitude of the
XTAL2 signal. With this the board has been running stable for a few hours
now; but I'll let it run through the night.

I'll keep the list posted.

This is good progress, but I wonder why lifting IC6 kills the issue dead? I
see that programmable input T0 is pin 1 and clock is 2 and 3. You don¹t
suppose we have a track routing issue here? Perhaps there is noise from T0
being inducted into the clock pulse train from the adjacent track? Just got
to finish a few issues and shall have a proper look at the problem. I now
have my errant board in the lab.

Cheers,
Andy




[Non-text portions of this message have been removed]

At 9:01 Uhr +0100 01.09.2010, Andrew Jury wrote:

> This is good progress, but I wonder why lifting IC6 kills the issue dead?

I don't know.

My preliminary hypothesis is as follows:

It's a capacitance issue. The stray capacitance between XTAL1 and XTAL2
(Pin 2 and 3) is different. Of course we're dealing with very small
capacitance here (C22 and C23 are 10pF and 22pF originally). But looking at
the datasheet for the 8048 it appears to me that those two caps are
specifically tailored to the stray capacitance of the board.
The replacement KLM-367A has a very slight difference in routing for the
traces to Pin 2 and 3. On the original board the traces are further apart;
they diverge somewhat. On the replacement board those traces are closer and
mostly in parallel.
Depending on the tolerances of the components one may end up with a
situation where the difference between XTAL1 and XTAL2 gets too large. On
my replacement board there was a difference of about a volt between the
amplitudes - much more than on the old boards. I don't really KNOW if
that's the problem - however with a changed value for C23 the board did run
the whole night without a single hiccup. Perhaps the whole clock circuit
becomes unstable (and hence vulnerable to other interference) if the two
signals are too different?

I will test this further, probably this evening. Once again - so far this
is only a hypothesis.

Food for thought:

I have a datasheet for the 8048H and the 8048AH. Both contain examples for
the clock circuit but both are slightly different.

Both datasheets contain identical examples for the "Crystal Oscillator
Mode". Those examples use two different caps and the following formula:

C1 (equals C22): 5pF +- 0.5pF + (STRAY < 5pF)
C3 (equals C23): 20pF +- 1pF + (STRAY < 5pF)

(There is an optional C2 cap which apparently isn't used in the KLM-367.)

Interestingly the datasheet for the 8048AH has also an example for a
"Ceramic Resonator Mode". And here both C1 and C2 are rated 33pF (+- 5%).

From the looks of it I'd say that the KLM-367 design actually calls for a
Crystal Oscillator and not for a Ceramic Resonator. That service manual
also specifies a Crystal Oscillator.

Malte,

Very interesting! We certainly do have a stray capacitance issue here. When
I connect my scope to T0 and watch the pulse train going into the CPU the
instability on this board stops dead (a bit like removing IC6). I think the
probes I am using have a capacitance of 47pF. Hmmm...this is odd...more to
follow...

Andy


[Non-text portions of this message have been removed]

Andrew;

I'm still in the "theory" stage of everything.

It appears that my measurements as far as amplitudes went were probably
completely off. They were done with a "x10" probe setting since this seemed
to affect the operation of the KLM-367 least. However the waveforms and
values looked very different from what I am getting now with a "x1" probe
setting.

Using two 33pF caps (as suggested by the 8048AH datasheet) did not work out
well (instant lockup). Neither did using a quartz oscillator with the stock
value caps (very unstable).

I also experienced more stable behaviour when doing measurements yesterday.

Right now I am using a 10pF ceramic cap for C23 (instead the 22pF cap) and
the synth has yet to crash. I'll let it run for a few hours.

One thing that was odd (or may reflect on my measuring abilities...) is
that the signal for XTAL1 and XTAL2 is reading not 6 MHz but around 10 to
12 MHz. One of the new boards reads even 18 MHz!

Malte,

This is going to be a very tough problem to solve, I fear. However, I have
been doing some experimentation here and have noticed that every time I
probe T0 on the baord and leave the probes attached the MCU remains stable.
What I have done is simulated the capacitance of the probe by soldering a
22pF ceramic capacitor between T0 and ground. At present my board is holding
stable. I know this does explain much but it will give me time to think
about what on earth is going on here! What happens if you revert to the
original resonator and capacitor combo and try the trick with the T0 line.
Do you get the same effect or is the board just as bad? It might also be
helpful to note that this board is susceptible to random bank switching just
after the top panel controls are adjusted. Which goes back to this whole S/H
data/IC6/probes on T0 conundrum...

Cheers,
Andy


On 01/09/2010 20:05, "Malte Rogacki" <gacki@gacki.sax.de> wrote

>
> Andrew;
>
> I'm still in the "theory" stage of everything.
>
> It appears that my measurements as far as amplitudes went were probably
> completely off. They were done with a "x10" probe setting since this seemed
> to affect the operation of the KLM-367 least. However the waveforms and
> values looked very different from what I am getting now with a "x1" probe
> setting.
>
> Using two 33pF caps (as suggested by the 8048AH datasheet) did not work out
> well (instant lockup). Neither did using a quartz oscillator with the stock
> value caps (very unstable).
>
> I also experienced more stable behaviour when doing measurements yesterday.
>
> Right now I am using a 10pF ceramic cap for C23 (instead the 22pF cap) and
> the synth has yet to crash. I'll let it run for a few hours.
>
> One thing that was odd (or may reflect on my measuring abilities...) is
> that the signal for XTAL1 and XTAL2 is reading not 6 MHz but around 10 to
> 12 MHz. One of the new boards reads even 18 MHz!
>

[Non-text portions of this message have been removed]

Andrew;

I'm back to the previous configuration now (resonator, stock values) and
have experienced the first "lockup" (which is insofar not completely
correct as the machine isn't totally locked up).

Symptom is garbage settings for the sound, simultaneously lighting A, 1, 5,
Manual LED's.
Turning the Cutoff control makes the LED shift between 5 and 8 (but that
happens only on the last few degrees of the turning motion.
I can not call up "Manual" again - except sometimes. It's as if the machine
is in some random "hiccup" mode now.

It took about 10 minutes for the lockup to appear.

I'll try now the cap between T0 and ground.

Andrew;

a 22pF cap between T0 and ground does not cure the lockups and crashes on
my machine.

This problem is an utter stinker! I have now had the suspect board running
for almost 12 hours without a glitch. The owner was seeing consistent
problem after 20 minutes or so. There has to be some noise getting in
somewhere and this is specific to the Polysix it is located in. This board
was soak tested for almost three days without issue before it was shipped! I
will have a play with the cable routing to see if this is an issue. Oh, BTW,
the board has been tested with both a KLM-366 and 366A as I though this
might be an issue, but clone remained stable.

Cheers,
Andy

>

[Non-text portions of this message have been removed]

Andrew;

I've reverted to my previous version now (with the 10pF cap for C23) and
the board has been runing fine now for several hours again.

The problems with that particular KLM-367 also were manifest in another
Polysix so I doubt my case has something to do with a specific machine.

Hi Malte,

You are right. The problem I have had is reproducing the problem. As I said
earlier this fault wouldn¹t show itself for hours on end. I finally realized
the thing to do was to actually play the P6 which very quickly reveals the
problem. Especially if you bash out a few chords and sweep the filters and
VCO controls as you are doing it. In my case I used Œmanual¹ mode as the
control for the experiment. When the fault appeared bank lights A-D would
extinguish and lights 2, 4, 7 would light. What kind of patch is that?

The CPU is clearly very susceptible to noise on the clone board, so much so
than the old Œ367 for some reason. Even getting you hand close to the
oscillator for testing sends it into a spin. If you compare this with the
original Œ367 the circuitry is ultra stable. Whilst measuring the clock
input at C22 I realized that in doing this the whole CPU would appear less
likely to crash when I was probing. If I leave the scope probes in place and
then bash out a few chords whilst sweeping the filter, etc, I cannot get the
board to leave manual mode. I am guessing that as the scope probes are very
high impedance then it must be the capacitance of them which is stabilizing
the circuit. This is also the case with just the probes attached, not even a
scope! My probes are around 47pF, so that in parallel with the 22pF I am
testing would be 68pF (or as near the E series value as you can get). This
is much along the same lines as your initial experiment, but with a more
clear view of what might be happening. Perhaps we should suggest people
leave the scope probes attached before they screw down the lid! Can anyone
else think of any property that might be introduced the probes? In the mean
time I¹ll trying some more modifications to the oscillator.

Cheers,
Andy

On 02/09/2010 19:31, "Malte Rogacki" <gacki@gacki.sax.de> wrote:

> Andrew;
>
> I've reverted to my previous version now (with the 10pF cap for C23) and
> the board has been runing fine now for several hours again.
>
> The problems with that particular KLM-367 also were manifest in another
> Polysix so I doubt my case has something to do with a specific machine.



[Non-text portions of this message have been removed]

Andrew;

your observations go along with my theory that this is a question of the
balance of the clock signals. By attaching the probe to C22 (or XTAL1 at
the chip) you increase the capacitance of C22 (which should originally be
10pF, not 22pF).
On the other hand I had decreased the capacitance of C23 (originally 22pF,
now 10pF but I first tested this with a 22pF and 10pF cap in series which
should give less than 7pF).

As I said earlier my measurements probably don't amount to much. I again
measured with the 10x probe setting. So even if we ignore the actual values
the following was still apearent:

- the signals on XTAL1 and XTAL2 are closer in amplitude on the old boards
than on the new boards
- exchanging C23 for a smaller value brought up the amplitude for XTAL2 on
the new board

All this was only measurable with the 10x probe setting; the measurements
with the 1x setting did not show such a strong difference (but they almost
always locked up the processor and showed a completely different waveform).
If I'm reading this right my probe has a capacitance of around 100pF in the
1x range and around 15pF in the 10x range - which of course explains why
measuring with the 1x setting does no good.

The new board with the different C23 value has been running now for over 12
hours.

I need to find out how to make screenshots from my scope (it's a digital
one); or perhaps I'll simply take pictures with the camera.

Hi Malte,

@aj: I feel we are making some progress here (at last)...

>
> your observations go along with my theory that this is a question of the
> balance of the clock signals. By attaching the probe to C22 (or XTAL1 at
> the chip) you increase the capacitance of C22 (which should originally be
> 10pF, not 22pF).
> On the other hand I had decreased the capacitance of C23 (originally 22pF,
> now 10pF but I first tested this with a 22pF and 10pF cap in series which
> should give less than 7pF).
>
> @aj: So just to recap you new have both C22 and C23 at 10pF and all is stable?
>
> As I said earlier my measurements probably don't amount to much. I again
> measured with the 10x probe setting. So even if we ignore the actual values
> the following was still apearent:
>
> - the signals on XTAL1 and XTAL2 are closer in amplitude on the old boards
> than on the new boards
> - exchanging C23 for a smaller value brought up the amplitude for XTAL2 on
> the new board
>
> All this was only measurable with the 10x probe setting; the measurements
> with the 1x setting did not show such a strong difference (but they almost
> always locked up the processor and showed a completely different waveform).
> If I'm reading this right my probe has a capacitance of around 100pF in the
> 1x range and around 15pF in the 10x range - which of course explains why
> measuring with the 1x setting does no good.
>
> @aj: I have the same issue. If I go anywhere near XTAL2 with my test gear the
> CPU halts/crashes. This is also the case on the original Œ367. Even with the
> probes in x10 mode. That aside the amplitude of the clock signal is
> significantly lower on the XTAL2 side.
>
> The new board with the different C23 value has been running now for over 12
> hours.
>
> @aj: I am going to try your suggested values with this board and then run
> through my routines to and see if I can get it to behave in an erratic way.
>
> I need to find out how to make screenshots from my scope (it's a digital
> one); or perhaps I'll simply take pictures with the camera.
>
> @aj: My scope is so old it has an interface for an RS-232 line printer! I
> think in this case a photo might be best!
>
> Cheers,
> Andy

[Non-text portions of this message have been removed]

Thanks for covering so much ground so far. This thread has been an enormous help so far. I'm following along closely!

Andy can you clarify why the NJR4560 is swapped out for TL072P? I see that there are other 4560s out there that are available, though perhaps something in the specs makes it incompatible. Of hand here one form Mouser:

http://ca.mouser.com/ProductDetail/NJR/NJM4560D/?qs=sGAEpiMZZMtCHixnSjNA6EIiaWUSjOTKRvX1zNG7jh4%3d

Hi,

Glad you are enjoying the entertainment! Seems after a 12 hour bashing the
errant board I have been trying to diagnose is now stable. Replacing C23
with a 10pF ceramic seems to hold the clock circuit stable. Seems that the
noise/instability in the clock signal was indeed an issue and all the other
pointers described in previous threads were influential, but not the cause
of the problem. I would therefore suggest, if you are having with problems
with instability (i.e. The patch banks jumping back to A1) then address the
issue as described above. Thanks for your help with this one, Malte!

One other thing. If you have used a transplanted ceramic resonator from the
original Œ367 do check the legs are still attached to the device. They have
a habit of breaking off in a way that looks like they are still connected.
If possible use the blue Murata types listed on the group buy site.

Regards,
Andy

On 03/09/2010 16:51, "ML" <f115@rocketmail.com> wrote:
>
> Thanks for covering so much ground so far. This thread has been an enormous
> help so far. I'm following along closely!



[Non-text portions of this message have been removed]

Hi,

As before with other components the op-amp would have been selected because
of general availability and the fact that the spec of the TL072 meets or
exceeds the spec of the original device. I would be interested to here if
anyone has used this device in their circuit and whether it makes any
difference.

Cheers,
Andy

On 03/09/2010 21:20, "ML" <f115@rocketmail.com> wrote:

>
>
>
>
>
> Andy can you clarify why the NJR4560 is swapped out for TL072P? I see that
> there are other 4560s out there that are available, though perhaps something
> in the specs makes it incompatible. Of hand here one form Mouser:
>
> http://ca.mouser.com/ProductDetail/NJR/NJM4560D/?qs=sGAEpiMZZMtCHixnSjNA6EIiaW
> USjOTKRvX1zNG7jh4%3d
>
>
>



[Non-text portions of this message have been removed]

Oh ok - fair enough. I'm just trying to differentiate which parts are attempting to remain true to the original and if there's a technical reason why one might use a modern equivalent. Thanks again!

Well let¹s put it this way, without the corporate press release. If you use
a TL072 it will work and it won¹t change the authentic sound of the P6 we
all seek!

Cheers,
Andy

On 03/09/2010 23:16, "ML" <f115@rocketmail.com> wrote:

>
>
>
>
>
> Oh ok - fair enough. I'm just trying to differentiate which parts are
> attempting to remain true to the original and if there's a technical reason
> why one might use a modern equivalent. Thanks again!
>



[Non-text portions of this message have been removed]

Just read this post below! I am having problems with instability with my patch banks jumping around. When I hit the switches for Programs 5-8 the patch changes to A1, B2, C3, or D4 (for my selection of 5, 6, 7, 8 respectively).

I just picked up a Korg Polysix that the previous owner informed me had work done on it. I'm wondering if a new board was installed? How can I tell whether or not I have a clone board in there? Everything looks super clean.

--- In PolySix@yahoogroups.com, Andrew Jury <andy@...> wrote:
>
> Hi,
>
> Glad you are enjoying the entertainment! Seems after a 12 hour bashing the
> errant board I have been trying to diagnose is now stable. Replacing C23
> with a 10pF ceramic seems to hold the clock circuit stable. Seems that the
> noise/instability in the clock signal was indeed an issue and all the other
> pointers described in previous threads were influential, but not the cause
> of the problem. I would therefore suggest, if you are having with problems
> with instability (i.e. The patch banks jumping back to A1) then address the
> issue as described above. Thanks for your help with this one, Malte!
>
> One other thing. If you have used a transplanted ceramic resonator from the
> original Œ367 do check the legs are still attached to the device. They have
> a habit of breaking off in a way that looks like they are still connected.
> If possible use the blue Murata types listed on the group buy site.
>
> Regards,
> Andy
>
>
> On 03/09/2010 16:51, "ML" <f115@...> wrote:
> >
> > Thanks for covering so much ground so far. This thread has been an enormous
> > help so far. I'm following along closely!
>
>
>
> [Non-text portions of this message have been removed]
>

Hello,

The KLM-367 clone is signed "djn & jed 2009" at the back of it (see picture) :

http://img839.imageshack.us/img839/6033/clonesignature.jpg

--- In PolySix@yahoogroups.com, "ieinzz" <ieinzz@...> wrote:
>
> Just read this post below! I am having problems with instability with my patch banks jumping around. When I hit the switches for Programs 5-8 the patch changes to A1, B2, C3, or D4 (for my selection of 5, 6, 7, 8 respectively).
>
> I just picked up a Korg Polysix that the previous owner informed me had work done on it. I'm wondering if a new board was installed? How can I tell whether or not I have a clone board in there? Everything looks super clean.
>
>
> --- In PolySix@yahoogroups.com, Andrew Jury <andy@> wrote:
> >
> > Hi,
> >
> > Glad you are enjoying the entertainment! Seems after a 12 hour bashing the
> > errant board I have been trying to diagnose is now stable. Replacing C23
> > with a 10pF ceramic seems to hold the clock circuit stable. Seems that the
> > noise/instability in the clock signal was indeed an issue and all the other
> > pointers described in previous threads were influential, but not the cause
> > of the problem. I would therefore suggest, if you are having with problems
> > with instability (i.e. The patch banks jumping back to A1) then address the
> > issue as described above. Thanks for your help with this one, Malte!
> >
> > One other thing. If you have used a transplanted ceramic resonator from the
> > original Œ367 do check the legs are still attached to the device. They have
> > a habit of breaking off in a way that looks like they are still connected.
> > If possible use the blue Murata types listed on the group buy site.
> >
> > Regards,
> > Andy
> >
> >
> > On 03/09/2010 16:51, "ML" <f115@> wrote:
> > >
> > > Thanks for covering so much ground so far. This thread has been an enormous
> > > help so far. I'm following along closely!
> >
> >
> >
> > [Non-text portions of this message have been removed]
> >
>

Hi,

You can easily tell if you have a clone board fitted. You¹ll find that there
are six trimmer controls in the centre of the board. If they are round and
have cross heads this is an original board. If they are rectangle and have
slotted adjustment heads this a clone board. In order to fix this problem
you will need to replace C23 on the board with a 10nF ceramic disk
capacitor. Any electronics tech or even you should be able to do this if you
can solder!

Let us know how you get on!

Cheers,
Andy

On 04/09/2010 20:52, "ieinzz" <ieinzz@yahoo.com> wrote:

>
>
>
>
>
> Just read this post below! I am having problems with instability with my
> patch banks jumping around. When I hit the switches for Programs 5-8 the patch
> changes to A1, B2, C3, or D4 (for my selection of 5, 6, 7, 8 respectively).
>
> I just picked up a Korg Polysix that the previous owner informed me had work
> done on it. I'm wondering if a new board was installed? How can I tell
> whether or not I have a clone board in there? Everything looks super clean.
>



[Non-text portions of this message have been removed]

Turns out I don't have the clone board. Underneath it looks worked on, isn't signed, and I have the cross head trimmers on the topside. My problem seems to be exactly the issue with the new boards however. I'm getting instability in the programmer section with the write switch writing to different programs and program switches 5-8 selecting banks and different programs. Are there any known issues with C23 on the original boards? I'm looking at C23 on my board and it's a small ceramic reading "22". I'd have no trouble doing basic soldering/desoldering but it appears I'm close to getting over my head!

--- In PolySix@yahoogroups.com, Andrew Jury <andy@...> wrote:
>
> Hi,
>
> You can easily tell if you have a clone board fitted. You¹ll find that there
> are six trimmer controls in the centre of the board. If they are round and
> have cross heads this is an original board. If they are rectangle and have
> slotted adjustment heads this a clone board. In order to fix this problem
> you will need to replace C23 on the board with a 10nF ceramic disk
> capacitor. Any electronics tech or even you should be able to do this if you
> can solder!
>

Hi,

No, this issue does not apply to the original board. This could be caused by
all sorts of different problems. Something you might want to check is to
make sure the ceramic resonator, marked X1 on the board, is still connected
to it¹s legs. These have a tendency to come apart especially during transit.
Otherwise it is back to the usual disintegrating tracks caused by the
battery leak problem. If you are at the end of your tether and can¹t find
anyone to have a look at this problem then drop me a private message.
Otherwise any good synth tech should be able to deal with this.

Cheers,
Andy

On 05/09/2010 21:36, "ieinzz" <ieinzz@yahoo.com> wrote:

>
>
>
>
>
> Turns out I don't have the clone board. Underneath it looks worked on, isn't
> signed, and I have the cross head trimmers on the topside. My problem seems
> to be exactly the issue with the new boards however. I'm getting instability
> in the programmer section with the write switch writing to different programs
> and program switches 5-8 selecting banks and different programs. Are there any
> known issues with C23 on the original boards? I'm looking at C23 on my board
> and it's a small ceramic reading "22". I'd have no trouble doing basic
> soldering/desoldering but it appears I'm close to getting over my head!
>



[Non-text portions of this message have been removed]

On a battery-damaged reconditioned board, there are endless possibilities.
The synth was working when I repaired it (for whoever you purchased it
from) -- however, it's not uncommon for new problems to surface.
The easiest thing to do is send the board back to me, so I can see
what's what & what isn't.
Also, the serial number of the synth would be helpful, so I can check my
notes & find out when I worked on this & what was done (I've done about
sixty of these things, and I can't keep them in memory).

~GMM
analogsynthservice.com


ieinzz wrote:

>
>
> Turns out I don't have the clone board. Underneath it looks worked on,
> isn't signed, and I have the cross head trimmers on the topside. My
> problem seems to be exactly the issue with the new boards however. I'm
> getting instability in the programmer section with the write switch
> writing to different programs and program switches 5-8 selecting banks
> and different programs. Are there any known issues with C23 on the
> original boards? I'm looking at C23 on my board and it's a small
> ceramic reading "22". I'd have no trouble doing basic
> soldering/desoldering but it appears I'm close to getting over my head!
>
> --- In PolySix@yahoogroups.com <mailto:PolySix%40yahoogroups.com>,
> Andrew Jury <andy@...> wrote:
> >
> > Hi,
> >
> > You can easily tell if you have a clone board fitted. You¹ll find
> that there
> > are six trimmer controls in the centre of the board. If they are
> round and
> > have cross heads this is an original board. If they are rectangle
> and have
> > slotted adjustment heads this a clone board. In order to fix this
> problem
> > you will need to replace C23 on the board with a 10nF ceramic disk
> > capacitor. Any electronics tech or even you should be able to do
> this if you
> > can solder!
> >
>
>

[Non-text portions of this message have been removed]

Sixty! And here I thought I was getting to be an old pro at these after
dealing with eight.
Don B.

_____

From: PolySix@yahoogroups.com [mailto:PolySix@yahoogroups.com] On Behalf Of
klosmon
Sent: Sunday, September 05, 2010 5:35 PM
To: PolySix@yahoogroups.com
Subject: Re: [PolySix] Re: Problem with new KLM-367A




On a battery-damaged reconditioned board, there are endless possibilities.
The synth was working when I repaired it (for whoever you purchased it
from) -- however, it's not uncommon for new problems to surface.
The easiest thing to do is send the board back to me, so I can see
what's what & what isn't.
Also, the serial number of the synth would be helpful, so I can check my
notes & find out when I worked on this & what was done (I've done about
sixty of these things, and I can't keep them in memory).

~GMM
analogsynthservice.com





[Non-text portions of this message have been removed]

G,

I'll send it in to you. When I was looking at the underside I noticed one of the jumper wires had broken off at an end. Lifting it up the other end broke off! I soldered it back to the two points but this didn't fix/change anything.

I've got your email and will shoot you something in a minute.

Ian

I just have to give credit to Andy and Malte for figuring this error out.

I am very impressed. You guys are true pros.

I have updated my web page and the parts list to note this error and the
fix.

Best regards,

Jed


[Non-text portions of this message have been removed]

Hi Everyone,

check the reset line and do the reset adjustment per the service manual


On 2017-04-25, 06:16 rnw6ueo2ru3grihf5iabauttna56367yo4n5umk4@yahoo.com
[PolySix] wrote:

> [Attachment(s) <#TopText> from
> rnw6ueo2ru3grihf5iabauttna56367yo4n5umk4@yahoo.com [PolySix] included
> below]
>
> Hi Everyone,
>
> I am currently rebuilding my Polysix and have replaced the battery
> damaged KLM-367 with one of Jed's new KLM-367A boards. Everything worked
> great for a while but recently the front LEDs have locked on Programs 1
> and 8. None of the Bank LEDs are lit, nor is the Manual button, and they
> do not respond when pressed. I tried resetting the circuit per the
> service manual, but to no avail as the LEDs will not return to Bank A
> and Program 1.
>
> Per this thread I also tried replacing C23 with a 10pF ceramic
> capacitor. While this did bring the amplitude of XTAL 1 and XTAL 2
> closer together, and I was able to confirm a 6MHz input at each point,
> it did not solve the issue. I then probed TP4 (Clock) and noticed there
> was a weak signal that did not resemble a clock waveform. Could this
> mean my RAM chip is defective? Has anyone else experienced this issue
> before? I have attached photos of the scope readings for reference.
>
> Thank You,
> Daniel
>
>
>
>

Thank you for the reply Johannes. I've just checked the reset line and all of the components within it test ok. I've also completed the reset adjustment per the service manual, but still cannot get Bank A and Program 1 to light up once I remove the 56k resistor from KLM-376. The D/A adjustment steps that follow work as they should, but once those are completed and the synth is powered off/on again the LEDs remain stuck on Programs 1 and 8.

Hi, where must be connected de 56K resistor to adjust reset circuit?
I have an battery leak KLN367 that i try to repair and need this info.
Thank you in advance.
Regards

There is an empty 2 pin male connector on the power supply board, KLM-376, and this is where you connect the 56k resistor. Under the "Files" tab you can download the Service Manual and it will take you through the reset procedure.

Ok, thank you very much, go to DL service manual.