Discussion about the Korg PolySix synthesizer group photo

Yahoo Groups archive

Discussion about the Korg PolySix synthesizer

Archive for polysix.

Index last updated: 2026-03-30 01:17 UTC

Thread

Another P6 with all LED's lit....

Another P6 with all LED's lit....

2016-07-02 by noddyspuncture@hotmail.com

Hi folks, been a while since I was here, but I have another question for the P6 experts amongst us.

I looked at this issue a while back and then shelved it, but now am revisiting it to hopefully solve the issue.


My P6 had the battery problem and was professionally repaired before I bought it and it worked great for a long time. Then is started acting up - finally settling down to having all LEDs lit and with the usual mangled sound.


I have confirmed that there are no broken tracks and scoped all IC pins and there is activity. I just bought myself a logic probe but my problem is that basically I don't understand multiplexing and although I can see whether or not there are high's and low's, I cant decide if that is what they should be... especially on the main 8048 processor.


On the main processor there is activity, both on my scope and using my new logic probe, I have a waveform on ALE and also the crystal is running with a nice waveform on pins 2 and 3.


WR & RD read high on my probe constantly and on TP4 (Pin 32) there is 2.76 volts and a very strange looking washed out waveform on my scope.


Pin 1 of connector-11 has nothing on it - as do all the other connector-11 pins - so the CPU seemingly isn't scanning the panel controls.


My main concern is around IC27 though. Pins 5,6 & 7 are all sitting at 1.995 volts (I know 6 & 7 are connected together) and both my logic probe and scope are dead when probing them. I am getting a healthy pulsing 'high' on pin 1 though... which feeds the inputs of the 4051's.


I get a feeling that because of a fault here I can't complete the setting procedure outlined for VR's 1,6 & 7 in the manual - as all LED's stay lit...!


Reset (pin 4 of the 8048) is working fine - probed with my logic probe. I can add the 56Kohm resistor, set VR1 for a 'low' on pin 4 and it goes to 'high' when the resistor is removed, so that looks fine.


I have swapped out IC27... albeit for a normal dual op-amp (I have a 4560 on order) but it's just the same.


So I am beginning to suspect the HA17408P D/A converter. Would anyone agree, going off my findings please? Has anyone ever had a D/A converter go on a P6...?


Cheers,

Tom











Re: [PolySix] Another P6 with all LED's lit....

2016-07-02 by <tomlinson6050@roadrunner.com>

First thought: Your tech did not find and correct ALL the battery gunk problems. So often the best solution (albeit, a time consuming one), is to build and install a'clone' board. Your time and money might be better spent going with the KIWI-6 upgrade board.
------------------------------------------------------




---- "noddyspuncture@hotmail.com [PolySix]" <PolySix@yahoogroups.com> wrote:

=============
Hi folks, been a while since I was here, but I have another question for the P6 experts amongst us.
I looked at this issue a while back and then shelved it, but now am revisiting it to hopefully solve the issue.


My P6 had the battery problem and was professionally repaired before I bought it and it worked great for a long time. Then is started acting up - finally settling down to having all LEDs lit and with the usual mangled sound.


I have confirmed that there are no broken tracks and scoped all IC pins and there is activity. I just bought myself a logic probe but my problem is that basically I don't understand multiplexing and although I can see whether or not there are high's and low's, I cant decide if that is what they should be... especially on the main 8048 processor.


On the main processor there is activity, both on my scope and using my new logic probe, I have a waveform on ALE and also the crystal is running with a nice waveform on pins 2 and 3.


WR & RD read high on my probe constantly and on TP4 (Pin 32) there is 2.76 volts and a very strange looking washed out waveform on my scope.


Pin 1 of connector-11 has nothing on it - as do all the other connector-11 pins - so the CPU seemingly isn't scanning the panel controls.


My main concern is around IC27 though. Pins 5,6 & 7 are all sitting at 1.995 volts (I know 6 & 7 are connected together) and both my logic probe and scope are dead when probing them. I am getting a healthy pulsing 'high' on pin 1 though... which feeds the inputs of the 4051's.


I get a feeling that because of a fault here I can't complete the setting procedure outlined for VR's 1,6 & 7 in the manual - as all LED's stay lit...!


Reset (pin 4 of the 8048) is working fine - probed with my logic probe. I can add the 56Kohm resistor, set VR1 for a 'low' on pin 4 and it goes to 'high' when the resistor is removed, so that looks fine.


I have swapped out IC27... albeit for a normal dual op-amp (I have a 4560 on order) but it's just the same.


So I am beginning to suspect the HA17408P D/A converter. Would anyone agree, going off my findings please? Has anyone ever had a D/A converter go on a P6...?


Cheers,
Tom

Re: [PolySix] Another P6 with all LED's lit....

2016-07-02 by noddyspuncture@hotmail.com

As I have checked and quadruple checked all tracks, traces and connections - and all are OK, I am assuming it's component failure..!? I had been suspecting the 8048 but after confirming the reset function is working I am hoping that it is some other component.

Vref + & - ... pins 14 & 15 of the A/D converter both  read millivolts... I'm thinking this might be a clue...?

By the way would anyone kn ow what I should be seeing on TO... pin 1 of the 8048 ...?

Cheers,
Tom

Re: [PolySix] Another P6 with all LED's lit....

2016-07-03 by Bob Grieb

Actually, I think essentially 0V at pins 14 and 15 of the DAC is correct. Pin 14 is a
current input, not a voltage one. Pin 5 of the 4560
should be at 2.0 volts DC, so check that. Probably it's fine.

T0 is used to sense the comparator output. When the comparator is high,
it should be at 5V. When the comparator is low, D9 clamps it at 0.6V below
ground, and D10 add 0.6V to that so you get 0V. Those are the two levels you
should see. The comparator is part of a feedback ADC. The DAC output is compared
with the front panel voltage being measured, and adjusted until its the same. At that pt,
the front panel voltage is known.

Bob

--------------------------------------------
On Sat, 7/2/16, noddyspuncture@hotmail.com [PolySix] <PolySix@yahoogroups.com> wrote:

Subject: Re: [PolySix] Another P6 with all LED's lit....
To: PolySix@yahoogroups.com
Date: Saturday, July 2, 2016, 3:11 PM

As I have checked and quadruple checked all tracks,
traces and connections - and all are OK, I am assuming
it's component failure..!? I had been suspecting the
8048 but after confirming the reset function is working I am
hoping that it is some other component.
Vref + & - ... pins 14 & 15
of the A/D converter both  read millivolts... I'm
thinking this might be a clue...?
By the way would anyone kn ow what I
should be seeing on TO... pin 1 of the 8048
...?
Cheers,Tom

Re: [PolySix] Another P6 with all LED's lit....

2016-07-03 by noddyspuncture@hotmail.com

Many thanks for that Bob,

So let me get this right - when turning front panel controls I should be seeing either 0v or 5v on 'T0'- pin1..?

I am getting no waveform on R37, the input to IC6 - which I think is wrong.

One other thing I forgot to mention is that on the 8048 DB0 - DB7 lines, my logic probe shows activity on all - except DB0  & DB3 (pins 12 and 15)... would you expect to see this...?

Cheers,
Tom


---In PolySix@yahoogroups.com, <bobgrieb@...> wrote :

Actually, I think essentially 0V at pins 14 and 15 of the DAC is correct. Pin 14 is a
current input, not a voltage one. Pin 5 of the 4560
should be at 2.0 volts DC, so check that. Probably it's fine.

T0 is used to sense the comparator output. When the comparator is high,
it should be at 5V. When the comparator is low, D9 clamps it at 0.6V below
ground, and D10 add 0.6V to that so you get 0V. Those are the two levels you
should see. The comparator is part of a feedback ADC. The DAC output is compared
with the front panel voltage being measured, and adjusted until its the same. At that pt,
the front panel voltage is known.

Bob

--------------------------------------------
On Sat, 7/2/16, noddyspuncture@... [PolySix] <PolySix@yahoogroups.com> wrote:

Subject: Re: [PolySix] Another P6 with all LED's lit....
To: PolySix@yahoogroups.com
Date: Saturday, July 2, 2016, 3:11 PM

As I have checked and quadruple checked all tracks,
traces and connections - and all are OK, I am assuming
it's component failure..!? I had been suspecting the
8048 but after confirming the reset function is working I am
hoping that it is some other component.
Vref + & - ... pins 14 & 15
of the A/D converter both  read millivolts... I'm
thinking this might be a clue...?
By the way would anyone kn ow what I
should be seeing on TO... pin 1 of the 8048
...?
Cheers,Tom

Re: [PolySix] Another P6 with all LED's lit....

2016-07-03 by Bob Grieb

Didn't you mention having an oscilloscope? If so,
why would you use a logic probe? The scope gives
you a lot more information if you know how to interpret it.

When the MCU is converting voltages, it executes a
"successive approximation" routine. This involves stepping
through each bit of the 8-bit digital value and then looking
at the comparator output to see if that bit should be 1 or 0 in
the digital result. This is done at CPU speeds, so the waveform
on T0 should be toggling between high and low at a high rate
as the mcu is constantly scanning the front panel, or at least
it should be. BTW, isn't it annoying that the Yahoo mail font
changes zeroes to look like an o.

It seems that your MCU may be getting stuck somewhere in
the code. It may be waiting for something that never happens,
in which case the DB0 and 3 values you see may be correct
for that piece of code. Is the reset signal sitting at a high level?
It should not be pulsing low.

Bob


--------------------------------------------
On Sun, 7/3/16, noddyspuncture@hotmail.com [PolySix] <PolySix@yahoogroups.com> wrote:

Subject: Re: [PolySix] Another P6 with all LED's lit....
To: PolySix@yahoogroups.com
Date: Sunday, July 3, 2016, 7:23 AM


Many thanks for that Bob,
So let me get this right - when
turning front panel controls I should be seeing either 0v or
5v on 'T0'- pin1..?
I am getting no waveform on R37, the
input to IC6 - which I think is wrong.
One other thing I forgot to mention
is that on the 8048 DB0 - DB7 lines, my logic probe shows
activity on all - except DB0  & DB3 (pins 12 and 15)...
would you expect to see this...?
Cheers,Tom

Re: [PolySix] Another P6 with all LED's lit....

2016-07-03 by noddyspuncture@hotmail.com

Yes Bob... pin 4 Reset is sitting 'high' (according to my logic probe) and I also measure 4.85v on it ... I can do the procedure successfully - adding the 56K resistor, setting VR1 so reset is just 'low' and when I remove the resistor it then jumps to 'high'.

All that works... so would you then say that my 8048 processor is good?
Could the MCU be getting stuck because it's faulty or would another fault elsewhere cause that?
Should I be looking elsewhere do you reckon?

I am seeing a sine wave of about 50mV on 'T0' - 8040 pin1 - that's not right is it?

Cheers,
Tom


---In PolySix@yahoogroups.com, <bobgrieb@...> wrote :

Didn't you mention having an oscilloscope? If so,
why would you use a logic probe? The scope gives
you a lot more information if you know how to interpret it.

When the MCU is converting voltages, it executes a
"successive approximation" routine. This involves stepping
through each bit of the 8-bit digital value and then looking
at the comparator output to see if that bit should be 1 or 0 in
the digital result. This is done at CPU speeds, so the waveform
on T0 should be toggling between high and low at a high rate
as the mcu is constantly scanning the front panel, or at least
it should be. BTW, isn't it annoying that the Yahoo mail font
changes zeroes to look like an o.

It seems that your MCU may be getting stuck somewhere in
the code. It may be waiting for something that never happens,
in which case the DB0 and 3 values you see may be correct
for that piece of code. Is the reset signal sitting at a high level?
It should not be pulsing low.

Bob


--------------------------------------------
On Sun, 7/3/16, noddyspuncture@... [PolySix] <PolySix@yahoogroups.com> wrote:

Subject: Re: [PolySix] Another P6 with all LED's lit....
To: PolySix@yahoogroups.com
Date: Sunday, July 3, 2016, 7:23 AM


Many thanks for that Bob,
So let me get this right - when
turning front panel controls I should be seeing either 0v or
5v on 'T0'- pin1..?
I am getting no waveform on R37, the
input to IC6 - which I think is wrong.
One other thing I forgot to mention
is that on the 8048 DB0 - DB7 lines, my logic probe shows
activity on all - except DB0  & DB3 (pins 12 and 15)...
would you expect to see this...?
Cheers,Tom

Re: [PolySix] Another P6 with all LED's lit....

2016-07-03 by Bob Grieb

Concerning your sine wave question, you didn't tell me the DC level,
and I don't know that the inputs to the comparator are doing, so it's hard to
say what that might mean. It could mean that you are not providing a good
ground to the scope and nothing more. 50mV is not enough to change a 0 to a 1
so the CPU will ignore it if the DC level is correct. Knowing the frequency of
the sine wave would really help to understand why it's there. If it's 60 Hz,
that's one thing. If it's 10 MHz, that's something else.

As I said before, I don't know why anyone would use a logic probe if they have
an oscilloscope. You would just be removing what could be useful information.

I can't say for sure that the 8048 is good, but if I had to simply guess based on
previous experience, I would say that it probably is fine.

Did you try using the test switch? If the CPU reacts to a change in that switch,
then that would tell us something. Also, turning that switch on puts the code into
a fairly tight loop, which might enable testing some other things.

What do you have on pin 32 of the MCU? Is there any sort of pulse on it?
(With the scope, not the logic probe)

Bob

--------------------------------------------
On Sun, 7/3/16, noddyspuncture@hotmail.com [PolySix] <PolySix@yahoogroups.com> wrote:

Subject: Re: [PolySix] Another P6 with all LED's lit....
To: PolySix@yahoogroups.com
Date: Sunday, July 3, 2016, 10:06 AM


Yes Bob... pin 4 Reset is sitting 'high'
(according to my logic probe) and I also measure 4.85v on it
... I can do the procedure successfully - adding the 56K
resistor, setting VR1 so reset is just 'low' and
when I remove the resistor it then jumps to
'high'.
All that
works... so would you then say that my 8048 processor is
good?Could the MCU be getting stuck because
it's faulty or would another fault elsewhere cause
that?Should I
be looking elsewhere do you reckon?

I am seeing a sine wave of
about 50mV on 'T0' - 8040 pin1 - that's not
right is it?

Cheers,Tom

Re: [PolySix] Another P6 with all LED's lit....

2016-07-03 by Bob Grieb

I would also check the INTb pin, pin 6 and see if it's sitting low.

If you turn on the test switch, and then power cycle the synth, if the CPU is working
it should go into a tight loop of code where it just converts channel 0 (Eff Spd/Int) and
chan 1 (filter cutoff) over and over in a loop. You should see the DAC output change to follow
those two voltages as you change the pots. If that is working, then the MCU is OK.

If it's not, then troubleshooting just that loop should be easier than the main control loop.

Bob

--------------------------------------------
On Sun, 7/3/16, noddyspuncture@hotmail.com [PolySix] <PolySix@yahoogroups.com> wrote:

Subject: Re: [PolySix] Another P6 with all LED's lit....
To: PolySix@yahoogroups.com
Date: Sunday, July 3, 2016, 10:06 AM

Yes Bob... pin 4 Reset is sitting 'high'
(according to my logic probe) and I also measure 4.85v on it
... I can do the procedure successfully - adding the 56K
resistor, setting VR1 so reset is just 'low' and
when I remove the resistor it then jumps to
'high'.
All that
works... so would you then say that my 8048 processor is
good?Could the MCU be getting stuck because
it's faulty or would another fault elsewhere cause
that?Should I
be looking elsewhere do you reckon?

I am seeing a sine wave of
about 50mV on 'T0' - 8040 pin1 - that's not
right is it?

Cheers,Tom

Re: [PolySix] Another P6 with all LED's lit....

2016-07-03 by noddyspuncture@hotmail.com

Thanks for the reply Bob,

On pin 32 with the scope there is "something" but it looks like a small washed out or smudged waveform. By that I mean not a nice crisp square or sine - more like a blurred 'audio' signal would. If I then switch the 'test' switch it gets smaller in amplitude and then seems to dance around a bit...!

One thing I did notice when investigating data from the front panel is that CN11-6 (INH 1) is constantly low and CN11-7 (INH 2) is constantly high. Should these not alternate between low and high...?

Cheers,
Tom

Re: [PolySix] Another P6 with all LED's lit....

2016-07-03 by Bob Grieb

In test mode, only one would be active, since both channels being measured are in the same chip.
In normal mode, both signals would be toggling, as the CPU samples all of the pots.

Terms like washed out and smudged are not helping me to understand what you are seeing.
Also, simply not triggering your scope properly could generate waveforms that would fit these
descriptions.

If you trigger the scope properly, a waveform won't "dance around". It may get wider, or have
its amplitude vary.

Did you look at the DAC output when in test mode?

Starting to think I am not helping much here.

Bob

--------------------------------------------
On Sun, 7/3/16, noddyspuncture@hotmail.com [PolySix] <PolySix@yahoogroups.com> wrote:

Subject: Re: [PolySix] Another P6 with all LED's lit....
To: PolySix@yahoogroups.com
Date: Sunday, July 3, 2016, 2:41 PM


Thanks for the reply Bob,
On pin 32 with the scope there is
"something" but it looks like a small washed out
or smudged waveform. By that I mean not a nice crisp square
or sine - more like a blurred 'audio' signal would.
If I then switch the 'test' switch it gets smaller
in amplitude and then seems to dance around a
bit...!
One thing I
did notice when investigating data from the front panel is
that CN11-6 (INH 1) is constantly low and CN11-7 (INH 2) is
constantly high. Should these not alternate between low and
high...?
Cheers,Tom

Re: [PolySix] Another P6 with all LED's lit....

2016-07-04 by noddyspuncture@hotmail.com

Hi Bob,

I uploaded a few photos of my oscilloscope shots into a folder called "Tom's" in the photos section here. I named them to include the time & volts settings but unfortunately the names were cut off... so I've included the time & volts division settings for each one below.

I'm not sure what you mean about "power cycling"... do you mean repeatedly turning the synth off and on in between turning the controls...?

Nothing happens to the DAC output waveform when I turn those controls in TEST mode.
And Pin 6 on the 8048 is sitting HIGH

Pin 32 has a waveform - the one I called 'washed out' or 'blurred'. It isn't an actual pulse like the one on pin 11, as you can see... it's not a single line but a mush of weak looking lines all together. That is how I tried explaining the difference.

When I switch the test switch it stays much the same just becomes smaller in amplitude.

The oscilloscope shots I've uploaded are -
Pin11: ALE (2v/.5uS)
Pin1: T0     (10mV/.5uS)
Pin3: Xtal   (50mV/.5uS)
Pin4: Reset (20mV/.5uS)
Pin6: Inh     (50mV/.5uS)
Pin32: in normal mode  (50mV/.5uS)
Pin32: in test mode    (50mV/.5uS)
DAC pin4: output  (1V/2uS) 


Cheers,
Tom



---In PolySix@yahoogroups.com, <bobgrieb@...> wrote :

I would also check the INTb pin, pin 6 and see if it's sitting low.

If you turn on the test switch, and then power cycle the synth, if the CPU is working
it should go into a tight loop of code where it just converts channel 0 (Eff Spd/Int) and
chan 1 (filter cutoff) over and over in a loop. You should see the DAC output change to follow
those two voltages as you change the pots. If that is working, then the MCU is OK.

If it's not, then troubleshooting just that loop should be easier than the main control loop.

Bob

--------------------------------------------
On Sun, 7/3/16, noddyspuncture@... [PolySix] <PolySix@yahoogroups.com> wrote:

Subject: Re: [PolySix] Another P6 with all LED's lit....
To: PolySix@yahoogroups.com
Date: Sunday, July 3, 2016, 10:06 AM

Yes Bob... pin 4 Reset is sitting 'high'
(according to my logic probe) and I also measure 4.85v on it
... I can do the procedure successfully - adding the 56K
resistor, setting VR1 so reset is just 'low' and
when I remove the resistor it then jumps to
'high'.
All that
works... so would you then say that my 8048 processor is
good?Could the MCU be getting stuck because
it's faulty or would another fault elsewhere cause
that?Should I
be looking elsewhere do you reckon?

I am seeing a sine wave of
about 50mV on 'T0' - 8040 pin1 - that's not
right is it?

Cheers,Tom

Re: [PolySix] Another P6 with all LED's lit....

2016-07-04 by noddyspuncture@hotmail.com

By the way Bob,

The trigger options I have on my scope are AC/DC/HF/LF/Line

I am trying them all each time I scope a pin and the only ones I see of any use are the AC and the HF ones...!

Cheers
Tom

Re: [PolySix] Another P6 with all LED's lit....

2016-07-04 by Bob Grieb

ALE looks great.

OK, the signal on pin 1 is 6 MHz, so it's probably simply capacitive
coupling from the adjacent pin 2, which is the main oscillator signal.
In other words, this signal is not changing state.

With crystal oscillators, unfortunately, many times the schem doesn't
show which is input and which is output. One will be a smaller signal
closer to a sine wave (that's the input to the osc) the other will be a larger
signal, closer to a square wave. That's the output. I suspect you are on
the input. Better to look at the output. Also, scoping the output is usually
safe, but putting a scope on the input can stop the circuit from oscillating
or change its frequency. Anyway, I think the oscillator is probably OK.

Power cycling just means turning the power off and then back on once. Not
over and over or anything.

What are you using as the scope ground?

Also, when displaying waveforms, I think it's best to first show the waveform
in a scale that matches how the CPU sees it. In other words, pick a place on
the scope for ground and use that same place for all of your pictures, so that
we know where ground is. Then set the scope to 1 volt/div, since these are
5V logic levels. That way we know right away if the signal will be considered
a "1" or a "0". If the picture shows a low amplitude high freq noise, I can't tell if
that's at 0V DC (low) or if it's at 4V (high). THe hf noise is just the 6 MHz of the
main oscillator, which makes me wonder how you are grounding the scope probe.
Also, we wouldn't normally look at a digital signal at 20mV/div unless we are
interested in the noise level. At this time we are more interested in whether
signals are sitting low or high, so zooming in to show us the noise is not really
that helpful. All digital systems will have some noise on the signals, but it's
usually ignored.

Pin 6 is INT, not INH. Looks like it is not switching, but I can't tell from your
photo if it's 0 or 1.

Pin 4 of the DAC is a current output, so you can't really scope it. We are
interested in the voltage output of the DAC, which would be the output of the
op amp that is fed by pin 4. The number isn't on the schem that I can see,
but I think it's pin 1 of IC27. That would be the one to look at with the scope.

So anyway, sorry but I can't tell much from your photos as they are sort
of focused on the noise in the system, which is the thing that digital systems
are designed to ignore. We need to zoom out and look at these signals from
a 0-5V distance, not see how much noise is on them.

We need more waveforms like ALE, even if they are flat lines. That will tell us
what the logic levels are.

If you do take more photos, please include the DAC voltage output, and P20 and P23.
Those are signal names, not pin numbers.

The "normal" trigger mode in many situations would be DC, IMO. THen you should have
a control called trigger level, and also a +/- selector. WIth these controls, you select
either a positive-going or negative-going section of the waveform as the trigger point,
and also the exact voltage level at which to trigger. If you don't use the trigger level
control, then AC might be the only mode that would work, since you are not setting
the level to match the DC level of the waveform.

Some of your photos look like the result of not triggering the scope at a consistent
place in the waveform.

Bob


--------------------------------------------
On Mon, 7/4/16, noddyspuncture@hotmail.com [PolySix] <PolySix@yahoogroups.com> wrote:

Subject: Re: [PolySix] Another P6 with all LED's lit....
To: PolySix@yahoogroups.com
Date: Monday, July 4, 2016, 7:25 AM

Hi Bob,
I
uploaded a few photos of my oscilloscope shots into a folder
called "Tom's" in the photos section here. I
named them to include the time & volts settings but
unfortunately the names were cut off... so I've included
the time & volts division settings for each one
below.
I'm not
sure what you mean about "power cycling"... do you
mean repeatedly turning the synth off and on in between
turning the controls...?
Nothing happens to the DAC output
waveform when I turn those controls in TEST
mode.And Pin 6 on the 8048 is sitting
HIGH
Pin 32 has a
waveform - the one I called 'washed out' or
'blurred'. It isn't an actual pulse like the one
on pin 11, as you can see... it's not a single line but
a mush of weak looking lines all together. That is how I
tried explaining the difference.
When I switch the test switch it
stays much the same just becomes smaller in
amplitude.
The oscilloscope shots I've
uploaded are -
Pin11: ALE (2v/.5uS)
Pin1: T0     (10mV/.5uS)
Pin3: Xtal   (50mV/.5uS)
Pin4: Reset (20mV/.5uS)
Pin6: Inh (50mV/.5uS)
Pin32: in normal mode (50mV/.5uS)
Pin32: in test mode (50mV/.5uS)
DAC pin4: output  (1V/2uS) 

Cheers,Tom

Re: [PolySix] Another P6 with all LED's lit....

2016-07-04 by Bob Grieb

Hopefully my other post will show up here. I don't see it yet.
One more comment. If you do manage to get a decent picture of the
DAC voltage waveform, in test mode please, change the horizontal
timebase to see all of the waveform before it repeats. If you are too
zoomed in, you won't get a stable waveform. If the CPU is working,
you should see a number of voltage levels in a repeating pattern.
You should set the scope so that we can see the start of the pattern
at the left, then the whole pattern, then at the right edge of the scope
we should see it starting the next loop. This way we can see how many
voltage levels there are in a loop, and how long it takes the loop to complete.

You should use DC trigger mode, and set the trigger level to a level that
only occurs once in the loop, if you can. So either near the highest
part of the waveform, or near the lowest part. If you pick a voltage near
the middle of the waveform to trigger on, the waveform will cross this value
many times in the loop and it will be impossible to get a nice clean scope
image.

On a simple waveform like ALE, it's much easier to get a nice stable scope
image.

Oh. If you do get DAC voltage showing different
steps, we will want to see what T0 input is doing at the same time. Do you
have two channels on your scope? To show two signals, you would need to
change to 2V/div so that each signal has half of the screen.

Bob


--------------------------------------------
On Mon, 7/4/16, noddyspuncture@hotmail.com [PolySix] <PolySix@yahoogroups.com> wrote:

Subject: Re: [PolySix] Another P6 with all LED's lit....
To: PolySix@yahoogroups.com
Date: Monday, July 4, 2016, 7:33 AM

By the way Bob,
The trigger options I have on my
scope are AC/DC/HF/LF/Line
I
am trying them all each time I scope a pin and the only ones
I see of any use are the AC and the HF
ones...!
CheersTom

Re: [PolySix] Another P6 with all LED's lit....

2016-07-04 by Ingo Debus

> Am 04.07.2016 um 13:33 schrieb noddyspuncture@hotmail.com [PolySix] <PolySix@yahoogroups.com>:
>
> The trigger options I have on my scope are AC/DC/HF/LF/Line

Surely it also has a switch to toggle between Normal and Auto trigger mode? On digital scope sometimes it’s hidden in a menu. Often you get better triggering results in Normal mode. You have to fiddle a bit with the trigger level knob to get a picture at all in Normal mode.

Ingo

Re: [PolySix] Another P6 with all LED's lit....

2016-07-04 by noddyspuncture@hotmail.com

Thanks for all the info Bob,

I am using the "big black wire" coming from near the battery area for all my grounds, so I think that should be OK...?

You've given me a lot to digest and go at here... I shall try and get it all sorted and upload more shots.

Many thanks for your help..!

Cheers,
Tom

Re: [PolySix] Another P6 with all LED's lit....

2016-07-04 by Bob Grieb

Yes, that should be a good ground. BTW, if I am interpreting the DAC pin 4 photo
correctly, It certainly looks like some data is being written to it, which would be a good
sign. We can tell more when you get a good picture of the op amp output (in test mode).

I suggest that the first priority should be trying to determine if the 8048 is OK. If it is,
then we can try to determine where the problem lies. If it's not, then you will need to
find a replacement. I like the idea of using the test switch as it greatly simplifies the code
loop, which will help us to understand what's happening. The main loop is much more complicated.

Trying to troubleshoot from a distance is pretty tough, but maybe we can make some progress.

Bob

--------------------------------------------
On Mon, 7/4/16, noddyspuncture@hotmail.com [PolySix] <PolySix@yahoogroups.com> wrote:

Subject: Re: [PolySix] Another P6 with all LED's lit....
To: PolySix@yahoogroups.com
Date: Monday, July 4, 2016, 2:30 PM


Thanks for all the info Bob,
I am using the "big black
wire" coming from near the battery area for all my
grounds, so I think that should be OK...?
You've given me a lot to digest
and go at here... I shall try and get it all sorted and
upload more shots.
Many thanks for your
help..!
Cheers,Tom

Re: [PolySix] Another P6 with all LED's lit....

2016-07-05 by noddyspuncture@hotmail.com

Hi Bob,

I made two more photos... of the DAC (IC27 pin1) and the T0 at the same time.

Dual trace, scope set to 2v and using DC trigger and tweaking the level until the waveforms showed up. There is only one spot on the level control where they show, it's quite critical.

There are two shots... one at 1uS time div and one at 50uS time div.

Switching the test mode switch doesn't make any difference to these at all.

On my scope I cannot find any 'horizontal timebase' switch so I couldn't do that part... and the slope +/- switch didn't seem to do much either, it only seemed to change the brightness...!

I also measured the voltages on these pins with my meter - IC27 pin1 has 4.67v in normal and 7.72v in test modes. The T0 pin1 has -9mV

Is this info any good...?

Cheers
Tom



---In PolySix@yahoogroups.com, <bobgrieb@...> wrote :

Hopefully my other post will show up here. I don't see it yet.
One more comment. If you do manage to get a decent picture of the
DAC voltage waveform, in test mode please, change the horizontal
timebase to see all of the waveform before it repeats. If you are too
zoomed in, you won't get a stable waveform. If the CPU is working,
you should see a number of voltage levels in a repeating pattern.
You should set the scope so that we can see the start of the pattern
at the left, then the whole pattern, then at the right edge of the scope
we should see it starting the next loop. This way we can see how many
voltage levels there are in a loop, and how long it takes the loop to complete.

You should use DC trigger mode, and set the trigger level to a level that
only occurs once in the loop, if you can. So either near the highest
part of the waveform, or near the lowest part. If you pick a voltage near
the middle of the waveform to trigger on, the waveform will cross this value
many times in the loop and it will be impossible to get a nice clean scope
image.

On a simple waveform like ALE, it's much easier to get a nice stable scope
image.

Oh. If you do get DAC voltage showing different
steps, we will want to see what T0 input is doing at the same time. Do you
have two channels on your scope? To show two signals, you would need to
change to 2V/div so that each signal has half of the screen.

Bob


--------------------------------------------
On Mon, 7/4/16, noddyspuncture@... [PolySix] <PolySix@yahoogroups.com> wrote:

Subject: Re: [PolySix] Another P6 with all LED's lit....
To: PolySix@yahoogroups.com
Date: Monday, July 4, 2016, 7:33 AM

By the way Bob,
The trigger options I have on my
scope are AC/DC/HF/LF/Line
I
am trying them all each time I scope a pin and the only ones
I see of any use are the AC and the HF
ones...!
CheersTom

Re: [PolySix] Another P6 with all LED's lit....

2016-07-05 by noddyspuncture@hotmail.com

Hi again Bob,

I've now also added three photos... Pin6 Int - with the scope at 2v div
Also Pins21 & 24 (P20 & P23) both also at 2v/5uS divisions... DC trigger.

Hope they are now OK...?

Cheers,
Tom

Re: [PolySix] Another P6 with all LED's lit....

2016-07-05 by noddyspuncture@hotmail.com

I made a typo Bob... that should be 4.67v and 4.72v (not 7.72v) on IC27 pin1...!

Cheers,
Tom


---In PolySix@yahoogroups.com, <noddyspuncture@...> wrote :

Hi Bob,

I made two more photos... of the DAC (IC27 pin1) and the T0 at the same time.

Dual trace, scope set to 2v and using DC trigger and tweaking the level until the waveforms showed up. There is only one spot on the level control where they show, it's quite critical.

There are two shots... one at 1uS time div and one at 50uS time div.

Switching the test mode switch doesn't make any difference to these at all.

On my scope I cannot find any 'horizontal timebase' switch so I couldn't do that part... and the slope +/- switch didn't seem to do much either, it only seemed to change the brightness...!

I also measured the voltages on these pins with my meter - IC27 pin1 has 4.67v in normal and 7.72v in test modes. The T0 pin1 has -9mV

Is this info any good...?

Cheers
Tom



---In PolySix@yahoogroups.com, <bobgrieb@...> wrote :

Hopefully my other post will show up here. I don't see it yet.
One more comment. If you do manage to get a decent picture of the
DAC voltage waveform, in test mode please, change the horizontal
timebase to see all of the waveform before it repeats. If you are too
zoomed in, you won't get a stable waveform. If the CPU is working,
you should see a number of voltage levels in a repeating pattern.
You should set the scope so that we can see the start of the pattern
at the left, then the whole pattern, then at the right edge of the scope
we should see it starting the next loop. This way we can see how many
voltage levels there are in a loop, and how long it takes the loop to complete.

You should use DC trigger mode, and set the trigger level to a level that
only occurs once in the loop, if you can. So either near the highest
part of the waveform, or near the lowest part. If you pick a voltage near
the middle of the waveform to trigger on, the waveform will cross this value
many times in the loop and it will be impossible to get a nice clean scope
image.

On a simple waveform like ALE, it's much easier to get a nice stable scope
image.

Oh. If you do get DAC voltage showing different
steps, we will want to see what T0 input is doing at the same time. Do you
have two channels on your scope? To show two signals, you would need to
change to 2V/div so that each signal has half of the screen.

Bob


--------------------------------------------
On Mon, 7/4/16, noddyspuncture@... [PolySix] <PolySix@yahoogroups.com> wrote:

Subject: Re: [PolySix] Another P6 with all LED's lit....
To: PolySix@yahoogroups.com
Date: Monday, July 4, 2016, 7:33 AM

By the way Bob,
The trigger options I have on my
scope are AC/DC/HF/LF/Line
I
am trying them all each time I scope a pin and the only ones
I see of any use are the AC and the HF
ones...!
CheersTom

Re: [PolySix] Another P6 with all LED's lit....

2016-07-05 by Bob Grieb

Timebase is just another way to say horizontal time per division, which
you are adjusting. BTW, your scope probably has an adjustment called
"trace rotation". Usually a hole where you poke a small screwdriver to make
the trace level instead of sloped like yours is.

I am not sure what to tell you. Your pictures show that nothing much is happening.
You should cycle power after switching into test mode. I assume that you did that.
If the CPU was hung up in the main loop, and then you change the switch, it might
have no effect.

If reset is high, and not glitching low, and the oscillator output pin has a decent
square wave or something close to it at 6 MHz, with good logic levels like 0 and 4V,
then the CPU should be running, and a lot more should be happening. Did you check
the power and ground pins to the CPU? That would have been the first step. I forget
to suggest that. Pins 40,26,5, 20, and 7 should all be checked with your scope since you
have one. And check the oscillator output pin with the scope to insure a good main
clock to the CPU. I think that is pin 2 (not the pin that you scoped earlier).

Bob

--------------------------------------------
On Tue, 7/5/16, noddyspuncture@hotmail.com [PolySix] <PolySix@yahoogroups.com> wrote:

Subject: Re: [PolySix] Another P6 with all LED's lit....
To: PolySix@yahoogroups.com
Date: Tuesday, July 5, 2016, 10:17 AM


I made a typo Bob... that should be 4.67v and 4.72v
(not 7.72v) on IC27 pin1...!
Cheers,Tom


---In PolySix@yahoogroups.com,
<noddyspuncture@...> wrote :

Hi
Bob,
I made two more
photos... of the DAC (IC27 pin1) and the T0 at the same
time.
Dual trace,
scope set to 2v and using DC trigger and tweaking the level
until the waveforms showed up. There is only one spot on the
level control where they show, it's quite
critical.
There are
two shots... one at 1uS time div and one at 50uS time
div.
Switching the
test mode switch doesn't make any difference to these at
all.
On my scope I
cannot find any 'horizontal timebase' switch so I
couldn't do that part... and the slope +/- switch
didn't seem to do much either, it only seemed to change
the brightness...!
I
also measured the voltages on these pins with my meter -
IC27 pin1 has 4.67v in normal and 7.72v in test modes. The
T0 pin1 has -9mV
Is
this info any good...?
CheersTom

Re: [PolySix] Another P6 with all LED's lit....

2016-07-05 by Bob Grieb

If you want to mail me your 8048, I can test it in my P6.
You would need to pry it out carefully, a little bit at a time,
alternating between the two ends with a small screwdriver, and
have some anti-static foam, or some styrofoam wrapped in a
single layer of alum foil, to plug the chip into as soon as it is
free of the socket. And try not to touch any of the pins. Just
the body of the chip. Ground yourself before prying out the chip
by touching ground on the synth.

Or maybe there is someone on the list who is near you?
Not sure what else to suggest. My gut feeling is that the chip is
probably good, but the waveforms certainly don't seem to imply that.

If the EA pin is high, that would totally mess it up. Your ALE
waveform pretty much means that the crystal oscillator is fine,
I think.

I replaced the KLM-367 in my P6, even though I had fixed the old board
and it was working OK. Hard to know how long the replacement boards
will be available, and the price is very attractive. It's a fair amount of work
to build up the new board, since you have to move some parts from the old
one. But then you never have to worry about battery damage again.

Bob

Re: [PolySix] Another P6 with all LED's lit....

2016-07-06 by noddyspuncture@hotmail.com

Yes Bob, all those power pins to the 8048 are fine...!

Although I am actually getting waveforms and "activity" around the 8048, it all seems to be in the millivolts and not at logic levels as you say.

The Xtal osc is fine... but there isn't anything on the 'clock' TP pin 32. And Reset is sitting high.

I have checked each and every 8048 pin to see of it connects to all destinations on the schematic and it's all fine so I think it must be component failure somewhere, the problem is finding it.

I am assuming that other IC's having failed elsewhere could also give the symptoms I have...?

I suppose if I buy one of those new boards and transfer all my IC's onto it I'll still have the same problem which would be annoying to say the least...!?

I have a friend with a Polysix - I'll see if he'll let me swap the 8048's.

Cheers,
Tom









---In PolySix@yahoogroups.com, <bobgrieb@...> wrote :

Timebase is just another way to say horizontal time per division, which
you are adjusting. BTW, your scope probably has an adjustment called
"trace rotation". Usually a hole where you poke a small screwdriver to make
the trace level instead of sloped like yours is.

I am not sure what to tell you. Your pictures show that nothing much is happening.
You should cycle power after switching into test mode. I assume that you did that.
If the CPU was hung up in the main loop, and then you change the switch, it might
have no effect.

If reset is high, and not glitching low, and the oscillator output pin has a decent
square wave or something close to it at 6 MHz, with good logic levels like 0 and 4V,
then the CPU should be running, and a lot more should be happening. Did you check
the power and ground pins to the CPU? That would have been the first step. I forget
to suggest that. Pins 40,26,5, 20, and 7 should all be checked with your scope since you
have one. And check the oscillator output pin with the scope to insure a good main
clock to the CPU. I think that is pin 2 (not the pin that you scoped earlier).

Bob

--------------------------------------------
On Tue, 7/5/16, noddyspuncture@... [PolySix] <PolySix@yahoogroups.com> wrote:

Subject: Re: [PolySix] Another P6 with all LED's lit....
To: PolySix@yahoogroups.com
Date: Tuesday, July 5, 2016, 10:17 AM


I made a typo Bob... that should be 4.67v and 4.72v
(not 7.72v) on IC27 pin1...!
Cheers,Tom


---In PolySix@yahoogroups.com,
<noddyspuncture@...> wrote :

Hi
Bob,
I made two more
photos... of the DAC (IC27 pin1) and the T0 at the same
time.
Dual trace,
scope set to 2v and using DC trigger and tweaking the level
until the waveforms showed up. There is only one spot on the
level control where they show, it's quite
critical.
There are
two shots... one at 1uS time div and one at 50uS time
div.
Switching the
test mode switch doesn't make any difference to these at
all.
On my scope I
cannot find any 'horizontal timebase' switch so I
couldn't do that part... and the slope +/- switch
didn't seem to do much either, it only seemed to change
the brightness...!
I
also measured the voltages on these pins with my meter -
IC27 pin1 has 4.67v in normal and 7.72v in test modes. The
T0 pin1 has -9mV
Is
this info any good...?
CheersTom

Re: [PolySix] Another P6 with all LED's lit....

2016-07-06 by Bob Grieb

So far, the only signals we have seen that looked reasonable
were ALE and the oscillator input. You have lots of signals
that show very small wiggles that look like capacitive coupling
from either the oscillator or ALE. So far, I haven't seen pics of
anything else that's changing. If the CPU were stuck in reset,
that would explain it. You did say something about the data bus pins
changing, except for 2 of them. But didn't post any photos of that.
Were they swinging 0-4V? Of course with the data bus, you need to
see who is driving it. If the RAM is enabled, that data may be coming
from it, and not from the MCU.

Anyway, hope you can try that 8048 in another Polysix. That would
certainly be a useful test.

Not sure what other chips could have failed that would cause the MCU
outputs to be so quiet. In the test mode code, it should be writing values
to the DAC over and over in a loop. The only thing it checks is the T0 input,
and that won't cause it to stall, just changes the digital value.

--------------------------------------------
On Wed, 7/6/16, noddyspuncture@hotmail.com [PolySix] <PolySix@yahoogroups.com> wrote:

Subject: Re: [PolySix] Another P6 with all LED's lit....
To: PolySix@yahoogroups.com
Date: Wednesday, July 6, 2016, 8:00 AM

Yes Bob, all those power pins to the 8048 are
fine...!
Although I am
actually getting waveforms and "activity" around
the 8048, it all seems to be in the millivolts and not at
logic levels as you say.
The Xtal osc is fine... but there
isn't anything on the 'clock' TP pin 32. And
Reset is sitting high.
I have checked each and every 8048
pin to see of it connects to all destinations on the
schematic and it's all fine so I think it must be component
failure somewhere, the problem is finding
it.
I am assuming that other
IC's having failed elsewhere could also give the
symptoms I have...?
I suppose if I buy one of those new
boards and transfer all my IC's onto it I'll still
have the same problem which would be annoying to say the
least...!?
I have a
friend with a Polysix - I'll see if he'll let me
swap the 8048's.
Cheers,Tom

Re: [PolySix] Another P6 with all LED's lit....

2016-07-06 by jw_dewdney@yahoo.com

that sounds remarkably similar to the problems i was having a few months back! in my case there was at least one slightly out of spec resistor in a critical place ... i was using a 22k in place of a 20k for some unknown reason... goes to show that sometimes the problem can be more subtle than one suspects...well its not something i would have suspected.