Korg Polysix MG Issue

[cwc]

After many weeks refurbishing my "new" old Polysix I am very nearly there! I have one last issue before the final tune up and I wondered if anyone had any ideas. All knobs, switches and buttons on the synth work and do what they should. The KLM-367 board is a brand new clone with new clone CPU so I assuming that it is ok.

The issue is that when I turn the Polysix on from cold, the MG frequency is properly set as designated by the preset but "leaks through" at a higher level than the preset's actual values causing a significant wobble in the tone. This happens even when changing presets.Turning the MG frequency and MG level controls to 0 will not entirely eliminate this. At the same time, the arpeggiator pulse leaks into the tone causing "catches" in the tone that can be varied by turning the Arp Speed control. However, all these issues gradually fade as the synth warms up and after a minute or two and the synth works perfectly. On switching the synth off and on again when warm there is a slight MG wobble in the tone on pressing a key which disappears in seconds. Then all is good.

This suggests to me that a component(s) is faulty and is “leaky” but starts to function correctly when it has warmed up. Any ideas?

One other thing I would like to know is whether how long the synth typically takes to initialise. I have noticed that when cold the buttons light for around 8-10 seconds before the synth initialises to preset A1, but when it is warm it initialises instantaneously.

[Mooger5]

Sounds like cold solder joints, or solder bridges, or residue. Onlyafter materials expand they finally make, or break, cintact. The mg is in the klm367. Check the area.
The thick green wire must be soldeted to grounf the 367 and the upper board. Keeps the noisefloor low.

[cwc]

Hi Mooger 5. Many thanks for the response. I am hoping KLM367 is ok as I bought it as a populated board. Since my original post I have noticed that if I disconnect the harness from CN05 from the clone KLM367 (i.e. the connector comes from the patch select, write enable/disable and tape enable/disable part of the KLM371 Programmer Board), patch A.1 comes up without the problematic modulation when I turn the synth on.

[Mooger5]

Makes sense. Does that 367 have new headers. The ones I replaced mine with dont fit as tight as the originals.. In particular thr L shaped 90 degree headers. That sit horizontally next to the oscillatorc board must be pushed hard for the best grip.

[Mooger5]

You should also test if the fault gets stored in memory I think. Save a patch while the unit is cold. Then when the unit warms up and the fault disappears, call up the same patch. I think thereś nothing wrong with the MG itself, but somewhere between the processor and the oscillators.
Does it affect the vcf and the vca as well?
Is the mod wheel potentiometer clean?

[cwc]

Thanks once more, Mooger5. I removed the headers from the original KLM367 with my desolder gun and soldered them in to the clone board. I have checked and all the connectors are fully seated and the connections are good. As you suggest I will test if the fault gets stored in memory when the unit is cold and report back. Since my original post I have noticed that if I disconnect the harness from CN05 on the clone KLM367 (i.e. the connector that comes from the patch select, write enable/disable and tape enable/disable part of the KLM371 Programmer Board), patch A.1 comes up without the problematic modulation when I turn the synth on. I can't quite see how that board could cause the modulation symptom unless (as someone suggested elsewhere) something on that board is somehow pulling voltages down on startup. So, I will be double checking the voltages with and without CN5 harness connected and will report back on that too.

[cwc]

Update: I can see from monitoring the power supply voltages that the "modulation" symptoms occur while the voltage is increasing up to 5v after turning the synth on and disappear when a stable 5v is reached. This takes a significant amount of time (1 minute+) when turning the synth on from cold and a split second when turning the synth on when warm. I tried the KLM boards connected in various combinations and found that the 5v supply is being dragged down by the KLM366 voice board. There are no obviously faulty electrolytic caps on that board (no noticeable bulges, corrosion or goo) but as they can drag voltages down and replacing them is cheap and easy for me (I have a vacuum desoldering gun) that is what I will do next. I will then report back.

[Mooger5]

It´s a good effort but frankly a serious reason for alarm would be if the unit acted up after warmup, not before, right? That would be a sure sign something was definitely not OK. This is being optimistic, but don´t forget it´s a vintage analogue synth. It´s an old and temperamental Korg and that´s part of the charm. Even with some new parts in, don´t expect it to behave like it´s fresh from the factory. Even then the service manual warns that it takes around 10 minutes of warmup time for every circuit to stabilize. Take a look at the PSU. Both digital and certain analog sections such as the filter chips are powered by the same 5V rails.
Anyway, you could use freezing spray to detect any possible culprit. If something is sucking up more current than it should, then it´s dissipating that current into heat.

[cwc]

Thanks Mooger5 for your support so far. Recent death in the family, hence no updates for a while. It certainly puts life and hobbies into perspective. Enjoy it while you can, I say. Anyway, I solved the Polysix MG issue. Replacing the electrolytic capacitors on KLM-366 Voice Board did not work – replacing the Key Assigner IC was the answer. Just a hunch that paid off. Strangely, as per Mooger5's suggestion I had tried freezer spray including on the original Key Assigner IC when the synth was warm working as it should, but it did not provoke the issue to reappear. Anyway, the hunch paid off and it's sorted now. One last question: The synth's EG/Gate switch seems to work fine, but I have been a little thrown by the User Manual which says: “This [gate-type] envelope is not related to the EG in any way”: When I set the switch to gate, there is definitely a change and a gate effect. However, I can sometimes alter the sound with the ADSR when the switch is in gate position. Is this right? Or should the User Manual be taken literally? If there is something wrong, does anyone have any ideas what the cause might be?

[Mooger5]

Hello. Sorry to hear about your loss. Life goes on, man.

Bad solder joints can happen on recently assembled boards. I've done a few in my projects. Sometimes we trace a fault down to some component thinking it's bad and promptly replace it, redoing the joints that were the actual cause.

The gate bypasses the EG CV to the VCA. The CV to the VCF remains active. What you're probably hearing is the filter opening and closing with slow attack and decay, giving the impression of changes in amplitude, so it seems to be normal behaviour. Only if you set and hear a long release sound is a sign the gate function is not right, as the sound should end abruptly.

All the best.