Q On Chassis Ground vs Digital Ground vs Analog Ground - Help!

[Paul Marossy]

So I'm enjoying this old 1982 Korg Polysix synth that my mom gave me a few weeks ago. I did some minor repairs to it and it's working great with one exception. After about an hour, the sound at the output jack just mysteriously suddenly cuts out completely or starts to distort before it cuts out, in either case there is no sound at all. BUT the headphone jack still works fine, and they are both receiving the same signal from the same source (Signal Out from KLM-368 board, 10th page of PDF file linked below, lower right corner).

So I looked at The Schematic, and on the KLM-369 board (9th page of the PDF, upper right corner) there is the output jack directly connected to a 10K volume pot thru an attenuation switch. Now I replaced this output jack with a standard mono jack because it was damaged and everything seemed to be good until this weird problem started happening. I was looking at the schematic this morning I noticed there is a symbol for a chassis ground for the headphone jack and on the output jack the ground is marked as "GA", which I assume is analog ground. All of the IC chips in the unit appear to be connected to chassis ground as well. So I assume that the problem lies in the replacement output jack that I used. I did check the power supply voltages, they were all right on. And the fact that the headphones work (and everything else on the keyboard) does not seem to indicate that it has anything to do with a power supply issue.

All that to ask: does that mean that the output jack's ground should be isolated from the chassis ground? And if so, why would it cause the output jack to suddenly cut out after about an hour? Just trying to understand what the heck could cause this bizzare (to me) behaviour...

[CynicalMan]

If you look at Page 11, the power supply, GND and GND A are connected. So I'd be testing that connection at the output jack, making sure that they are still connected. My guess is that there's some sort of intermittent joint that's breaking the connection between the two. It's weird that it happens after an hour, maybe a heat thing? Anyway, that does mean that you should be able to connect the two grounds together elsewhere, you might just get noise from the loop or from the chips. So you could wait for the problem to happen, then try jumpering together the output and headphone grounds.

[Paul Marossy]

If you look at Page 11, the power supply, GND and GND A are connected. So I'd be testing that connection at the output jack, making sure that they are still connected. My guess is that there's some sort of intermittent joint that's breaking the connection between the two. It's weird that it happens after an hour, maybe a heat thing? Anyway, that does mean that you should be able to connect the two grounds together elsewhere, you might just get noise from the loop or from the chips. So you could wait for the problem to happen, then try jumpering together the output and headphone grounds.

On page 11, Connector CN-13 has a chassis ground for the +/-5V & +/-15V stuff, but the output jack is labeled differently, as "GA" on the schematic. And I just noticed that the attenuation switch ground is also connected to "GA", as well as IC1. Maybe this is just for reference only? I guess a few minutes with a continuity checker could confirm all that. It appears that the only place that GND & GND A should connect is at that power supply board.

I don't understand why it would do this after about an hour. If it was a dodgy connection it should be a problem all the time, not just after an hour of operation. I was initially thinking about it being a heat related thing, which made me think power supply, but that's apparently not it. It works fine again after being off for maybe ten minutes, which makes this a really strange problem.  

My original hunch, which I haven't mentioned thus far, is that perhaps there is a bad electolytic cap at IC1 on the KLM-369 board that somehow affects the signal at the output jack but still allows for operation of the headphone jack. Does that make any sense? That opamp for the headphone jack is a 4556 Dual "Decompensated" opamp, not sure what that means in the overall scheme of things.

[brett]

Hi
my 2c says that you have got it with "perhaps there is a bad electolytic cap at IC1 on the KLM-369 board that somehow affects the signal at the output jack but still allows for operation of the headphone jack."
Messing with old, "dry-ish" electros is not for the feint-hearted.
good luck

[Paul Marossy]

I took the keyboard apart again last night and looked at the output and headphone jacks. The headphone jack is a plastic type which is isolated from the chassis ground. The broken output jack I replaced was obviously a replacement and not the original - I could tell by the size of the hole in the mounting bracket that it was not the original jack. The original one apparently was a plastic isolated type jack too. So I replaced it with a similar type of isolated jack and so far so good.

I think probably the only reason they have those "GA" grounds seperated from the chassis/common ground is to keep the noise level as low as possible.

Anyway, if it does the signal cutting out thing again, then those caps are my next suspect and I will change them to see if that fixes it.

[R.G.]

Remember: Ground is all about controlling where the current goes.

[Paul Marossy]

Remember: Ground is all about controlling where the current goes.

Yes, with respect to keeping noise levels low, I learned that especially by building a few tube amps from scratch and star grounding existing ones. It's amazing what a difference it can make to pay careful attention to grounding in these types of equipment.