Uh-oh

jamos

Just had a strange thing happen.. second time in the past two months, though on a different patch.

Basically: On a couple of patches if I push polyphony too far, the synth will let me create additional voices: but then I get nothing from the output but harsh digital noise. With the perf I'm working on now, 5 voices in one slot are OK, but if I go to 6, chaos. The scary thing is, last night I had it running on 7 with no problem.

Has anyone ever heard of this? I'll upload the perf later. OS 1.4, of course.

blue hell · Posted: Wed Feb 13, 2013 8:34 am Post subject:

Interesting Shocked

Yup, I'd like to see a patch for that.

Tim Kleinert · Posted: Wed Feb 13, 2013 11:12 am Post subject:

Are delay lines involved?

I got hideous noise once, but only sometimes in specific situations. I was able to pinpoint it down to a defective RAM memory chip. Every DSP has its own RAM memory chip for delay stuff. And if a patch has delay lines in its audio path and it (or some of its voices) are allocated to a DSP with faulty RAM, this can produce loud digital garbage. If you reload stuff, or change voice count, it can change or disappear, as the patch is allocated elsewhere.

jamos · Posted: Wed Feb 13, 2013 7:12 pm Post subject:

{sigh} I guess there's no point in uploading a patch, since now Welcome G2 exhibits the problem. Polyphony up to 13, fine; polyphony 14 or up, no good. I have recorded the audio output and attached, for what it's worth.

Tim, yes all of these either use delay, reverb, or string oscillators. It sounds like you were able to fix yours - how did you isolate the RAM chip and how did you do the repairs?

How f'ing ironic. My G2s have been basically unused for a year, I sign up to do a noodle installation at an art show as a way to restart my engine and now this happens.

Dying G2.mp3

Description:

Output from my G2 when polyphony is too high.

Download (listen)

Filename:

Dying G2.mp3

Filesize:

39.18 KB

Downloaded:

1179 Time(s)

varice · Posted: Thu Feb 14, 2013 1:18 am Post subject:

Ouch! That sure sounds bad.

Tim was relatively lucky. It turned out that he had a defective RAM chip on the expansion board, much better than a problem with the main board! A replacement expansion board fixed his problem.

Since you hear the problem only when trying to use maximum polyphony, your problem *may* also be on your expansion board. Probably the best way to make sure is to remove your expansion board and then use a test performance patch to force the main board to max DSP cycles *and* max RAM usage. If you’re willing to do this test but need help with a test patch, just let us know.

There may be a way to test your G2 without removing the expansion board, but I am not sure that there is a way to isolate where the fault is (main board or expansion) just by loading a perf.

A quick way to isolate the fault to a RAM chip though would be to remove the delay and reverb modules from the Welcome G2 patch and then check if there is any noise when all voices are sounding at the max poly setting.

Good luck jamos…

Tim Kleinert · Posted: Thu Feb 14, 2013 1:40 am Post subject:

I pinpointed the problem by making a performance with a 2.7 seconds delay line in each VA and FX area, 8 in total. Thus I could see that one of those was generating noise. (It was also interesting that by reloading the performance, the offending delay module often got "shifted around", seems that the DSP assignment has some kind of round-robin type scheme.)
Then I further isolated the issue by removing the expansion board, which made the problem disappear in my case.

As I say, this is only my experience, and in my case it was related to a faulty RAM chip. The performance described above will easily reveal if your problem is the same.

jamos · Posted: Thu Feb 14, 2013 12:54 pm Post subject:

Thanks, guys. I'll try that patch and possibly pull the expansion card tonight.

3phase · Posted: Thu Feb 14, 2013 1:23 pm Post subject:

jamos · Posted: Thu Feb 14, 2013 9:26 pm Post subject:

Well, bullocks. Poopie. Etc.

I created a test patch (attached), and determined that yes, it is coming from the delay module. I suppose I could have been a bit more rigorous, but I'm going on Tim's idea here. Unlike his case, I found that the failed slot does not move around; it is always the FX area of Slot 1.

The bad news: the problem does not go away when I pull the expansion board. So it is on the main board.

Feces. Fornication!

Time to call Clavia and see if they have any thoughts on repair. Or does anyone else have an idea?

RAM test.prf2

Description:

Download (listen)

Filename:

RAM test.prf2

Filesize:

3.32 KB

Downloaded:

4184 Time(s)

blue hell · Posted: Fri Feb 15, 2013 7:57 am Post subject: