The restoration of this ARP Omni 2 was the first thing I posted when I started this blog. I wasn’t quite finished yet when I posted it, but I thought I was, so I decided to just go ahead and do the post. Ha, ha.
Now I’m finally actually done.
Basically, the Omni had a power supply fault that went undetected during most of my work time because it was so intermittent, but would cause the voltage to spike by 3 volts on both the positive and negative rails (that means an additional -3 on the negative rail). CMOS chips already don’t really like the +/-15 they’re being given here, and +/-18 can be fatal for them. I had replaced the electrolytic capacitors and performed the ARP field notice change (replacing R12 with a 47Kohm resistor per the notes here) and had repeatedly tested the PSU output, monitoring it with the scope, and it appeared fine, for long stretches of time, even under load.
Here are some (not even all) of the things that failed, and that I fixed, during that time, with their symptoms. I want to put this here in case anyone else ever has any of the same problems, so that I can spare someone a little bit of this madness.
- LM301 (Z2) on PSU board – when this op amp failed (twice), it caused nearby PSU resistors to start smoking, so I also had to replace some resistors including a 0.68 ohm 1 W one.
- LM1458 (Z50) on upper left area of lower voicing board – this is the summing op amp that sums the lower voices. 8’ disappeared on both the string and synth section. This actually burned itself out twice… when the LM301 in the PSU failed, this would fail.
- 4001 (Z51) on left side of lower voicing board – bass voice disappeared entirely (both footages)
- LM1458 (Z3) on string control board – you wouldn’t get any sound until you moved the attack slider at least ¼ of the way to the right; if the release slider was too far to the right, you wouldn’t get any sound; if you did get any sound, the release would always be long.
- 2N2907 transistors (Q28 and Q29) on lower left of lower voicing board – one or both of these was shorting negative voltage through; correct voltage was not being sent through string section release slider (range was -13V to -2V instead of -15V to + 8V), leading to release always being long
- 2N3904 transistor (Q1) on bass voice board – shorting negative voltage through transistor caused one shot input into 4007 Z1 to only go between -15V and -13V or so, instead of 15V and 0V, preventing bass ADR from triggering
- 4025 (Z38) on upper right of lower voicing board – after I got the bass mostly working again, this NOR gate was allowing two signals to pass through two of its pins, causing two bass notes to be sent to two of the keys, causing those keys to trill between the two notes
- 4007 (Z5) on bass voice board – ADR was not being applied to bass VCF or VCA
- 4011 on synth board – synth ADSR stopped working entirely
- 4069 (Z45-47) hex inverter on lower voice board – a chunk of 6 consecutive bass section notes stopped working – if a group of 6 stops working this is probably why. This is part of a clever/stupid series of logic gates that runs the “bass note priority” function.
So after I fixed all this stuff, there was still one issue remaining — an intermittent loud humming noise and the sound of all notes ringing at once that would suddenly enter the output. I left it for last just because I was procrastinating about working on an intermittent issue. I hadn’t though it was the PSU because I had watched it with the scope under load and it had seemed fine. But I finally caught it in the act with its voltages spiking and everything made sense. And I realized this was probably the reason this Omni had so many problems to begin with. The culprit was the MJE 1100 power transistor, which I hadn’t replaced because they are no longer available, and because as far as I had discerned they’re not one of the “best practices” things that people are advised to replace in Omnis.
At this point though, taking into account what Tim from Retrolinear had told me about how they ALWAYS entirely replace Omni power supplies and just wanting to give this poor synth a fresh start, I decided to cut my losses and replace the entire PSU board (minus the transformer) with one we had already built for modular synth testing, a low-noise, LM337/317 regulator, +/-15V adjustable one from a PCB designed by Ray Wilson. I tuned it down to +/-14 to give the CMOS a little bit of breathing room. I only regret not doing it sooner… I’ll be replacing the PSUs in all Omnis from now on.
I can’t believe it’s done, and when I took its goodbye photo I was sad that it doesn’t have slider caps. I wish it looked as good as I feel about finally being done with it.
That said… bring me your Omni to fix! I now know how to fix dozens of specific Omni problems and I want to see if I can beat my previous time. Meanwhile, I’ll be working on three more ARPs, two Avatars and an Odyssey.