It’s been almost a year since I began working on the first ARP Axxe I restored, and I was amazed by how much more smoothly this second one went, though it’s not really surprising as I have worked on over 70 synths including a half dozen other ARPs in the year since then. Still, it’s always cool to have a reminder like this of how much I have learned in that short time. I have fixed several times as many synths in the year since I officially launched Bell Tone as I did in the 7 or so years I was studying/building/fixing electronics before that.
I’m learning more electronics theory all the time, but a lot of what I’m learning is practical, synth-specific repair strategies, and about the idiosyncrasies of specific synths… things you don’t see when you look at a schematic.
Some of my less-obvious tips for restoring any ARP synthesizer, things I didn’t read elsewhere and had to figure out myself:
- I think it’s well known common practice by now (as it should be) to replace ALL tantalum capacitors in ARPs and replace filtering electrolytics in the PSU, but I also suggest that you increase the value of the filtering capacitors if it’s one of the synths where they were only 220 uf (I do 470 uf instead). You will also probably see something like two 22 uf tantalums in parallel on each board being used to decouple the power rails on ARPS with the “bigger” power supply. You can replace these with a single electrolytic and increase the value to 100 uf.
- If it has the “bigger” power supply design that the Avatar, Omni, Quadra, etc. have, seriously consider replacing the PSU altogether. The transients on power up/down are nuts and will just continue to cause mayhem. There is an official ARP field change that you can do to improve this situation, but I haven’t been satisfied with it.
- When desoldering, use the lowest temperature that you can manage (as I described in my desoldering tutorial post). Be very careful never to touch the tip of the soldering or desoldering iron you’re using directly to the traces of the PCB.
- Replace the large, cylindrical yellow polypropylene capacitor if it has one. This is the sample-and-hold capacitor in Axxes, Odysseys, and Avatars. A half dozen ARPS later, we’ve seen failed ones of these so many times that we always replace them now. If this is bad, you’ll hear each note’s pitch start sliding out of tune as soon as you play it. Illinois Capacitor part number 474MPW400K is a great replacement because it even looks like the original! Digikey and Mouser both carry it.
- Replace CD4011s used in envelope generators and possibly any other RCA A-series CMOS, LM723 (used in smaller PSU), LM1458s and LM741s op amps if you are feeling like being really thorough. The CD4011s have to be replaced with the unbuffered version that is available now, CD4011UB. If you use the buffered version CD4011BE you will hear a click on every key press. In my experience, all of the above have a fairly high chance of failing. You will likely have some bad ones and more may fail during the restoration because of all the times you have to turn it on and off while testing. It used to be I would find a few dead op amps in every “big ARP” I worked on, and figured out it’s more efficient to just replace them up front.
- If you are getting double triggering-on your two-buss Pratt Read keyboard and struggling to get all your J-wire key contacts lined up, I figured out what helps the most is gently pulling the lower part of the key contact outward a tiny bit to change the angle of the J-wire relative to the bar… not just changing the distance of the J-wire from the bar by bending the top.
- Reflow ALL pin connectors and add more solder to them. At first I would find that the synths would sometimes work and sometimes not work every time I reconnected a board… I’d get confused and think there was a serious issue… it was just these stupid pin connectors having cold solder joints, so now I do them all preemptively.
- Don’t freak out if you get a loud thump on pressing each key after doing the recap… you just need to adjust your VCA CV reject trimmer.
Work done: full recap and PSU rebuild with new regulator and thermal compound, replaced above-mentioned unreliable ICs, reflowed solder on pin connectors. Cleaned Pratt-Read keyboard contacts and buss bars, straightened all J-wires (this is the worst part). Cleaned all sliders (this is a tedious process that I won’t get into here), replaced one broken slider, calibrated all parameters.