ARP power supply repair

[Jim Alfredson]

I recently purchased a collection of ARP gear from an estate. Included in the lot are four ARP Omni II, four Odysseys, three Avatars, one Solina, one Pro Soloist II, and one ARP Sequencer. I also got two Chroma Polaris, a Korg MiniKorg, and a Korg MonoPoly. And the cool Hammond Auto-Vari 64 drum machine!

 

I started working on the Sequencer since I will sell that first to recover some of my investment. Almost all the ARPs of this era use a similar power supply that produces +/- 15vDC.

 

Before turning the Sequencer on, I opened the top and disconnected the power supply from the main board. I then tested the +15v and -15v rail to ground for shorts. No power was applied to the PSU at this time. I found that the -15v rail was shorted to ground. My first suspicion is always the tantalum capacitors on the PSU. They are notorious for dying and when they die they short.

 

Sure enough, I removed the tantalum on the negative rail and tested with my meter and it was shorted. In the picture below you can see the PSU and the two bright blue tantalum capacitors.

 

 

I went ahead and replaced both tantalums because I don't trust them, even though the +15v rail was not showing a short. I also replaced the electrolytics with slightly higher values (330uf instead of 250uf). I then plugged the PSU in and hooray! No shorts. I confirmed that both rails were outputting the correct voltages.

I then tested ground to both rails on the connector of the main board and didn't discover any shorts. But every single slider was gunked up to the point that they wouldn't move. This is also a common issue on ARP gear.

 

I decided to desolder all the sliders, soak them in isopropyl alcohol for an hour and then put them in the ultrasonic cleaner. This process worked wonderfully on the recently restored EML ElectroComp 400 sliders. Unfortunately these ARP sliders were just too gross. I even ran them through the ultrasonic cleaner again and they were still nasty. So I decided to open them all up and manually clean them by hand, lubricate, re-assemble, and finally re-attach to the main board. It took me about 6 hours total to do that work on 19 sliders.

 

After reinstalling the sliders, I attached the PSU and fired it up and it worked! I confirmed all the outputs were working, all sliders, the quantizer, everything was working great! Relatively easy fix!

 

The next day a friend came over and I wanted to show him the sequencer, so I plugged it in and got nothing. The front light was on but no other LEDs and no activity. Bummer.

 

I opened it back up and tested the power supply. It was outputting 0 volts. I checked the fuse and sure enough it was blown. The front light was working because it is wired before the fuse. Kind of dumb on ARP's part but whatever.

 

I replaced the fuse and tested the PSU, once again disconnected from the mainboard. The +15v rail was fine but the -15v rail was outputting almost -36v!!!! Ouch! I was testing the voltages coming from various parts of the board when the fuse blew again. No surprise.

 

I traced the issue to a bad TIP29 transistor on the negative rail. Replacement yielded a working PSU again. I replaced the TIP29 on the positive rail as a precaution. They are almost 50 years old after all.

 

I am waiting for a new fuse holder to arrive. The original fuse was soldered to the board. I temporarily soldered a chassis mount fuse holder in place for testing but I have a PCB mount holder coming that will work much better.

 

Another problem with this Sequencer is that all the slider caps are missing. In fact, most of the slider caps on all my ARPs are missing. My overall goal is to replace the sliders with new high precision Bourns sliders, which will need new caps anyway. But what caps to use?

The most common replacement caps for the last 10 years or so are the Mountain Switch 10TB series. Unfortunately those are currently in “end of life” status and suppliers like Mouser only have black and minimal red ones available.

 

After much searching, I found these little rubber mini toggle switch caps and wondered if they would work. As you can see by the pictures, they work great for both the original CTS sliders and the Bourns! I used a heat gun to get them to work on the CTS but it took only 10 seconds of gentle heating. No big deal.

 

You can find these at CE Distribution for 19¢ a piece. A great deal, imo. And they come in all colors.

 

 

 

 

Once the fuse holder arrives, I will install and leave the sequencer on for several hours to make sure it is stable. If so, I'll list it on Reverb and move on to the next ARP!

[The Real MC]

ARPs do not age well.  Lot of work to restore but they sound great.  That's a heck of a pile of ARPs!

Beware the Omni and Avatar - they both use the defective power supply that overshoots +/-15V rails on power up, known to blow CMOS ICs.  There's a factory bulletin to add voltage clamps to the outputs, highly recommended.

 

I don't trust original slide pots at all.  Yes they can be cleaned but their resistive element has a short life.  I had two fail in my ProSoloist and wound up replacing them all with rotaries.  Can't do that on other models, but brand new Bourns PTL slide pots and adapters that some people are making are a good substitute.  Phil Cirocco has restocked his slide pot replacement kits for Odyssey and Avatar.

[Jim Alfredson]

2 hours ago, The Undead MC said:

Beware the Omni and Avatar - they both use the defective power supply that overshoots +/-15V rails on power up, known to blow CMOS ICs.  There's a factory bulletin to add voltage clamps to the outputs, highly recommended.

Thanks for the heads up. I'll research that. Or maybe I'll just use this inexpensive Meanwell I found at Mouser. I ordered one to test out. https://tinyurl.com/y7nv3ujh

 

2 hours ago, The Undead MC said:

I don't trust original slide pots at all.  Yes they can be cleaned but their resistive element has a short life.  I had two fail in my ProSoloist and wound up replacing them all with rotaries.  Can't do that on other models, but brand new Bourns PTL slide pots and adapters that some people are making are a good substitute.  Phil Cirocco has restocked his slide pot replacement kits for Odyssey and Avatar.

 

I don't trust the originals either. During the pandemic, I spent hours refurbing the sliders in my mkII Odyssey. That was what, two years ago? And they are already acting up again. I have some adapter PCBs on the way. GMUSynth uploaded his gerbers to DirtyPCBs on the Vintage Synth Repair group and I ordered some. I wish he had just uploaded the actual gerbers so I could've ordered them from PCBWay or some other, faster company. As it is, they are coming from China and taking forever to get here. But they were incredibly cheap. So we'll see.

 

I should probably use them in the sequencer to make it more reliable. 

[Jim Alfredson]

I received my fuse holders today and installed one in the ARP. The required fuse is a 0.125A (1/8A) slow-blow fuse. The only one I could find at Mouser that fit the bill was a 5mm x 20mm fuse and it was almost $12!!! I thought it came in a pack of at least 5 but nope... $11.84 for ONE fuse. Insane.

Anyway, it works. I also bought that Meanwell PSU linked above (which is only $4 more than the fuse!) and amazingly the mounting holes are EXACTLY the same as the stock ARP PSU. So I think that might be the best solution going forward. Just replace the whole damn thing with a much higher efficiency, more reliable PSU.

[Jim Alfredson]

Ok so in testing the sequencer I found that the quantize outputs do not work. They just output a static CV. As I was poking around, trying to find out what was going on, the -15v died again. WTF? So I started poking around that and the fuse blew. Dammit. I did find a cheaper alternative so I'll order some of those. In the meantime, I put a 1/4a slo blow in there so I could continue testing.

 

After poking around some more, the -15v came back on its own. So I think there must be some kind of solder mishap or maybe bad solder joint(s) on the PSU. And maybe that's the problem with the quantize outputs as well because I replaced two chips that SHOULD have fixed the problem (a 4016 that tested bad in my Retro Chip Tester and a LM301 opamp) but they didn't.

More testing...

[Jim Alfredson]

Success! I found the cold solder joint on the PSU. I also found the culprit in the quantize circuit. It was the MC1458 opamp that drives the comparator circuit that then drives the counter circuit. The MC1458 output was stuck at -14v. I replaced it with a TL072 and it's working!!!

[Jim Alfredson]

Ha! Spoke too soon. The damn -15v rail keeps pooping out. I'm thinking its the rectifying diodes. I'm getting a solid 35vdc out of one side of the full-wave diode bridge and almost nothing out of the other side. 

[RABid]

(sigh) If I could deal with repairs the way you do I would own an Omni II and a Chroma. My first rig was an Omni II and a MiniMoog. Great for playing Cars. Not so great for covering Bob Seger or Lynyrd Skynyrd.  

[Jim Alfredson]

9 hours ago, RABid said:

(sigh) If I could deal with repairs the way you do I would own an Omni II and a Chroma. My first rig was an Omni II and a MiniMoog. Great for playing Cars. Not so great for covering Bob Seger or Lynyrd Skynyrd.  

 

It's handy to know how to fix these things because you can get them cheap. I bought this whole lot for less than what I'll sell the Sequencer for. And I just got an EMU Emulator II for cheap because it's non-functioning, but I've worked on them before.

 

Anyway, it was indeed the rectifying diodes. Evidently to save money, ARP used 1N4448 signal diodes as rectifiers. I replaced them with 1N4007 and the PSU is working great now. I'll do some long-term tests soon to make sure it stays stable.

[Jim Alfredson]

After further testing the PSU is totally stable and the sequencer is working perfectly. 
 

I think I will replace the faders however. 
 

My order of adapter pcbs came today and I built my first one. These adapters allow the use of Bourns faders to be used in place of the original CTS faders. 
 

As you can see, the adapters make the Bourns the same height as the CTS. Perfect!

 

 

[mate stubb]

Dang, my 2 Chromas didn't have that many problems after 20 years of being silent. One advantage to having few sliders I guess. I also had shorted PS rails due to tantalum caps. Chroma has a jillion of them.

 

Both now have modern Meanwell switching supplies. 

[The Real MC]

You're jamming in your studio.  You notice a very funky burning smell and you shut your studio right down.  No visible smoke.  The proverbial needle in the hay stack.  How do we find where that smell is coming from in this pile of gear?  Follow the smell with your very sensitive nose...

 

Narrowed it down to one of two rack mount Kurzweil ROMplers (in the BOTTOM rack, of course). When I opened the rear lid on this rack I was met with a very strong whiff of the same odor.  Phew.  Fortunately the module won't be hard to get out.  Spot the yellow "X" just left of center of picture.

That smell was familiar... from other restoration work I had done, I had a tantalum capacitor emit the magic smoke while a synth was on my bench and powered on... same awful smell.  These Kurzweils are from the era of tantalum caps that tend to die with age.  If that is indeed the culprit, all those tantalums are getting replaced (I own six of those Kurzweils).

[Jim Alfredson]

Tantalums in old gear get automatically replaced. My understanding is that the manufacturing process has changed and modern tants are far more reliable. The old ones supposedly grow crystals inside which short and go boom.

[The Real MC]

Found the culprit of the burning smell, it wasn't any cap. There are no tantalums in these Kurzweil modules.

This is the connector that carries raw voltage sources from the transformer to the power supply. Note the scorch mark on the connector. That wire I pulled out is the 5VDC rail which is high current.

The connector insert I pulled out started losing its tin plating, exposing the brass core material underneath. With the brass exposed, it oxidized quickly. The tin plating is supposed to prevent oxidation.

As the brass oxidizes, it becomes less connector and more resistor. Not good. Ohm's law states that power is the product of current squared and resistance. As resistance rises on that insert, it produces power and more HEAT.

This sets up thermal runaway - as resistance rises, the oxidation gets worse, the insert gets hotter, more heat causes more resitance causes more oxidation. Eventually it burned the connector shell, that was the source of the funky smell. Not the first time I have seen this type of failure.

Only one module had this failure, the other one was fine.

The good news is I have spare shell and insert in my parts bin. I'm going to order more replacement parts to do all my Kurzweil modules.

[AWkeys]

I don't get this deep into equipment repair, but I appreciate you guys sharing it - very interesting reading!  Really wish I had the knowledge you guys have, but thankfully I know a bunch of techs who are willing to do the work.

 

My ARP Pro-Soloist has recently been restored to full functionality - what an awesome synth! Just yesterday I dropped off my old String Ensemble to a tech friend to get the power supply rebuilt. Also seems like one of the 3 delay lines to the ensemble effect stopped working.

 

Next, I have a headcase of a white-face Odyssey that a previous tech could not bring back to life. I had brought it to him years ago and he just followed problems around the circuit board. To complicate matters, I also purchased an aftertouch kit from the now-defunct New England Analog which was installed, but I can't verify that it will ever work because of the state of this Odyssey. Ugh, old synthesizers....