Odyssey, fixed again <3

A bit of a late post this as I’ve had my Odyssey operational for some months now, but at least there is a lesson to be learned here. To recap the situation a bit, the power supply failure with my Arp Odyssey caused something on the VCF circuit on board C to fail. This board also hosts the AD/ADSR generators and the output amplifier, but they weren’t affected by the PSU problems.

Repairing the VCF section turned out a trip far more winding than expected. I made a false diagnosis in the process and thus ended up “downgrading” back to the original 4023 VCF module AND even (unnecessarily) removing the potting material in order to be able to measure the entire VCF circuit in detail! I won’t go into details about de-potting, but if you want to read about the process, check out this excellent page @ ArpTech.

Anyway, as I was getting nowhere poking around the 40235 dual VCF module, the idea with the VCF downgrade was to restore the synth to factory spec. Less complex design, calibration procedures corresponding to the original service spec and a factory etched board should help removing possible DIY errors from the problem solving equation.. And besides, the 40235 is bit of a bitch to measure thanks to its compact part layout :)

Assembled 40235 Dual VCF board.

Whereas in the end, the _real_ root cause for the VCF failure turned out to be A1 on board C (output driven near the positive voltage rail), being too much into diagnosing the dual VCF board instead I failed to look elsewhere and spot the real fault. This effect is pretty common with electronics design / troubleshooting and I fell right into it, despite having been in a similar situation countless of times. You know, stare at a circuit or schematic long enough and you sort of become blind to even obvious errors. Just as an example..

Not too long ago I faced a case where the motherboard of a device had to be revised because of some (end-of-life) component changes. Problem was that the original board design had been around for so long that it’s PCB layout files were lost. So, the board had to be drawn from scratch in order to fabricate the new version. When I got the first batch of these new boards for testing, it turned out that half of the internal address bus routing between the onboard EPROM and the CPU was missing altogether! No wonder the boards had problems booting up ;). It might feel surprising that such an apparent error can escape the eyes of the engineer who drew the new layout but rest assured these things do happen! Even the engineer who peer-reviewed the new design failed to spot this nasty short-coming.

Anyway, coming back to the Odyssey.. I performed the diagnosis and repairs during a rather long period, as I had to wait for some replacement parts to arrive in the mail (and whatnot). The funny thing is, once I had the Odyssey running with the 4023, it became apparent that I had stashed the 40235 VCF some place “too safe” and was unable to find it anymore (getting maybe just a bit too absent-minded? :).. Oh well, I’ll definitely take a look at it again once I find it.

The de-potted 4023 VCF module installed to board C.

Whilst de-potting the 4023, I started thinking maybe I could mod it to be installed ‘components up’. If pin headers were soldered in-between, the module would become removable and in turn make it easier to switch modules.. or just removing the installed one for repairs. Not much to it once the module was de-potted, I’ll let the pictures do the talking:

Header connectors on the 4023.

Pin headers for filter module on board C.

Modded 4023 installed.

Despite what it looks like, the headers make the module sit tight enough for it not to fall off. No idea how much mechanical vibration or shocks this plug-in version might endure, but it’s not like I’m going to expose the synth to any.

As for fixing the snapped slider, I decided to test inserting a support rod inside the plastic shaft. For the rod, I used a length of component feet. My magnificient insertion method:

Soldering in the support

As some of the plastic formed a small bulge around the component feet, the halves didn’t fit together too good. In an attempt to make the joint more solid, I added a small dab of epoxy in-between. Not that it helped much..

Both halves joined

So there. Not the most solid fix ever, but the slider does withstand careful use. Pushing it from the very tip would probably break the joint eventually. All in all, it seems that my Odyssey getting closer to slider swap. It’s not like my slider cleanup routine was meant as a lasting repair anyway :)

Tags: , , ,

7 responses to “Odyssey, fixed again <3”

  1. François says :

    Hi Arto,
    the filter on my whiteface Odyssey seem to be dead. Unfortunately, it’s one of the very first (build Oct ’72) and potting material was not gum-like, but rather stone-solid; unpotting involves nasty solvents and looks like a long operation.

    So I decided to build a new one from scratch. I already have the schematics and some components (OTA). However, before designing the PCB, I want to place the components similarly to the original design. I was wondering if you have more pictures of your unpotted filter module.


    • Arto says :

      I was wondering if you have more pictures of your unpotted filter module.

      Found only this one “isometric” shot I hadn’t uploaded, might be of some coupled with the photo of board traces.

      De-potted Arp Odyssey 4023 VCF module

  2. François says :

    Thanks Arto. I notice on this shot your filter does not make use of “tin can” OTA. This is good as they are now rare to find.

    • Arto says :

      I notice on this shot your filter does not make use of “tin can” OTA. This is good as they are now rare to find.

      Iirc, my 4023 used the canned type but unlike on the main boards (check post here), the CA3080 had it’s pin spacing readily matched for DIL package.

  3. Gabor says :

    Hi! My Arp ody 4023 filter generate white noise witch is high passing out when cutoff opening… The filter worling as expected, but noisy like hell… Can someone help me out here?

    • Arto says :

      Hi! Have you checked the condition of the switch which is used to route either white noise or ring mod to the VCF input mixer? The part is labeled S1 in Board C schematic. How about the slider P7 which sets the noise/ringmod level to VCF input?

      • Gabor says :

        The noise ringmod to vcf input works as expected! And a pedal, S&H to VCF cutoff as well! I tested everything. It seems it is generate noise whitout those sources! I Refurbished it many times. That problem came randomly and come back randomly more often. Now it is ollways there. I am sad because i figured out to refurbish kez bushings and all the elevtronic rods are cleaned. Second voice is stable now!And the duophonic and a keyboard works like a dream. But that noise

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

<span>%d</span> bloggers like this: