I have this synth in the shop for service that has overwhelmed me for a while. I finished a report on the first phase of this job today. I used the NOTES hypertext editor on Facebook and was very pleased until I published the report (note). In edit / preview mode, the embedded photos functioned as hyperlinks to the full resolution images. After publishing they do not.
A few things I can share after seeing your post above.
The key interface board uses many A style CMOS chips. Later ones won’t always work, but some UB (unbuffered) will.
I have a good photo and other info for that board, so if you need to know what chips are on it, please ask. (PM me or email.)
There’s a bug in the PCB layout of these keyboard interface boards. In the schematics they show (as they should) unused CMOS inputs being tied to ground. But on the boards, they forgot to include some of these grounding traces. For years these circuits would work fine, defaulting to a logic state that was correct. But after 35 years now, these chips either need to be replaced or have their unused inputs actually tied somewhere.
I know Casey Young. He was heavily into SEMs and had quite a few, but don’t know much more than that nowadays.
He used to play with Yes and Gary Wright.
Fwiw, I’ve come across many SEMs in the last ten years that made pops and clicks and have traced it to defective filter offset trims.
Sounds weird I know, but sharing this with you so you don’t have to hunt and search the same way I did.
It’s now very common to have to replace all electrolytic caps, many foil types (especially in the programmer), trims, etc.
Things that in past years rarely needed replacing.
Ah, Oberheims are funky machines to repair. I bought a basketcase OB-X that I am in the process of resurrecting. I managed to get the embedded system living and breathing again when I fixed the power supply and found a bad 80C98 corrupting the data buss. Front panel switches were bad due to nicotine contamination (found filmy residue with stains underneath when I pulled them out) but I have piles of NOS replacements. Have several dead CV MUX to fix, new caps, new trimpots, new CMOS/TTL, bad voicecards, etc. Every tantalum cap got removed (one started smoking while I had the hood open!), except the ones in the audio path on the voicecard system. Tantalums degrade with age and their maximum voltage drops, so any tantalums on the power rails should be replaced. Found a design error at the DAC that can kill 4051s downstream (small wonder that many OB-Xs turn up dead). This is a big job but the sound is worth it. That project has been on hold due to work demands, going to start it up again after I finish my Memorymoog project.
Larry those yellow ICs in the programmer were early CEM chips, the predecessor of the CEM3310. I do not know if they are compatible. Tom got Doug Curtis to develop a VC EG that didn’t suffer from voltage feedthrough. I believe the later version of the OB-1 also used them.
Thank you so much for that info Kevin. It sounds like you’ve had your face glued to these circuits quite a bit more than I have! Before I proceed with the work I need to discuss the situation with my customer.
The key interface board uses many A style CMOS chips. Later ones won’t always work, but some UB (unbuffered) will.
I have a good photo and other info for that board, so if you need to know what chips are on it, please ask. (PM me or email.)
Any info would be helpful I’m sure. Do you have any schematics other than the not so good blue print copies of free hand drawn programmer docs? I’ve got professionally drafted schematics for the keyboard control but not the programmer. Was Tom paranoid about industrial espionage? Why would anybody go to the trouble of sanding off those chips?
There’s a bug in the PCB layout of these keyboard interface boards. In the schematics they show (as they should) unused CMOS inputs being tied to ground. But on the boards, they forgot to include some of these grounding traces. For years these circuits would work fine, defaulting to a logic state that was correct. But after 35 years now, these chips either need to be replaced or have their unused inputs actually tied somewhere.
I was thinking about re-chipping the whole logic section. The functions are totally FUBAR and I would probably save time by NOT tracing each problem out with scope probe in hand. I tend to get carried away looking the circuit design when I’ve got that probe in hand. The memory and memory control on the programmer overhead logic could be more trouble than the keyboard control. The last time I needed 6508s I pulled some old ones out of something somewhere as I could not find them to purchase.
I know Casey Young. He was heavily into SEMs and had quite a few, but don’t know much more than that nowadays.
He used to play with Yes and Gary Wright.
I was foolishly mistaken about that FRAGILE 1977 GO TOUR road case stencil . Fragile is in reference to the gear, not the tour. YES toured the Album Fragile in 1971. GO was Al De Miola, Steve Winwood, and Klaus Schulze with Stomu Yamashta. I thought that Casey might know something about this synth if it was part of a YES production although I think it was a couple of years later when he moved to LA and then hooked up with YES.
Fwiw, I’ve come across many SEMs in the last ten years that made pops and clicks and have traced it to defective filter offset trims.
Sounds weird I know, but sharing this with you so you don’t have to hunt and search the same way I did.
It’s now very common to have to replace all electrolytic caps, many foil types (especially in the programmer), trims, etc.
Things that in past years rarely needed replacing.
I know those trim pots. The blue ones. The same ones that cause jittery VCOs and power supplies in the OBX product line right? It is a drag that we have to think about replacing so many caps that are still working.
Great blog! I’ve recently adopted an fvs with a few issues that I’ll be diving into here soon, and the fellow I bought it from pointed me here. I really appreciate all the timely info (and big thanks for sharing the schematics on your site too, Kevin).
The biggest problems on mine are that the kbd controller only fires the first two SEMs, a couple of oscillators have gone silent, its crying for new keyboard bushings, and I’m sure it could benefit from a new battery and recaping the usual suspects. One part of me wants to blow thru and just preemptively replace all the old CMOS, but I’m going to err on the side of caution and just scope out the problems and do targeted repairs where needed. This is probably the most challenging unit I’ve been involved with so far (I’m just a hobbyist), next to my OB-8.
Some of the issues on this one might be attributed to a mod that was installed to give triangle waves on the ext inputs of the SEMs. There is a PCB inside labeled “Mediamix TriangleMod”, and an associated rats nest of wires that needs some TLC. I’m thinking I might keep this mod in there, but go thru and document it, draw up a schematic, clean up the messy wiring, and add molex connectors to interface it with the SEMs (its currently hardwired in).
Here are some photos of it for your enjoyment:
A couple wires have found their freedom:
Alligator clip ftw
I didn’t notice any of the chips in the programmer or keyboard controller on this unit being sanded, and so if you need macro photos of any boards I’d be happy to help there.
There is also a photo of the keyboard I/O main board provided along with a listing of which chips they used.
I do not recommend that someone replace all the CMOS chips at once though.
It’s better to replace a few at a time and then recheck the functionality of that module.
If you replace all 23 chips at once, it’ll be a nightmare to locate any problem(s) afterward.
I’m including a video in there that is just me improvising, but it’s notable because the Oberheim is totally dry and it still sounds very fat.
(http://www.synthfool.com/fvsstuff/MVI_5951.mp4)
The other video in that folder is simply showing off a multicolor LED I installed in that client’s programmer.
Fwiw, I replace all electrolytics on all vintage gear.
Even if they work they tend to have a rated life of about 15-20 years and most vintage is far older than that.
It’s just good sense to afford an instrument another decade or two of operation.
Besides, even if the caps work and are close to the correct values, they can have a higher ESR than new ones.
Cap meters usually don’t measure ESR either (tho some do.)
In power supplies and audio circuits, new caps can make a difference.
Here is a quote from Wikipedia’s entry on ESR:
“A very serious problem, particularly with aluminium electrolytics, is that ESR increases over time with use; ESR can increase enough to cause circuit malfunction and even component damage,[1] although measured capacitance may remain within tolerance. While this happens with normal aging, high temperatures and large ripple current exacerbate the problem. In a circuit with significant ripple current, an increase in ESR will increase heat dissipation, much accelerating ageing.”
See: http://en.wikipedia.org/wiki/Equivalent_series_resistance
Thanks for the heads-up on the CMOS (buffered vs un-buffered). I was just about to fire off a Mouser order… really thankful that topic came up prior to me stocking up on a bunch of potentially incompatible chips.
That unit sounds great in your video! Is that the same machine as seen here? (the blue LEDs looked familiar…)
At a minimum, make sure the CMOS you acquire have overvoltage diode protection at the input pins. The CMOS in early Oberheims don’t have them (I’m talking to YOU, RCA!!!), which is one reason why they turn up malfunctioning!
Btw, after you’re done with all necessary repairs, you might consider adding a noise source.
Despite the 4 and 8 voice Oberheims being their largest SEM-based instruments, only their small 2 voice model featured noise.
Here’s the way I addressed the need-
A separate noise source with wires that terminated into the molex pins used on the SEM.
The power connector for it being compatible with the stock power supply.
In other words- it plugged right in.
One question- while I was in working on the keyboard controller, I noticed there is a lot of CMOS on the channel logic boards with no decoupling caps. While there is decoupling at the power connection on the kbd decoder board, I was just curious if I should consider adding decoupling caps at the individual chips on each board while I’m in here… Curious what you all would recommended?
Also, I wasn’t kidding about enjoying the blog format posted above! It inspired me to put up a little Wordpress site on a domain that I had parked, and I’ve starting consolidating some of my projects there - including this Obie. There are lots of details on the progress on my FVS restore there, and any feedback is welcome. If nothing else, enjoy the photos!
Back in the days of the FVS, coupling caps at the power pins of digital logic was not yet a convention. Given that CMOS imparts switching noise during logic transitions, it is not a bad idea to add coupling caps.
@Doug - thanks! Just saw your post here, and I’m still slowly making progress on this thing.
@MC- thanks re: the info on adding decoupling caps. I just refurbished the keyboard assembly last week and did this on the decoder board that’s tucked underneath. The chips there were already socketed, and so I figured I’d re-chip and add these caps to save having to disassemble the enclosure again later to get at this board.
The reason that I’m kicking this up again is for another quick question or two:
With respect to building Molex interconnect cables for this old Obie gear, I was wondering if anyone knows of connectors for external-facing applications (e.g. connectors with an integrated hood)? I don’t recall seeing any (or seeing any old Obie cables as a reference, for that matter), but thought I’d ask and solicit suggestions on making a decent looking cable to connect the MS-1a to the FVS.
My plan now is to leave the MS connected to the synth, and re-purpose the two holes that had been drilled in the FVS enclosure for that triangle-wave mod… using them to mount switches to toggle between keyboard or sequencer control of SEMs 3 & 4. I’ll also bring the MS’s noise source and s/h into the synth (ala TVS) via that 12-pin Molex.
2) Memory storage/recall is funky on the programmer - what goes in does not come out. Everything I traced up-to and after the A-D/D-A seems fine, but beyond that I’m kinda stuck… I’ve not yet owned/used a logic analyzer to sort this type of failure out, and I figure this will be a good excuse to start
I just wondered what you all might be using or would recommend in that space? I have access to a Nicolet 800c w/probes via my dad (he collects/restores old test and radio equipment as a hobby), but since he’s out of town it will be a while before I can get at it. Still trying to decide if I should just buy one.
With respect to building Molex interconnect cables for this old Obie gear, I was wondering if anyone knows of connectors for external-facing applications (e.g. connectors with an integrated hood)? I don’t recall seeing any (or seeing any old Obie cables as a reference, for that matter), but thought I’d ask and solicit suggestions on making a decent looking cable to connect the MS-1a to the FVS.
I’ve seen original cables for the OBs, but don’t recall any with hoods.
I’m not sure they exist, but don’t quote me.
The cables were basically raw wires run through non-heatshrink tubing with a wire tie around them at each end.
The actual connectors, pins and crimpers are still available around.
Other connectors might also fit if you’re going to have to make cables anyway.
Memory storage/recall is funky on the programmer - what goes in does not come out. Everything I traced up-to and after the A-D/D-A seems fine, but beyond that I’m kinda stuck… I’ve not yet owned/used a logic analyzer to sort this type of failure out, and I figure this will be a good excuse to start > >
I just wondered what you all might be using or would recommend in that space? I have access to a Nicolet 800c w/probes via my dad (he collects/restores old test and radio equipment as a hobby), but since he’s out of town it will be a while before I can get at it. Still trying to decide if I should just buy one.
I don’t own a logic analyzer. There have been times I wished I did, but managed without one, one way or another.
Sometimes it’s just easier to probe things manually, one by one and document the signals found.
An analyzer is much more necessary when things are at high speed or running clocked logic where comparing timing between signals is of great importance.
I can’t really provide much other help other than to suggest if everything is well with the A/D conversions, look towards the output sections and and their opamps.
Also, inspect pins and connectors closely all over.
The female pins used can break internally. They’ll look fine from the outside and their soldering will look fine too, but they can crack in the center from fatigue or corrosion and it’s not always obvious upon first glance.
Of course, make sure there’s a good battery in it, but it should work without one if powered up and left on while testing.
I have been restoring a four voice and the keyboard interface is missing the two 1120A boards. I think my chance of finding these boards is pretty slim so I suspect I will have to create new ones. The schematics available for this board are not very legible and there are many parts I cannot determine the exact part number. Does anyone have a detailed parts list for the 1120A board and what parts may be difficult to get?
Hi - just wandering back in here and happened to see your message. If it helps, Kevin Lightner shared a good photo of that board with parts (chips) list here: http://www.synthfool.com/fvsstuff/fvskeyio1120a.jpg.
I have a question that I was hoping to get some feedback on…
VCO2’s frequency control on one of my SEMs had no effect in the middle 2/3-or-so of its travel. I finally got time to scope it out, and traced it to… this
So… I’m thinking about trying to find a new 50K donor pot (maybe CTS?) that I can use to scavenge the resistive element and restore this freaky fine-tuning-ball-bearing pot. Have any of you been down this road? Any words of wisdom?
Almost done with this one! I’ve recapped the entire programmer (the “programmer” part works great), replaced all the power caps elsewhere, and this thing is sounding mighty fine! Just have this and the memory store/recall quirks to sort. I can live with the memory issues for a while… heck, I live with them everyday!
I agree that if you can locate the same CTS resistive wafer and rebuild the pot, it should work.
I have a bunch of CTS pots in various values, but not sure I have that particular value.
It may not matter much regardless- a 25K or 100K may work.
You may need to change the resistor that pot feeds, but it would probably work.
I’d also suggest checking the condition of the other VCO pot and filter pots.
Other pots may not have much life left in them.
Fwiw, I DO have that exact dual concentric pot in stock.
NOS, never used condition, but only one or two left and so it won’t be cheap.
Contact me via PM if interested. Thanks.
