Jump to content


Please note: You can easily log in to MPN using your Facebook account!

Prophet 5 (Rev2) Refurb


Recommended Posts

Back in the day - read early 1980s - a Prophet 5 would have set you back around half a years wages here in the UK...Completely out of reach of most (all?) amateurs.

The only answer was to build your own !

At the time I was a keyboard tech 'by day' .. so I gave the project a try, and built a 5 voice poly synth

Using CEM 3040 oscillators and SSM2040 filters - and with some help using a microprocessor to make the keyboard scanner - it actually worked.... Really hard to tune - but when it worked it sounded not bad !

 

Loved the filter sounds -- and used the SSM2040s for other projects at the time. (Quite cheap and easy to get - then! ) ended up with a few spares, which I kept for years, always thinking I'd use them again one day.

In the end I had 8 unused devices left, and sold them to a dealer about 10 years ago -- At the time he didn't believe I actually had 8 unused SSM2040s -- so I had to send him a photo. (copy attached)

He paid me what I asked .. but I think they might be worth a lot more now -- hey-ho!

The homemade 'Prophet clone' synth got broken up years ago -- all I have left is a couple of short tracks recorded onto a cassette recorder.... Not very good either musically or recording wise, but it brings back some happy memories

950.thumb.jpg.eb583764f04c79ea754c373698a24e30.jpg

Yamaha - YC61 - P105 - MOX6 - HC2 -- Neo Vent 2
Link to comment
Share on other sites



  • Replies 58
  • Created
  • Last Reply

Top Posters In This Topic

Top Posters In This Topic

Thanks for all the kind words. Apparently there's a whole lot'ta love for the Prophet5 Rev 2 around here :cool:.

 

On with the refurb story;

7YNigF2m.jpg

 

 

Last I reported I was getting the board to boot up and soon after got audio out. Turned out it wasn't much of a sound although it did respond to the key played. Only 2 of the front panel knobs were responding; the Filter cutoff and the Mixer Noise which sounded like it was controlling the filter resonance. Did some probing of the Demultiplexer circuit that resides on the processor board . This circuit basically sample and holds the voltage across each potentiometer (or recalls those values from memory) and connects them to the right place on the voice cards. Each one of those voltages is updated at a periodic clock frequency otherwise known as the sample and hold clock rate (different from but same concept as an LFO S&H waveform).

 

That clock rate didn't look so good and I suspected something might not be right with the processor timing (again). But I turned my focus to the front panel boards:

 

8X8CW0hm.jpg

 

I probed the front panel VMUX signal which is a composite of all the pot voltages. Each pot is allocated a unique time slice in that composite signal (referred to as time multiplexed). You can actually observe on a scope a small slice of that composite signal change its voltage level when moving a pot. Originally there were only 3 or so time slices (there should be 24 - one for each pot). Still I could see when moving the filter amount it would change one of the slice levels . But moving the noise knob shifted the entire signal.

 

Only a dozen or so logic chips on the left fp board and I decided to socket these just like the CPU board. Somewhere in my probing something looked funny. I changed out the chip and now the noise was no longer responding.

 

Focused my troubleshooting on the ADSR pots, learned a whole lot about the timing scheme but still wtf, are the EPROMs screwed up? And then by some fluke I happened to hit the preset button in the Filter section and now I'm getting more activity on that VMUX signal coming from the front panel boards. At this point I was seeing 4 time slices iirc with the filter amount, resonance and envelope amount responding. But still no activity from either ADSR.

 

Cutting to the chase I eventually hit the preset button in the LFO section which resulted in 8 time slices. I then decide it was time to read the Operation Manual (not the service manual dummy). Found out that you have to be in manual mode as opposed to preset for most of the knobs to respond. I thought I was in manual when I turned of the filter preset but that only controls 3 knobs. You have to turn off the preset button in the programmer section.

 

And now all 24 time slices are showing up (top trace):

AzuxEdPm.jpg

 

While the pots are now doing something the voltages produced are anemic as seen on the bottom trace. And I'm back to no sound. I was able to trace this to a bent pin on one of the processor board ICs.

 

Much better:

2qaHy1Cm.jpg

 

From here I could start playing with everything and amazingly almost everything seemed to be working. I did have substantial leakage on VCO 1 when both waveforms were off; or so I thought. Again another IC bent pin was keeping the 'off' control signal from transitioning to OV. I love easy problems even if they're my fault :blush:.

 

I also noticed the boot-up time is much much faster now (6 seconds or so) as is the auto-tune. So on to the calibration routine in which I completed the first 6 steps or so. This thing is responding great and no need to mess with the VCO tuning. There is one voice that is noticeably louder than the others so I'll have to go back and look at that. Otherwise this thing is sounding fantastic. And I had it powered on for a good 6 hours yesterday after completing everything. The tuning stability is...stable; way better than what I expected. Power supply heat dissipation - back panel is warm to the touch, not hot. I think this has a lot to do with it.

 

rvxFMQxl.jpg

 

I'm going to re-grommetize the keybed next and that's about it. Other than I now have to lay the unfortunate news on Paolo that this thing burned up beyond recognition and I had to throw the smoldering mess in the trash.

:keynana:

Link to comment
Share on other sites

Congratulations Mark! I love happy endings to troubleshooting mysteries. Gotta love digital. It actually responds to ... Logic! That and a lot of due diligence and hard work that is.

 

Well done and a great post!

Whenever you find yourself on the side of the majority, it is time to pause and reflect.

-Mark Twain

 

Link to comment
Share on other sites

Somewhere back a year or so ago, someone posted a history/documentary of the Dave Smith stuff. I don't care for the current product--too fizzy sounding for me--but the older stuff was a different bag. It turned out that the fork in the road (to my ears) was the Prophet 5. With that in mind I've been watching this thread with interest. Fate sometimes throws me a wounded item that I can repair, like that Kurzweil K2500. Granted, that one went sideways, but I have to say that the keybed made a dandy MIDI controller once I built a pretty chassis for it out of some red oak and black cherry...which reminds me, I ought to post a picture or two now that we have the ability to upload pictures to KC.

 

I'd like to thank those who helped with that project again. You eased my path considerably and I've had a lot of fun with the keyboard since then.

 

Grey

I'm not interested in someone's ability to program. I'm interested in their ability to compose and play.

Link to comment
Share on other sites

Thanks gents!

 

Otherwise this thing is sounding fantastic.

 

Lets hear it.

 

:thu: Should have something this weekend.

 

KjHbiydl.jpg

 

Ummm, make that next weekend or soon after. Waiting on some j-wires.

 

But it does sound great. Or did. :cry:

Link to comment
Share on other sites

  • 4 weeks later...

Crap - almost a full month to get the j-wires from the Netherlands. I blame RudyS :evil::wave:.

 

I did replace the 7 "repaired" wires and with a whole lot of tweaking got this puppy working really well. I will go back one more time and make some final adjustments on a few keys.

 

0jdmeRnm.jpg

 

I do not like these j-wires. But interesting that the "single" bus bar is split into 4 sections unlike the continuous bar used in the mini and others . You can see where 2 of the pieces are connected by a piece of black heat shrink tubing. Not sure why they did it this way although it did make it easy to slip out one really worn section and replace it with a new piece.

 

xYO63MGm.jpg

 

The single bus bar means that mono mode is very strange (at least to me). If you hold a note and play a second note the second note will play. But when you release the second note it does not return to the first note. There's no sound. Also there was one late discovery looking into why one of the voices sounds so different from the others. Turns out the resonance doesn't work due to a failed SSM 2020 Dual VCA. Well there had to be one I guess :(. But If anyone happens to have a spare...

 

Anyway she's just about there and man does it match up looking all nice and pretty in this spot.

 

liiZBYwm.jpg

 

OK - who asked for some sounds? No apologies for the noodling but I do apologize for the raw video. I ran it through the MS video editor adding some text notes but the sound turned to poo. I'll see if I can find another editor but in the mean time...

 

 

[video:youtube]

Link to comment
Share on other sites

Thanks Prof and Eric and Dave! Last night I ordered 2 SSM 2020 chips (1 spare) from Wine Country. I was surprised they had any. They have extremely limited supply with a price to match but no complaints. These guys (or guy/gal) are great!

 

Coincidentally there was announcement of a reissue SSM 2040 filter last week as shown here.

 

Hopefully the other SSM chips will follow as they did for the CEMS. And at a much more palatable price.

Link to comment
Share on other sites

I don't think you'll see a reissue of the SSM2020. Ron Dow made a lot of advances in VCA technology since the 2020 such that there's not much compelling reason to reissue them.

 

It would make more sense to make an adapter board with modern VCA in SMT package.

 

Great news on the SSM2040 reissue though!

Link to comment
Share on other sites

For those interested in more electronics related stuff...

 

A day or 2 after doing the demo above I noticed that things didn't sound so good. Turns out all of the Oscillator Bs (all voices) had this kind of harmonic modulation thing going, like really bad FM programming. It was part noise and part tone on top of every waveform only it didn't change pitch with the keys.

 

In scoping around on a couple of the voice oscillator B SSM chips I could see the selected waveforms but the scope was having trouble syncing due to jitter on the waveforms. I wasn't sure whether this had anything to do with the problem or not. And then I spent way too long troubleshooting this and coming up empty. I nearly threw in the towel attributing the problem to inherent noise and not wanting to mess too much with these very expensive SSM chips. The Osc B path does have an extra gain stage due to splitting off for the poly-mod functionality with Osc A and well I figured this might account for the extra schlopp.

 

So back to my studio it went for 5 minutes before I couldn't take it any more - back to the bench. This time around I got smart and looked at the CV inputs before their in-line series resistor. If you look directly at the SSM chip the inputs all looks clean. I should have known this but oh duh...

 

It turns out that the initial osc B pitch CV generated by the CPU and added in to all the other pitch control sources was dipping down 100 mV or so each sample and hold period. This was effectively modulating the VCO frequency. Like I said, bad FM.

 

oMCLH6km.jpg

 

Here's a better look, AC coupled for looking at just the variation:

 

phBzHSsm.jpg

 

The sample & hold circuit consists of just a low leakage capacitor that's recharged each cycle to maintain it's value and fed into a very high impedance op-amp. Just about every front panel knob function has one of these circuits to maintain the current patch. It was a 50/50 crap shoot whether bad cap or bad op-amp. And of course I'm wrong EVERY TIME!:mad:

 

With op-amp replaced here's that same Initial CV:

YkCLtgFm.jpg

 

I'm convinced now that this problem came about with better than 15 hours of burn-in time. Something folks should request/confirm when having repair work done.

 

Anyway back to sounding awesome. I was also able to acquire an SSM2020 from Wine Country and fixed the no resonance issue on V5.

Link to comment
Share on other sites

Thanks Richie. Who knew when I started this back in January the hand-over of Paolo"s refurbished P5 would be our first somewhat nervous social event in 3 months. Even better that our get together with Dave Doerfler joining us turned out to be Paolo"s surprise birthday celebration. Only we were the ones surprised having no prior knowledge that the very next day was his birthday. :wacko:

 

Thanks for posting the video.

Link to comment
Share on other sites

  • 5 months later...

It was interesting to view the repaired P5 test video, all the more with the current attention for the new redo of the synth. Interesting to hear all kinds of partial signal errors in the machine, leading to a recognizable Prophet sound, but lacking some lushness and warmth and having some remarkable kinds distortion. I've been working some on yet another way to get the sound back of these kinds of synths, by trying to correct also the virtual analog buildup of the PC3 (Kurzweil) and that leads to strange sound errors as well, not the standard ones every sound set of the kind has, and also not like the software emulation package. It's like when pro signal path elements come together they produce properly DAC prepared waves that are recognizable as analogue sounds of high quality, but with yet other modulation issues than your example.

 

T.

Link to comment
Share on other sites

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.

Guest
Unfortunately, your content contains terms that we do not allow. Please edit your content to remove the highlighted words below.
Reply to this topic...

×   Pasted as rich text.   Restore formatting

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.

×
×
  • Create New...