Juno-106 chip trouble

[bendragon]

Two (or possibly three) chips in my Juno106 have turned sour.

Firstly, there is a clone chip in place for voice 2. Its not a very good clone, even by ear it doesn't sound right and for years I thought it was just me.



Secondly is voice 4. This Chip managed to go from perfect to completely dead in 2 hours flat.
With the saw and pulse waves switched off, the noise and sub set to 0, gated VCA, both filters fully open with the res on the VCF set to 0 and the chorus switched off, this is the noise voice 4 gives when a key is pressed.

The square wave 'pop' at the end occurred when I released the key.

During calibration (to try and get voice 2 to play nice with the rest and to see if I could eliminate the noise on voice 4) the resonance on voice 4 died completely. Turned off the synth, turned it back on, voice 4 makes a sound, but not as expected; constantly gives loud pops, shrieks and sputters - random silence occurs as well.

It sounds like the 80017A has gone bad, but I just want to confirm the 5534A hasn't died as well judging by the noise given off in the MP3 I posted.


Lastly, I know these are perfect clones of the chips - I've heard them in person: http://www.analoguerenaissance.com/D80017A/
However, I just want to know if the shop is still active. It hasn't been updated since 2009.



Thanks a bunch!

[TrygveS]

Yes the shop is active.
I received a couple of chips from him this week, and they sound great (exactly like the original chips)!

[pflosi]

Replace all 6. Analogrenaissance has a pay-5-get-6 policy. All originals will die sooner or later.

[cornutt]

Your voice 4 problem is almost certainly the 80017A. The 5534's rarely fail.

[Synthead]

Try soaking your suspect 80017A in a sealed jar of acetone overnight (maybe even a couple of nights) and chip off the epoxy coating. The Ebay user "synthspa" gets his name from doing this to other peoples' chips and many people have had success fixing their 80017As this way. Rumor has it that the epoxy becomes conductive over time, but I believe it's because of a separate issue, not sure what though.

I second Analogrenaissance, I installed one in my '106 as my plan B and it isn't bad. Tuning it is a nightmare because it sends loud sine waves through TP8 through TP13 instead of the normal pulse(ish) waves the stock chips produce when you're tuning the VCF. Also, the VCF saturates and becomes slightly detuned when many notes are played versus when it's not being played, so when tuning it, you need to play tones with a really low duty cycle to get it right. In performance, it is really only noticable when you turn your oscillators off, turn the resonance to 100%, the frequency of the filter to about 30% or so, and the filter keyboard tracking to 100%, but if you aren't sticking your ear out for it, you probably won't notice it then either (after the VCFs are tuned). But don't let this scare you away from them! The engineer did a wonderful job making these chips as close to the original as possible, is on his seventh revision, and they are the best clones you can buy bar none.

[phesago]

Try soaking your suspect 80017A in a sealed jar of acetone overnight (maybe even a couple of nights) and chip off the epoxy coating. The Ebay user "synthspa" gets his name from doing this to other peoples' chips and many people have had success fixing their 80017As this way. Rumor has it that the epoxy becomes conductive over time, but I believe it's because of a separate issue, not sure what one.

This is true. It's a surprisingly awesome fix! I didn't believe it myself at first! But it is a legit fix.

[bendragon]

Maybe I mistyped my email address. I got ZERO notifications of the replies in this thread.
That aside, I have went and ordered two new chips. One to replace the rather shitty clone and another to replace the dead one. I know all the chips will eventually die but I can't afford to replace all of them right now - low pay, broken car etc. etc. thou knoweth the drill.

I'll try the acetone trick if another chip goes iffy. I have doubts weather its a long-term solution though, planning to keep this Juno for as long as physically possible.

[bendragon]

Try soaking your suspect 80017A in a sealed jar of acetone overnight (maybe even a couple of nights) and chip off the epoxy coating. The Ebay user "synthspa" gets his name from doing this to other peoples' chips and many people have had success fixing their 80017As this way. Rumor has it that the epoxy becomes conductive over time, but I believe it's because of a separate issue, not sure what though.

I second Analogrenaissance, I installed one in my '106 as my plan B and it isn't bad. Tuning it is a nightmare because it sends loud sine waves through TP8 through TP13 instead of the normal pulse(ish) waves the stock chips produce when you're tuning the VCF. Also, the VCF saturates and becomes slightly detuned when many notes are played versus when it's not being played, so when tuning it, you need to play tones with a really low duty cycle to get it right. In performance, it is really only noticable when you turn your oscillators off, turn the resonance to 100%, the frequency of the filter to about 30% or so, and the filter keyboard tracking to 100%, but if you aren't sticking your ear out for it, you probably won't notice it then either (after the VCFs are tuned). But don't let this scare you away from them! The engineer did a wonderful job making these chips as close to the original as possible, is on his seventh revision, and they are the best clones you can buy bar none.

I did a pretty good job of tuning the pathetic excuse for a clone thats already in my Juno. Kept my face glued to an oscilloscope for neigh on an hour to do it but it got there in the end. The raw waveform still comes out in stereo, though. It doesn't pan left and right as if the chorus was on, its just... really wide stereo. That confuses me to no end.

[Don T]

I'll try the acetone trick if another chip goes iffy. I have doubts weather its a long-term solution though, planning to keep this Juno for as long as physically possible.

If I were you, I wouldn't wait until the other voices have problems, I'd strip the coating now while they still work.

I've repaired a few of these beasts now, and the success rate of repair when stripping the coating after problems have started is about 60%. You will also have much better luck by letting the chips soak for about four days, until the coating visibly starts to separate from the chip on its own. The trick is to be able to pull the coating off without using anything but your fingers. I've never had a good chip fail doing it this way (looks for wood to knock on).

[synthparts]

You need to be careful with that. Soak them too long and the legs can start falling off. Also in my experience chips that were going to fail have already failed long ago so I don't think I'd recommend pulling out good ones and stripping them. I've seen a lot more damage done to the board traces by people removing chips without the proper technique/tools.

[cornutt]

The raw waveform still comes out in stereo, though. It doesn't pan left and right as if the chorus was on, its just... really wide stereo. That confuses me to no end.

Interesting... I'm trying to figure out how it would do that. Perhaps the subosc isn't shutting all the way off when the control is at zero. That would definitely be a problem with the clone. The other thing, though, would be that the PWM isn't shutting off, and that's not a function of the 80017A. Have you tried looking at it on a scope?

[Don T]

You need to be careful with that. Soak them too long and the legs can start falling off. Also in my experience chips that were going to fail have already failed long ago so I don't think I'd recommend pulling out good ones and stripping them. I've seen a lot more damage done to the board traces by people removing chips without the proper technique/tools.

I dunno, I'm thinking if the legs fall off, that would be the reason the chip failed, broken solder joints on the pins.

You're right about the traces, those suckers must be one molecule thick. This kind of leads me to my theory that causes these things to fail: The coating not conducting enough heat away from the circuit. It doesn't kill the chips, it breaks those amazingly fragile solder joints between the components and the traces, due to expansion/contraction from heat. It even appears to me through a magnifier that the traces may literally be painted on, and terminated with the usual solder cream used for surface mounting components. I know you're already aware the resistors themselves are painted on, so there is precedent.

One thing I have tested: I don't think the coating becomes conductive at all. After removing it on a few chips, I tested quite a few of the "shells" with my best DMM set on the highest resistance range, and even with the probes almost touching each other, the coating still reads infinite resistance. Even the little bit of coating that usually remains in between chip pins after removal (I don't dare try to pick it out after making that mistake once, it broke solder pads away from traces, even being careful and using a fine surgical tweezers and a big magnifier) shows no conductivity when touching the probes to adjacent chip pins. So, that's why I think it's due to heat flexing, combined with the unusual method of putting the traces on a very unusual substrate. I'm not even sure a hot-air tool, or even an infrared rework station would do any good trying to do any repair on a chip that's already failed, assuming a trace has already separated from a solder pad.

At least, that's my theory and I'm sticking to it . I mean, I've been wrong many times before, and I'm sure the trend will continue.

By the way, did you get my PM? There's something I've been working on that I cannot find a couple parts for to save my life...

[bendragon]

The raw waveform still comes out in stereo, though. It doesn't pan left and right as if the chorus was on, its just... really wide stereo. That confuses me to no end.

Interesting... I'm trying to figure out how it would do that. Perhaps the subosc isn't shutting all the way off when the control is at zero. That would definitely be a problem with the clone. The other thing, though, would be that the PWM isn't shutting off, and that's not a function of the 80017A. Have you tried looking at it on a scope?

Its a digital osc on a computer and its in mono only No stereo input.

[bendragon]

Voice 2, which is one of the new chips, has just died on me. Its whisper quiet.

I think I might have overloaded it - I discovered that I've calibrated the synth incorrectly and the resonance frequency is off the scale. Push the res slider 3 quarters of the way up and the entire synth becomes inaudible. Oops :/

Either that or I've not soldered one of the pins correctly. I'm going to have to open it up again to take a peak

I've fixed my own gear in the past, why does the JU106 have to be the pest of the lot?

[THEODICY]

Just take the voice board out and send it to the synth spa. Cheap, easy and he will calibrate it for you as part of the deal. Alan is a good guy to deal with!