Acid Damage Issues

The Last few months have been strewn with synths with Battery acid damage issues, we’re not entirely surprised by this, as the price of any analogue synth climbs, the more dross comes in from attics, sheds and garages with totally unbelievable sob stories attached as to how it got in this condition. You aren’t fooling anyone guys, you got it for next to nothing and now want it sorted so you can cash in. If it isn’t too bad and the desired result is obtained reasonably quickly, then you may get your wish, but if it is bad and a lot of tedious work has to be done then the costs will rise very sharply and to high levels and your returns may not meet your expectations, especially for a rare but undesirable item that has an unknown inflation path.

The list of equipment that we have seen over the last few months with NiCad battery backup has seen several new additions, one unexpected item was the Cheetah Master Series controller keyboard, in this case the 88 key version. It has a very nice Fatar hammer action keyboard and a lot of buttons for a lot of control and is a very well built piece of equipment but the NiCad killed it.
We had it back in September for a first crack at it and there were a lot of tracks to rebuild and a couple of IC’s to replace but it worked for a week after it left us, but now the CPU has died, and it is a weird one only available according to our best searches from littlediode.com, It is a 64 pin chip that is the same size as a normal 40 pin IC, the pins are of necessity closer together, (the beginnings of the miniaturisation that plagues us today), how many synth service agencies can handle large BGA chips?
The old CPU will have to be cut out, there is not much chance of desoldering a 64 pin chip with some acid damaged pins without further risk to the pcb, a suitable 64 pin socket is available but can’t be used due to height constrictions as the CPU is mounted on the same side as the front panel switches, and space is very limited. Checking out the Internet seems to confirm that very few of these Cheetah’s remain in the wild (just like the feline version!) It’s probably due to the battery problem and the complete lack of any service documents for what is a very sophisticated controller.
As of early March 2015, the new CPU has been fitted and the Cheetah finally working properly and back with it’s owner, the last hurdle was that controller wheel 3 (for MIDI volume control) wasn’t working whereas it did before, and this turned out to be yet another acid eaten PTH (plated through hole). The whole thing had to be dismantled yet again to sort this. A new switch mode regulated Wallwart was converted for use with the machine, as the original was a big transformer with a Euro two pin mains power connnector which was a pain to plug into an Irish (or UK) socket.

The Teisco SX400

We got one of these in a couple of weeks ago which was traded on a local auction site. It’s provenance is not certain but it’s complete lack of functionality is beyond doubt. It’s dead, it has ceased to be, it is an ex synth!

This synth did have a Ni-Cad battery originally but has a Coin Cell in it now that the recent seller thought might be enough to bring this monster back to life. There are some signs of acid damage on the pcb but nothing as severe as you would see on a Polysix or Poly 61. We suspect that this synth was dead a long time before the battery got a chance to leak, and that may explain it’s lying up in storage for so long. Its front panel was also covered in the residue of very old Gaffa (or Duck) tape across most of it’s width, which had dried out to become a horrible brown scab over much of it. White Spirit seems to be the only thing that works on old Gaffa glue, Isopropanol just won’t make a dent, and the only scraping tool that does no damage is your own fingernail, you can’t use screwdrivers or anything metal, it will scratch the paint permanently. Other possible scraping implements are wooden cocktail sticks or large cactus spines, We have tried all of them and unfortunately the fingernail is the quickest and best. The picture below is obviously after several hours of cleaning, and a severe need for a manicure!

sx400-front-panel

This synth is definitely a strange beast, very clever in some ways and extremely dumb in lots of others. It’s keyboard has Aftertouch using circuitry derived from an ARP patent, but it only has 4 Oscillators which are assigned in ways more reminiscent of the Korg Monopoly than a polysynth, but it does have 4 filters and 8 envelope generators making it a bit more like the Jupiter 4. It also has 8 preset tones, and 8 programmable ones, again similar to the Jupiter 4. Unlike the JP4, none of the Oscillator functions are stored, while almost everything else is. The other intriguing thing is the tremendous amount of electronic circuitry used to achieve so little result.

Gordon Reid lambasted this synth in a Retro review in Sound on Sound magazine as a complete turkey, but one that he owned and had a grudging like for. http://www.soundonsound.com/sos/oct01/articles/teiscosx400.asp
It has discrete oscillators using uA726 chips (as does the JP4), discrete 24dB ladder filters using the same transistor types as the TB303, the only other Polysynth we can think of that has discrete ladder filters is the Memorymoog. According to Gordon Reid, the SX400s’ filters don’t sound good, but if this beast ever comes close to working we have a few ideas to improve them. It also has discrete Voltage controlled Envelope Generators, and uses loads of 3080 and 3280 IC’s for it’s various VCA functions.

So is it worth spending a huge amount of time on? Logic says no, the values of its rare components are more than enough to recompense everyone, but it is a unique if weird machine, and quite rare these days on Ebay, synthfind.com says the last one sold in 2012, which is a long time ago in synth bartering terms but there were a few more in 2011, but none in a perfectly working state.

cpu-board

The above picture is obviously of the CPU board, the little white stickers indicate the first candidates for further testing and/or replacement. None of the CPU board behaved anything like it should. The Z80 CPU had clock and activity on all it’s Data and Address lines, as well as the 4 main control lines. For the time being we had to give the Z80 the all clear, an EPROM image obtainable from the Internet was burned to an EEPROM, and tested but made absolutely no difference, so the original EPROM was reinstated. The initial issue was as follows, the only function of the synth that seemed to work was the Preset and Memory buttons on pcb Syn 36. These weren’t flawless but the right buttons lit up the right LED’s, so a hint again that the CPU and EPROM were not culprits.

Unfortunately nothing else does, the ADC and DAC are completely non functional, the multiplexers for the sample and holds are not getting any control signals at all, the keyboard didn’t work and none of the analog switch drivers (used for selecting waveforms, modulation options etc.) worked. The way all the external circuitry is driven is a little unusual. The CPU’s address lines mainly just talk to the EPROM and RAM, only 12 of the 16 available address lines are used, almost all other addressing is done by a few address lines and the I/O and Read and Write lines of the CPU along with a large and complicated system of decoding IC’s to scan and drive all the functions of a Polysynth.

This is March 2015, but for continuity the story will continue here for a bit. It seems as though the synth was stored on it’s end with the pitch bender towards the ground, and the battery acid has wiped out every IC in its path to the left of the CPU board which includes all the Pot and Slider reading, and all the Voltages for the Sample and Holds. We replaced a total of 15 IC’s on the CPU board which were obviously damaged and/or behaving suspiciously. This is a quarter of the IC’s on the CPU board and the only net gain has been that the keyboard scanning now works. Each of these new IC’s lives in a Gold plated turned pin socket, so the time and costs are already mounting. Some of the IC’s are not common items and will have to be obtained from various and far away sources. This synth has gone back to it’s owner due to time and space considerations while the replacement IC’s arrive, and the next batch of 15 IC’s is ready to go into it, which will be to replace everything in the path of the acid leak.

We think that the acid itself does not kill the IC’s but power being applied to the synth when battery acid is present on the PCB certainly does. Our reasons for this are twofold, the behaviour of Polysixes and Poly61’s while being used and the rapid and almost logical progression of failures until the machine is dead. The other was a GEM S2 Turbo which had it’s battery replaced while the synth was seemingly working fine (it had been in storage for a time) and within a day the chip that buffered the Display died, cutting the old chip out revealed a tiny droplet of acid under it which normal cleaning could not get to. The S2 worked fine after that for another couple of days and then we noticed severe Digital distortion which we think was due to problems with the Sample address decoders which live very close to the previously replaced chip, another drop of acid under another chip? Probably!

Battery acid problems are certainly some of the nastiest and difficult to diagnose and fix of all the problems that synths can have, and are becoming the principle reason to declare a synth dead, and not fixable at all or not economically viable to fix . This Teisco is perilously close to the latter. When it comes back and the next round of chip replacements is installed, that will be crunch time, if there is still no noticeable improvement in functionality, then we will have to “call it” and dispose of the body parts to save other synths with a better chance of survival.

dirty-bent-sliders

This picture is obviously of the right hand side control PCB, the sliders had taken a battering over time and were very stiff and notchy in their action, so they were all removed from the PCB, rebuilt, straightened out, cleaned and lubricated with new grease and then put back in. They measure perfectly and feel fine, the only reason the slider caps are not back on them is in case one or both of the Multiplexer chips on the board has failed, and somebody in the past Superglued the slider caps on, making them very difficult to remove without risking damage to the sliders themselves. When (and if) we are happy with the results, they will go back on with a soft (Evostik type) adhesive if necessary.

poly61-cpu-sml

The above image is of a Poly 61 CPU board, the synth doesn’t work and there is a green patina over most of the components from one end to the other of the board, the pins of several of the plug-in connectors have been eaten away and it seems a complete loss. It is possibly fixable given enough time and money, but would a €1200 repair bill for a 61 seem like good value?

jd800-keybed-sml

The above picture is of the worst condition Roland JD800 we have ever come across. The infamous red goo has not only leaked from the black keys (as is usual) but from all the white keys as well, leaving an incredible mess all over the keyboard and base of the synth. We went to a great deal of effort to rebuild the top octave and a half of the keyboard to ensure that the keyboard’s contact systems were still OK, and no they weren’t. It took many hours over a few days to clean the sticky keys thoroughly, apply new Epoxy to the weights and to clean the rubber contacts, and the keyboard still just doesn’t work. A replacement keyboard from Roland has been quoted to us at €850 from Roland Ireland, and to add insult to injury most of the tact switches don’t work either, so they will all have to be replaced, the Editing sliders have problems too, quite a few of those would have to be taken out, stripped and rebuilt. This synth is economically unviable to fix, with no guarantee of perfection even if a large amount of money is spent, so like many others lately, this one is beyond the call of duty!

Even before the process of trying to rebuild the keyboard, the output Jack Board had two smashed output jack sockets which were replaced and a dual op-amp in the DAC section which was seriously gone shorting out the analog voltage rails. We got it to make sounds, but only over MIDI, and as for the rest, it’s just too much.

Facebooktwittergoogle_plusredditpinterestlinkedintumblr

The Crumar Multiman-S

This was sent to us by an existing client who likes to buy rare and unusual things on Ebay and then inflict the mess upon us in the hope of a big profit further down the line.

This was another one of those, like all Italian machines of the mid and late seventies, it sort of worked but not properly. There was no smoke or bangs but parts of the synth would switch on and off in functionality at random, and others would just be flaky all the time. The synth was in pretty good cosmetic condition, it had all it’s original sliders and even the slider caps, the chassis wasn’t rusty and only a few details of the screws holding it together were incorrect which was easily rectified. Finding out why many parts of it were flaky led to something akin to Alice in Wonderland, we went through the rabbit hole and kept on discovering more and more issues. The connectors keeping it all together are of a type we have not seen before, they are like normal multipole connectors but without the plastic housings, think of a Molex but no plastic, These were all black as they are Silver plated and Silver tarnishes, and needs cleaning, and there are hundreds of them, and Isopropanol doesn’t clean this, Silver polish and hundreds of cotton buds and Isopropanol to clear the residue of the Silver polish does, and this is massively time consuming, we went through 500 cotton buds on this alone. This improved things but a couple of problems wouldn’t go away, and this was starting to look like bad solder joints on the aforementioned connectors to their circuit boards, as without the plastic connector housings on modern connectors, all blades and housings were soldered to either mini pcb’s or the main circuit boards they were connected to.

multiman-s-fp-sml

There were intermittent problems with the Upper Piano and the lower strings which were cured by resoldering all of the connectors, both male and female ends, so for a short period of time we had a fully working synth. When we say resolder we don’t mean heating up the solder joints and adding more solder, our experience is that this is an unreliable way of dealing with the issue, we remove all the old solder, reseat the connector and solder as if we were building it for the first time and this usually works. Until after another switch-on we had a cacophony of noise from the left side of the split keyboard, as if all notes were playing simultaneously when one key was pressed.

Hunting this problem down became something of a nightmare, it defied all logic, the Piano and Clav worked fine, so it couldn’t be the Tone generators, but the Brass and Strings exhibited this issue and rechecking all the work that had been done showed that no mistakes had been made, or at least no obvious or apparent ones. All the tone generating boards (there is one under each key) were removed three times, it had to be here logically but as soon as more than one TG card was replaced the issue recurred depending on how many TG cards were replaced. While checking out all possibilities we found a dead transistor on one of the string chorus delay line cards, it was in the output low-pass filter so it made the Strings a bit brighter, certainly not unpleasant but just not quite right although it sounded fine, a nice spikiness on the strings that showed up on a scope, sounded like a Juno with the filter opened up fully.

Finding the issue with the left side of the keyboard took a tremendous amount of time, we were looking for a single point of failure, one failed component or assembly, it worked fine one day and not at all the next so it had to be something big and obvious, had a loose screw found it’s way under the keyboard mechanism, had a transistor failed in one of the tone mixing sections? All of these possibilities were investigated and found to be not the case. We were very close to admitting defeat on this one even after the tremendous amount of time spent on it. We had the service documentaion for this synth but it is badly laid out and very confusing, every in and out of a pcb is referred to another drawing number, and every connection has an unexplained acronym, so navigating through it all is very tedious and difficult.

multiman-s-tg-sml

The two boards on the middle right of the photo are the tone generator boards. They take the two frequencies per note from the TOG’s and do some signal processing on them, the Piano and Clavi sounds combine the square waves from the the TOG’s and accept the keyboard Gate signals along with the Sustain voltage which is fed in parallel to all 61 tone generators. The rest of the circuitry turns the TOG’s square waves into Sawtooth waves for the Strings and Brass, each board has active filter circuitry to do the waveshaping, but the component values change across the keyboard as the frequencies increase to maintain the correct wave shape, so this required labelling each board with where it came from. Actually the component values change every three boards or so, but marking is cheap and you know where you think you are.

The left of the TG boards is badly tarnished, the right hand one has been polished, these are the female connectors, the male connectors are like a bent back L shape to fit into the females, they are not a bad connector apart from being silver plated, and the fact that if you don’t mate them correctly, the male blade can slip sideways and short to the next socket of the connector. This didn’t happen but it was something we kept a very close eye on, and was one of our chief suspects in the hunt for our problem, but it wasn’t that…

Getting to the point, we took out the tone cards for the left hand side of the keyboard for the third time, and checked through all seven diodes and three transistors on each card, and lo and behold, we found shorted D7’s on a lot of the cards, D7 is the component pointed to with the worlds worst yellow arrow. From the 26 tone cards on the left side of the keyboard, twelve of them had gone short, and replacement with 1N4148’s sorted the problem, we found an out of spec transistor on one card and replaced that. We still don’t know why a whole bunch of diodes decided to commit suicide on switch on one day, but the signs were that this might be a common failure on the Multiman as some of the diodes had been replaced previously, and we had to replace some of these again.

Now the synth worked again, but a weird artifact on the lower strings had us hunting back in the chorus system again, and we eventually tracked down a faulty transistor in the 3-Phase LFO board which was replaced. Strangely enough all transistor failures were due to open circuit base-emitter junctions rather than the usual short circuits that failed transistors normally exhibit, perhaps something that late 70’s transistors do.

multiman-complete-sml

This is the finished result, the Strings are glorious, the Piano surprisingly useable, and as for the rest, it’s a matter of taste.

The Crumar Performer

We have had two of these in the last couple of months, both in reasonable physical condition but with a differing variety of faults. The Performer is a stripped down version of the Multiman, it doesn’t have the Piano or Clav and doesn’t have a tone generator board under each key, and only a 4 Octave keyboard, but a lot of the circuitry is similar, the standout feature for us is the String chorus circuitry which while being a bit smaller is of an identical electronic design to the Multiman and still lives on 4 pcb’s screwed to the chassis under the keyboard assembly. One interesting thing (amongst many others) is that the Service Documents describe a Moog style Ladder filter and discrete Contour Generator for the Brass filter, but both of the Performers that passed through recently had a Brass Filter board with some rare and valuable SSM chips on it, an SSM2040 for the filter and an SSM2050 for the EG. The late and great Jurgen Haible posted a hand drawn diagram of this circuit as no Crumar documentation seems to have ever included the changed pcb, as at one time he owned one. The Crumar Schematics are a little more intelligible than the Multiman but not by much.

The first Performer wasn’t too bad but wouldn’t stay in tune. It became immediately apparent that the Master Tuning pot on the rear panel was totally shot. Not a problem you would think but this pot was a low resistance value and was physically huge, the rear pot body was more than 30mm in diameter, the size of the mounting threads was also large and not Metric, as the Pot was British made. Finding any kind of Pot to replace this in the 9mm and 16mm Pot age was difficult, and it took some time to find anything suitable. The pot was a sealed type and couldn’t be rebuilt so a substitute had to be found, and eventually was. There were a few other issues with the first Performer, but the faulty tuning pot was the biggie on this one.

The second Performer had a few more problems, the main one was a Brass On/Off switch that wouldn’t latch,and the other obvious problem was that the Octave transpose switch was missing. This is a particular kind of slide switch that is still being manufactured by C & K, but is inordinately expensive for what it does and is. Unfortunately nothing else would fit properly so we had to buy one. The Brass on/off switch was a completely different issue, the latching system was just missing and as this is a push button switch with 18 pins behind it (it is a 6 pole double throw latching push switch) and of a type which seems to be completely unobtainable from any source, this seemed like a game-stopper, the thinking was about how to find a way of replacing this switch with something that worked and didn’t compromise the overall useage of the synth, crazy things like multiple Relays or CMOS analogue switches on Veroboard were considered, and then rejected as far too much hassle. Our solution wasn’t perfect but it was partly elegant as most of the extra parts of the switch need never have been switched at all, if you have the Brass turned off, no Brass appears at the outputs, so why were all the gate signals made switchable on a few poles of the switch along with a few other Brass gating things. To cut a long story short we fitted a normal 6 pin latching switch and rewired the rest of the switch mounting board accordingly.

The only downside of our mod is that the Brass LED is on regardless of switch setting, but in every other way it works perfectly. This particular synth had dirty sliders which is slightly unusual for the Spanish made Piher components used, as the slider track is perpendicular to the circuit board. To envisage what we mean, if you look at most synths close up that have sliders, you can see the tracks of the slider from above and dust and dirt can fall in there quite easily, but Piher sliders (as well as the sliders in most Russian equipment) are at right angles to the horizontal and while not immune to contamination problems are at least a bit more resistant to them.

These faders look and feel very similar to the CTS brand sliders used in ARP products, and indeed in the past they may have been just blatant copies, but there are some differences. CTS sliders are a real pain to remove from an ARP circuit board due to the metal support guides which also hold the two sides of the slider together. These seem to be made of an alloy which is solderable but at this age is as brittle as die-cast zinc alloy, so taking it apart will destroy the tangs of alloy which hold it together, and other means have to be found to hold the rebuilt and cleaned slider together. The Piher sliders are ultrasonically welded together, breaking them apart will ruin them with no remedy. Thankfully a Pipette full of Isopropanol drbbled gently through the left hand side of the sliders and some consequent excercising of the sliders made these good again.

This machine was good to go but the usual Italian synth things had to be dealt with, tarnished connectors, including all the output Jacks, the keyboard Bus bars etc…

Amazingly we took no photo’s of either of these, so we apologise for that omission, just Google the synth and you’ll have all the pictures you need of how a Performer should look, both of these left here looking like that.

And now for something completely different- the Juno 106

It’s not Italian, it is not rare, it is not exotic, but it is a much loved and popular Analogue hybrid synth and we still see one at least every two weeks with failing voices (usually, but by no means exclusively) and as the information on many Forums is sometimes confusing and incongruous, we thought we might try to make the testing for failure as simple as possible, so you can say with confidence that Filter module X is faulty and these days Oscillator module Y may be faulty. There are 9 hybrid modules on a 106 Voice board, 6 of these are the Filter and VCA’s for the synth, and the other 3 are Oscillator modules (each of which handles 2 Voices). The Oscillator modules convert the Digital note signals from the CPU into analogue style waveforms to feed into the filters. They generate the Square and PWM, Sawtooth and SubOctave for each voice, and when they fail, any one of those three outputs can be lost, and which of these die seems to be quite random.

Module failures manifest themselves in a variety of ways, for example our latest Juno 106 had two voices totally dead on switch on, but after 10 minutes two others were burbling away and producing random noises without any keys being pressed, and further testing showed the Sawtooth waveform missing on one of the voices which was one of those with a faulty filter module (sometimes identifying the fault can be tricky).

Everyone who has a 106 knows how to put it in test mode, hold the Key Transpose button down while powering up the synth, hold the button down long enough until the letter C shows on the central digital display and then release it. The display should show something like this — _ in test mode the patch buttons serve no purpose but the bank select buttons do, and the usual most useful button to select is bank 3, this slams all the filters into resonance where most Filter faults will be spotted, burbling or crackling or wild frequency fluctuations can all be heard in this setting. In basic test mode, the keyboard assigner is Mono, all 6 voices are played with each key. The solution is to hold down both Poly1 and Poly2 buttons in the keyboard assigner section until both Poly LED’s light up. Now the synth will play in normal mode and show the voice number on the right hand digit of the display, you can play and listen to individual resonance notes and identify them. The pitch of the notes heard will normally vary by as much as a semitone as you play a key and cycle through the voices, but unless the differences between voices is huge, this is fixed by the calibration routines described in the Service Documents.

Anyone considering buying a Juno 106 should run the following sequence of tests to check for bad voices, it is not the only issue that these can have, but it is the most common and also most likely to render the synth unuseable.

Switch on the synth holding down the Key Transpose button until the letter C appears and release the button a couple of seconds later, you should see the offset bars in the display as described above, then press the Poly 1 and Poly 2 buttons until both LED’s light, then walk away for at least ten minutes, as some voice faults take a little time to appear as the modules warm up. The Bank buttons 2 through 6 all give useful information as to the condition of the synth’s voicing circuitry, Bank 2 turns on just the Suboctave oscillator, a common failure mode of the MC5534 Oscillator module, if all 6 voices sound the same here, then move on to Bank 4 which turns on the Sawtooth waveform, failures here are rarer but not unheard of, Bank 5 switches on the Square wave from the DCO, again a rare failure but we have seen it, all voice Square waves should be at similar levels and have a 50% duty cycle, moving the PWM slider with the modulation switch set to manual should change the Pulse width from 50% to about 95% for all voices. The Bank 6 button turns on the Noise generator, not usually a problem but worth checking. The Bank 7 button just seems to produce a low level sine wave and Bank 8 just produces a blip noise when keys are pressed. These are not part of the calibration routine and don’t seem useful for testing functionality. Bank 1 switch is used for the VCA voltage offset calibration and you won’t hear anything playing keys in this setting. The most important Bank is 3, this shows up faults in the 80017A filter and VCA modules, any notes you hear when no keys are pressed are due to faulty voice modules (probably the VCA section) any warbling or rapid fluctuation in the tone produced when pressing a key indicates that the filter part of the voice is faulty.

We received a 106 a few weeks ago which we were told had voice issues, when we ran the test routine two voices were immediately dead, and 10 minutes later two more were burbling away at random, and the suboctave on voice 5 was missing. Filter problems were on voices 2,3,4 and 5. With 5 dodgy modules out of 9 we recommended the client to replace them all with analoguerenaissance modules, which he agreed to and we promptly did. This is expensive but at least the spectre of an unreliable 106 when you need it most is put to bed for good (we have never had a problem with these modules and have installed hundreds).

juno106-new-modules-lhs-sml

The above picture is of the aforementioned 106 with 9 new modules fitted and full recalibration done, the volume pot stripped and cleaned along with the Bender sliders. This one is done and going back to it’s owner shortly.

The analoguerenaissance modules are on a white pcb whereas the original Roland modules have a black epoxy coating on them, so they should be instantly recognisable. That’s all for August !

Facebooktwittergoogle_plusredditpinterestlinkedintumblr

Norlin Moogs and their issues

During the latter part of May last month and the beginning of June we had a Moog Rogue in for a rebuild, which ostensibly just had the old tar from the front panel liner syndrome. The Rogue, as does the Opus 3 has a panel wide Bitumen based foam which is meant to keep dust out of all the sliders and switches, and was meant to mean that the switch and slider bodies were hidden from view with a nice neutral black surface under the panel.

We have done a couple of Opus 3’s before and while it was a bit messy to remove the now sticky foam liner from them, nothing had prepared us for the state of this Rogue.

roguepanel1_sml

The above image may give a clue but not a real idea of the extent of the difficulty in removing this stuff from the board, the tar, and it certainly was now tar, was everywhere, on all the cables, all over components on the board, deep in the sliders (but not the rotary pots as it turned out). Even the inside of the front panel required scraping with screwdrivers and a lot of Isopropanol to get it clean, the Tar literally ate into the metal of not only the front panel but also the metal parts of the switches and slider bodies, the tar also removed the coloured stripes from a lot of resistors and other components on the main pcb.

roguepanel3_sml

The above picture is from part way through the process, most of the major tarry bits had been removed and obviously the fader stripping and cleaning had begun, tar is still all over the wiring and all of the switch and rotary pot bodies, all the switch bodies had to be scraped to even try and make them decent, and in the end all the switches had to eventually be removed from the pcb to be thoroughly scraped and cleaned in order to make them function correctly. Tiny fragments of tar got everywhere, all over tools, hands, clothes, everything.

roguepanel_done_sml

The above photo is of the board getting close to being complete, the wiring is all clean, the pcb is pretty clean, and only one slider is due for its rebuild, but the fun was only just beginning.

The keyboard wasn’t working well, the majority of the keys in the centre of the frame were not working properly, so we thought a good cleaning would sort this out, if only it was that simple. Someone in the past had spilt tea down into the keyboard and it had wreaked some havoc with the contacts, to the extent that for one key the contact bar had been eaten away and was just gone. A piece of Pratt-Reed J-wire was inserted into the contact retaining plastics with some difficulty, and hard wired to the common bus bar. The keyboard mechanism in the Rogue is a Panasonic one with a contact system of twin leaf switches per note, one for pitch and one for gate. Very thin and very delicate and difficult to clean and realign if necessary, which it was. This keyboard mechanism is shared by the Moog Prodigy, which we will discuss in the next article.

When we got the synth and had to take it apart, we marked the connectors in the usual way, permanent marker applied to both sides of a pluggable connector, different connectors are marked in different places so that mistakes should be impossble. Somehow or other we had tremendous problems with the CV voltages from the keyboard which would plunge to -10V on key release, and weird behaviour from the envelopes and VCA, and it was… yep the keyboard connector was in upside down, we did mark it and put it back together as marked, but it was obviously wrong. The reason that this took up far more time than is decent is because the scaling and tuning of the keyboard were still fine, we got 1V/Oct across the keyboard without problem. We replaced every chip in the envelope and VCA section trying to hunt this one down and it was only in desperation that we went back to the the details of the Service Manual and checked wiring colour codes against pin numbers on the connector. Obviously the penny now dropped and we rectified this. The irony is that the synth’s tuning didn’t need to be recalibrated at all, we had proper envelopes and still perfect tuning, no wonder that this was a tough one to find. At least the two envelope chips (CD4007UBE) are now in good quality sockets and they are well known failures in old Moogs, and some Korgs, so replacement in the future will be just a chip swap, if you can get them. They are not like CEM’s or SSM’s but they have been out of production for decades as their day has gone. They are not particularly expensive, just proving more elusive by the year.
Obviously we could not leave the toggle switches naked and exposed to view, so we made black cloth tape covers for them all that supplied all of the visual benefits of the original foam.

The Prodigy-Smack my synth up

The title line isn’t entirely a joke, the woodwork around this synth was entirely broken up, it apparently had been knocked off a keyboard stand and fallen heavily, but this wasn’t its first fall, there were signs of woodwork repairs internally from many years ago, and whatever glue was used, it was white and very brittle, as from the synths second dive, most of this had broken up. The synth was originally sent in to be tuned, and to have the woodwork sorted but even this one had a lot of hidden gremlins. Oscillator 1 was supersonic most of the time, creating a high pitched whistle which defied keyboard control, or oscillator octave switching, or anything else. It was indeed the obvious CA3046 which had failed, it’s a well known phenomenon with these mid-time Moogs. The keyboard once again became the main focus of this repair, it was awful, again a lot of the keys in the middle of the frame behaved intermittently and badly, a full day was spent on this mechanism trying to make the gating and CV contacts behave correctly, and even then the next day there were keyboard issues.

Skipping those for a while we had to investigate the very poor tuning reliability, all attempts at calibration were fine at the time of doing but futile minutes later. The power supply was solid, no ripple or dropouts so that wasn’t it, but testing led to suspecting the octave select switches. It turned out that the contacts in ALL the slide switches were as black as the ace of spades, and whereas some were accessible to solvent cleaners directly, the Octave select switches had to be removed from the PCB to be cleaned properly, Isopropanol and Q-Tips were good enough to clean these, as seems to be the case for most open slide switches on any synth.

So far so good, the next gremlin was an intermittent massive modulation problem which would badly affect the power supply, and modulate everything, it seemed like a dying IC and indeed it was, the clue to which one was the fact that the synth began to drone endlessly if the sustain was turned up, yes it was one of the 4007’s in the filter ADR, and a day later the amplifier ADR chip also failed. We were not massively surprised, these IC’s are very delicate and very old, and it is better that they die here than 200 miles away and back with the client. The reason for this IC suicide we think is due to the large number of power cycles that occur when a synth is being repaired, during a difficult repair a synth may be powered up and down a hundred times or more, which is probably far more than most users would do in a year, and old and weakened IC’s will give up under the onslaught.

I apologise for not having any Prodigy pics for this one, it was very time consuming, and no pictures were taken apparently. We got the keyboard working well enough but didn’t trust it for the long term, so we fitted CV and Gate Jack sockets on the back, but beware, there is a technical mistake on all the documentation for the official updates from that era, if you connect to pin 2 of U12 as the keyboard in/out as it states, when you feed an external CV in, any change in the filter keyboard tracking switch reduces the Voltage and detunes the synth and by a considerable amount, due to the 4k7 resistor in the feedback loop of the Op-Amp. Use pin 1 for the keyboard switching Jack and all will be good.

As a small aside, the power LED had failed, so we replaced it, but we tried an experiment where we used a Tricolour LED with the red side wired for power as per normal, but used the green side of the led to receive drive from the LFO comparitor so the power LED flashes at the LFO rate, the LED is on the PB/Mod panel so this wasn’t entirely out of place, and no extra holes had to be drilled. The client hasn’t complained so perhaps he likes it too, all synths should have blinky lights for LFO rates!

Jupiter 8, When Good Synth’s go Bad

Out of the blue we got an e-mail stating that a client had a JP8 which had died a number of years ago, and which he would like repaired, fair enough, no one wants to own a dead JP8. This one was fitted with a Groove MIDI interface from 1987 and he suspected this might be the issue. We have a lot of experience with JP8’s, having owned one in the past and having serviced a surprising number from this small island of ours, many people thinking they have the only one in the country and being shocked to hear that we have worked on a dozen at least.
They have a reputation for being a very reliable synth, and this is well deserved, they are very well engineered and a quantum leap from it’s predecessor, the Jupiter 4, the engineering leap between the two still staggers us.

But they are getting old, and random IC failures plague this beauty as much as it’s contemporaries, and in a similar manner to the OBX(a)’s and the Prophet 5’s, old chips will die.

Due to circumstances that occured several years ago, we have a spare CPU board, Interface board and one Module Controller board in our spares section, all of which were fully functional last time we checked them in a JP8, so we replaced the CPU board in this Jupiter with ours, and it booted fine, Autotuned and basically worked, at this time we diagnosed and fixed op-amp failures on two voice cards, one to do with the VCO modulation section on Voices 1 and 2, and another which concerned the filter frequency control on Voice 3. At this point the 8 voices all worked and sounded similar enough to each other, so attention was now drawn to the front panel, and the scratchy and intermittent sliders, and broken pots and switches. Full recalibration of a JP8 requires a fully functional front panel, it doesn’t have the test programs of the later Juno 60, and as there are about 220 trimmers in a JP8, this is no small job. Apart from being scratchy, many of the slider stems were bent and some were distorted due to being bent many times, you can’t take a pliers to these in situ and hope for a good result, so the entire front panel set of pots and sliders were taken out of their pcb’s, dismantled, fixed, cleaned and rebuilt, and new lubrication applied, as the old stuff had turned to a sticky jelly, and wasn’t smoothing anything.

And then the unlikely happened, the JP8 refused to boot in a similar way to the original problem with our CPU board in it, all scoping seemed to point to the Z80 CPU having failed, as it had with the synth’s own CPU board. We reckoned that the Sharp Z80 CPU in the original board was the main problem in the synths own CPU board so we took it out and put in a Mostek Z80 from our Sequential Tom drum machine, we will replace this with a Zilog Z80 in time but it did work, the only issue was Autotune failed miserably every time, the tuning comparitor LM311 on the synths original CPU board had also failed. Replacing this gave us full Autotune success and almost full JP8 functionality, We only qualify this because only full calibration can reveal any nasties that remain in this synth, and we do expect a couple more, dying NEC4558’s have been the main other problem so far.

The Groove MIDI interface has been unplugged throughout all of this, so it wasn’t a failure mechanism, but obviously we have to get the synth as perfect as possible before reintroducing the Groove, just in case there was an issue with this. I do have some photographs of this process, but actually nothing that looks good on a website

We are into mid July at this point in time, but for continuity we’ll continue this story here, this Jupe is a gift that keeps on giving, problems that is. Suddenly we lost Osc 1 on voices 5-8, this was due to a blown 4558 in the VCO frequency summing section on the Module Controller card (Mod Con or Con in Rolandspeak). The Op-Amp went to full voltage rail output sending all the Osc 1’s supersonic so it seemed like they were dead but were just oscillating at 50 kHz. Fixing this gave us full Autotune goodness again. A day was spent calibrating the upper 4 voices, it all looked good, most of the trimmer settings were still fine after 30 odd years but the odd ones were well awry, all looked good for finishing this the next day, but then all of a sudden Voice 3 Osc 1 had failed Autotune, and scoping it showed the Oscillator itself to have gone supersonic, we’ve seen this before many times in Moog synths (the Micromoog, Prodigy, Rogue) and it almost inevitably means the 3046 matched transistor chip has failed, theoretically no problem but for the fact that in a JP8 the two Tempco’s are tiny and irreplaceable and glued to the 3046, and due to the short leads of the Tempco’s, putting a nice gold plated DIL socket is not a viable option, so with some very careful surgery to safeguard the Tempco’s, cut out the 3046, replace it and find a good way of thermally bonding the Tempco’s back to the new chip this can be made good. We have worked on a lot of JP8’s and this is the first time we’ve seen an Oscillator failure like this, but in the grand scheme of things these failures have happened elsewhere and we are aware of the failure mechanism.

Fixing the Oscillator has proven to be more problematic than we thought, the obvious suspect (the CA3046) was not the culprit, it usually is on a Moog. It was quietly oscillating at 65kHz and doesn’t seem to respond to voltage control. As the VCO is a closed loop system with many components. The Oscillator timing cap C6A was taken out and tested, IC3A was replaced, TR1A was replaced but with no change, the NF510 FET was taken out and checked and seems fine, we will probably swap it with Voice 4 Osc 1’s FET to confirm it is good, if it is the faulty item then Voice 3 Osc 1 will tune and Voice 4 Osc 1 will not , but I doubt that will happen. All that is left are IC1 (which seems ok, voltage summing seems fine but it is a C4558) and IC4 which is the comparitor. Removing IC’s from this Voice board is very difficult indeed, it just doesn’t want to let them go and as TL080’s are not readily available, cutting them out is not an option, time and patience are required. If IC4 is at fault then we will take one from the Envelope buffer as that can be replaced with a TL081 without any consequences, as the 80’s higher slew rate in not needed in an EG, I suspect the only reason the Envelope circuits have TL080’s is because it kept inventory simpler at Roland, TL082’s are used extensively elsewhere.

The answer did turn out to be IC1, the C4558, the voltage summing half of the chip was fine but the Servo/ Integrator was obviously faulty (with hindsight) as the bad readings on this op-amp could have been the result of any problem in the Oscillator system and logically this one seemed the least likely, the only reason we changed this before IC4 (the TL080 comparitor) was because our run of luck with this synth seemed to have C4558’s dropping like flies, and indeed this was another one. As of this time there haven’t been any further casualties and the synth seems perfect after calibration. By the way, the Oscillator tuning is easy and very reliable, DO NOT mess with the Filter calibration, the service manual is very cryptic, relying on pictures of an Oscilloscope readout which is difficult to achieve even with a good scope and which in every JP8 we’ve seen are just plain wrong, the resonance settings on every Jupe we’ve seen are far stronger than the manual indicates, and while the JP8’s filter doesn’t self oscillate as standard, it could but I suspect the filter sound in response to programming would change the nature of the synth too much. This JP8 was an earlier one which didn’t have the two little switches on the Module Controller board to short out Noise and to make all the filters resonate (used by Roland later for filter calibration to simplify it in production) so the Resonance boost update was added to both Module Controller boards (shorting out the noise is simpler to do during calibration with a jumper wire). We fitted jumper links like you would see on the back of computer hard-drives, there are spare pads designed into the pcb for such weirdnesses along with space for a spare TL082 which seems to have never been needed.

After checking everything out again many times, stripping and rebuilding the output board (volume balance and Arpeggio speed) and testing again, it was time to reinstate the Groove Electronics MIDI Interface. This went well, we half expected the worst but it didn’t happen, the JP8 booted normally, tuned very quickly and behaved flawlessly including accepting MIDI in and producing MIDI out. The Groove was unusual (and great) for the time in that it sent keyboard information and program change information to MIDI out, something that has become commonplace only in recent times, but for 1987 this was forward thinking. Probably this was just a consequence of integrating the MIDI interface with the original Operating System on a very intimate level, a MIDI interface replaced the DCB interface, and the new firmware swapped DCB stuff for MIDI compatibility, the Europa upgrade for the Jupiter 6 seems the closest analogy, and Lazlo’s upgrades for the JP4, the Juno 60 and hopefully soon for the JP8 all seem to embrace this ethos, as have many other modern updates of recent times. The Kenton approach of older times of one interface fits all doesn’t seem efficient or cost effective these days, and this is probably why most have been discontinued. We have a few old Kenton interface boards here and they were hand crafted for each individual model of synth from a common template, it must have been very time consuming and expensive for Kenton to do the individual mod’s for each type of synth, not to mention writing the Firmware for them all. Kiwitechnics have taken a no holds barred approach to upgrades, some require just processor changes, some entire pcb changes but there is no doubting the quality of the supplied parts or boards even if they are not cheap, but they are good!

jp8-voice1-4-sml

The above picture gives some idea of the number of Trimmers on each voice card, the pic shows voices 1 to 4, voices 5 to 8 are in a similar position underneath.

jp8-modcon

The above picture is of the the upper Module controller board, the jumper link for the Resonance boost is quite discrete but it is the small green jumper to the right of the Env1S trimmer, all wiring was done on the underside of the pcb for neatness, and didn’t affect signal flow as this was what Roland did on the later ModCon boards.

This synth is finished and back with it’s owner, and all seems to be well.

Facebooktwittergoogle_plusredditpinterestlinkedintumblr

Business as usual

It’s a funny thing, the last two months have been some of the busiest ever, and the most successful in terms of the synth comes-in, gets-fixed and goes-out ethos.

The ironic thing is that there are no horror stories, no disasters and no spectacularly interesting problems that were solved. Most of them were Digital synths, and most of these have onboard diagnostics as they date from the 1990’s or later, and the problems with Digital synths are normally Human interface things like the keyboard, the sliders, encoders, displays and the like, all things that are bread and butter to us. Getting spare parts is still tricky but not impossible, and as Digital synths have become almost a disposible item if they break, our spares collection of user interface parts is quietly growing, although the storage of the carcasses is somewhat of a pain.

Facebooktwittergoogle_plusredditpinterestlinkedintumblr

Couple of Prophet 5’s

We have a couple of Prophet 5’s here at the moment, one is a Rev 2 and the other a Rev 3.0. They both have quite different problems as to why they came to us. The Rev 2 passes Autotune perfectly well but then a few oscillators detune rapidly afterwards rendering the synth unusable. Checking with some of the world’s experts on the Rev 2 basically ended up with the idea that it really could be anything. Truly anything, from the power supply, the Sample and Hold systems, the Autotune Multiplexers, the Oscillators themselves, the sockets the chips sit in, the timing capacitors, any of the CMOS switches that do the Polymod functions, the Tuning and Scaling Trimmer pots, the list goes on. Also the problems could be a random mixture of all of the above for different voices. The normal run of events is to try and track down a fault by methodically going through the circuitry and hunting down anomalies, but with the Rev 2 there are no obvious measurable anomalies other than a Tuner telling us things are not right, a few millivolts of any leakage or droop is difficult to monitor in a 10 Volt CV system.

Back in the early days of programmable Polysynths, there was a mindset that they had to be designed to almost mimic their non programmable modular cousins, the 0 to 10 Volt control was the norm, with one Volt per octave being the standard. The Rev 2 Prophet 5 was one and the Oberheim OBX was the other, so all the Sample and Hold circuitry sent out 0 to 10 Volt signals by default, and the circuitry around the Oscillators, Filter and VCA’s were designed to match up with the IC Datasheets which were also marketed as a way to make synth modules with a reduced parts count due to these miracle chips.

With remarkable speed, synth designers realised that it wasn’t necessary or desirable to adhere slavishly to the datasheets, and both Dave Smith and Tom Oberheim redesigned their flagship products to make them more reliable and easier to produce. Those products being the Rev 3 Prophet 5 and the OBX-a. Both designers realised that a 0 to 5 Volt DAC was more feasible and reliable, but also made much of the interface circuitry much simplified, and scaling the resistor values around the expensive Curtis IC’s didn’t change the sound, as they were for the most part current controlled.

We are aware of the anomally that all Analogue synth modules are described as VCO’s, VCF’s and VCA’s but the truth is that they have always been current controlled, voltages from the control interface are converted into currents to work the actual active circuitry, currents charge the capacitor for the Oscillator, control or strangulation of current into the filter dictates the Cutoff Frequency, and so it is with the VCA.

p5rev2voiceindicatorsml

The above image is of a Voice indicator we built for the Rev 2, the tuning drifts made the Sequential technique of putting your finger on the Oscillator chips to see which voice was playing completely futile. Why on earth did no Prophet 5 ever get one of these built in, the Jupiter 8 had them, the OBX-a eventually got them, the Korg Polysix had them. The LED’s are low current 1 mA types and the resistors are 10 kOhm so they don’t load the gate signals to the EG’s. It works well and will be left in the synth. Voice 2 is being triggered here, and the LED’s are much brighter than the photograph shows.

We spent some considerable time with this Rev 2 a while ago, endless scoping and Voltage measurement didn’t get us anywhere, so this time we stocked up on lots of spare parts and just started replacing things.

The logical first step if Autotune doesn’t do it’s job when it thinks that it does is to replace the 4 Multiplexers that handle the Oscillator tuning biases (the error correction voltages) and the oscillator frequencies themselves that are fed to the CPU for comparison. These IC’s are MC14051 chips which are not expensive, the gold plated turned pin sockets we put them into are far more expensive. There was a definite improvement after these were replaced, as the Oscillator drift occured after a few minutes as opposed to seconds, and was not as severe. For some strange reason, it was always Oscillator 1 on a few voices that gave trouble. The next port of call were the MC14016 quad analog switches that handle the waveform switching of Osc 1 and also the Sync, a little bird told us that weird behaviour in these chips can cause detuning even though they are not obviously involved in the frequency control system, so we replaced them all in all 5 voices. There is a current internet wisdom that old CMOS multiplexers and Analog switches will degrade over time, and new ones are much better in a number of ways, such as higher speed of response, lower On resistance and better protection against overloads.

Previously we had gone for the obvious culprits if oscillators won’t stay in tune, bad SSM chips, bad sockets, bad capacitors, bad Sample and hold capacitors… All of these possibilities were tested and components replaced with no improvement. Replacing the CMOS was giving us far better results but not perfection yet.

The main issue for us with Prophet 5’s is that they seem to die a little more while you work on them, changing chips and other components (usually one at a time) requires a great many switch off and then subsequent switch on’s, which on a weirdly ill machine may require hundreds of power cycling events, and as soon as you have finally isolated a problem another one appears, as in when switched on the synth is dead, no lights, no autotune, nothing ! So while you were trying to resolve an analog issue, a whole bunch of the CPU section was waiting to commit Hari-Kiri, and as soon as you have succeeded with the first, the latter kicks in, leading to another very time consuming round of Digital signal analysis to find out what chip has died now. Only one P5 has been exempt from this in our experience (and yes Andrew it’s yours!).

On a seriously sick P5, we will by default change all the Tantalum decoupling capacitors on the power systems for low ESR electrolytics, they are on all the PCB’s and there are around 25 of them. Old Tant’s can go short at any time and besides causing one of the Voltage Rails to self protect, which in theory causes no damage (although the synth won’t power up) the endless Internet jury is still out on the possible side effects of bad power sequencing on the exotic SSM or CEM chips on the voice cards, and potentially the multiplexers also. There are Tantalum caps in other parts of the system but as they have resistors in series with them and many are in the Audio path they are very unlikely to go short and they have not been problematic and thus never been touched, the Audio artifacts attributed to Tantalum’s in Audio systems may be a contributing factor to the Prophet sound, so we leave them alone, and if by chance one were to fail we would replace it with another Tantalum capacitor.

On this particular Rev 2, the object of the rebuild ( it was originally meant to be just a repair, if only it was that simple) was just to get the voices back in tune but even as this process was ongoing, Dying Prophet Disease started to rear its head, one of the SSM2050 EG chips died, and swapping this with a known good one confirmed this, and several switch on’s later a second Envelope died although this wasn’t the chip, it was a ceramic decoupling cap surrounding the chip. After the EG’s were sorted and we thought normality had returned to proceedings, now the autotune tuned everything a Semitone high, leading to the possibility of a stuck bit in the DAC on the CPU board, although it could be almost any part of the master Oscillator tuning section on the CPU board. We have the chips and we have the test gear, but any way you slice it this is going to be very time consuming.

Facebooktwittergoogle_plusredditpinterestlinkedintumblr