The No.1 Website for Pro Audio
 All  This Thread  Reviews  Gear Database  Synths for sale     Latest  Trending
Troubleshooting an Alesis A6 Andromeda - No Boot
Old 19th February 2018
  #301
Gear Maniac
 
toobdude's Avatar
 

Quote:
Originally Posted by No.7 View Post
Unfortunately it's unlikely that Mouser or Digikey would stock the 60w ribbon for the front panel cards, because it has three connectors rather than just being a simple 'end to end' ribbon.

60w ribbons are more difficult to DIY than, for example, 25w ribbons because there are many more pins to pierce the insulation. A lot more force is required and that can distort the connector. I'll be making up a brand new ribbon for my A6. I should have enough parts for two ribbons. I'll let you know how it works out and maybe I could send you the extra one.

I'd note that KNOWING a ribbon is defective is really helpful. It can explain symptoms. Replacing a ribbon without testing it might not explain anything. So testing it is still a very good idea.

I've included a photo of my ribbon. This was factory fitted! Three pins are shorted (11,12 & 13) and one is entirely missing (6)! That's how bad these ribbons can be.
Damn!! I have some homework to do .......
Old 20th February 2018
  #302
Here for the gear
 

Quote:
Originally Posted by No.7 View Post
Unfortunately it's unlikely that Mouser or Digikey would stock the 60w ribbon for the front panel cards, because it has three connectors rather than just being a simple 'end to end' ribbon.

60w ribbons are more difficult to DIY than, for example, 25w ribbons because there are many more pins to pierce the insulation. A lot more force is required and that can distort the connector. I'll be making up a brand new ribbon for my A6. I should have enough parts for two ribbons. I'll let you know how it works out and maybe I could send you the extra one.

I'd note that KNOWING a ribbon is defective is really helpful. It can explain symptoms. Replacing a ribbon without testing it might not explain anything. So testing it is still a very good idea.

I've included a photo of my ribbon. This was factory fitted! Three pins are shorted (11,12 & 13) and one is entirely missing (6)! That's how bad these ribbons can be.
My Andy's front right pcb was wrong sometimes due to the faulty ribbon cable. The led matrix did not work properly and the pot mux sent wrong data to pcb (the pot values randomly changed). A new cable can be the solution. Now i have no problem with it after i moved the cable, but i will order this, and i will make a new cable just in case
Old 20th February 2018
  #303
Gear Maniac
 

@ anteelope . Excellent! Perfect kit. Pretty good value too. 60w connectors can be difficult to crimp though.
Old 21st February 2018
  #304
Gear Maniac
 
toobdude's Avatar
 

Quote:
Originally Posted by anteelope View Post
My Andy's front right pcb was wrong sometimes due to the faulty ribbon cable. The led matrix did not work properly and the pot mux sent wrong data to pcb (the pot values randomly changed). A new cable can be the solution. Now i have no problem with it after i moved the cable, but i will order this, and i will make a new cable just in case
Nice! I'll check my cable and get one of these too. Thanks.
Old 21st February 2018
  #305
Lives for gear
 
oldgearguy's Avatar
 

Quote:
Originally Posted by toobdude View Post
I will add checking the pin continuity of the ribbon cables to my list of "things to check". That would suck if that had been the problem all along. I actually had a similar thing happen when restoring a Memorymoog. I could not figure out why the power supply which I totally rebuilt was not working. I spent SO MUCH TIME sorting through the PSU and the problem ended up being the wire connector to the circuit board was just not making contact - a couple of the little spring-pins inside the connector had lost their "spring" after 30 years and was not contacting the pin on the circuit board. Once I bent the clip back such that it made contact with the board pin all was well. A 1 minute fix took me weeks to figure out!!!
When I owned some pinball machines and was constantly repairing them, I was surprised to learn that those typical Molex connectors had a durability rating of 25.

Yep - 25 connect/disconnect cycles before the plating and tension was starting to be compromised.

So it's worth checking all those mechanical connections first before diving into the circuitry, especially on connectors (like the power supply) that get pulled and reattached often.
Old 21st February 2018
  #306
Gear Head
 
Genesis92's Avatar
yes and it depends of the environment too.
25 cycles is for clean air. if there is acid gases in the air this can be much lower.
I know a project used near roads who had lot of problems caused by car exhaust on such connectors because they are not sealed.
Old 21st February 2018
  #307
Gear Maniac
 

And 'smoke' machines. They'll fill your gear with gunk.
Old 14th March 2018
  #308
Here for the gear
 

Andromda A6 yet another splash screen problem

Just bought an Andy "as is" off Ebay and trying to resurrect it. The synth is in mint condition and very clean on the inside.

When I power it on it gets stuck at the splash screen. From reading this thread my first attempt was to replace the 3 Mhz crystal which didn't bring the synth to life. I did not replace the SMD caps next to the crystal but will do that soon. The PSU seems to be ok although the voltages do vary +/-1V relative to the documented ones.

As for the next iteration I'm not sure what to do and I'm hoping there is someone here who can give some advice.

I thought about replacing the CPU but I'm not sure if getting to the splash screen argues that the CPU is functional. Can someone please comment on this?

Next, I will try to replace the FLASH with one holding the image that was shared earlier in this thread.

Does this seem like a reasonable approach?

Best
Old 14th March 2018
  #309
Gear Maniac
 

If you get a splash screen the front panel processor is okay and the crystal is most likely NOT the problem. You don't need to swap the load caps for the crystal. What you really do need to do is check your ribbon cables for shorts and open circuits, especially the 60w ribbon that spans the three front panel boards. Plenty of people have found fault with them.

After that it's best to make sure of the power supply, any extra ripple or noise on the 7v (and 5v) rail can cause spurious operation. Recapping it is not a bad idea. Swapping the Coldfire CPU is fairly drastic and best done after ribbons and PSU have been thoroughly checked out.

If you have a decent cap or LCR meter it's really worth checking all the caps that decouple the logic supplies. With a bit of luck your problem is down to ripple and noise and not a blown CPU.

Check BCLKO Pin 170 on the Coldfire. It isn't used in the A6 but you should still be able to see the Bus Clock on this pin. If you do then the Coldfire isn't entirely dead.
Old 14th March 2018
  #310
Gear Addict
 
adhmzaiusz's Avatar
Quote:
Originally Posted by tkandersen View Post
Just bought an Andy "as is" off Ebay and trying to resurrect it. The synth is in mint condition and very clean on the inside.

When I power it on it gets stuck at the splash screen. From reading this thread my first attempt was to replace the 3 Mhz crystal
Hmmm.... before doing anything else, i would suggest holding soft button 1 on boot and see if you get the option of initializing memory. See if any boot commands work before shot-gunning any more parts!
Old 16th March 2018
  #311
Here for the gear
 

Quote:
Originally Posted by No.7 View Post
If you get a splash screen the front panel processor is okay and the crystal is most likely NOT the problem. You don't need to swap the load caps for the crystal. What you really do need to do is check your ribbon cables for shorts and open circuits, especially the 60w ribbon that spans the three front panel boards. Plenty of people have found fault with them.

After that it's best to make sure of the power supply, any extra ripple or noise on the 7v (and 5v) rail can cause spurious operation. Recapping it is not a bad idea. Swapping the Coldfire CPU is fairly drastic and best done after ribbons and PSU have been thoroughly checked out.

If you have a decent cap or LCR meter it's really worth checking all the caps that decouple the logic supplies. With a bit of luck your problem is down to ripple and noise and not a blown CPU.

Check BCLKO Pin 170 on the Coldfire. It isn't used in the A6 but you should still be able to see the Bus Clock on this pin. If you do then the Coldfire isn't entirely dead.
Hi

Thanks for the nice information. That surely gives me a direction. I will let you how this project progress.

Much appreciated
Tom
Old 16th March 2018
  #312
Here for the gear
 

Quote:
Originally Posted by adhmzaiusz View Post
Hmmm.... before doing anything else, i would suggest holding soft button 1 on boot and see if you get the option of initializing memory. See if any boot commands work before shot-gunning any more parts!
Hi

Thanks for the advice. I forgot to mention that I did try to solve the synth problem using the soft buttons - no luck there!

Best
Tom
Old 22nd March 2018
  #313
Here for the gear
 

Quote:
Originally Posted by adhmzaiusz View Post
Hi Barry, you could check out the Mean Well catalogue on Jameco to just replace the switching power supply with something equivalent. They've really come down in price... for another project I picked up a +/-15v 5V for something like $14usd.
My power supply seems not healthy. I replaced the electrolyte caps and at no load I'm measuring:
v1 = +7.7V
v2 = +12.4V
v3 = -17.7V
If someone knows where I can buy a PSU that fits into the synth I'm very interested.

I will probably replace the ICs on the PSU as a next step.

Best
Tom
Old 22nd March 2018
  #314
Gear Maniac
 

No load? The PSU won't work properly with no load. Many SMPSUs don't, it's just damned difficult to tell which ones! If you really want to test it 'outside' the A6 knock up a dummy load. A 100R 3w-5w resistor across +7V and Dig GND and 300R 3w-5w across each audio rail and Audio GND. That'll draw 50mA or so from each rail and guarantee that the PSU starts up properly. It's well worth having a load like this to hand, useful for almost all 'logic + split rail audio' type PSUs. Fix the resistors to a heatsink for prolonged 'soak' testing.
Old 24th March 2018
  #315
Here for the gear
 

Quote:
Originally Posted by bklein View Post
In my experience with 3 systems it looks like the flash is getting corrupted as a result of an attempt to update it due to problems already experienced. All 3 units needed the cpu swap. I have programmed replacement flash parts if interested.
Hi

I’m interested in buying flash replacement parts. Please PM me.

Thx
Old 12th May 2018
  #316
Gear Nut
I finally almost finished repairing my Andromeda. Replaced around 70 caps on the analog output board, added the resistor bridge to the SRAM chip, replaced the battery with a battery holder and replaced the display with a really nice blue one from eBay (still need to modify the housing of the Andromeda as this new display sits about 3mm below the surface).

Most importantly, thanks to BKlein, the Andromeda now fully boots for the first time since I've had it (11 years). I used to have to turn it on, off, on to fully start. This must indeed have been a dead ceramic resonator on the original display. I'm super happy about that.

My Andromeda seems to forget its tuning now though it seems, after a restart everything except VCA says nottune. I don't remember if this is a known bug. Probably something is not working correctly with my new battery holder.

Also the problem of effects sounding endlessly with some voices remains. Strangely enough it is now only voice 6 that does this, I remember it being voices 3 and 4 before. I wonder if this could be caused by a bad ribbon cable.

Does anyone here have a clue what could be causing this strange problem? How can the digital effects have a problem with only specific voices?

Anyway, many thanks to the awesome people in this thread providing much needed information for fixing this fantastic synth.
Old 12th May 2018
  #317
Gear Maniac
 

Great news to hear that it's booting. But problems remain with tuning and FX?

A while ago I did a comparison of the PCBs by extracting JPEGs of the layouts of different PCB versions from the manual and superimposing them to check for any differences. There wasn't much but on the Analog Master PCB I noticed a significant change, diodes added to U31 to clamp it's output to the supply rails.

I thought I might add these diodes so started working out where exactly the relevant tracks were and where I could fit the diodes. While doing that I noticed that U31 had changed it's power rails from +5VA to +/-12VA. The clamping now made perfect sense.

So I scrutinised every detail of the Analog Master PCBs power layout on it's schematics (Pages 62 and 68) and discovered that three opamps had changed their power rails over the various versions of the Analog Master PCB (Versions A to D).

U29 and U30 changed from +/-12VA to +5VA between versions B and C. U31 changed from +5VA to +/-12VA (and the diodes were added) between versions C and D.

U29 and U30 are the FX Send outputs from the Analog Master PCB. U31 is the OSC/FILT_ASIC_TUNE and the OSC/FILT_ASIC_TEMPERATURE outputs from the Analog Master PCB.

Clearly there were problems with FX levels and ASIC tuning that Alesis tried to resolve by changing the scaling of these output and control signals. It seems likely that they also had to change the software interpreting the control lines. It may be that there's a slightly different version of software required for versions B, C and D of the Analog Master PCB.

I don't know if this explains your problems, but it does relate to the areas concerned.

What to do? It'd be really useful to know if there are different versions of software/firmware for the different versions of Analog Master PCB.

Sorry to be long winded, but I thought it best to post all the info I have.
Old 12th May 2018
  #318
Gear Maniac
 

Testing Ribbon Cables:

Testing ribbon cables can be a pain in the output port. Not only must open circuits be tested for but also short circuits to adjacent pins. With a 60 way ribbon this is no fun.

Ribbon testers are available but they can be expensive. Network cable testers are cheap. Take a 2m Cat5 patchcord (flat cable cords are best), cut it in halves and terminate each half to a 0.1" pitch IDC shrouded header, cut off the shroud. Plug the patchcord halves into a network cable tester.

With this you can test 8 ribbon connections at a time, with opens and shorts indicated.

Note that, depending on the connector used, when you cut off the shroud the IDC bar may fall off. It can simply be glued in place.

There are other termination styles, connectors, and test adapters that could be used, but I reckon a flat Cat5 patchcord and shrouded header are easiest to source.
Old 12th May 2018
  #319
Gear Nut
Thanks for your quick and detailed replies No.7! That is very useful information.
Turns out I didn't have the thread subscription turned on, fixing that now.

I've never heard about those analog board revisions before. That could indeed be related to the problem I'm having. I haven't come across any different types of firmware.
It is a really strange problem, the voice keeps on sounding, but only through the digital effects. Then when the voice is played again the problem changes to the new note. It seems to get progressively worse when switching between programs (not entirely sure what exactly makes it worse, the problem isn't there initially), not when you just leave the Andomeda on.
Even stranger that it changed from voice 3 and 4 to voice 6, this indicates there is something that easily affects the problem.

It tunes fine and did so before. However now that I fixed it, it only remembers the VCA being tuned after a reboot, and not anything else. I'm not sure if that could be caused by the battery holder, or if it is a new problem.

The ribbon cable tester solution sounds interesting, I'll look into that.

Thanks again for the help!
Old 13th May 2018
  #320
Gear Maniac
 

Thanks for your thanks. A response like that makes the whole idea of forums worthwhile.

That really is a weird one! Maybe the presets have been corrupted and need to be cleared.

I'd guess that the FX Send mod (U29 and U30) had to do with distortion and the OSC/FILT mod (U31) had to do with pot jitter (the OSC/FILT lines share the Main MUX and ADC with the pots) and tuning problems . The FILT_ASIC_VCA_CAL also shares that MUX, so it may be that noisy or 'clipping' OSC/FILT lines can upset tuning and calibration just as they upset pot reading. This is mentioned briefly in the manual, but it's significance isn't obvious.

There is an ASIC_TUNE_PULSE that goes to the main CPU, so I wonder whether constant activity (or noise or 'clipping') on this line might upset the main CPU. But, having scanned the manual and read your posts carefully, I doubt a relation to your FX problem.

The manual does mention bad ASICs causing problems and does list test procedures to identify them. If your problem stayed with Voice 3/4 you might suspect a bad voice or filter ASIC, DIG_DATA, ANALOG_DATA, or Enables to these ASICs. Voices 3/4 share ASICs and various control lines.

But the problem swapping to Voice 6 contradicts that suspicion. Voice 6 does not share ASICs or much else with Voice 3/4. The only commonality (apart from power) is in the Voice Address and VGER_DATA that both come from the main board. So I think you're right to suspect a memory problem. Doing the 'Slow SRAM' test seems a good idea.

DIGFX1(2)_BUS are separate outputs from the Filter ASIC and the only way of controlling them seems to be the DIG_DATA, ANALOG_DATA, Enables, or VOICE_ADDR lines. VOICE_ADDR and Enables are common with other ASICs, leaving the DATA lines which are handled in blocks of four voices so they too are common. Affecting one voice or filter seems to be something that can only be done in 'software', so most of the hardware must be innocent.

The only other things to suggest are making sure the PSU voltages on the ASIC board are good and clean, and that there isn't a soldering problem or a solder ball rolling around the PCB. But that sort of problem should affect more than one voice at a time so it's straight back to a failed ASIC or corrupted memory.

I'd guess you've run through exactly the same chain of logic, but sometimes it's good to bounce it off someone else and a relief when they come to the same conclusion. At least your A6 boots, that'll brighten any day! Mine's still awaiting a renewed attempt.

I did have a bad ribbon, the one spanning the three front panel boards, so it really is worth testing them just to be confident they're not at fault.

Regards and the very best of luck with your A6.
Old 13th May 2018
  #321
Gear Maniac
 

Sorry folks, one more thing.

U31 also changes from a TS912 to a TL082 when it goes from having a single supply rail to a split supply rail. So it's not just a matter of adding a few diodes and swapping the PSU rails, which isn't trivial to start with.

If I was working on the Analog Master and/or ASIC boards this is the mod I would do. It may clean up various tuning, calibration, and pot jitter problems. Alesis produced a whole new PCB (Version D) for a good reason.

I'd also give U25 (the LM2940CT +5V regulator) a heatsink and make sure C231, C232, C238, C241, C413 and C85 were perfect.
Old 13th May 2018
  #322
Gear Nut
Hi No.7, thanks for a very detailed answer yet again I'll need a little bit of time to parse all that information

I've done the Slow SRAM check and luckily it passed. Always cool to see the 'spaceship' mode with all the LEDs on.

I've just opened up my Andromeda again (left all the screws out, so that was easy). Upon close inspection of the Fast SRAM Lower chip with a strong magnifier, I think there's a short with the adjacent pin that I soldered the resistor bridge to. That should most likely explain the tuning not being remembered.
I remember checking that thoroughly after I soldered it, don't know how I didn't notice that.

I'm going to fix that and see if it helps. As for the ribbon cables, I think I'll make a replacement cable for the one going from the analog board to the main board to see if that will fix the problem with the effects.

I'll check the analog board for bad contacts and will check for solder debris as well. Thanks for all the useful tips and the support, it's a real comfort when working on an expensive and relatively rare machine like this.

I hope you will be able to fix yours soon, you've put an amazing amount of work into it already. Once you get it working it should be in a fantastic state.
Old 13th May 2018
  #323
Gear Maniac
 

Ha ha, if my spaceship had a control panel like the A6 I'd ask ESA/NASA whether I could swap it out for something with pedals and handlebars. The moon shot craft had the brains of a calculator and they got back okay.
Old 13th May 2018
  #324
Gear Nut
Yeah pedals and handlebars would probably be handy lol.

Well I looked at the soldering again, it wasn't a real short, but there were a lot of solder specks between the pins. Cleaned that up using a toothbrush, but nothing changed.

The original battery has three contacts, the battery holder has two. I remember checking if the remaining contact was already connected to the other and it was, but maybe the tuning memory problem does lie there. It does remember my global settings and user presets.
Also it does seem to be out of tune when rebooting.

Also removed the ribbon cable between the analog and main board and placed it back to see if anything would change, but it didn't.
Found my post on this forum that mentioned the FX problem with voices 3 and 4, so I remembered that correctly. Thanks internet for being my backup memory Really strange how it changed to voice 6.
Old 13th May 2018
  #325
Gear Maniac
 

I'm not so sure about the pull-up on the SRAM chip select pin. Alesis released a main board version that had component pads for various pull-ups but apparently didn't 'stuff' those components. If the pull-up was critical I reckon they'd fit it.

I guess they included the pads just in case they were required, but found that they didn't cure whatever problem they were targeting, so they didn't 'stuff' them on production boards. I'm not sure I'd risk damaging the chip with too much soldering just to fit that pull-up.

50 and 60 way ribbons aren't easy to make. It takes a lot of force to push all the tines through the insulation of the cable. I have a dedicated bench press for ribbon making and I still have to be careful. If it proves difficult give me a shout and, depending on where you are, I may be able to make one up and send it to you.
Old 19th May 2018
  #326
Gear Nut
Quote:
Originally Posted by No.7 View Post
I'm not so sure about the pull-up on the SRAM chip select pin. Alesis released a main board version that had component pads for various pull-ups but apparently didn't 'stuff' those components. If the pull-up was critical I reckon they'd fit it.

I guess they included the pads just in case they were required, but found that they didn't cure whatever problem they were targeting, so they didn't 'stuff' them on production boards. I'm not sure I'd risk damaging the chip with too much soldering just to fit that pull-up.

50 and 60 way ribbons aren't easy to make. It takes a lot of force to push all the tines through the insulation of the cable. I have a dedicated bench press for ribbon making and I still have to be careful. If it proves difficult give me a shout and, depending on where you are, I may be able to make one up and send it to you.
I didn't know these cables were so hard to make. Thanks for your your very kind offer. I live in the Netherlands. I'm currently working on the boards again, to see if there is any other way I can solve the problems. If it does seem the ribbon cables are the cause it would be great if you could help out.

I just remembered I re-capped the whole analog board as well, as it was using the same capacitors as the one that went bad on the analog master board. That might explain the problem moving from voices 3 and 4 to voice 6. Interestingly the chip for voice 5 and 6 has two capacitors instead of one. I don't see anything wrong there though.

I did find some very small solder specks near contacts and removed them, who knows this might fix something.

My pull-up resistor was touching the via next to the capacitor it was soldered to. I assume this via is connected to that same contact, but I can't see it well enough.

The contacts that are related to the battery all measure 3 volts, so nothing seems to be wrong there. This makes me think the problem with the tuning not being saved has something to do with the pull-up resistor and not the battery holder.
Old 19th May 2018
  #327
Gear Maniac
 

You're welcome. The Netherlands are 'local' after all.

I'm sure you could make a jig using a G-clamp and a few pieces of steel, but it may be more effort than is warranted for a single ribbon cable. I'd be inclined to test your existing cables carefully. You'd have to test any new cable, so checking the old ones may save you a lot of time and effort.

I'd also test the caps you removed, just to be sure you have cured a problem. I had some caps that were clearly 'domed', so they got replaced, but I tested them so as to know just how bad they'd failed. It's nice to know that you really did replace defective components rather than just swapping out a whole pile of caps.

Yes, the pull-up isn't easy to solder properly. I used an 1/8w resistor and a microfine tip on the iron and still had to be very careful. With all those tiny pins so close together ANY solder debris is a problem.
Attached Thumbnails
Troubleshooting an Alesis A6 Andromeda - No Boot-alesis-a6-pull-up-1a.jpg  
Old 19th May 2018
  #328
Gear Nut
That is very cleanly done. My local electronics store only had 1/4 Watt resistors unfortunately.

Well, I removed the resistor because the one of the leads came off the SRAM pin and I wanted to test if it did anything. Now the synth doesn't boot at all anymore. Not even a single flicker of the logo. Turning it on-off-on also doesn't make it work anymore. Not a good sign. Working on this machine is terrifying.

So I'm going to solder it back on. Not looking forward to that Hopefully that will make it work again.

As for the capacitors, I replaced all capacitors of the same value that was bulging at the main output. So I'm sure most of them hadn't gone bad yet, but I didn't trust them anymore.
Old 19th May 2018
  #329
Gear Maniac
 

It turned out neat but I bent and discarded a few resistors before I had one with the perfect shape!

Lots of people, including me, would love to know whether that pull-up makes a significant difference. So right now YOU are the cutting edge, the vanguard, of the worldwide Andromeda A6 repair effort.
Old 19th May 2018
  #330
Gear Nut
Wow, I feel honored

Welp, I soldered it back in, not touching the via anymore. Still doesn't start properly, but now I'm back at turning it on-off-on to get it to start. So the new display definitely didn't fix the startup problem. Guess I need to check again if I really soldered the resistor back correctly. Not going to be doing any more soldering tonight though.

Tuning still gets lost after a restart, this is such a mystery. FX on voice six are still acting up.

EDIT: And now it started up properly in one go again. I'll have to keep an eye on it, but it seems the resistor might indeed be a cure for hanging at boot time.
I wonder if leaving one of the battery soldering points empty somehow prevents power from going to one of the layers of the board. It is a multi-layer board right? Also I saw a faint circular stain around the two capacitors near the battery. Would that mean they have leaked? They are not bulging.

Last edited by LamerDeluxe; 19th May 2018 at 09:17 PM..
Topic:
Post Reply

Welcome to the Gearslutz Pro Audio Community!

Registration benefits include:
  • The ability to reply to and create new discussions
  • Access to members-only giveaways & competitions
  • Interact with VIP industry experts in our guest Q&As
  • Access to members-only sub forum discussions
  • Access to members-only Chat Room
  • Get INSTANT ACCESS to the world's best private pro audio Classifieds for only USD $20/year
  • Promote your eBay auctions and Reverb.com listings for free
  • Remove this message!
You need an account to post a reply. Create a username and password below and an account will be created and your post entered.


 
 
Slide to join now Processing…
Thread Tools
Search this Thread
Search this Thread:

Advanced Search
Forum Jump
Forum Jump