Having an odd issue with my SQ-80 where the power supply is popping fuses 1 and 2, but only sometimes. Going off an old schematic I found, these fuses are for the +/-12V lines for the audio, disk drive, and VBB for the display via some complex circuitry.
The digital side of the keyboard seems to work perfectly. 5V checks out fine, and the front panel and display are both responsive, though the display is dim when the fuses go.
If I remove all connections to the power supply the fuses are fine. They’re also fine if I connect the display and disk connections.
If I connect the main board connector the fuses will pop, but if I power the +/-12V lines from on the connector with my bench PSU there’s nothing out of the ordinary with their current draws. Presumably that means the issue lies with the VBB portion of the circuit. I also replaced the two regulators a couple years ago.
I’ve attached a schematic of the power supply from a previous post on this forum, though it’s a pretty rough scan. If anyone has any suggestions or things I might test, please feel free to mention them as I’m not entirely sure of my next step. I’m also happy to answer any questions I can.
(ignore the circled resistors, they were from a previous discussion)
Its the electrolytic capacitors. They are past time expired, and are drawing too much current. That’s why the fuses are randomly blowing. the power supply is probably extremely hot too.
If there are tantalum capacitors they are dead shorts when they fail, and many times they give no outward appearance if they do.
Start with the power supply, change those caps with quality ones, not mysterious Chinese brands. Then do one board at a time.
Oh, I should have mentioned that the caps were replaced with Nichicons at the same time that I did the voltage regulators a couple years ago. There are also no tantalums anywhere in the synth.
Right now I’m waiting on a pack of fuses so I can continue testing it, but in the meantime I’m working on cleaning up the PSU schematic above so it’s actually readable.
Since the fuses are blowing only when the main board is connected then I would suspect your problem to be on the main board. Do the fuses pop immediately after powering on when the main board is connected?
Since the board powers fine off your bench PSU I suspect it could be an aged electrolytic cap on the mainboard. A capacitor naturally will pull a large amount of current when first exposed to a voltage as it charges. If the capacitor is aged then the initial current pull could be high enough and a long enough duration to blow the fuse in the synth’s power supply but your bench supply can handle it.
If the synth runs fine for a while then the fuses blow it could be a leaky capacitor on the mainboard (current leakage, not necessarily physically leaking any electrolyte). Have you measured the voltages of the power supply with the mainboard connected to see if anything is being dragged down? There could be many other things causing this but these are my first thoughts.
Also here are some better scans of the schematics on the web SQ-80 PSU
remember the 5v rail on the Sq80 is a kludge design choice. They have attached a resistor on it to get more current out of the regulator and its always under stress.
Replace with Traco TSR 1-2450E
The whole PSU section with heatsink will run cooler and be under less stress. You keep the audio power rails on linear.
Wow thanks, these are fantastically helpful! I’m familiar with Buchty’s site but I guess I never delved deep enough to discover these schematics, just the technical documentation pdf he made. I wish I had known about the PSU schematic before spending so much time redrawing it!
As for my SQ-80, the replacement fuses arrived and after sequentially connecting each board and power testing it, none of them have blown. Which is nice since now I have a booting SQ-80 and not nice in that I still don’t know what was causing that issue in the first place. Annoyingly a fault-free PSU is kind of hard to test for faults.
Sadly since my last post it seems to have developed a new, exciting fault, and it’s probably due to my testing. None of the voltage-controlled parameters (filter cutoff, resonance, panning, and ENV4 response) are responding to their program controls. According to the datasheet for the SSM2300 ICs they should not have signals applied to their inputs unless they’ve first received power. Thanks to these schematics I can see that they’re powered by +7V/-5V supplies taken from the +/-12V rails, which I was powering up after the CPU during testing, so that one’s on me.
Given this information I suspect this either means that all four SSM2300 ICs have failed, or the AD7524 CV DAC has failed (it and its associated TL081 do not seem, to be generating voltages either). I have a replacement DAC on the way for testing (since it’s about 1/5th the cost of an SSM2300), and I’ll be creating a test circuit for the SSMs to see if they respond properly outside of the synth.
I remember your advice from my previous thread on the SQ-80 power supply woes.
I also remember that the schematics you posted for it were wrong (they were for the ESQ-1), the resistors you advised to remove were wrong (they disabled the Vbb supply rail), and your “fix” didn’t even refer to the right 5V rail in the first place.
The LM7805 on the SQ-80’s power supply is for the display board, and there is no “kludge” resistor connected to it to “get more current out of it”. It’s output pin runs straight to the +5VDIS rail and off to the display board, with a couple capacitors connected in parallel (not uncommon for the 7805). The main 5V digital rail is regulated by an LM2925. Had I known more about electronics at the time I might have raised these points sooner since resistors limit current, not raise it.
Now granted, the Traco TSR 1-2450E did end up working in place of that LM7805 but my issues then and now were never related to the display in the first place. At best any issues with the display were a symptom of something else.
As such, I can’t recommend people follow your advice.
there are multiple versions of the ESQ and SQ80 power supplies. The schematics on the internet are the best available. I could not have known precisely what PSU board was in your possession. Indeed, you knew this and you cannot therefore put the blame on me.
this what you said in the other thread:
‘We_Robot’s advice seems sound but the resistor values he listed are different from the schematic and what’s on my SQ-80 PSU so I was waiting for his response before moving forward…’
'Well, I’ve discovered we_robot’s mistake.
He’s using a PSU schematic for the ESQ1, not the SQ80.
My fault for not checking first. Do your research, folks.
But you seem to be doing well on your own so go on my friend… You will find nothing from me on this anymore. I have fixed plenty of ESQ1s and SQ80s from PSU, display, logic, A/d and i’m offering to share my insights freely.
The way you are going you will end up with something to prop open the door but maybe along the way you will learn something.
Alright, fixed my doorstop. Turns out I was right about the DAC. Popped a replacement in and almost everything went back to normal save for one unresponsive SSM2300. I’ll be taking the opportunity to test the Analog Devices SMP18 as a replacement. Still an eye-watering $30 per chip but at least it wasn’t manufactured in the 80s.
Sadly still no joy on what was burning out the fuses. I’ll have to wait until the next time it acts up to begin the troubleshooting process anew.
Alright, it’s been a couple months and I have a couple updates on the SQ80:
The Analog Devices SMP18 would be a perfect drop-in replacement for the SSM2300 if it weren’t an SOIC package. It needs an SOIC to DIP adapter which is inconvenient but other than that the chip functions perfectly in the SQ80.
The power issues are a little more ambiguous. I found that the insulation tape on the voltage regulators had failed and two of the heatsinks were shorted. After I replaced the insulation it stopped popping fuses but it does still suffer from the occasional soft reset. But at least for the time being it’s functioning more or less as it should.
