Bring Roland SH2000 back to life? [SOLVED]

[zorgitron]

Hi everyone,
I used to be very active on this site when I was a high school student, and now I've let myself get obsessed with synths again!

So I have an SH2000 that went silent. I'm doing all I can to figure it out and fix it myself, but I'm just wondering if very experienced and knowledge gentlemen have some pointers to help me out.

When I got it, it worked, but was very crackly and unstable. Sometimes presets would work and sometimes they wouldn't. Then it went silent, but it was an easy fix, the output jack was just shattered. Easily fixed it.

As I played it more, it seemed to even out, and almost all the presets worked all the time. The only malfunctioning bit was the portamento. Maybe 1 out of 10 times it somewhat worked, but it really didn't behave as it should have. It just would drone out a note without moving or do something so strange I can't really describe.

Then when I turned it on one day, nothing that used the VCO worked. I could hear the filter and noise only by the "wind" preset.

Soon after that, while trying to fix it, the "Wind" preset also went silent.

Now, the only sound I can get is the sound of the keys' voltage tappy taps. I guess that means the keys are still sending their messages...

I've got the 40 page manual, and I've got an oscilloscope, so I've been probing around. Every one of the probing points that should be a waveform are just straight voltages, such as Pin 3 of the 555 IC.

(I'm still new to using an oscilloscope, so I often wonder if I'm using it correctly. I could see clear waveforms while working on other projects though.)

Please, if you've read this and got any ideas, share with me!
I'm willing to take my time and do whatever it takes to get it working again. It's really the best sounding instrument I've owned.

[Yekuku]

Hi and welcome to the board.
Can you please confirm that you are getting +15V and -15V on the power supply ?
If the power supply is ok, it would be a good idea to wiggle the filter and VCA boards, and check if it produces any sound. You should clean the connectors both on the cards and the mainboard, if there is oxidisation on the connectors , it is possible that the problem lies there.
Let us know how it goes, good luck !

[zorgitron]

Hi Yekuku,
Thanks for the reply. I checked the power supply again, and it was 15v +- 0.5v all around.

I appreciate that you know about the 2 boards that are stuck to the big matrix board in this synth. Anyway, I've removed them and checked the connections, and nothing looks very oxidized (just smells moldy). I've unplugged and checked every connector on this synth. The connector trick is what brought my Korg Lambda back, so that's one thing I try to check.

I've removed and checked that big main matrix board, and a lot of the soldering worked looked bad so I re-flowed some of the joints, but that didn't solve anything.

One thing caught my attention during some research: apparently Tantalum capacitors go back/short, and they could destroy other components near them.

There are 2 of those tantalum capacitors on the VCO board at least. I keep seeing on ARP repair websites like synthchaser.com that you should replace Tantalums without hesitation... it might have already killed the 555.

Also, I removed and checked the Polystyrene capacitor serving the 555 on the VCO, and it tested fine...

[Yekuku]

Hi Zorgitron,
can you please confirm that the negative voltage line -15V is also present ?
If there is no sound coming from the synth, this could be culprit.
Other than that, the SH2000 has 2 sound sources, the oscillator and the noise generator.
Check VCO connectors pin 1, are you getting a pulse waveform ?
Check VCO connectors pin 10, are you getting a key on / off pulse waveform ?
Check VCF connectors pin 1, are you getting white noise ?
If you are getting all these and no sound from output, then it should be the VCA.
Clean the trimmers on the vca board, and check again.
Let us know how it goes.

Edit: concerning the tantalums, yes they are the usual suspects but when they fail , they short that means that at least one voltage rail should have shorted to ground. You can replace them with electrolytics, if you are still in doubt.

[zorgitron]

Hi again Yekuku, thanks for the pointers.

I probed again on the VCO and VCF pins and didn't get any waveforms anywhere, just straight DC voltage. I've probed probably every point on the schematic that shows some waveform and just get a straight line every time. Take a look at the pictures:

VCO Pin1


VCO PIN1 non-blurry closeup, such low voltage... almost nothing at all!


VCF Pin1, no noise just a straight line across the scope at 5.5v.


Here's VCO pin 10, as far as I can tell, the keypress voltage seems to work as it should.


The power supply output 14V + and -
I guess that's a little low.

[madtheory]

First off, read again carefully what Yekuku said. I think your PSU is definitely faulty as you don't have -15V. Recap it and check it again ideally with a scope so you can see what the ripple is like.

I don't think there are any tantalums in this?

Second, get the service manual. Is that the manual you meant?

After you've done the PSU you probably need to check other stuff it might have damaged. More good diagnostic info here:
http://mezzoauto.blogspot.com/2019/05/r ... -2000.html

I have bits of one if you need anything, ask

[Yekuku]

Hmm, thats strange.
Ok, check if the vco board is working:
- VCO Board check connector pin 2 , it should be 15V
- VCO Board check connector pin 4, it should be -15V
- VCO Board check 555 Ic pin 8 , it should be 15V
- VCO Board check Q105 collector pin for pulse wave

[madtheory]

Hmm, thats strange.
Ok, check if the vco board is working:
- VCO Board check connector pin 2 , it should be 15V
- VCO Board check connector pin 4, it should be -15V
- VCO Board check 555 Ic pin 8 , it should be 15V
- VCO Board check Q105 collector pin for pulse wave

As you'll see from the blog, the PSU causes all sorts of strange problems if it's not on spec.

[zorgitron]

First off, read again carefully what Yekuku said. I think your PSU is definitely faulty as you don't have -15V. Recap it and check it again ideally with a scope so you can see what the ripple is like.

I don't think there are any tantalums in this?

Second, get the service manual. Is that the manual you meant?

After you've done the PSU you probably need to check other stuff it might have damaged. More good diagnostic info here:
http://mezzoauto.blogspot.com/2019/05/r ... -2000.html

I have bits of one if you need anything, ask

Thanks for your kind insistence that I properly check the PSU. I checked the manual more closely to see if I was probing it correctly, and it definitely looks like I found something out of order!

The PSU looks okay visually... but...


Okay here is the positive voltage:


And now this this should be negative, but it's totally off!!! It should be -15v on these pins.


Yes, I am consulting the full 40 page manual. It's the best scan I could find, but it's still quite hard to read.

I've been reading the mezzoauto blogspot since I've bought this synth. That Sh2000 post had lots of really interesting analysis. It's kind of hard to find info on this synth, it's really overshadowed by the other Rolands in its family.


So I shall be recapping this as soon as I can and I'll let you know what the result is. Thanks again!

[Yekuku]

Disconnect the psu from the mainboard and check the voltage again, if the voltage is right , a shorted component might be dragging the negative voltage line to ground.
If the voltage is still down check the bridge rectifier ( green diode boxes), the caps and the regulators.
But I have an idea, the power switch is a dual pole / throw, my bet is that one contact is dirty thus not powering the negative voltage line.
CHECK AND CLEAN THE SWITCH !

[madtheory]

Disconnect the psu from the mainboard and check the voltage again

Yessssss!!!!!!!!

[zorgitron]

Disconnect the psu from the mainboard and check the voltage again, if the voltage is right , a shorted component might be dragging the negative voltage line to ground.
If the voltage is still down check the bridge rectifier ( green diode boxes), the caps and the regulators.
But I have an idea, the power switch is a dual pole / throw, my bet is that one contact is dirty thus not powering the negative voltage line.
CHECK AND CLEAN THE SWITCH !

Ok I've checked the PSU voltage again after unplugging everything, and it's totally the same voltage readings.
Interesting idea about the switch.
The switch looks like kind of a pain to remove though. I'll have to be careful and take my time.

Lots of new recapping parts on their way already...

[Yekuku]

To test the switch, unplug the synth from the mains, engage the switch and use a multimeter's continuity to check the connectors pins
(1) & (5) should be 0 ohms
(3) & (7) should be 0 ohms

Spray a bit of WD 40 inside the switch to clean it.

Attachment updated.

[zorgitron]

To test the switch, unplug the synth from the mains, engage the switch and use a multimeter's continuity to check the connectors pins
(1) & (5) should be 0 ohms
(3) & (7) should be 0 ohms

Spray a bit of WD 40 inside the switch to clean it.

Attachment updated.
SH2000_PSU_3.jpg

Ok I tested the switch pins for continuity and got beeps for both poles. No ohms showed up on the multimeter. The solder joints look quite crusty, but I've seen worse. I think I'll re-flow them.

I believe I know some equivalent to WD40 in Japan. WD-40 is lithium based right?

[Yekuku]

I am not sure if it is lithium based, any contact cleaner with lubricant will do , no need for WD40.
If there is continuity , the switch works, no need for losing your time with this.
OK, now that you have ruled out the switch, you should check the transformer.
Based on the previous attachment , take VAC readings on the 2 secondaries, you should read near 20V on both windings.
CAUTION: playing around with transformers could be lethal !

Edit : have you checked the fuse F602 ? if it is good , remove it and check the holder for oxidisation, then reinstall. Then remove D603 and D604 and check them, they are double diodes , the orientation is printed on them, they are easy to test. Also check diode D606 for short.