Roland MKS-30 Blinking Lights Problem

[olafwagner]

Hi Folks, long time fan of the Syntaur Youtube channel, but the first time that I am posting here.

When I last used my Roland MKS-30 it worked perfectly, but it has been in storage for a while, and when I plugged it in today, it has developed a problem.

After turning on, the bank display briefly shows ‘08’ for a split second, but thereupon both the bank and patch displays show ‘88’ and 88’ but they appear to be flickering, and the ‘MIDI Message’ and ‘Dynamics’ lights are constantly blinking as well. None of the front panel buttons seem to have any effect on this. (I have uploaded a picture, but in the moment the camera snapped the image, the bank display is off, which further shows that the display is flashing).

I have checked the power rails, and the only anomaly that I could find is that the silkscreen on the power PCB refers to a ‘+7’, but I get 11.5v on those pins. Apart from that all voltages look good. The battery is very old (probably never replaced), so I don’t know whether that would cause this?

Any advice would be much appreciated.

[olafwagner]

I have checked the battery and it is still good (slightly over 3v).

Is it possible that this behavior could be caused by the infamous failing voice chips issue? (I would think that logically it wouldn’t, it would just make one or more voices sound weird and not that the entire synth would act up).

Are these MKS-30’s prone to failing capacitors?

[olafwagner]

OK, I have done some more diagnosing, and something is pulling down the voltages massively. This is the output from the 7805 regulator:

[refer to ‘01’ in the attached image]

(Note: As a ‘new’ user I can only upload a single image, so I had to jam all my images into a single image - sorry about that)

The measurement seems very rippled, and is wavering between 2 and 3v (scope is 1v/div and I adjusted it that 0v is at the bottom for simplicity) I also measured the input to the 7805 and that sits at just around 5v, which is way too low.

Then I disconnect the power connector to the main board (this connector contains just the +15v,-15v,7v and 5v rails (and GND):

[refer to ‘02’ in the attached image]

When I then power up the synth again and measure the rectified output, it is a super-clean 5v (I also measured the input voltage to the 7805 and that sits at around 11.8v):

[Refer to ‘03’ in the attached image]

Now I am trying to figure out whether something on the board is shorting a bit, and thus pulling down the voltage rail? Or does it mean that the caps on the power supply are bad?

Looking at the power supply schematic, both the 5v and 7v rails come off the same winding on the transformer and then the same bridge rectifier, just that the 7v rail goes through resistor R9 (that gets nice and hot) and the 5v rail is rectified by the 7805:

[Refer to ‘04’ in the attached image]

I will probably pull the caps on the power supply board and test them (specifically C5 which sits right beside the super-hot resistor R9 , but I am curious whether bad caps can cause this? Surely it is not a bad bridge rectifier or transformer right - then it would either work completely or not at all.

One the main PCB I also measured the resistance from each power rail pin to ground (to see if there are any obvious shorts on any rail) but they were all either in the K-Ohm or M-Ohm range respectively, I just don’t know whether that is right or not.

[olafwagner]

OK, I checked a few capacitors on the power supply, and they were all good (measured both capacitance and ESR after removing them and they were good as new, which was surprising).

I then started disconnecting power rails (I removed one pin at a time from the molex-style connector), and have discovered that my problem is squarely on the 5v rail. In other words some component on this rail must be shorting to ground and pulling the entire rail down.

The problem now is how does one go about finding the culprit. It could be an IC, but there are MANY IC’s on this mainboard (and all are soldered except the ROM).

There aren’t any tantalum caps, from what I can tell the bulk of them are polystyrene - are those known to go bad and short?

Lastly, is it possible that the epoxy on the voice chips (yep, mine are still encased, haven’t had any problems in the past yet) would short enough to pull 5v line down? Surely not?

Any guidance would be much appreciated.

[olafwagner]

OK, after more searching, I definitely have some kind of power problem.

• With the 5v rail (on the main PCB) disconnected, the 5v rail measures at 4.95v (measured at the output of the 7805 regulator - the input measures around 11v)
• When I reconnect the 5v rail (on the main PCB), the 5v rail drops to 2.5v (measured at the output of the 7805 regulator - the input has now dropped to around 5.5v!)
• I then connected my bench power supply set to 5v @ 600mA to the main pcb, and with everything switched on, the MKS-30 works perfectly. (the schematic for the power supply (included below) shows 800mA for the the 5v rail, so for the synth to work at 600mA makes me believe there is no short on the main PCB, which only leaves the power supply…

So it almost seems like the power supply cannot provide enough current on the 5v rail to power the synth. I have taken the following steps:

• I removed and tested the large capacitor C5 (6800uF, 25v) and it measures good for both capacitance and ESR
• I have replaced the 7805 regulator
• I have replaced capacitor C7 (100uF, 16v) even though it also measured good
• I removed and measured C8, and while I get around 50nF (instead of the expected 100nF), I guess it is still OK.

That leaves me with only 3 thoughts:

• Is it possible that while capacitor C5 measures good out of circuit, it is shorting a bit under load?
• Could there be a problem with the bridge rectifier D2 (both D2 and D1 get damn hot, but I think that is expected)
• Could there be a problem with the transformer?

I am at a loss because I am used to voltage rails either working completely or not at all. Hopefully someone on this forum has more power supply experience and will have come across something like this before.

[synzoneau]

I would replace the lot in the 5v part of the PSU invluding the bridge rectifier you can get the 2W04G or 2W01G. I have seen bridge rectifiers fail under load before, recently in a juno 106. It sometimes would take an hour for it to happen but it would blow the -15v fuse causing the audio to disappear, I’d replaced everything else but until I changed the bridge rectifier it kept happenig. Look for burn on the board near the legs.

[olafwagner]

Thank you very much. I ordered some 2W01G rectifiers as well as the large capacitors from DigiKey. I replaced the rectifier on the 5v rail, and also replaced the 6800uF cap. When I first fired it up, it didn’t work.

I was suspicious of the 7805 voltage regulator. (I had previously replaced the 7805 regulator with a L7805CV - thinking that this was the cause and that it would be the same spec.) Well, I guess it isn’t because when I swapped back to the previously removed 7805 regulator, everything now works.

Thereupon I also replaced the bridge rectifier on the 15v rail, as well as the 2 large 4800uF caps for the +15v and -15v rails just for good measure as the rectifiers seem to work hard, and the caps are right beside the really hot rectifiers.

At first I thought that everything was perfect, but when I crank up the volume I can hear a slight note ‘lingering’ (which after turning on the diagnostic mode shows me its chips 3 and 6). So now onto the next steps, desolder the voice chips and bathe them in acetone for 48 hours to remove that dreaded epoxy…

Thanks again for all your help, this really helped me.