I usually have my psr sx700 routed through the ad inputs to a mixer leading to my monitors. Just set this back up and exactly the same fault exists - with max gain on the ad inputs and full volume from the psr, the distortion starts at about 2/10 on the master volume and increases from there. Reducing gain and or psr volume allows the mox volume to go a bit higher but distortion still occurs from half way at least. The fault very much seems linked to the master volume slider on the mox.

DavidF - 26 December 2020 11:41 AM

I usually have my psr sx700 routed through the ad inputs to a mixer leading to my monitors. Just set this back up and exactly the same fault exists - with max gain on the ad inputs and full volume from the psr, the distortion starts at about 2/10 on the master volume and increases from there. Reducing gain and or psr volume allows the mox volume to go a bit higher but distortion still occurs from half way at least.

OK, that’s what I expected.

 

DavidF -

The fault very much seems linked to the master volume slider on the mox.

What we’ve determined so far is that the analog stages on the JK board are unable to cleanly handle signal beyond a certain level. The Master Volume control setting determines that level, but it itself is innocent. Testing so far has eliminated certain possible causes. We’re narrowing the possibilities, but there will be a few more tests needed. Here are a couple more…

1) Measure DC voltage from pin 1 of CN003 (analog ground) to pin 12 of CN002. CN002 can be seen on the attachment I previously posted, and pin 12 is its last pin, nearest CN003. Take the measurement as soon as the MOX is powered on, and again when it’s fully booted up.

2) See the attached diagram. This section of the JK board is near the front edge, opposite JK002 (L/MONO jack). Place the negative meter probe on pin 1 of CN003 as before, and the positive probe on the lead of R002 as I’ve marked with a dot.

The first measurement is to verify that the muting transistor control voltage is correct for turning the muting off once the MOX boots. The second one is to check that the voltage determining biasing of the op-amps is right - if normal, it should be very close to half of the supply voltage, so about 4.4-4.5 volts.

As before, please let us know what you find.

Click thumbnail to see full-size image

OK for 1 was about 7 on boot up dropping to about
-2.5, but latter number fluctuated down to 0 quite a bit. Not sure if it’s the needle connections but tried several times.

For 2, I couldn’t get a reading other than 0 = it looks like a brown striped transistor or resistor, and I had to put the needle against the leg entered the pcb. I tried with 200m setting and got a reading of 00.9 or less most often 00.0 or 00.1

DavidF - 26 December 2020 05:48 PM

OK for 1 was about 7 on boot up dropping to about
-2.5, but latter number fluctuated down to 0 quite a bit. Not sure if it’s the needle connections but tried several times.

Not precisely what I expected, but it may be OK. For now, let’s ignore this because the next reading is more interesting.

 

DavidF -

For 2, I couldn’t get a reading other than 0 = it looks like a brown striped transistor or resistor, and I had to put the needle against the leg entered the pcb. I tried with 200m setting and got a reading of 00.9 or less most often 00.0 or 00.1

R002 is a resistor - according to the schematic and parts list, its value should be 10kohms (10,000 ohms) +/- 5%, and with the standard color code the bands would be brown-black-orange-gold.

Presuming you were making good contact with both CN003 pin 1 and the R002 lead I indicated, there’s a problem at that point in the circuit.

Before “we” (meaning mainly myself) get too excited, please check once more, making sure the positive probe is on the R002 lead I pointed out. The color bands should match those above.

The voltage at that point should definitely be around 4.4-4.5 volts. If not, the problem you’re experiencing would be explained. Once you verify the near-zero volts reading there, we can discuss the limited possible causes.

OK double checked and same outcome

Cn003 pin 1 to pin 4 8.92
Cn003 pin1 black to cn002 pin 12 - 7 on boot up dropping to 0 after a few seconds. Seems a bit unreliable getting consistent reading even when needles are fully in contact.

Cn003 pin 1 black to r002 positive (left hand side as I look at the mox from the keyboard side nearest - always zero, never give any sort of reading on 20 on the meter

DavidF - 27 December 2020 09:20 AM

Cn003 pin 1 black to r002 positive (left hand side as I look at the mox from the keyboard side nearest - always zero, never give any sort of reading on 20 on the meter

OK, that indicates a definite problem. Now we have to pin it down further. This will take a few more steps…

There’s another 10kohm resistor in this bias circuit, R001. It’s perpendicular to R002, and just a bit further inboard (towards the rear). The lead of it I’m going to refer to next is the one at the end furthest from R002…

1) Negative probe to CN003 pin 1, positive to the R001 lead as I just described above. There should be about 9 volts there. Let us know what you measure, and if not close to 9v, stop at this point.

Otherwise, some resistance checks will be next. These are done with NO power to the MOX - unplugging the AC adapter is a good idea to ensure that’s the case.

2) Set your meter to the 20kΩ resistance range. Positive probe to the R002 lead I pointed to in the previous attachment. Negative to the other end of R002. (You’re measuring across R002.) The reading will likely change with time - wait for it to settle to a fixed value, or at least changing slowly. Make note.

3) As in step 2, but swap the positive and negative probe positions.

4) Positive meter probe to R001 lead as I described for step 1. Negative to the other end of R001. (You’re measuring across R001.)

5) As in step 4, but swap the positive and negative probe positions.

=======================================================

Extra credit for those wondering what this is all about:

R001 and R002 form what is known as a “voltage divider”.
See https://learn.sparkfun.com/tutorials/voltage-dividers/all
Because the two resistors are the same value (10k), the division should result in one half the supply voltage appearing at the junction of the two resistors. However, instead of the expected approximately 4.5 volts (half of 9) being found, it’s been measured at close to zero. That’s an obvious fault condition.

I’ve attached the relevant part of the schematic so that the curious can see how R001 and R002 divide the 9 volts. Obviously there’s more to the bias circuit than those resistors, and I may explain further as the diagnosis proceeds.

Click thumbnail to see full-size image

5pinDIN - 27 December 2020 12:38 PM

DavidF - 27 December 2020 09:20 AM

Cn003 pin 1 black to r002 positive (left hand side as I look at the mox from the keyboard side nearest - always zero, never give any sort of reading on 20 on the meter

OK, that indicates a definite problem. Now we have to pin it down further. This will take a few more steps…

There’s another 10kohm resistor in this bias circuit, R001. It’s perpendicular to R002, and just a bit further inboard (towards the rear). The lead of it I’m going to refer to next is the one at the end furthest from R002…

1) Negative probe to CN003 pin 1, positive to the R001 lead as I just described above. There should be about 9 volts there. Let us know what you measure, and if not close to 9v, stop at this point.

About 8.92

Otherwise, some resistance checks will be next. These are done with NO power to the MOX - unplugging the AC adapter is a good idea to ensure that’s the case.

2) Set your meter to the 20kΩ resistance range. Positive probe to the R002 lead I pointed to in the previous attachment. Negative to the other end of R002. (You’re measuring across R002.) The reading will likely change with time - wait for it to settle to a fixed value, or at least changing slowly. Make note.

0.01 then 0.00

3) As in step 2, but swap the positive and negative probe positions.

0.02 then 0.00

4) Positive meter probe to R001 lead as I described for step 1. Negative to the other end of R001. (You’re measuring across R001.)

3.3

5) As in step 4, but swap the positive and negative probe positions.

3.3

=======================================================

Extra credit for those wondering what this is all about:

R001 and R002 form what is known as a “voltage divider”.
See https://learn.sparkfun.com/tutorials/voltage-dividers/all
Because the two resistors are the same value (10k), the division should result in one half the supply voltage appearing at the junction of the two resistors. However, instead of the expected approximately 4.5 volts (half of 9) being found, it’s been measured at close to zero. That’s an obvious fault condition.

I’ve attached the relevant part of the schematic so that the curious can see how R001 and R002 divide the 9 volts. Obviously there’s more to the bias circuit than those resistors, and I may explain further as the diagnosis proceeds.

OK. The previous zero-voltage reading along with the zero-resistance reading across R002 indicate a short to ground. R002 has two capacitors in parallel with it, C002 and C004 - they’re located next to R002. If either is shorted the result would match what you’ve measured. The only other possibility is a short in op-amp IC001. While the op-amp could be the culprit, in my experience one of those two capacitors having failed is more likely.

It’s now time to put your soldering skills to use. I’d suggest starting by removing C004, a 47µF/16v electrolytic. Once removed its resistance can be checked to see if it’s the cause of the short. However, sometimes heat will change things, so even if it doesn’t read near zero ohms, check resistance across R002 again to verify that the resistance has gone up.

If removing C004 doesn’t get rid of the short, remove capacitor C002, a 0.1µF/25v ceramic, next. As with C004, check its resistance out of circuit, and again look across R002.

If it’s shown that one of the caps is bad, you might be able to temporarily run the MOX without it. I said “might” because there’s a possibility that the circuit could become unstable without the capacitor. That’s not likely to cause any damage, but things might not sound “right”.

On the other hand, if removing one or both of those capacitors doesn’t bring the resistance reading across R002 up from zero, then IC001 becomes the guilty party.

My word I think you may have fixed it!

As instructed, I used my soldering iron for the first time ever and removed capacitor C004. Took a while for the solder to liquify on the underside of the pcb but when it did the capacitor pulled out quite easily, with no solder mess left to disturb other components/circuits. The capacitor resistance measured 0.16 with the meter set to 20k. I then measured the resistance of R002 which has gone up to about 6.0, and R001 measured the same value almost. I also checked the voltage to R002 and it now reads 4.45v. I switched the mox back on and the CRACKLING SOUND HAS GONE! It sounds perfect, like it always has before! I sat here in my office on the floor stunned with amazement and joy, how simply fabulous! I can’t thank you enough for your support and advice /and patience with this complete novice in electronics!

I have reconnected the jk board but left the bottom screws out as I assume there may be a need to replace the capacitor? Is this necessary as the mox sounds totally fine, and I don’t want to mess up what I have managed to achieve with your help so far! That said it looks fairly straight forward to solder in a replacement capacitor as a reverse of the process of removing it (famous last words!).

I’m simply overjoyed and await your advice and next steps.

Awesome..

DavidF - 27 December 2020 04:56 PM

My word I think you may have fixed it!

As instructed, I used my soldering iron for the first time ever and removed capacitor C004. Took a while for the solder to liquify on the underside of the pcb but when it did the capacitor pulled out quite easily, with no solder mess left to disturb other components/circuits. The capacitor resistance measured 0.16 with the meter set to 20k. I then measured the resistance of R002 which has gone up to about 6.0, and R001 measured the same value almost. I also checked the voltage to R002 and it now reads 4.45v. I switched the mox back on and the CRACKLING SOUND HAS GONE! It sounds perfect, like it always has before! I sat here in my office on the floor stunned with amazement and joy, how simply fabulous!

Well, that’s a really nice outcome! The smile on my own face likely parallels how you’re feeling.

 

DavidF -

I can’t thank you enough for your support and advice /and patience with this complete novice in electronics!

You’re quite welcome. I have to admit that I’m sometimes reluctant to get involved with “remote control” repairs, especially when the other person isn’t well versed in electronics. However, it was a pleasure working with you - even though you aren’t experienced, you’ve been willing to follow directions and provide useful feedback. Take it from me, that’s too often not the case.

 

DavidF -

I have reconnected the jk board but left the bottom screws out as I assume there may be a need to replace the capacitor? Is this necessary as the mox sounds totally fine, and I don’t want to mess up what I have managed to achieve with your help so far! That said it looks fairly straight forward to solder in a replacement capacitor as a reverse of the process of removing it (famous last words!).

I’m simply overjoyed and await your advice and next steps.

Awesome..

C004 acts as a filter. Without it the MOX output might be slightly noisier than otherwise. However, the 9 volt analog supply feeding the op-amp bias circuit is already pretty well filtered before it gets to where C004 is, so there might not be an easily discernible change with/without that capacitor.

If you decide that you want to replace C004, I can provide some guidance as to what to get, and some tips on technique. One pointer that’s absolutely important - electrolytic capacitors are polarized devices, and they and the circuits they’re used in don’t take kindly to putting them in backwards.

Let us know what you decide to do.

By the way, was the place you took the MOX to authorized by Yamaha? If so, Yamaha might want to know the level of “service” you received there. If they charged you for the “diagnosis”, perhaps they’d like to refund that fee to you. (Of course, given current health concerns, you might not want to visit anyone unless absolutely necessary.) In order to avoid any possible legal issues, it would probably be best to not mention the name of the service organization here on Motifator.

If you decide that you want to replace C004, I can provide some guidance as to what to get, and some tips on technique. One pointer that’s absolutely important - electrolytic capacitors are polarized devices, and they and the circuits they’re used in don’t take kindly to putting them in backwards.

Let us know what you decide to do.

Yes please, I would like to complete the task and bring it back to original condition. I had a quick look for capacitors and don’t understand the market so advice would be great.

By the way, was the place you took the MOX to authorized by Yamaha? If so, Yamaha might want to know the level of “service” you received there. If they charged you for the “diagnosis”, perhaps they’d like to refund that fee to you. (Of course, given current health concerns, you might not want to visit anyone unless absolutely necessary.) In order to avoid any possible legal issues, it would probably be best to not mention the name of the service organization here on Motifator.

Not sure if it is an authorised service centre. Chap said he had access to yamaha parts ordering service. Can’t understand how they diagnosed a DM board fault and clearly were only able to swap out whole components. The learning for me is to do my due diligence first - I did not research the market for synth repairs and this company focus on general band repairs like amps etc. They were within a short drive so I went in hopeful expectation rather than with a recommendation. A quick Google and other synth owners have had the same issues - gear declared un repairable. The charge to look at is it is annoying now but I did agree it up front. Would rather have had honest advice about what their capability was, and a recommendation for someone more specialised who could help. I felt a bit helpless to begin with so didn’t take the time to research fully. My first proper synth was a roland u20, had it for 25 years and virtually gave it away as the keys stopped working and was afraid of high repair costs on a low value instrument. Guy who bought it for a song fixed it easily so was determined not to do the same again. I love romplers and the mox6 remains a top versatile keyboard, whereas the u20 was very limited so more nostalgia than anything else.

I have thoroughly enjoyed tinkering with your advice, and to get it working again is most satisfying. So yes would like to finish the task and replace the faulty capacitor.

Thanks again for the help :-)

DavidF - 28 December 2020 07:38 AM

Yes please, I would like to complete the task and bring it back to original condition. I had a quick look for capacitors and don’t understand the market so advice would be great.

The original C004 is a 47µF 16volt electrolytic capacitor ("cap" from now on). Such caps are available in several variations, so it’s not surprising that what to get isn’t obvious. Here are some pointers…

Since 47µF (47 microFarads) is a standard value, it shouldn’t be difficult to match that. The voltage rating, 16 volts, is a maximum the cap should be used at. Since in-circuit the cap will see less than 5 volts across it, 16v is plenty. Caps are also typically rated for temperature - common specs are 85degC and 105degC. The MOX doesn’t tend to get particularly hot, but 105degC caps tend to be longer-lived, so I’d go with one of those.

So 47µF/16v 105degC would be the specs I’d suggest.  It should, of course be “radial” type - that is, with both leads exiting at one end. While there are several manufacturers of such caps, Panasonic is a brand I trust.

Something like this…
https://www.mouser.com/ProductDetail/Panasonic/EEU-FC1C470H/?qs=/C1U95aQ15vhemYBctOGEw==
...or similar should work. Note the size and lead spacing, and make sure that what you get is close enough to the original so it fits easily. The above link is to a US firm, but you should be able to find a somewhat local supplier for such a part.

As to mounting the new cap…
I suggest that the holes in the board be cleared before trying to insert the leads. That can be done by heating the original solder while pushing a needle/pin through from the top of the board. Once through, remove the iron and wiggle the pin so that the hole remains open as the solder cools and solidifies. (There are certainly more sophisticated methods, but this technique usually works.)

Electrolytic caps have their polarity indicated - typically it’s the negative lead. For C004, that would be the lead closest to the edge of the board.

Soldering…
Do you have solder? If so, tell us what it is. I’ll let you know if it’s appropriate, or make suggestions.

 

DavidF -

I have thoroughly enjoyed tinkering with your advice, and to get it working again is most satisfying. So yes would like to finish the task and replace the faulty capacitor.

Thanks again for the help :-)

You’re welcome, I’m glad to help.

Soldering…
Do you have solder? If so, tell us what it is. I’ll let you know if it’s appropriate, or make suggestions.

 
I bought the soldering iron many years ago, much like the multimeter, just to have in as I do like my tools…

The solder I have is savbit alloy multicore 5 cores ersin flux 13swg 1.22mm, says OK for electronics, must be 20 yrs old at least. Soldering iron is OK but tip is a bit large but seems OK for the purpose as there is enough space to avoid the other components.

DavidF - 28 December 2020 01:04 PM

The solder I have is savbit alloy multicore 5 cores ersin flux 13swg 1.22mm, says OK for electronics, must be 20 yrs old at least. Soldering iron is OK but tip is a bit large but seems OK for the purpose as there is enough space to avoid the other components.

That solder should be fine for the purpose, and age isn’t an issue. The MOX may have been assembled with lead-free solder, so the melting temperature may be slightly higher than that of your solder. Therefore, make sure the solder already on the board melts before adding additional solder, or the two might not want to readily combine.

BTW, not that it really matters, but 1.22mm is 18swg. :-)

Good luck, I hope the cap replacement goes well.

Well after 3 months of thinking about it, finally soldered in a replacement capacitor and all is well, mox is working perfectly 😁