Roland TR-626 noise problem

[mckenzie]

Hello
My 626 has a constant annoying noise from the output & phones. When it's powered from a switch mode psu there is a high pitch whining, associated with the smps I presume. When I use a linear psu there is a low pitch hum instead. With batteries there's no noise. Turning the vol pot up/down turns the noise up/down. There is a large piece of aluminium that shields the main board from the switch board & it's grounded well.

I replaced the electrolytic caps in the power supply section but no good. Does anyone have any clues what to look for? Schematic is on page 11 attached.
Thanks all

 

[reirei]

Try a PSU with really low rating - like 100mA or less!
Somehow the 626 (and 505) are really picky on this...

 

[ElSmurf]

Looks like the batteries are connected to the switched PSU input, so there's a good chance the PSU is the problem, like @reirei said.

 

[NoisyIndividual]

Did you replace C74, across the Zener D24?
I think I would replace both these, since any lack of regulation here would be very apparent with an external PSU but non existent with batteries.

And maybe even Q9 as well if you have something at hand to try.

 

[Audio1Man]

This is a power supply problem. SMPS (Whine) need a min load to stop any burst/pulsing mode of operation, and may also need additional filter cap.

The transformer supply is 50/60/100 or 120Hz (Hum) needs extra filtering (Larger Cap) to reduce the ripple on the dcv.

Duke

 

[Chandra]

These units were produced in the late 80s and may not be suitable for SMPS PSU.
SMPS were not popular those days for small devices.
Try with a regular transformer type PSU and if it works, just go with it.

Regards.

 

[mckenzie]

Did you replace C74, across the Zener D24?
I think I would replace both these, since any lack of regulation here would be very apparent with an external PSU but non existent with batteries.

And maybe even Q9 as well if you have something at hand to try.

Thanks. Yes I replaced C74. I'll try changing the zener as well then & see if I can find a similar spec substitute for Q9 also.

This is a power supply problem. SMPS (Whine) need a min load to stop any burst/pulsing mode of operation, and may also need additional filter cap.

The transformer supply is 50/60/100 or 120Hz (Hum) needs extra filtering (Larger Cap) to reduce the ripple on the dcv.

Duke

Thanks. Yes I found another thread where someone had the same issue & it was resolved by using the recommended (BOSS - PSA-120 | Power Adapter) Boss PSA120 psu. However like I said, I've also tried it with a similar linear psu (Korg KA186) which has an almost identical spec to the Boss & I get a low pitch hum instead. It looks like the filtering in the circuit is lacking a somewhat from what I've read, but id still expect the linear psu to perform better than it is.
As you suggest I'll try adding a larger cap, was thinking 1000uf would be a good start ?
Would also be good to try to filter the SMPS noise if possible, what would be a good capacitor value for this?

 

[NoisyIndividual]

The hum may possibly be due to excessive current draw somewhere.
The batteries won't hum, they will just run out quickly without you noticing.
Maybe you could measure the DC current draw on batteries, and/or with a linear supply, and see whether this is an issue or not.

Or see if anything is getting warm, the Zener for example.

 

[mckenzie]

Thanks for the help.
Current draw is ok at around 30mA. I tried a 1000 & a 680uf across the input from the ext power supply but it actually made the noise worse with both type power supplies.
I found another power supply I had lying around, just another old random linear type & it seems to have cut the noise quite a bit but still had a slight hum. I added a 220uf parallel with C76 & the noise has pretty much gone

 

[Audio1Man]

Don't modify the circuit. By add the cap to C76 this will not cure the problem. Use a male & female cable, split the cable and add a new big cap to it. Connect the PS 2 wires to the CAP and connect the 2 wires going to 626 to the cap leads below the 2 wires from the PS.

We need the CHARGE CURRENT to go directly to the CAP TERMINALS and the DISCHARGE CURRENTS to come from the CAP TERMINALS. This keeps the common charge and discharge currents very low.
Duke

 

[mckenzie]

Don't modify the circuit. By add the cap to C76 this will not cure the problem. Use a male & female cable, split the cable and add a new big cap to it. Connect the PS 2 wires to the CAP and connect the 2 wires going to 626 to the cap leads below the 2 wires from the PS.

We need the CHARGE CURRENT to go directly to the CAP TERMINALS and the DISCHARGE CURRENTS to come from the CAP TERMINALS. This keeps the common charge and discharge currents very low.
Duke

You mean basically putting a large cap in parallel with the psu +/- ? I tried that & it made the noise worse. Increasing the value of C76 seems to have quietened things down pretty well & its been working ok for a few hours at a time. Is there a potential problem with the way I've done it ? Thanks

 

[Hubbub]

I note that Q11 feeds +9Va to a slew of op-amps, it could be that C64 is not doing its job as a noise/hum attenuator. I would replace C64 with a good low-ESR capacitor. And why the C63 area is fed from 9V instead of +9Va makes no sense. I would have connected the top of R181 to +9Va to get lower noise/hum into the output amplifiers bias node. Leaving R181 as-is and possibly replacing C63 with a good low-ESR capacitor may help. C31 (and maybe C32, but it is less critical) might also be a candidate for replacement, it filters the bias point for a bunch of sections.

 

[NoisyIndividual]

30mA current draw really shouldn't be producing much in the way of problem ripple. Maybe the problem is in the grounding.
I notice the power input ground first goes to JG (presumably Jack Ground) which is indeed where the outputs get their ground. Everything else uses a different ground via the inductors.
Try running a wire from the jack output socket to various other grounds, or direct to the DC socket, there may be some current running through a poor grounding layout causing noise enerywhere.

 

[mckenzie]

I note that Q11 feeds +9Va to a slew of op-amps, it could be that C64 is not doing its job as a noise/hum attenuator. I would replace C64 with a good low-ESR capacitor. And why the C63 area is fed from 9V instead of +9Va makes no sense. I would have connected the top of R181 to +9Va to get lower noise/hum into the output amplifiers bias node. Leaving R181 as-is and possibly replacing C63 with a good low-ESR capacitor may help. C31 (and maybe C32, but it is less critical) might also be a candidate for replacement, it filters the bias point for a bunch of sections.

I replaced these but no joy unfortunately but thanks for the suggestions

30mA current draw really shouldn't be producing much in the way of problem ripple. Maybe the problem is in the grounding.
I notice the power input ground first goes to JG (presumably Jack Ground) which is indeed where the outputs get their ground. Everything else uses a different ground via the inductors.
Try running a wire from the jack output socket to various other grounds, or direct to the DC socket, there may be some current running through a poor grounding layout causing noise enerywhere.

tried with different ground points but no joy am afraid.

Thanks for all the help with this. Im pretty happy with how it sounds now with the extra 220uf in parallel with C76 - is there any potential problem with this addition ?

 

[Philbo King]

How'd you find the schematic? Roland service docs are notoriously hard to get hold of.

After a drunk spilled beer in my GR-30 guitar synth, I hunted for a schematic quite a while with no luck. Luckily I found a fried trace under the microscope and was able to fix it with a length of 36 gauge wire, staked down with CA glue (TacPac) every 1/2".