Korg SMPS dead

Carlosraj

Member
2010-01-17 11:12 am
Hi


I am working on a SMPS (Mean Well PT-65-R11) from a Korg PA800 keyboard. The power supply is dead. I checked all the component and they are OK. I changed the PWM IC (UC3842) and tried but still no power. On closer look on the underside, I realised one smd resistor (R8) which looks dark and the 3 digit marking on it gone. Taking the resistance reading with my DMM, it reads approx 800 ohms, goes down to about 400 and then goes up again. I suspect this smd is blown or partially blown. It connects Pin 6(output) from UC3842(PWM) to the gate of 2SK3562(switching mosfet). I could not get the schematic of this SMPS. Can anyone guess what could be the value of this resistor.


Thanks
 

Carlosraj

Member
2010-01-17 11:12 am
OK, I put in the 10 ohms resistor, replaced UC3842, replaced shunt regulator AZ431 with TL431 and powered the SMPS without any load. Now I get +14.8VDC, -14.8VDC and +5VDC, which are almost the exact voltages for this power supply. I then plugged the power supply into the keyboard and powered it. There was no display, no power indicator but slight motoboating sound from the keyboard speakers. Measuring the voltages with the keyboard plugged in, the +15V and -15V have dropped down and fluctuating between 9V to 6V. When unplugged from keyboard, they measure normal voltages again. Can it be due to some short circuit within the keyboard or a problem with the SMPS itself. Please assist.
Thanks
 
When a switch-mode mains PSU fails lots of components can be destroyed due to the high voltages, replacing one at a time is likely to just burn the new components..

Identify all the failed components first, replace them all, much better chance of success, but usually these supplies are not economically repairable. Shotgunning all the active components is possibly worth trying (don't forget the opto couplers), if you can get part numbers for all of them, but you still have to test capacitors for shorts and for burnt-out resistors, and then there's the transformer windings to check - buy a new supply!

Perhaps test the keyboard with an old ATX supply (+/-12V should be enough for the 15V rails for basic functioning.)


Better to test the PSU into dummy loads than the keyboard, if it goes haywire you won't risk the keyboard.
 
Try loading the power supply with a pair of resistors that simulate the expected keyboard load. That might involve a bit of a guess but would prove the supply is up to the job.

Otherwise UC3842 is a current mode control IC with pin 3 being the sense. It will/should be connected to a current sense resistor in the source of the Mosfet with an RC network, C at the IC R to the sense resistor. See if you can trace/find it.

Having previously blown up it is possible that that resistor has been a bit toasty and had its value increased. If that is the case then being too high will send the supply into current limit earlier than expected.
 

PRR

Member
Paid Member
2003-06-12 7:04 pm
Maine USA
www.diyaudio.com
Jameco.com

Search Keyword: Mean Well PT-65
Here is +/-15V,5V model for $14, not worth fixing if you can buy it at that price.

Yours has more letters so you need to check connectors, physical layout, size, etc, and any current specs you can scrape up. KORG may have specified more performance, or a different mix of output currents, OR may have negotiated a down-price model of less performance or smaller size.
 
Sounds like the supply is basically working again but still suffers from bad caps, usually on the secondary side. The ESR meter may know more if there aren't any obviously bulging candidates. Maybe check mains power consumption if you are unsure about excessive current.

Assuming a decently equipped stash of caps, this may still be cheaper than getting a new ones.. besides, I wouldn't bet on anything first-rate being in there to begin with.
 

Carlosraj

Member
2010-01-17 11:12 am
As I wanted to check the regulation is working or not, I switched ON the power supply and checked the voltages. I have +15V, -15V and +5V. I then connected a 24V 24W bulb (the only one I had) across the +15V and turned ON the smps. I also connected a DMM across the bulb to monitor the voltage. The bulb was dead and the voltage has dropped to approx 2V and fluctuating. I realised that there is a small preset near the output connector. I slowly adjusted and then all of a sudden the bulb lite up and the voltage up to 14.8V. I adjusted it to exactly 15V. When I switch off and on, it takes about 2 secs and the voltage is up to 15V and remains steady. Next I plugged in the smps into the kaeyboard. Turn ON and nothing happens. The 15V has dropped to about 7V. I tried adjusting the preset, but the voltage does not seem to exceed 9V and it is fluctuating. Is the problem with the smps or the keyboard. Any suggestions.


Thanks
 

PRR

Member
Paid Member
2003-06-12 7:04 pm
Maine USA
www.diyaudio.com
.. besides, I wouldn't bet on anything first-rate being in there to begin with.

I have used and abused a lot of MeanWells. They always gave good service for many years. Often they out-lived what they came with and I re-cycled them into other gear, including a campus network switch.

I don't know if KORG might have negotiated a lesser level of reliability.

I wish we knew just how much the keyboard demand was, and had proper load resistors for that demand. This poke-and-see repair makes me nervous.
 

Carlosraj

Member
2010-01-17 11:12 am
After putting aside, today I turned on the keyboard again. At first there was a slight ticking sound from the keyboard speakers. After about 4 to 5 mins later, the ticking disappeared and the display screen started to flicker. It keeps pulsating on and off at approx 2 to 3 times a second. I left it on for more than an hour and its still flickering. There are something things I am doubtful of. The R8, the smd gate resistor between the output of UC3842 and the gate of 2SK3562 was blown and cannot identify the value. One of the forum member told me it is a 10 ohm resistor, I put that in but am not sure whether it is the correct value. There is another smd resistor that looks suspisious , R41 has only 1 digit left, the other 2 digits are missing, it reads 3.3K. I do not know it is the correct value or partially burnt. I have replaced UC3842 and TL431 and the other components check OK. The only other option I have is to replace all the electrolytic caps to 105*C . Please assist.

Thanks
 

pitbul

Member
2011-07-29 9:34 am
MorbidFractal sugested that You check current sense resistor, current sense resistor is the one on the source leg and connected to gnd (negative side of the bridge or diode rectifier).
That resistor should be low ohmic value, depends on current requirement of the SMPS, it is not easy to find correct value, because this resistor defines when current through the mosfet will be switched off.
Maybe it is for testing purpose easier to connect good power supply with current meter on the positive rail and determine how much current is needed for correct work of the equipment.
To me it looks like you don't have enough power to supply to the Korg.
After that you should decide which resistor to put. If you put too low value you will enter selfdistruct region of the SMPS, current will be too high and mosfet will blow, or be shut by over voltage sense (TL431).
After writing this above I concluded that maybe there is something wrong with resistors around TL431 and you have lower voltage margin which switch off the smps. Determine with some graphical instrument did you reach your wanted voltage.
 

Carlosraj

Member
2010-01-17 11:12 am
Hi


As a last attempt, today I decided to change all the electrolytic caps. Yes, the smps started working. I tried leaving it on for 5hrs with the sequencer running, and so far there was no problem. I guess it is solved. Thanks to all you guys for your advice and support.


Thanks