Main SMPS Transformer Overheated
A client sent me his keyboard’s power supply via postage to see if I could repair it. It is from a very old Roland Keyboard.
The power supply was made by KARTIS in Italy and bears the model nr SWM30. It has output voltages of +5V, -12V and +17V. Scanning over the board, I could see some discoloration of the pcb and also a blown fuse. Further investigation with a DMM showed that the switching Mosfet was shorted. Removed it to confirm. Part no BUZ90.
I decided to check all the remaining components but couldn’t find anything else faulty. I then decided to replace the Mosfet and see if I could power it up successfully. Normally I would replace the controller IC too, but decided to try my luck. Brought the unit up slowly on a variac and dim bulb tester in place of the fuse. Bulbs started to light up brightly.
So, this obviously meant that there still was still a fault somewhere. Replacing the controller IC at this point, made no difference. I quickly drew up a schematic for the primary section, as there were no schematics available for this particular power supply. This made thing slightly easier.
I then decided to pull out my trusty RING TESTER, Guess what. Primary of transformer was not ringing any LED’s. This means that we either have a faulty transformer or there is faulty secondary component somewhere. I then desoldered the transformer and confirmed that it was indeed faulty.
I also used the Ring tester to test this. Then the issue became to find another transformer. I decided to contact the local agents for Roland. They informed me that this unit was discontinued but that they would see if they perhaps had a scrap transformer in their inventory. A week later the transformer arrived in the mail.
I could see that it was a used transformer, but luckily when ringed it gave me 5 LED’s. This meant the transformer was good!
Comparing it to the old transformer, I could clearly see that the original transformer had overheated.
I replaced the transformer and also decided to upgrade the controller IC from a TDA4605-2 to a TDA4605-3. This has a higher current handling capability. When powered up again, the unit was cycling slightly. Loading the +17V output rail with a 100ohm 3W resistor, stabilized it.
Confirmed voltages on all the other output rails good.
I will package the power supply carefully and send back to customer. I LOVE MY RING TESTER!!
This article was prepared for you by Riaan Diedericks. He runs his own electronics repair shop in Pretoria, South Africa. He specializes in Pro Audio repairs. Please give a support by clicking on the social buttons below. Your feedback on the post is welcome. Please leave it in the comments.
If you would like to get a Blue Ring Tester and Blue ESR meter you can click on the photo below:
P.S- If you enjoyed reading this, click here to subscribe to my blog (free subscription). That way, you’ll never miss a post. You can also forward this website link to your friends and colleagues-thanks!
You can also check his previous repair article below:
https://jestineyong.com/must-read-italian-made-powered-speaker-improvements/
Justice
May 21, 2019 at 9:36 pm
Great article Riaan. I am excited to hear someone from my home country, great job .
John Shadbolt
May 21, 2019 at 9:54 pm
Great repair, lucky that Roland had the matching transformer. Did you check the output capacitors? It might be the camera, but the one right next to a heatsink looks bulged, a replacement would probably be smaller so less likely to get dried out.
Ghashaan
May 21, 2019 at 10:04 pm
Good job!
I learned something important from your article. If the output voltage is fluctuating, I should try with a load in order to stabilize the output. But I still wonder why this trick is required for some power supplies and not for all!
I enjoyed reading your article!
abdellah kamli
May 22, 2019 at 5:33 am
Hi Ghashaan,your clever comment lets pointing a sensitive question.In attempt to bring a response I would like to share this modest explanation: when the output is fluctuating this means it oscillates at some frequency, by loading the output with the resistor it cancels the oscillations. The output capacitors, it seems, act with the resistor 100 Ohm as a low pass filter RC in which the equivalent capacitor has a high impedance at low frequencies so that the major energy is directed to the mass and the current flows through the resistor allowing a stable output voltage.
CHEERS
Ghashaan
May 22, 2019 at 7:39 pm
That's a sweet explanation.
Albert van Bemmelen
May 21, 2019 at 10:38 pm
Interesting repair with useful schematic.
Since the transformer was internally shorted, replacing the controller IC at first obviously made no difference.
Since the BUZ90 mosfet already had gone, most likely the original TDA4605-2 controller IC had gone too. Why it would be best to replace that controller IC to a new TDA4605-3 controller IC after you had replaced the bad transformer. This way taking no chances in blowing up the mosfet and the transformer again.
Parasuraman Subramanian
May 21, 2019 at 11:04 pm
Very Good article packed with very useful hints and ways. Getting a replacement for an SMPS Tx is not even heard of in our area. So, I am amazed to see the kind of support you all get there!
Yogesh Panchal
May 24, 2019 at 10:29 pm
Schematic is very helpful to understand working process path.
Good Job!
Thanks for sharing.
Rhey
June 4, 2019 at 3:30 pm
Excellent
Sean McKinnon
December 17, 2022 at 12:07 pm
Hi Riaan,
I am currently trying to fix my friends keyboard with the same failed power supply - but I'm stuck. I did learn from the TDA4605 manual that there is a no-load "safety function" that limits the pulse duration on pin 5 to "render the circuit absolutely safe under no-load conditions"... which could explain why you needed the 100 ohm resistor without the keyboard. Anyways, I've checked almost everything on the board, and I'm thinking I need to replace the TDA4605 chip and was wondering where you were able to source one? FYI, I see the output voltages rising and falling kinda like you described, but even with the PS plugged into the keyboard.
One thing I was hoping you could explain is the function of pin 2, the manual states the following: "A voltage proportional to the drain current of the switching transistor is generated there by the external RC-combination in conjunction with the primary current transducer. The output of this transducer is controlled by the logic and referenced to the internal stable voltage V2B. If the voltage V2 exceeds the output voltage of the regulations amplifier, the logic is reset by the stop comparator and consequently the output of pin 5 is switched to low potential. Further inputs for the logic stage are the output for the start impulse generator with the stable reference potential VST and the supply voltage motor"
Shouldn't the voltage be inversely proportional to the drain current??
Thanks for your time and effort!!
Sean