Dead power unit (blowing fuses)

[LektroiD]

I have a dead synthesizer power supply which keeps blowing fuses. I replaced the 0.68r resistor as it had also smoked, this has not rectified the problem.

The service manual oddly doesn't have the schematic for the PSU, only the silkscreen overlay...

http://www.synfo.nl/servicemanuals/Korg/PROPHECY_SERVICE_MANUAL.pdf

I'll add a couple of photos from the dead PSU, in case I've missed anything obvious.

[oldway]

Probably a shorted Mosfet (perhaps a TOP series ?)
Check also the bridge rectifier and if there is no shortcircuit in the +12V and +5V outputs.

[LektroiD]

Checked for continuity across the primary and it is short. So I removed the transformer and the 100nF/275V cap which is in-line with it. The cap has an ESR of 16r and the transformer beeps short on my meter at both the primary and secondary. Not having tested transformers before, I also checked out the other transformer on the board and that is also short at the primary and secondary. Either a continuity test is not how to test a transformer, or both transformers are dead?

I have also tested continuity on the power outputs as suggested, it seems most of that side of the board is grounded. I have attached a photo of the ground short (the darkened areas of the photo)

[AmmoJammo]

As mentioned, mosfet is 99% most likely shorted.

Reading what you've tested so far, I'd be inclined to think that you probably shouldn't be playing with mains powered devices 

[Vito_R]

Test the Mosfet (most likely failed)and ALL diodes, bridge rectifier, and other transistors for shorts or leaky conditions. Pay particular attention to that small transistor next to the mosfet in the heatsink.  You need to remove the Mosfet in order to test it properly and trigger the gate on by flipping one of the probes on one side and then back to the other.  I would also check for any other open resistors in that circuit.  Better to be sure you do not miss any other components that will just cause a cascade failure and burn out all your new replacements.  Whatever you do, do not replace the fuse again until you are sure there are no more faults.  Rule of thumb for next time, before you replace the fuse check the board for any other faulty components.  Good luck

[Vito_R]

Checked for continuity across the primary and it is short. So I removed the transformer and the 100nF/275V cap which is in-line with it. The cap has an ESR of 16r and the transformer beeps short on my meter at both the primary and secondary. Not having tested transformers before, I also checked out the other transformer on the board and that is also short at the primary and secondary. Either a continuity test is not how to test a transformer, or both transformers are dead?

To test the transformer properly you would need a ring tester.  I wouldn't be concerned with it at this point as your issue is most probably caused by what was discussed above.

[Riotpack]

+1 Test the MOSFET or TOP switcher (can you identify the part number for us?) depending on what that power device is. It is taking out the R68 current sense resistor so alot of current is passing through the TX, MOSFET/IC and resistor.

[sean0118]

...the transformer beeps short on my meter at both the primary and secondary. Not having tested transformers before, I also checked out the other transformer on the board and that is also short at the primary and secondary. Either a continuity test is not how to test a transformer, or both transformers are dead?

Keep in mind that both the primary and secondary are literally coils of wire. If you are checking the continuity of the first primary terminal to the second primary terminal you would expect it to be a short. But the primary should not be shorted to the secondary side coil because they are only linked by the magnetic core.

edit: and are you sure the problem is the power supply and not a fault with the synthesizer?

[miguelvp]

No wonder that area is all ground, seems you have a bridge between GND and the pin next to it in the upper right corner of the picture
Also the clearance on the last two pins is iffy.

Edit, the center part is 0V so all that area you marked is probably all ground. but the pin that looks that is shorted is supposed to be 15V.
but the silkscreen in that service manual doesn't really align well with the actual circuit. Actually the alignment is horrible and it's offset low and to the left by quite a bit.

And thirdly, what is that pin in the middle doing by itself just floating in there? (nevermind that's just the transformer probably not used)

[tom66]

This being a self-oscillating power supply (no controller IC) be sure to find and locate any output shorts.
If the output is shorted, in many cases it can damage the primary side too.

[TheBay]

OP,

You have been given advice on checking the Mosfet/switchers and other silicon devices, that would be a very quick check with a multimeter, why are you checking the transformer. I have never seen a short from primary to secondary windings on a transformer like that!

Check what has been said, also check the bridge rectifier, nice to see they actually used one not 4 diodes lol.

[LektroiD]

I tested the MOSFET both in and out of the board with my DCA55 and it certainly is short on all pins. I also checked the shorts as mentioned above, but the board seems OK, it was possibly some photographic aberrations making it look worse than it is. Despite that, I scraped between all the tracks in that area but to no avail. My guess is that because they are all tied to various taps from the secondary side of the transformer, it has created a short that way, and is normal.

I also checked out the diodes, which all seem ok according to the DCA55.

I've ordered a new MOSFET, so when that arrives I'll swap it out and hope for the best.

[Andy Watson]

I also checked the shorts as mentioned above, but the board seems OK, it was possibly some photographic aberrations making it look worse than it is. Despite that, I scraped between all the tracks in that area but to no avail.

I think whoever suggested that was pulling your leg - s/he was referring to the fact that the registration of the board overlay given in the manual is not 100% accurate.

I've ordered a new MOSFET, so when that arrives I'll swap it out and hope for the best.

You could be lucky but more likely you're on a fast track to empty pockets if all you do is swap-out components and hope for the best. What test equipment do you have? Do you have the facility to run the board with some form of current limit?

[Vito_R]

I tested the MOSFET both in and out of the board with my DCA55 and it certainly is short on all pins. I also checked the shorts as mentioned above, but the board seems OK, it was possibly some photographic aberrations making it look worse than it is. Despite that, I scraped between all the tracks in that area but to no avail. My guess is that because they are all tied to various taps from the secondary side of the transformer, it has created a short that way, and is normal.

I also checked out the diodes, which all seem ok according to the DCA55.

I've ordered a new MOSFET, so when that arrives I'll swap it out and hope for the best.

Ok good job but make sure you test the other transistors, especially the little one next to the mosfet inside the heat sink.  Use your DCA55 and/or the diode function on your multimeter to check all of them.  If you miss any components you run the risk of burning out your replacements.  Don't forget to also test the resistors to see if they are open, especially the low value ones.  Desolder one leg of each to be sure.   Did you test your bridge rectifier?  Use your diode test function on your multimeter to look for proper forward voltage drop and no reverse leakage.  Good luck

[LektroiD]

The diodes all tested ok, but all the transistors & rectifier were out...

Both 2SD734 are dead.
The 2SK1459 is dead
The HA17431P is showing as a "Digital Transistor" on the DCA55, but showing dead on other testers.
The bridge rectifier is dead, but I can't find a datasheet from it's marking; "J5.j".

If I swap out all these parts it will be fairly expensive (not to mention the 0.68r resistor that I already replaced), so I just hope there isn't something else causing them all to blow...

[N2IXK]

Nothing really expensive ON that board.  The FET might be a few dollars, but certainly less than $10 The bipolars are cheap NPN transistors,  Less than $1 each from most distributors.

The HA17431P is an integrated circuit (shunt regulator). You aren't going to get a meaningful test with an ohmmeter or transistor tester.

The bridge rectifier is fairly generic. Probably good for an amp or 2. Any 2A unit that has the same package/pinout and is good for 400PIV or more should work.

[Vito_R]

The diodes all tested ok, but all the transistors & rectifier were out...

Both 2SD734 are dead.
The 2SK1459 is dead
The HA17431P is showing as a "Digital Transistor" on the DCA55, but showing dead on other testers.
The bridge rectifier is dead, but I can't find a datasheet from it's marking; "J5.j".

If I swap out all these parts it will be fairly expensive (not to mention the 0.68r resistor that I already replaced), so I just hope there isn't something else causing them all to blow...

Good job in finding those transistors, I suspected they would be dead especially the one near the mosfet.  The HA17431 appears to be a voltage reference or what we call a shunt regulator IC and the DCA will not be able to test it, only the DCA PRO version probably can.  Just get the datasheet and pin configuration. You can try testing it with a current limited power supply by injecting voltage to the input within the specified input range and reading the output.  If any doubt just don't take the chance and swap it.  Any way you can get a schematic for this so you can determine the part numbers for the bridge?  If you can't then you might have no choice but to estimate it and just use a higher voltage and amperage rated bridge then what you think you have there.  Try to find more info on it. Good luck and keep up the good work. 

This is some data on the HA17431


Description
The HA17431 series is temperature-compensated variable shunt regulators. The main application of these products is in
voltage regulators that provide a variable output voltage. The on-chip high-precision reference voltage source can
provide ±1% accuracy in the V versions, which have a VKA max of 16 volts. The HA17431VLP, which is provided in
the MPAK-5V package, is designed for use in switching mode power supplies. It provides a built-in photocoupler
bypass resistor for the PS pin, and an error amplifier can be easily constructed on the supply side.

datasheet download

[Vito_R]

I would also make sure you check for any other open or out of spec resistors just to be on the safe side.  That's just me though.    I don't like surprises after I power things up.

[LektroiD]

Does the bridge rectifier need to be DIP, or can I replace with an upright, like this...

[wraper]

HA17431P is very unlikely to be dead as it is on secondary side. Even if it is, it is completely safe try to power PSU without changing it. One suggestion, when you will power it first time, find ~100w incandescent light bulb and connect it in series with PSU when connecting to mains. If something is not ok, it will light up and save parts you changed from blowing one more time.

[LektroiD]

HA17431P is very unlikely to be dead as it is on secondary side. Even if it is, it is completely safe try to power PSU without changing it. One suggestion, when you will power it first time, find ~100w incandescent light bulb and connect it in series with PSU when connecting to mains. If something is not ok, it will light up and save parts you changed from blowing one more time.

Basically use the bulb as a fuse? Couldn't I simply use a lower amperage fuse instead? That's also in series on the input. Just that I have an abundance of fuses in stock, and no incandescent bulbs...

What's your thoughts on using an upright rectifier in place of the DIP (as pictured above)? I'll insulate the AC pins with heat shrink to be safe. That one is 200V, 2.5A...

[wraper]

HA17431P is very unlikely to be dead as it is on secondary side. Even if it is, it is completely safe try to power PSU without changing it. One suggestion, when you will power it first time, find ~100w incandescent light bulb and connect it in series with PSU when connecting to mains. If something is not ok, it will light up and save parts you changed from blowing one more time.

Basically use the bulb as a fuse? Couldn't I simply use a lower amperage fuse instead? That's also in series on the input. Just that I have an abundance of fuses in stock, and no incandescent bulbs...

What's your thoughts on using an upright rectifier in place of the DIP (as pictured above)? I'll insulate the AC pins with heat shrink to be safe. That one is 200V, 2.5A...

Not as a fuse but as a current limiter. Lower value fuse likely will blow on switching on because of the inrush current. And even if don't blow, likely won't protect the parts.

[N2IXK]

Different rectifier package should be fine, as long as it is installed correctly.

[mjrpainless]

I actually built a current limiter the other day for about 25 bux, and thought to my self... why didn't I build one sooner!

Heres a short vid:

[LektroiD]

I'm back on this...

I built the current limiter, it works great, thanks for the heads up on that, excellent piece of safety test gear.

The Power supply is still not working. 

I get around 150V on the primary side of the 2nd transformer, and zero out of either +5V or +15V. I have only tested topside components because the supply is safest powered inside the housing.

I powered the machine up from my bench PSU, bypassing the dead supply, and everything is fine, so I'm looking for a cheap power board I can swap with the dead one, as it looks to me like the repairing it more hassle than it's worth. So, where's the best place to get someting like that? I did a search on ebay for "+15V +5V" but it's not really bringing much up. I wonder if there's any old products such as old DVD players, etc that would use the same power config. If I can get one of those cheap, I'd be happy to rip an old DVD player apart and salvage the PSU, or whatever products may use this +5V/+15V... Any ideas?