Maker Pro
Maker Pro

Sparkle Power switching power supply

jhh

Jan 28, 2010
12
Joined
Jan 28, 2010
Messages
12
HI I have a FSP200-61GT SPI switching power supply with a burned out Mosfet on the high voltage side. There is (I think) a 2 watt resistor on both sides of the mosfet. But they test good, my problem is the Mosfet has a 1/8 watt resistor 0n the left leg and I think it had one on the middle leg but it is melted (that's left and middle looking at the mosfet from the front). I need to know the value of the middle resistor also information on how to get a schematic for this supply. Thanks Jim
 
Last edited:

Mitchekj

Jan 24, 2010
288
Joined
Jan 24, 2010
Messages
288
Getting a schematic would probably be painful, at best. Most likely impossible. Never hurts to ask the manufacturer though.

My best guess is that it's the current sense resistor that's metled. These are usually very low resistance, high power guys. (Something like 0.1 - 1.0 Ohms at 2-3W, depends on the design.) It would probably be connected between the source pin on the FET and ground. This is a 200W supply? Probably going to be a push-pull type architecture, but I can't say anything for sure unless I saw the thing.

Getting a datasheet for the FET (there would be more than one if it's indeed a push-pull) may help.

Keep in mind though, the FET and resistor melting are probably a side effect, not the root cause. Something caused a current spike through the primary of the transformer, but I can't really begin to speculate what caused it without actually troubleshooting the thing.

You've probably got a few more components which are bad, as well...snubber circuit, bridge, fuse, choke, etc.... all would be suspect.
 

jhh

Jan 28, 2010
12
Joined
Jan 28, 2010
Messages
12
RE Sparkle Power switching power supply

I'm sorry I didn't explain very well. The left and middle legs of the mosfet have been cut off and very small resistors( 1/8 watt) have been soldered to the shortened legs of the mosfet to replace the legs.Hard to explain ha ha If you unsolder the mosset and look at it, left leg is a resistor middle leg is a resistor and right leg is normal mosfet leg.(no kidding) I read the left resistor and got a 230K OHM reading (digital meter) The big resistors (2 watt ) I was talking about on either side I think are for soft start??? I know the mosfet is (sss 4n60A) left resistor is around 230K OHM but the middle one is melted to bad to get a value. Yes the fuse was shot but the other resistors ( the 2 big 2 watt ones) and diodes test good ( out of circuit tested) .The power was off and on a couple times that day maybe a power surge? Thanks for the help Jim
 
Last edited:

Mitchekj

Jan 24, 2010
288
Joined
Jan 24, 2010
Messages
288
230k Ohm on the gate sounds feasible. If these are the actual switching FETs, then the second resistor must have been some kind of current sense resistor. There's really no other reason to put any type of resistance in series w/ the FET, the idea is to get the least amount of resistance possible from ground to the transformer winding.

Looking at the specs for that guy, I'm not sure that's what the FET in question is being used for. Usually you'd see something rated 20+ Amps, especially on a 200W supply. If possible, can you post pictures? Whole system, both sides, close ups of areas in question, etc.?

Power spikes, brownouts and the like very well could have caused failure if the supply wasn't designed to keep the effects to a minimum. (Good TVS diodes, MOVs, etc.) The fact they've soldered resistors to the FET leads, makes me want to say the design wasn't all that top-notch to begin with.
 

jhh

Jan 28, 2010
12
Joined
Jan 28, 2010
Messages
12
re power

https://www.electronicspoint.com/members/jhh-albums-power-supply.html Don't know if this will work but this should be the pictures. The FET and fuse is laying on the meter and as you can see left and middle legs was where the 1/8 w resistors were. The big (I think 2 w resistors) pink left and blue to the right of where the FET goes read good. The pink one shows signs of being hot but reads good. I don't need this power supply but I'm trying to learn about electronics but maybe old people should just watch TV ha ha that will be the day. ha ha Thanks Again Jim P.S There is a short jumper running from the left leg of the FET over to between the resistor and the glass diode that just in front of the FET????
 
Last edited:

Mitchekj

Jan 24, 2010
288
Joined
Jan 24, 2010
Messages
288
Aaaah, I see what you're saying. I had it pictured differently.

I'm all for people trying to learn electronics... I'm learning every day. Anyone can do it with the proper time and effort.
Unfortunately, the switch mode power supply is one of the more complex architectures in the power world. It may be trickier to learn from it than with, say, a linear supply.

Not sure if this is a flyback or push-pull SMPS. Normally you'd see flyback in <150W supplies... but anything's possible. I can't tell if the TO-220's in the back are the switching FETs or maybe even secondary side rectifiers.

I can't say for sure what the FET in question is doing, based on his location on the PCB. If you can get a good picture of the traces on the back (the entire PCB,) and a top-down view of the whole primary side (with all the components on it) maybe I can make a guess, and we can figure out your mystery resistor. :) Who knows, someone else on this board may be able to say "Oh, that FET is the ______, and it's doing ______." But it's not standing out to me. Don't give up hope. hehe.

The black char mark on the PCB doesn't look promising either. Usually not a good sign when things are producing that kind of heat. Don't know if that was a one-off when the supply blew up, or if it's an "over time" deal. What looks like a diode, just below the resistor, has also started to char the PCB.

I would definitely expect some cascading failures with this one. But it may be as simple as a shorted switching FET and a diode or two. :)
 

(*steve*)

¡sǝpodᴉʇuɐ ǝɥʇ ɹɐǝɥd
Moderator
Jan 21, 2010
25,510
Joined
Jan 21, 2010
Messages
25,510
I'd be worried that if there's that much charring of the board that you may be getting conduction paths where you don't want them.

It's possible that the charring occurred originally because of either poor design, or excessive load and when the charring got bad enough it changed the circuit characteristics on such a way as to destroy it. Charring like that won't happen in the short time between catastrophic failure and the fuse blowing.

Is there a part number on that TO-220 device? <-- OK, I found it.

Considering the mosfet is switching rectified mains (probably) the 600V 2.5A@100C rating sounds about right.

I can't imagine that 230k resistor is the gate resistor. The specs suggest you'd be likely to find something like 15 ohms. Did you get that value by measurement, or from reading the bands?

I'd be tempted to start looking for a replacement PSU -- assuming I knew the specs of the current one.

If you do try to repair it, the first time you try it (smoke test) I would advise wiring a standard light bulb in series with the mains. That will act as a very effective current limit which *may* prevent huge amounts of damage if there's still a fault. I'd start with a 15W globe, and increase it if the power supply seems unwilling to start. Place a light load on the PSU (for example, 100 ohm resistors).

Switchmode power supplies are a bit like thoroughbred racehorses. They're fast, light, powerful, sexy, and in some cases a sure bet. But if they fall over while they're running, they are often taken out the back and shot.
 
Last edited:

Mitchekj

Jan 24, 2010
288
Joined
Jan 24, 2010
Messages
288
Edit: Oops, your post disappeared. Sorry if some of the below doesn't seem to fit in with the thread. lol

Ok, best I can tell, that FET is likely to be part of an active PFC circuit. It looks like the traces fit in with that theory as well. (Active PFC in an SMPS would probably be a boost converter between the bridge and the bulk storage caps.) It looks like the bridge has traces running over to the area, and it also makes sense that a power outtage/brownout would kill your active PFC.

That FET being rated at a relatively low current sounds about right. The other FETs, the E13007-2, look like the main switching FETs. 8A sounds a bit better for that application. All those FETs should be rated 400-600V.

Mostly speculation of course, a schematic would be ideal! :)

As for the resistors not matching the color codes (and yes, Grn Blk Blk would be 50) they could be partial opens from the heat/failure.

Once you start replacing things and want to turn it on, as Steve said for smoke testing, the bulb in series will be a good idea if you don't have a variac handy. Have to agree, much easier to put this thing down and just replace it. It would be really satisfying to get the thing running again though. :)
 
Last edited:

jhh

Jan 28, 2010
12
Joined
Jan 28, 2010
Messages
12
re power

I didn't think the pictures I did today come up and deleted the post. Well I went to youtube last night and watched a video on the correct way to repair a switching power supply, so the guy turns on the power and it starts to smoke so he kills the power and said aha theres the problem now I'll show you how to fit this thing ( I thought wow now I'll learn something) so he's looking for his tools and finally here he is with a big hammer and proceeds to beat the thing the pieces. That's one option I guess haha thought I'd die laughing for being so dumb haha But than maybe I did learn how to repair a switching power supply the correct way??? hahaha:eek: just kidding and thanks a lot for your help, I'll set it back for a while maybe I'll find a schematic or something. Thanks a Lot Jim:) I have to work on getting these pictures on the post as I'm having trouble getting the last ones to post.
 
Last edited:

(*steve*)

¡sǝpodᴉʇuɐ ǝɥʇ ɹɐǝɥd
Moderator
Jan 21, 2010
25,510
Joined
Jan 21, 2010
Messages
25,510
Edit: the bulb in series will be a good idea if you don't have a variac handy.

Variac is not a good idea for a switched mode power supply. They are essentially constant power devices. As you reduce the voltage they draw more current until either something blows up, or you reach some other physical limit.

You could even damage your variac.

I was operating a computer many years ago, and I noticed the fluorescent lights had gone out and the incandescent lights were a dim red glow. I measures the mains voltage (nominally 240V) and it was around 70V. The power supply was clearly still producing the same power output.
 

Mitchekj

Jan 24, 2010
288
Joined
Jan 24, 2010
Messages
288
Ooo, good catch! I've been working with dimmable supplies the past couple years, it's getting ingrained. The variac is a great tool for those. (LED drivers, 2-wire dimmable.)

No idea how an ATX would react to that, other than not well. Lol, thanks for pointing that out, Steve.

Edit: Jhh, Those links you've posted aren't working. I saw the photos by clicking your user name and looking at the photo albums. You can try to attach them directly to your message when you post, and they'd show up just under it. If they're too large, you can always .ZIP them and attach that file.
 
Last edited:
Top