OWON SPE 6103 Blowing Fuses Roasted Diode

[clearchris]

This was a great little power supply, until it let out the smoke.

I believe it's the output diode (pic of mayhem attached).

Could I remove the diode, power it up (no load) and see if the rest of the PS is functional? Right now it pops the fuse every time.

Also, does anyone have a diode that they like better than this one? (ER2004CT)
https://www.panjit.com.tw/upload/dat...0CT_SERIES.pdf

If anyone wants to know more about this power supply there's a pretty good teardown video here https://www.youtube.com/watch?v=bRWdQESwXMA

[R_J]

Looks like the insulator failed from heat and the tab/screw shorted, remove the screw and then check the dual diode, if it is shorted it needs replacing, if the dual diode checks ok, you could try the power supply without load and without the diode's tab touching the heatsink
I would still replace the diode and new insulator.

[petehall347]

looks cooked .maybe it was assembled badly

[clearchris]

Looks like the insulator failed from heat and the tab/screw shorted, remove the screw and then check the dual diode, if it is shorted it needs replacing, if the dual diode checks ok, you could try the power supply without load and without the diode's tab touching the heatsink
I would still replace the diode and new insulator.

I didn't realize the tab wasn't electrically connected to the heat sink, thank you for the tip.

I tested the diode in circuit, all legs have continuity to the heat sink and to each other.

Looks like new diode, new pad, new insulator, at minimum.

[clearchris]

looks cooked .maybe it was assembled badly

I think I definitely have to test the fan circuit also, people complain about how noisy this PS is, I haven't found that to be the case. Perhaps something is wrong with the fan, and the insulator melted... The solder doesn't look great at all, but I suspect this failure put out enough heat to remelt the solder, my house still smells of burning circuits.

[stj]

is the fan controlled by the thermister bolted above the diode?
because that would explain it

[CapLeaker]

With a little leg bending you get a 30 amp diode in a TO247 case in there. You almost can put anything in there as long as it has minimum the same ratings as the old diode. You can put anything higher voltage, higher amperage and even faster recovery diode in there. Some are full pack (fully isolated). This thing died because it got wayyy to hot.

[clearchris]

is the fan controlled by the thermister bolted above the diode?
because that would explain it

Yes, it is. Is that a bad location for the thermistor?

[clearchris]

With a little leg bending you get a 30 amp diode in a TO247 case in there. You almost can put anything in there as long as it has minimum the same ratings as the old diode. You can put anything higher voltage, higher amperage and even faster recovery diode in there. Some are full pack (fully isolated). This thing died because it got wayyy to hot.

Yes, I looked for better diodes, but I noticed I couldn't get anything in a TO220. I can bend some legs, I just can't seem to find anything with better specs at any price. I can get higher voltage, higher amperage, but it seems that everything is slower, or has no recovery time (and is different type of diode [silicon carbide] that I don't know enough to substitute).

Also, TO220 and TO247 don't take the same insulator bushings, do they?

[CapLeaker]

Yes, I looked for better diodes, but I noticed I couldn't get anything in a TO220. I can bend some legs, I just can't seem to find anything with better specs at any price. I can get higher voltage, higher amperage, but it seems that everything is slower, or has no recovery time (and is different type of diode [silicon carbide] that I don't know enough to substitute).

Also, TO220 and TO247 don't take the same insulator bushings, do they?

TO220 you get about 20A and that's it. TO247 you get 30A. Problem is getting rid of the heat. A full pack doesn't need insulators as it is in plastic, tab and all, but wants a bit better cooling. Either the fan failed, the PSU was overloaded, the vents blocked etc.

[clearchris]

Either the fan failed, the PSU was overloaded, the vents blocked etc.

Well, that's interesting now.

The vents weren't blocked, it's basically a brand new PSU that I bought in early 2024. No dust inside. Maybe it has been used for a few days total, mostly charging batteries.
The PSU shouldn't have been overloaded, I don't really recall ever using it above 2 amps or so, I charge all batteries super slow.

I was plugging in a battery to be charged, the unit was switched on, but no power applied to the power supply terminals from the power supply side. I suspect that there was some backflow from the lead acid battery, and it overwhelmed the PSU protections. I guess that assumes that what I was doing at the time was the cause of the failure and the failure had not been brewing for some time aleady, or a factory defect.

So the question is, will changing the diode help prevent this in the future (if so, should I look at voltage specs, amps,or other spec?), or was this potentially a bad part that gave up too soon? Or should I attempt to tweak the fan circuit to give more cooling at the expense of noise? Maybe both?

[stj]

the fan circiot probably wasnt the cause.
check the fan actually runs incase their is a fault and then replace the diode properly

what may have happened if the 2 diodes (outer pins) are linked, is one side went open-circuit and the other side had to deal with double the current - until it couldnt.
and the excess heat it generated was what sped-up the fan trying to cope with it.

[petehall347]

poor materias shoddy workmanship . the insulator has burned . it should stand more heat than it did looks like it bubbled out from behind it .there shouldn't be a gap there . i bet the screw was left loose or there was a hole in the insulation allowing an arc to form . .might have been bad soldering too . but still the insulator should not melt like that .

[petehall347]

all this debate etc when it should still be in warranty for a free exchange or repair .

[clearchris]

all this debate etc when it should still be in warranty for a free exchange or repair .

I'm talking to OWON now. First response was "replace the fuse."

Expecting second response to be "ship it back to china at a cost more than you paid for it so we can tell you it was user error and you have to pay for repair and for us to ship it back to you at a cost exceeding the original purchase price."

[redwire]

The story does not make sense-
If connecting a battery (what voltage?) to the PSU when it was off, blew it up due to backfeed, this rectifier likely got roasted by the battery current. Which would also flow through the transformer, if this is what happened.
Or was it due to the PSU putting out too much (charging) current?
It's odd the diode ER2004CT is a high voltage part 400PIV 20A because the lower voltage parts dissipate 35% less heat i.e. the ER2001CT is 0.95Vf verses 1.3Vf.

Never connect a battery to a bench PSU when it is powered off. Most will croak when this is done.

[CapLeaker]

Ahh… plugged the battery in before turning on the power supply. Don't do that.

[petehall347]

so no protection against that sort of thing like a relay ? i suppose they arent sold as battery chargers .

[clearchris]

The story does not make sense-
If connecting a battery (what voltage?) to the PSU when it was off, blew it up due to backfeed, this rectifier likely got roasted by the battery current. Which would also flow through the transformer, if this is what happened.
Or was it due to the PSU putting out too much (charging) current?
It's odd the diode ER2004CT is a high voltage part 400PIV 20A because the lower voltage parts dissipate 35% less heat i.e. the ER2001CT is 0.95Vf verses 1.3Vf.

Never connect a battery to a bench PSU when it is powered off. Most will croak when this is done.

Transformer looks fine thankfully.

So what's the process here? Connect battery when power supply is powered up, disconnect battery while power supply is powered up? What happens if wall power is lost?

Is there a list of rules to operate a bench power supply? Learning by trial and error kind of sucks.

[redwire]

Most bench power supplies have design shortcomings that can cause damage to themselves or the gear you are powering.
But I don't know your specific PSU. If you have the schematic, you can look for the protection parts, or if the marketing blurb says "it can charge batteries" then there is a chance it's OK.

Batteries
Most PSU's have a reverse-polarity diode which makes sparks or smokes and shorts, if you connect a battery backwards. Or it could be what happened to yours, rectifiers cooked beyond what the PSU could do to them.

The Backfeed problem
If the PSU is off, or set to a low output voltage, or there is a power outage - and you've connected a battery, the battery discharges back into the PSU and can damage semiconductors in it.
I think about half the bench PSU's I repair have no coverage for this. There are pics of one linear PSU that totally burned up and caught fire - because there is no fuse involved at all on the low voltage/battery side.
PSU designers only think current is coming out of the power supply, never back in.

The Turn Off
Set your PSU to a nice 5V say, then power it off (or a mains power failure happens) and the output voltage spikes way up, a blip that blows whatever you were powering.
This is due to the control logic power collapsing faster than the main filter cap, and a 20V blip is easily spit out there and quickly falls off to not make any smoke.
But next time you power up, surprise! your circuit is dead.


So the one rule I follow is to have the PSU powered on, voltage and current set to near what the battery should get, and then connect the leads (not backwards).
I never leave the setup unattended, i.e. leave home and go shopping.

If you are really not sure, best is to add a series diode from the PSU output. This will prevent most of this drama but the voltage drop of the diode must be accounted for, which makes it kind of a hassle if you want exact charging voltage.