MPPSolar Inverter/charger

[BruceS]

The unit is (I believe?) to be a PIP-4048HC. 48V nom ... 4000w output.... I think 40A input for charging... either AC or solar.
These units have no model branded on them but info can be seen on the LCD IF/WHEN powering up.
I have given to me a non-functional MPPSolar inverter/charger that I've pulled down without even trying to power up.
Looks like this. (see below)




I took a picture of the main board after removing some sister boards & the long heatsink bolted to the mosfets etc.
(see below)


I can see all 8 mosfets or whatever are seriously damaged. On very close inspection I can see no other damage anywhere. The bigger ones are 8 of 1MBH75D-060S FE .... the ones blown are 8 of IXTQ96N20P.
Surely it wouldn't be as simple as replacing these 8 things?

 

[(*steve*)]

Well... you could just replace them. It might fix the problem, or they could just blow again immediately. Fortunately you have part numbers this time

With mosfet failure like this, I would be checking very carefully what the gates are connected to as they will probably have received the full drain voltage. If this is switching a low voltage you may be OK, otherwise not so much.

Much could be learned from knowing *why* it failed. Were the batteries connected backwards? Was a solar array of too high a voltage connected? Did someone drop a screwdriver into it while it was turned on? Any clues?

 

[BruceS]

Unfortunately not many clues. It's changed hands a couple of times in a dead format! It does have a fan facing upwards that is designed to catch screws, metal filings & bugs!! I personally suspect it was fed power in the outlet side but have no idea. Apparently the last owner was told it does attempt to fire up but says something like "output short needs fixing & restart unit". I didn't even try it.
I noticed it'll cost me around $180au to buy the 8 bits.......... minimum of 30 per purchase?
Might pass it on to the next person as well? lol (It's not MPPT...)

 

[(*steve*)]

Sounds like it's detecting shorted mosfets.

It might be worth removing the obviously failed mosfets, cleaning up the board, and reapplying power from a suitable low energy source to see what happens.

If the mosfets are all in parallel you could replace one with something similar and connect the output to a battery (with a series resistor) to see what happens.

 

[BruceS]

Removed all 8 mosfets. (snipped legs off for now)
Cleaned up the board with alcohol (isopryl ?? from Chemist)
Hooked up 48V DC supply with the pos wire fed through a 60w 240V A/C globe & a switch inline.
Turned on the switch & got a bit of a glow in the globe which petered out to almost nothing .... went out?
I guess next step is to put the mosfets in & try it?

 

[(*steve*)]

Did it report a fault when you powered it up without the mosfets?

 

[BruceS]

I wasn't game t leave power on for too long but long enough to see the light globe fading out........ I'm almost certain that the LCD screen never came on at all. No flicker, no message or anything. I was only using light wire though, like on a bedside light.

 

[BruceS]

A few more details.
This inverter doesn't have any mechanical switch on it at all. It has a button (digital? ) to hold in to switch off or on.
I didn't try using this. It appears like even with out 'turning it on' that a sudden use of current may be charging up a capacitor or something?
If I test it again through the globe & try pressing the "on" button, how long do you think it'll be OK to do the test?
Most of these units seem to do a test before firing up so I'd think it may need 30 seconds or so.
Or just wait till I get the mosfets back in........... or one of them?

 

[(*steve*)]

Try it without the mosfets. I don't think it will do much harm that hasn't already been done. I'm pretty sure you'll get some sort of error, it may be useful to know if this changes after you replace thee mosfets (and yeah, I'd probably just replace one -- or one in each bank -- at first).

 

[BruceS]

You were right Steve! Fired it up & after the check it still reports "output short".
Might be one for the bin!

 

[(*steve*)]

It's probably reporting output short because it sees no voltage across the output.

It's probably not capable of knowing that it has no output transistors

If all those mosfets are in parallel, try replacing one of them and powering it up with the series lightbulb again. The bulb may be enough to prevent the mosfet expiring immediately if there is still a problem.

There's a good chance that there are other cheaper mosfets than an exact replacement. This might be a reasonable replacement.

 

[BruceS]

Thankyou Steve. I've got 10 "original" ones on the way for $22.40US.
Worth it just for fun!
AliExpress..... Free postage from China.

 

[(*steve*)]

I wish you luck with the "originals"

 

[BruceS]

Yes, I'm not sure if they'll be 'rip=offs' or not but I've got the original markings to compare them to.
http://www.aliexpress.com/item/IXTQ96N20P-original-authentic-direct-sale/1946743105.html
Do you think they'll be 'no good'? Surely the originals would also be Chinese jobs?

 

[(*steve*)]

You can never be sure. I've got a lovely collection of LM337's with beautiful markings that don't seem to be LM337's.

I've also bought stuff that is exactly what it was labelled as.