EEVblog Electronics Community Forum
Electronics => Repair => Topic started by: obd.tech on July 30, 2021, 05:04:40 pm
-
Similar topic to here
https://www.eevblog.com/forum/repair/repairing-electronic-unit-for-bd35-compressor-output-mosfets-blown/ (https://www.eevblog.com/forum/repair/repairing-electronic-unit-for-bd35-compressor-output-mosfets-blown/)
Attempting repair of a Secop Electronic Unit for BD35 Fridge compressor, company formerly known as Danfoss.
Model 101N0650 for 12/24V BD35 compressor.
This failed control unit was blowing the 15A 12V DC rated inline Fuse (even with all the fridge components disconnected)
From my research it's a 3phase DC step up starter & control unit for the refrigerator pump?
Any ideas or help that I can use to test anything!? & hopefully repair, would be good to get from the great guys on here!
(I'm a relative newbie at this stuff, but interested to learn from the experts on here!)
Thanks :-)
Update :- after inspecting with my microscope,
top right mosfet ?? D2PAK type thing is STPS30M60SG-TR (Schottky Rectifier Diode)
7 Leg 1 near middle is STH240N75F3-6 .... if this helps??
Inverter MCU is ST7FMC2
-
Any ideas people on what to check (how to?) or unsolder to test?
Can the Caps draw too much, to blow the 15a fuse?
Any help appreciated!
Thanks
-
the caps are not the primary suspect
check with multimeter from motor wires to + and - rails for start
the 'power module' driving the compressor is maybe shorted. most likely the 7-pin component, ou didn't got the code right, recheck codes
replacing the power module IC, it's up to your skills and patience for removing the paint
-
on earlier models they had mosfet issues, you have a shot they haven't change the hardware designer and the mosfets keep dying from bad design, designing power stages is not as easy at it looks, especially for inverters living in eastern europe for example, where mains 220V supply is a joke
see https://forum.allaboutcircuits.com/threads/danfoss-compressor-speed-controller.118993/page-2
so identify correctly your mosfet power stage code and do the math if it's worth repairing, almost sure the power stage is dead, viewing the big fuse is unhappy
-
Thank you perieanuo :clap: Great! Some help from a kind person! :-+
Sorry, the 7 leg power stage chip number was my bad typing :palm: it's a STH240N75F3-6 :-+ (corrected now in original post)
The coating is rubber silicone like stuff, so comes off easy with a rub, so not too much of a problem.
So, you are saying use a multimeter to test for circuit resistance from the motor 3 wires and + - main power terminals? = to + is 109k ohms (open circuit to -)
Thanks
-
hi,
using diode mode of your multimeter test directly the mosfets on your board, for example for the 7-pin STH240N75F3-6 test between TAB (drain) and 2 (S), but you can test also G-S and G-D for shorting.
Usually is D-S going rogue, especially if you said big current fuse dies.
test first the transistors that are making motor driver bridge (i see maybe 6 mosfets in soic8 package next to motor wires pcb connection, they are also primary suspects here, give us the code or clear photos)
apparently the 7-pin power mosfet acts for all the inverter current, along with maybe the diode on the corner side near the electrolytic and the big filter. thad diode may be for Vcc rail protection, with A on - rail and K on + rail, so if the diode is shorted-> burnt fuse. check that too, give us correct code of that diode (i suppose...)
you have also like 2 mosfets at Vcc entry side, near the small electrolytic cap. i doubt they are shorted
if all mosfets/transistors are fine, then we have a problem, the microcontroller may be the problem, but in my experience he never dies
what says the multimeter if you do a diode-test between Vcc+ and Vcc-?
-
Today's update for perieanuo :-+ and others?
Please see attached the photo of the output MOSFET chips by the motor wires.
Nearest the wires the soic8 are I*RP704G but I can't find a data sheet on that :-//
Further away from the wires the soic8 are IR2301S (found that)
Thanks for the previous reply.
Top right mosfet ?? D2PAK type thing correct part no. is STPS30M60SG-TR (Schottky Rectifier Diode)
What says the multimeter if you do a diode-test between Vcc+ and Vcc-? The 2 input power terminals to the PCB? = .OL
This might be daft for some, but I've never used the diode mode on my :-DMM before, so a great learning curve for me!
Not sure what I'm doing so tested both ways polarity on the STH240N75F3-6 as per your instructions :-//
Tab (Drain) to 2 (S)
+ & - =163. - & + = 481. (0.188v. 0.492v)
G 1 to S 2
+ & - = 1. - & + = 633. (1.895v. 0.616v)
G 1 to D
+ & - = 1. - & + = 876. (2.161v. 0.788v)
Hope this makes sense? and if you can enlighten me any, even better :clap:
Update .... Those multimeter results are on an el cheapo 1 I had at home, so I thought I'd double check with my auto workshop Uni-T UT210E & got different results. I've put the Uni-T results above now in brackets.
What next?
Thanks
-
mosfet datasheet is here https://www.infineon.com/dgdl/irf7341pbf.pdf?fileId=5546d462533600a4015355f64f031b63 (https://www.infineon.com/dgdl/irf7341pbf.pdf?fileId=5546d462533600a4015355f64f031b63)
-
so the shottky isn't shorted, the 7-pin mosfet isn't and Vcc+ <-> Vcc- rail not shorted hmmmm
i was expecting short between Vcc+ and Vcc-, if you found none, start testing the IRF7341PbF (you have 6 of them in theory, 3-phase bridge), then if they are kinda not-shorted, test the mosfet drivers IR2301 between the outputs HO;LO and VCC;COM (diode test between each output and each power tail in the 2 directions)
remember, the diode tests made in-circuit are not always 100% reliable, meaning some mosfet or driver may be faulty even if you don't detect a clear short-circuit
but afair, the blown fuse points a big short, definitively you will find it, except if he's coming from microcontroller wrong switching logic, but this should happen once in a million cases.
-
Schottky Diode STPS30M60S test
K to A
+ to - = 0.50v
- to + = 0.19v
Now the IRF7341 saying there are 6 got me confused/thinking, until I realised that they are a dual package soic8, so only 3 physically!
The test results are all similar + or - about 0.02v for all 6.
D to S
+ to - = 0.55v
- to + = 0.57v
G to S
+ to - = 1.82v
- to + = 0.68v
G to D
+ to - = 1.82v
- to + = 1.18v
Now the IR 2301 driver Soic8, I'm not sure I understood you correctly, so please check my methodology!?
Com to Lo
+ to - = 0.63v
- to + = 1.82v
Vcc to Ho
+ to - = OL
- to + = OL on 2 ....... But the middle Soic8 reads 2.20v :wtf: :bullshit: :-//
(I retested a few times to make sure & now get consistent reading, fiddly with pins, haha)
Have we found something!?? Or have I tested it wrong? Please confirm/ reply my friend/teacher :clap:
Thanks
-
hi,
sorry for delayed answer
regarding IRF7341PbF, i wasn't attentive, they are 2 mosfets in one package, the bridge needs 6, so you got 3 ic's, it's fine. i counted mosfets, not ic's...sorry
seems they are ok, i expected clear shorting between D-S
for the 2301, you cannot really test them without removing the drived mosfets and replacing them with some resistor load (maybe 1kohm or something, datasheet may be consulted to decide that value for the load) and watching the scope output.
so 2301 may be problematic (the only test but not 100% percent conclusive that can be done is verify between HO/LO and VS/VB for shorts and symmetry and on input side HIN/LIN to VCC/COM also some symmetry between the drivers, careful not to burn processor output that drives those drivers). i mean diode test between HIN and VCC for exemple in both directions and same with LIN<->VCC/COM.
in theory all those tests are with driver extracted from pcb, but if done carefully it can be done 100% safe in circuit using standard diode tester (the injected voltage won't be bigger than your microcontroler accepts, you can check this yourself from multimeter and micro datasheets.
for a short like yours, it's either final mosfets either theit drivers either the micro, if you say on DC rails you got no short.
if you remove the 3 double mosfets and power the pcb, the DC rail should be nice with you and not pop the fuse, in this way you are sure the short is coming from final mosfets commute.
in industry pcb's i worked, in cases like yours usually technicians replace without thinking much the mosfets and their drivers, the micro is rarely burnt.
your testing was correct, but you gotta insist with drivers or just replace mosfets+drivers if you can't find the dead driver with the 'simmetry' method i described or maybe some mosfet that may be altered, but i'm surprised you can't find mosfets/diodes shorted, that points the drivers