My BC168 Charger went POOF!

[powersupply]

Interesting to see that it is mostly that channel failing, seems to be a systematic error.

Three things come to mind: the systematic (design?) error is still there, the driving chip is damaged now, and the "IRF7832 from ebay" could be fakes.

In such situations I'd always upgrade the FETs (if needed match them) or double them and what I've learned here about fakes from China, would get the parts from a reputable vendor.

 

[Tommy L]

It's been over three months. I contacted the vendor in china to see if they are sending me a replacement.

I'm being patient.

Tommy L sends....

 

[powersupply]

Looking at the differences it could be as simple as worse heat dissapation of that channel.
These FETs degrade in their ratings rapidly with heat.
If you look at the specs, it loses 20% current rating at 70 degrees.
I would definetly use a higher voltage (VDSS) and amp (ID) part as a replacement.

 

[powersupply]

I'd replace them (all) with something like RENESAS - RJK0454DPB - N CH MOSFET, 40V, 40A, LFPAK-5

You'd need to move the FETs a little norh (referencing DSC_0800.jpg) and scratch away the green screen beneath them.
Also scratch free beneath pins 1..3.
Solder those areas with a little solder (just cover them flat with solder basically).
then solder the FET in place and extend pin 4 with a wire to its pad.

Or use a TO220 FET right away and be done with it.
IPA037N08N3 G - MOSFET, N CH, 80V, 75A, TO-220FP

 

[dnmun]

so nobody ever bot the 7832 parts on ebay and tried to repair them? just the one where the channel was shorted? any luck in isolating the short? did the schottky diode on that channel short out or is it still a diode?

 

[meelis11]

so nobody ever bot the 7832 parts on ebay and tried to repair them? just the one where the channel was shorted? any luck in isolating the short? did the schottky diode on that channel short out or is it still a diode?

As I said, I replaced it and it burned out/shorted again. Where is shottky diode you refer to?

 

[meelis11]

I'd replace them (all) with something like RENESAS - RJK0454DPB - N CH MOSFET, 40V, 40A, LFPAK-5

You'd need to move the FETs a little norh (referencing DSC_0800.jpg) and scratch away the green screen beneath them.
Also scratch free beneath pins 1..3.
Solder those areas with a little solder (just cover them flat with solder basically).
then solder the FET in place and extend pin 4 with a wire to its pad.

Or use a TO220 FET right away and be done with it.
IPA037N08N3 G - MOSFET, N CH, 80V, 75A, TO-220FP

What is reason to replace them - is it really solution to a problem to replace them with more powerfut FETs? You think it is design problem and used transistors are too small? Maybe problem is in some other part and FETs act like fuses - if you put bigger FETs there, something else burns?

 

[powersupply]

Well, it is clearly a design problem given that so many fail right there.

SO-8 is not really good for handling much power. And I do think that heat is the cause there.
Considering that all channels are wired same way that is.

Well, the part is blown, replacing it takes effort and time and cost.
So why not use a better part and exclude the doubt and risk of failure again?

Again, I mentioned three reasons I could think of for failing, it could still be the driver IC, that might have been damaged, too.

 

[meelis11]

Well, it is clearly a design problem given that so many fail right there.

SO-8 is not really good for handling much power. And I do think that heat is the cause there.
Considering that all channels are wired same way that is.

Well, the part is blown, replacing it takes effort and time and cost.
So why not use a better part and exclude the doubt and risk of failure again?

Again, I mentioned three reasons I could think of for failing, it could still be the driver IC, that might have been damaged, too.

Other channel mosfets stayed cool, but channel #2 mosfets were getting hot - I touched them with finger when charge current was only 0.5A. There must be some design flaw that same mosfets always burn out.
What about adding small heatsinks on top of mosfets? Could that work?

Edit: forgot to say that when channel #2 was working, it made pulsation noise that I could hear. It was quiet but sounded like voice modem dialing a number
Now when that channel is dead and others are charging, it is quiet. Different PWM on that channel? Or bad capacitor?

 

[powersupply]

Edit: forgot to say that when channel #2 was working, it made pulsation noise that I could hear. It was quiet but sounded like voice modem dialing a number
Now when that channel is dead and others are charging, it is quiet. Different PWM on that channel? Or bad capacitor?

Now other possibillities open up, with transformers and as you mentioned capacitors all becoming culprits, too.

If you can measure those parts (exact resistance, inductivity, capacity etc) do that first.

Other channel mosfets stayed cool, but channel #2 mosfets were getting hot

Was this with the replacement FETs or originals?

Considering all being equal and that others had failures at the same channel, I still think this is a heat induced issue.

Adding cooling/heatsinks in hot areas is ALWAYS good.

 

[meelis11]

I has been a while and charger is just resting on its box...time to repair it.

I already have some IRFZ44N TO220 MOSFETS (55V, 49A, RdsON 22mOhms)
Can I use them (If I manage to physically solder them to right pins)? Is higher resistance can be issue somehow? (22mOhms vs 4mOhms)

Meelis

 

[parabellum]

Report please, is it a FAT, SHotkey or driver.

 

[sigimem]

I had the same issue with my BC 168 charger lately. Channel 2 and 3 were burned near that 8 leg thing Op Amp/Fet?
I got help from a friend who is good at repairing electronic devices we wanted to order parts but the pcb was too damaged.
In the end he soldered out the burned parts and the charger stayed on again (shut down just after getting power before).
I had false voltage readings on the cells after we opened it. I can now use the charger after I connected the removed channel ouputs to the previous channel.
So I am now using channel 1 4, 5 and 6 which is ok for my since I have 4s packs. The output of channel 1 is connected to 2 and 3 to make it work.
Maybe someone has an idea how to avoid the proplem? more cooling?

 

[whereswally606]

I have two of these failed through user error. Anyone here fancy having a go at fixing them?

 

[sigimem]

What does user error mean?
I would probably take them if you are willing to sell them.

 

[TheBeastie]

Well I am still going strong with my BC168, its my first and only Lipo charger. I have kind of treated it like someone would as I imagine people treated their first color TV set when color TV first came out.
The thing about this kind of charger is that its incredibly easy to connect 2 lipo packs in series then plug in this charger and ruin it.

Also if you dont know there is a 9cell charger out there the UNA9. If you have killed your BC168 maybe give the Una9 a try.
http://www.ebay.com.au/itm/360732877603?ssPageName=STRK:MEWAX:IT&_trksid=p3984.m1438.l2649
Google around for a better price.

Also the CB86 looks more capable then I initially thought, seems like you can plug in different lipos and instead of just parallel charging them like a simple breakout cable would it actually charges each pack separately and "cycles" through each pack after the previous pack is done, see below at 5:40 secs.
https://www.youtube.com/watch?v=Tw9fdp6fb6Y#t=340
http://www.ebay.com.au/itm/Revolutions-RadioLink-CB86-Professional-RC-Model-LiPo-Battery-Balance-Charger-/221471851711?pt=AU_Toys_Hobbies_Radio_Controlled_Vehicles&hash=item3390c044bf&_uhb=1

 

[whereswally606]

User error means I had them plugged into a battery 12s which I forgot to remove my connector to break it into 2 6s packs so it must have over bolted them both. I will sell them if you want them £10 a piece but I'm in the UK. And they are broke you will have to fix them.

 

[TheBeastie]

S6 on my BC168 just died. It still displays voltage but when I hit charge the current meter always just stays at 00, so its not going to charge via S6 any more.
I am not sure what killed it. I had charged up the first S1-6 to 4.07volts and it had been sitting there like that for 12hours. I unplugged the balance cable and after a minute I smelt a horrible toxic plastic burning smell.

I turned off the PS to the BC168 and turned it on again, I plugged on a S6 lipo and all volts were reading OK.. I hit the charge button and amps meter started ticking for all 6cells then more smoke came out of the BC168 and the amps on S6 meter went and is now always 00 but it still shows volts.

So I still got S1-5 still working as far as I can tell.

I wonder if it was age or if I unknowingly did the dreaded mistake plugged two balance leads to either ends of my battery (while in full serial) then into the BC168, I dont think I did though because thats always on my mind when I use it.
Been using it for about 2 years so I guess I had more mileage then most.

 

[fatty]

Looking at the differences it could be as simple as worse heat dissapation of that channel.
These FETs degrade in their ratings rapidly with heat.
If you look at the specs, it loses 20% current rating at 70 degrees.
I would definetly use a higher voltage (VDSS) and amp (ID) part as a replacement.

Well, it is clearly a design problem given that so many fail right there.

SO-8 is not really good for handling much power. And I do think that heat is the cause there.
Considering that all channels are wired same way that is.

Well, the part is blown, replacing it takes effort and time and cost.
So why not use a better part and exclude the doubt and risk of failure again?

Again, I mentioned three reasons I could think of for failing, it could still be the driver IC, that might have been damaged, too.

Although this poster is inactive, I thought his advice to switch the SO-8 MOSFET to a TO-220 to be a bit.. premature.

There are other MOSFETS in SO-8 package with lower conduction and switching losses, and higher power dissipation:
The IRF7832 referenced is rated at 25C for 20A and 2.5W dissipated, and at 70C for 16A and 1.6W dissipated, and 3.1mΩ resistance and 34nC gate charge.
The SI4126DY is rated at 25C for 39A and 7.8W dissipated, and at 70C for 31A and 5W dissipated, and 2.7mΩ resistance and 30nC gate charge. Pulsed drain current is lower though.

Without having one in hand to test, I'm curious if the Si4126DY would work.

 

[fatty]

I received a BC168, advertised as 2020 version with 4.35V LiHV support, board revision 1.3.
While board version changes are unknown, I hope to test/avoid the theory that high temperature kills the MOSFETs by upgrading the two fans from 0.5A to new Delta AFB0412VB-A 0.21A fans. I'll also be removing the JST daughterboard to hardwere 7/18 cable, so airflow across the MOSFETs will be greatly increased.