ebike kit advice needed

D

Deleted member 4366

Guest
I've just opened up mine to check measurements. It's tricky getting some because of the capacitors, which distort them. To properly test FETs, you need to remove them from the pcb. The only thing you can really determine is that you don't have any shorts. To be honest, I'd be surprised if any of the FETs are blown.

Looking at my pcb, the main are of risk of shorting is at the end opposite the one where the wires come out. The positive rail has a reinforcing wire soldered to it that runs very close to the yellow phase plane. It looks a bit dodgy to me. Is yours touching?

The ground rail is on the other side of the pcb under the heatsink, so not so easy to get a short there.

Is it possible to get a photo of your complete pcb back side to see if there's anything else?
 

flying-chase

Pedelecer
Sep 25, 2013
56
0
Letchworth uk
Hi d8ve an open circuit would be 1 on my dmm, I.e infinite resistance. The lower the number the less the resistance. So green is showing a circuit, while yellow is open.
 
D

Deleted member 4366

Guest
It seems strange. I get the same readings as you on the green phase, the same on the yellow and blue low-side FETs, but I can't really measure the high-side ones because the capacitors make it change. The high-side ones on the green phase start at 4k, then climb slowly if I keep the probes on. Different meters will give different results, I guess, so I'm not sure whether comparison is meaningful. I would expect all low-side FETs to be the same.

Unfortunately the photograph of the whole back of the pcb is no good for me. The bit I need to see is at the back angled away. I need to see it from above. Also, Flikr doesn't allow me to zoom or rotate the picture nor download it. Can you do another one and put it on photobucket? The photo is only clutching at straws anyway. The cause of the problem could be anywhere.
 
D

Deleted member 4366

Guest
One last thought. You measured the green shorted to ground before, but now you have thousands of Ohms of resistance between the green phase and the low-side connection to ground, which implies that the short is no longer there. Please check it again from the green phase to the ground to see if the short has gone. If so, check whether the fuse still blows when you connect the pcb out of its case. I still think that there's a possibility that the short id through the case.
 

flying-chase

Pedelecer
Sep 25, 2013
56
0
Letchworth uk
Hi d8veh, I'll do a photo when I get home to get better quality. If I use the diode resistance tester on the dmm. I get
positive to green phase .000
positive to yellow .526
positive to blue .526

ground to green .000
ground to yellow .526
ground to blue .526
 
D

Deleted member 4366

Guest
Iit doesn't make sense. On the low-side FETs one pin is joined directly to ground and the other directly to the phase wire. If you had zero resistance between green and ground, you'd get the same zero resistance between the two pins. Instead of using the diode test setting. Why don't you use the resistance scale?
 

flying-chase

Pedelecer
Sep 25, 2013
56
0
Letchworth uk
Hi d8veh,
measured using the resistance scale, meter set to 20k ohm.
ground to yellow 9.95
ground to blue 10.00
ground to green 0.00

positive to yellow 9.95
positive to blue 10.00
positive to green 0.00

Then for clarity tip of red probe connected to tip of black probe 0.00 at 20k ohm and 00.4 at 200 ohm.
 
D

Deleted member 4366

Guest
Be fore going any further we need to resolve the discrepancy I explained above. You have just conformed that the green is shorted to ground.
The wide track along the edge of the board is the positive rail. The negative is on the other side. If you look at the two high-side FETs on the green phase, you can see that one leg is connected to the positive rail and the other is connected to the green phase wire. It's the same for the two low-side ones but the connection to the negative rail is on the other side of the pcb. If green is shorted to ground, the two legs of the low-side FETs are both connected to ground, so you cannot measure resistance between them, but in post#59 you measured 4K. That's impossible. Between the two tests, one is wrong. Which one?

If the problem is an immediate short circuit from positive to negative, you should be able to see that by measuring the resistance between positive and negative wires. It must be zero. Please confirm that. If the connection is open, then the short must happen after connection, which could mean one FET permanently open (blown).

One other thing. If you connect the controller the wrong way round, you get something like a short circuit. Can you check the polarity of your battery leads with a meter?
 

flying-chase

Pedelecer
Sep 25, 2013
56
0
Letchworth uk
Probes in positive and negative give me 00.4 at 200 ohm (0.00 at 20k ohm).
I'll go and check the battery polarity now.

How can you tell which are the high sided fets?


green phase wire to right leg all fets measured 00.4 (at 200 ohm scale)

green phase wire to left leg fets measured 90.2 (number 5 looking at the fets, controller wires on the rhs), 90.2(number 6 looking at the fets, controller wires on the rhs), 4.5 and 4.5 (at 200 ohm scale) for numbers 7 and 8
 

flying-chase

Pedelecer
Sep 25, 2013
56
0
Letchworth uk
Battery leads polarity is correct and correct when connected up to each other.

The middle set of four fets are for the green phase wire. I have just measured again and I do get 00.4 measuring from middle leg to right leg. Measuring from middle to left leg I get 92.9(closest to the two caps), 92.9, 4.5 and 4.5(Fet closest to all the controller wires).

I put the probe in to the positive cable and measured to the left and right hand legs which gave the same results as above. I then put the probe in the ground and got the same results on all left and right hand legs of the four fets in the middlle.
 
D

Deleted member 4366

Guest
The high side FETs are the ones where the middle leg is connected to the battery positive (high voltage) and the low-side ones have it connected to ground (low voltage. The right legs are always connected to the phase wire. By switching on/off the high and low FETs, the CPU changes the voltage between the phase wires to make the correctly timed A/C for each phase to turn the motor.

If a high and low FET were open at the same time, there would be a short, but they would have to be stuck open because the CPU limits the current, and doesn't switch them on until you open the throttle, which is why I think that there is a physical short somewhere. I've not heard before of fuses blowing on connection due to blown FETs.

Your reading across the red and black battery wires indicates a dead short.

If you look on the solder side of the pcb where the red wire is soldered, it's an island surrounded by an insulated ground plane. Can you scrape around it just to make sure there are no bridges.
 

flying-chase

Pedelecer
Sep 25, 2013
56
0
Letchworth uk
Hi d8veh, there don't seem to be any bridges. However the section closest to the extra wire seems very close to the black and the same on the other side closes to the FETs. But the gap if there is very hard to see.
 
D

Deleted member 4366

Guest
I've been thinking about it for a bit. It did go for a couple of seconds, which seems to rule out a short on the PCB. Something must have failed. There's only two things I can think of that could blow: the FETs or the 12v regulator. On the opposite edge of the pcb to the FETs is another thing that looks like a FET, but it's on its own. It's the 12V regulator that is also connected to the battery voltage, but it's difficult to test without power and with the + and - shorted out.

I have one last question before I give up. I take it all the tests have been done with the LCD disconnected?
 
D

Deleted member 4366

Guest
Here's mine:


BMSBattery are trying to improve their customer service. It's worth an e-mail to see what they say, but they don't know what you connected to it and how, so I can imagine that they'd be quite suspicious. If you coulf find a fault with it, you'd be in a much stronger position. Have a close look to see if any of the surface mount components are not soldered properly.
 

flying-chase

Pedelecer
Sep 25, 2013
56
0
Letchworth uk
Finally, found some damage. Red wire for speed sensor slightly melted and one end of grey cable (ABS) not attached. Though this last could be after the fact.