Controller MOSFETS and question on failure.

[Fordulike]

A few questions actually, for any techies out there.

Take for example a BBS02 9 MOSFET controller. It is my understanding that a controller like this, has 3 MOSFETS allocated per phase wire.
If one single MOSFET fails, does that usually cause the controller to become dead? If this is not the case, could the controller work intermitently?

Also, If one MOSFET fails, does it normally take out the others in the same group?

 

[Deleted member 25121]

A few questions actually, for any techies out there.

Take for example a BBS02 9 MOSFET controller. It is my understanding that a controller like this, has 3 MOSFETS allocated per phase wire.
If one single MOSFET fails, does that usually cause the controller to become dead? If this is not the case, could the controller work intermitently?

Also, If one MOSFET fails, does it normally take out the others in the same group?

It all depends.....

If a single MOSFET fails and goes short circuit then the controller is basically dead since there will be no output on the associated phase.

If a single MOSFET fails and goes open circuit it will cause the 2 MOSFETs in parallel to be under increased load which is likely to result in them failing at some time. Also the controller performance could possibly be compromised with the associated phase output being unable to provide the required current under load.

Here's the datasheet for the MOSFETs:

https://www.infineon.com/dgdl/irfb3077pbf.pdf?fileId=5546d462533600a4015356153f0d1def

 

[Nealh]

Failure is often just one fet to start with , the BBS is sealed in potting compound.
My BBS01 failed first time round in the middle of some woods , power (current surged) intermittently then the controller eventually failed full stop.
I think one failed on the high side causing the others to handle more current, during the ride I tried to molly coddle the power to try and get me home but in the end after a bit more riding the controller failed completely.

Mosfets like the irfb3077 are N type and work by being open circuit in ebike controllers. The Gnd side is the low side and is the switching side, I believe it is this side that fails most often then not. A typical resistance is 10k on all three of the phases.
The V+ side is the high side and an open circuit and infinity or reading of 1 is registered when resistance testing on the wiring , any reading of 0 shows a failure of the open circuit.

The wire testing tells you which side of the circuit and which phase has failed but not which mosfet, the mosfets then have to be tested individually to see which one has failed. In practice it is best I think to renew all the fets on the failed phase.

 

[Deleted member 25121]

The IFRB3077 is an enhancement mode power MOSFET, this means that if you measure across its drain and source pins there should be a high resistance. In operation the MOSFET will be pulsed on during which brief time there will be a resistance of just an ohm or so across those pins, you'd need to power up the controller and motor and use an oscilloscope to properly observe what's happening when the motor is being driven.

FYI:
There are also depletion mode MOSFETs, the term enhancement mode is used to indicate that the current flowing between the source and drain pins is increased (enhanced) when a voltage is applied to the 3rd pin on the device, the "gate" pin. With 0V on the gate pin the current flowing between source and drain pins is very very small ie there's a high resistance.

With depletion mode MOSFETs the current flowing between the source and drain pins is decreased (depleted) when a voltage is applied to the gate.

 

[Fordulike]

Thanks for your knowledgeable replies ebiker99 and Nealh. Another question. I've just replaced my BBS02 48v 25A controller, after it fizzled out earlier this week (thanks to a forum member for supplying the replacement).

I've reprogrammed it with my original preferred settings, but dropped the total current limit from 25 amps to 20 amps. Would this likely protect the controller from future failure, or am I just being over-cautious? The original lasted around 5k miles with my style of riding!

 

[Deleted member 25121]

If I were you I'd certainly run if off 20A for a while and make sure the motor performs well, it's possible that a short or fault in the motor blew your old controller.

If all seems well then increase it to 25A.

What was happening when the controller went wrong and what's you style of riding? Do you still have the stabilisers on the bike (only joking)

 

[Fordulike]

No idea why the controller blew. I always ensure the bike is in the correct gear, so that from a standstill, there is minimal load on the controller. Plus, the motor is spinning nice and fast whatever the terrain.

Might have been condensation within, but I'll probably never know the real cause. Luckily, it's pretty easy to replace a BBS02 controller.

 

[Fordulike]

If a single MOSFET fails and goes open circuit it will cause the 2 MOSFETs in parallel to be under increased load which is likely to result in them failing at some time. Also the controller performance could possibly be compromised with the associated phase output being unable to provide the required current under load.

I'm sure this is what had happened for some time before the controller gave out.

After installing the replacement controller and reprogramming to exactly the same configs as my previous one, except the max current which I dropped to 20 amps, I took it for a ride.

It's far more lively than before, quicker off the mark and stronger torque! I've now dropped the max current to 18 amps and reduced PAS start current to 30%.

The only other time I experienced this situation, was years ago with a MAC motor/Infineon controller combination. After replacing the controller for exactly the same model, it performed stronger all-round. This leads me to believe that even though a controller appears to function normally, it can become weak due to component failure.

 

[tommie]

No idea why the controller blew. I always ensure the bike is in the correct gear, so that from a standstill, there is minimal load on the controller. Plus, the motor is spinning nice and fast whatever the terrain.

Might have been condensation within, but I'll probably never know the real cause. Luckily, it's pretty easy to replace a BBS02 controller.

Was yours the earlier BBS02 750w type or the later upgraded item with the different fets?

 

[Fordulike]

Was yours the earlier BBS02 750w type or the later upgraded item with the different fets?

Upgraded fets.

 

[Fordulike]

I hope this isn't going too far off topic, but how does a programmable controller adjust the current limiting. I can understand the old school controllers, which use shunts, but not so much software based?

Humour me

 

[Sturmey]

I hope this isn't going too far off topic, but how does a programmable controller adjust the current limiting. I can understand the old school controllers, which use shunts, but not so much software based?

Humour me

Pulse Width Modulation (PWM) is used. Plenty of explanations on line e.g. https://en.wikipedia.org/wiki/Pulse-width_modulation

 

[Nealh]

The shunts are solid state resistors and they resemble one of little black components on a pcb.

 

[Deleted member 25121]

PWM provides a means of control, not the sensing. The width of the pulse turning on the MOSFETs is modulated, ie varied, to control the energy being applied to the motor. Hence the name, Pulse Width Modulation.

Sensing is still done by measuring the voltage developed across series resistance.

 

[Fordulike]

The shunts are solid state resistors and they resemble one of little black components on a pcb.

Ok, so controllers still require these shunt resistors to function. I guess that the controllers I remember from years ago, had a shunt resistor that was a bit more obvious, i.e. a piece of coat hanger looking wire between two circuit board tracks.

I think I'm starting to get my head around this, thanks guys

 

[Nealh]

Yes the shunt resistors on the integrated controllers which get smaller and more compact have in the main done a way with U shaped metal shunts and the resistor/s and now use the solid state type.
D8veh/Vfr mentioned in the past one could still increase the resistance/current of these by piggy backing/soldering another one on top of the existing one, though the work would need to fine.

 

[Fordulike]

Anyone question

Is there a direct correlation between the battery current limit and the motor phase current. I.e. is it directly proportional to the value set in the' battery current limit' box, on the controller software page?

I'm specifically referring to the Bafang BBSxx controller software, which doesn't have a phase amps setting.