OPD, car rolls backwards when stopping on downhill slopes (merged thread)

[dave99]

I share with you a video of this issue:
It's very dangerous for people who are crossing behind the car.

wow - that cannot possibly be a desired feature!?

 

[DixieTGS]

The one time it happened to me it didn't seem too worrisome. In the split second before 0 MPH it felt inevitable that it was going to happen, and I had automatically braked to a halt within 10-20 cms of starting to go backwards. No idea how far it may have gone if left unchecked, but I suspect much less than in the video.

I think the brain could tell the deceleration wasn't changing and so knew that backwards was the inevitable outcome. ?

It felt just the same as when we used to muck around off road in my youth and go from reversing to forwards, using the clutch and accelerator but no brake.

 

[Gruntbuggly]

MG New Zealand have had my MG 4 for over a week on the pretext that it was only my vehicle thats affected. Clearly that's a lie. They've also claimed that this is designed behaviour in order to discharge excess regen power. Also bullshit. The bottom line is that this is a dangerous, undocumented "feature" (i.e. hazardous design flaw). I'm considering progressing with a class action lawsuit against MG as the vehicles should never have been allowed on sale. What kind of "feature" slams the car into reverse because you touch the brakes??!!! Anybody else keen to join in the litigation?

 

[NLMGSAN]

MG New Zealand have had my MG 4 for over a week on the pretext that it was only my vehicle thats affected. Clearly that's a lie. They've also claimed that this is designed behaviour in order to discharge excess regen power. Also bullshit. The bottom line is that this is a dangerous, undocumented "feature" (i.e. hazardous design flaw). I'm considering progressing with a class action lawsuit against MG as the vehicles should never have been allowed on sale. What kind of "feature" slams the car into reverse because you touch the brakes??!!! Anybody else keen to join in the litigation?

I agree that this should never happen. I hope you get enough support from fellow MG drivers in NZ. Good luck!

 

[Rockstar]

in my case, MG and local dealer have not investigated the problem.
Like always, it's very difficult to obtain something from global brand where there are only local feedback.
@Gruntbuggly have you share with your dealer my video on youtube? could you also make public your experience creating a video linked with mine?

 

[Gruntbuggly]

My MG4 has been into service twice for this to be looked at. No solution. MG New Zealand seem to be taking the same line as all other territories, that this behaviour is "normal" and is a "feature" of the vehicle. Utter bullshit. This is a dangerous fault and the car should never have made it to sale. I'm currently looking at the possibility of legal action against MG, maybe as a class action lawsuit. This might need to be something done for each sales territory. But in the meantime I'm gathering references to this fault as evidence. My next step will be to bring in the media. This needs to be made very public if we're going to get any action from MG. Talking about it on forums like this won't get a solution.

 

[JonnyG]

What happens if you brake harder and longer ?

 

[Rockstar]

My MG4 has been into service twice for this to be looked at. No solution. MG New Zealand seem to be taking the same line as all other territories, that this behaviour is "normal" and is a "feature" of the vehicle. Utter bullshit. This is a dangerous fault and the car should never have made it to sale. I'm currently looking at the possibility of legal action against MG, maybe as a class action lawsuit. This might need to be something done for each sales territory. But in the meantime I'm gathering references to this fault as evidence. My next step will be to bring in the media. This needs to be made very public if we're going to get any action from MG. Talking about it on forums like this won't get a solution.

Please , inform me what will happen in NZ about this problem and MG position if they react.

 

[Rockstar]

What happens if you brake harder and longer ?

a big noise of brakes againt wheels because of reverse movement and the car stops.
If you release before energy was empty, the car goes back (see video)

 

[OriginalBigAl]

YT , incarnation Australia, Porsche EV conversion episode 28 explains this pretty well. he has a MG ZS donor car without OPD and so a car with OPD will have a stronger response .

 

[Rockstar]

??????

YT , incarnation Australia, Porsche EV conversion episode 28 explains this pretty well. he has a MG ZS donor car without OPD and so a car with OPD will have a stronger response .

 

[DixieTGS]

YT , incarnation Australia, Porsche EV conversion episode 28 explains this pretty well. he has a MG ZS donor car without OPD and so a car with OPD will have a stronger response .

Fascinating video, and obviously shows a related phenomenon, but his description doesn't explain it to me at all. I need someone with good physics knowledge to give further detail to that explanation or maybe a different explanation entirely.

In particular his comment about reverse thrust suggests it is using energy from the battery in order to put energy back in (get max regen) ?.

Using reverse thrust on an OPD car to do the final bit of stopping does seem possible to me however, as I'm guessing the stopping force available from regen is proportional to the motor speed. I.e. almost nothing at very slow speeds.

The OP mentions hearing the brakes doing the final bit of stopping though ?.

 

[Coulomb]

All AC electric motors can operate in four quadrants, i.e. the torque can be forwards or backwards when the motion is forwards or backwards (4 combinations). When operating, zero torque is a balance; the motor controller can smoothly change the phase of the applied rotating magnetic field relative to the field from back EMF. It's a bit tricky explaining it if you're not familiar with AC motors.

So the motor controller has complete control of the torque, in magnitude and sign (direction). Thus, this is a control issue. The motor controller will be aware of the present torque, and it will have a torque set point (the torque that it's aiming for). It will probably use a PI control algorithm (Proportional and Integral components). This is needed to achieve the desired torque quickly and accurately, without excessive overshoot and continual corrections.

The integral part of the control is necessary among other reasons for preventing a small offset between the set and actual torque. But there is a thing called integral windup that can occur if you don't implement the integral control properly. Basically, when you need to apply a large change, as you would if the brake is pressed quickly, or if the OPD pedal is reduced quickly, then the integral term can become quite large. Without some limit to this term, you end up with overshoot that persists as the large integrsl term slowly reduces. In this situation, the result is creeping backwards for a time, or if the integral wind-up is worse, you get significant reverse torque.

 

[DixieTGS]

Ooh, I knew back emf was going to crop up sooner or later ?. That's great, it's given me enough to get a reasonable idea of what's going on, and what to look up on wikipedia if i want to understand better.

Is it fair to summarise as:

1) From the initial speed down to a few mph, regen is working as normal.
2) At a few mph the regen is still quite high due to the down hill.
3) Hitting the brake briefly (as done by OP) overloads the control software due to poor software.
4) The result is continued reverse force even after 0 mph reached which uses battery rather than charges it.
5) The control algorithm 'by its nature' sorts itself out after a few seconds and brings the car to a halt.

 

[Coulomb]

Is it fair to summarise as:

Yes, pretty much spot on, as far as I know, not being any sort of expert with motor control firmware.

 

[OriginalBigAl]

Well done guys, every day is a learning day.

 

[Rockstar]

Wonderful explanation but in the end, it's a wrong management from MG software of a normal physical mechanism.
It confirms what I thought because in my case, it appeared after an update. I had 2 theories: software bugs or sensor calibration issue.

The 1 million euros question, how to force MG to solve this problem because it's not a 'normal feature' and it's dangerous?

 

[barry7153]

I love OPD mode and use it continuously, however on a couple of occasions lately, I have come down from a reasonable speed to a stop by just lifting off the accelerator only to have the car stop and then start reversing backwards. I don't really know how far as I switch to panic and slam on the brakes. Just wondering if anyone else has a similar experience and if it's just a rock backwards after stopping or a potential problem I need to get looked at.

 

[AlC]

There's been quite a few people reporting that behaviour you're experiencing on my local MG4 Facebook page, so you're not alone.

It's not an issue specific to your car but perhaps a software issue. One theory is that the car applies reverse torque to bring it to a stop below the speed where regen is effective, and it sometimes overcooks it and ends up rolling backwards.

Some people claim it's more common if you're stopping on a small downward incline, where you might naturally expect the car to roll forward, but surprisingly you get pushed backwards instead.

I've not experienced it myself, my 51 Excite was delivered without the OPD option and I haven't had the software update applied to enable it yet.

 

[JonnyG]

I think its been mentioned before (mainly on gradients). It's defintely something to watch out for !!