Interesting/worrying story from Scotland - ZS with no brakes.

[johnb80]

The price of implementing regen I think. I assume all EVs have similar systems.

I've never heard of a system like this and frankly, it's ridiculous and in my view very poor design. All of the EV's I've driven simply increase the level of regen in the first part of pedal travel with full regen occurring just as the pedal pressure starts to build and mechanical braking starts. Simple to implement, cheaper than this very over complicated sounding system.

 

[Tim Green]

I cannot see why there should be any physical connection between regen and the brakes. When you take your foot off the accelerator, regen starts immediately ( as with an ICE), the level of regen is set manually, 1,2 or 3. As soon as you brake, it merely assists the regen at whatever level it is set. If it is felt necessary to have progressive regen, this could be achieved using the rate of change of speed, which is an available signal already.

 

[mg4mc]

All of the EV's I've driven simply increase the level of regen in the first part of pedal travel with full regen occurring just as the pedal pressure starts to build and mechanical braking starts.

I guess my question would be, how do you know?
I didn't know the MG worked like this until I read the technical manual. None of the user documentation explains it.

Apparently there are three different modes of category B regen (using the braking system rather than just the accelerator). This the the modified mode where physical braking is held off until required and is the most effecient, so if be surprised if other manufacturers weren't using it a well.

If it were just the first part of the pedal travel that did not connect to the hydraulics, how would you provide the "feel" feedback that is seamless as it transitions from regen only to regen plus brakes?

 

[Gadget Geek]

When I have gone downhill, if I lift off the accelerator pedal, the regen kicks in to the set amount to try and slow the car. When I touch the brake pedal, it has absolutely no effect on the level of regen. The two systems seem to be completely seperate. It should be called regen slowing, similar to how the engine can slow an ICE car.

 

[Verdigris]

When I have gone downhill, if I lift off the accelerator pedal, the regen kicks in to the set amount to try and slow the car. When I touch the brake pedal, it has absolutely no effect on the level of regen. The two systems seem to be completely seperate. It should be called regen slowing, similar to how the engine can slow an ICE car.

Depends on how fast you're going and how steep the hill is. If the power meter isn't showing maximum regen then when you touch the brake pedal the regen will increase until it reaches the maximum, only then the mechanical brake starts to assist. (there's probably a bit of crossover to make the transition smoother).

 

[johnb80]

I guess my question would be, how do you know?

Easy, foot off the accelerator and see the level of regen on the gauge, lightly touch the brakes and see the level of regen increase, further pressure you can usually hear the pads on the disks (especially so on my i3 when it's rained and the discs are rusty, you hear the rubbing noise from the surface rust).

I didn't know the MG worked like this until I read the technical manual. None of the user documentation explains it.

I do really wonder if it does really work like this or the chinenglish translation has not gone too well.

Apparently there are three different modes of category B regen (using the braking system rather than just the accelerator). This the the modified mode where physical braking is held off until required and is the most effecient, so if be surprised if other manufacturers weren't using it a well.

Tesla has several different modes of regen including coasting type situation, regen only occuring when you start to use the brake pedal, VW ID3 is similar, BMW i3 has no options on Regen it's on and thats it, but do see the level of regen increase when you move the brake pedal.

If it were just the first part of the pedal travel that did not connect to the hydraulics, how would you provide the "feel" feedback that is seamless as it transitions from regen only to regen plus brakes?

The master cylinder is an open cylinder with a fluid reservoir on top. Hydraulic fluid is free to flow into the piston. Operating the pedal causes the piston to move forwards, eventually closing off the supply from the reservoir and then increases pressure in the now sealed brake hydraulic lines. It's this travel that is used to increase the regen, hydraulic braking doesn't take over or transition from regenerative braking, its in addition. All the driver is aware of as more pressure is applied to the pedal is greater deceleration.

I cannot see why there should be any physical connection between regen and the brakes.

It's a better system. If you have the capacity for even greater braking effect than foot off the accelerator gives, why not add it as the brake pedal is pressed during the first few mm of pedal travel. It would reduce wear on brakes, recover more energy etc.

When you take your foot off the accelerator, regen starts immediately ( as with an ICE)

I know what you mean except in ICE there is now regen, only engine braking.

, the level of regen is set manually, 1,2 or 3. As soon as you brake, it merely assists the regen at whatever level it is set.

Maybe on MG but definitely not on many other EV's

If it is felt necessary to have progressive regen, this could be achieved using the rate of change of speed, which is an available signal already.

Surely it would be better to change the upper levels of regen (over and above what you get with accelerator and brake pedal off), in the first few mm of pedal travel, the more you press the greater the regen.

 

[Gadget Geek]

Depends on how fast you're going and how steep the hill is. If the power meter isn't showing maximum regen then when you touch the brake pedal the regen will increase until it reaches the maximum, only then the mechanical brake starts to assist. (there's probably a bit of crossover to make the transition smoother).

Not in my ZS it doesn't.

 

[Verdigris]

Not in my ZS it doesn't.

Ah, I assumed the ZS had the same system as the MG5. My mistake.

 

[aserraric]

So the MG4 uses a fundamentally different system for braking/regen than the ZS does? That would explain why the ZS seems unable to use regen in ACC mode.
It also means this whole discussion is moot, since the news article talks about a ZS, not an MG4.

Unless someone has access to the technical manual for the ZS and can prove me wrong, I'm going to assume that the ZS's brakes are permanently connected to the pedal. Which, again, doesn't give this story a lot of credence.

 

[Tim Green]

Ref JohnB8, using brake pedal movement is very difficult as each car is different, depending on brake pad wear etc. Using a pressure sensor is difficult for the same reason, it is only really possible for a pressure/no pressure switch. Thus a measurement that can be made digitally from an existing signal such as speed is easier and much more accurate.
Ref Asserraric, there is obviously some connection between the hydraulic system and the computer because the car uses the brakes to maintain correct speed in ACC mode. Thus it must be able to increase and decrease pressure in the system and is thus a potential cause of brake failure.

 

[johnb80]

Ref JohnB8, using brake pedal movement is very difficult as each car is different, depending on brake pad wear etc.

I disagree, worn pads and disks make no difference to the pedal travel, the pistons simply stay further out. Even allowing for differences, the system work just the same, start to press the brake regen increases, if there was more travel befor friction braking started so what ? regen stays at max and then the friction brakes start. This is exactly how other cars function, I can demo this on my I3 no problem.

Using a pressure sensor is difficult for the same reason, it is only really possible for a pressure/no pressure switch.

Not necessary, pedal travel works fine.

Thus a measurement that can be made digitally from an existing signal such as speed is easier and much more accurate.

But not telling the systems what the driver is wanting to do and is thus irrelevant.

Ref Asserraric, there is obviously some connection between the hydraulic system and the computer because the car uses the brakes to maintain correct speed in ACC mode. Thus it must be able to increase and decrease pressure in the system and is thus a potential cause of brake failure.

Maybe increase pressure but I doubt decrease pressure, no reason to that, if the driver is braking the ACC would be off anyway and therefore no modulation of brake pressure is required.

 

[aserraric]

Ref Asserraric, there is obviously some connection between the hydraulic system and the computer because the car uses the brakes to maintain correct speed in ACC mode. Thus it must be able to increase and decrease pressure in the system and is thus a potential cause of brake failure.

The computer can apply brakes without user intervention (otherwise ACC would not work), but I doubt the computer can prevent braking through the pedal (not to mention how unsafe that would be, I doubt the various government road safety agencies would allow that without a failsafe mechanism).

 

[Tim Green]

JohnB80, well neither of us know how it is done, and it will differ between cars (when reviewing a design, engineers usually comment "well I would not have done it like that"!). Also, what you are told (eg applying the brakes increases the regen) tells you nothing about HOW it is done. If it were me designing, I would start with a signal from the brakelight switch and the rate of change of speed. It might be possible with just the latter, as the computer will already know the level of regen.

As to brake failure, this can happen at any time due to flexible pipe leaks, rusty pipes etc and any extra items that could fail only increases the possibility. Hydraulic failures are rare nowadays, but computer malfunction is not, especially in a low cost system with inadequate backup. (Unlike planes for instance).

 

[RobinH]

I my experience gently pressing the brake pedal just increases the amount of regen from the lifted accelerator.

From the MG4 Systems Description and Operation manual...





So, I was wrong, even normal braking is "simulated" drive-by-wire.

Only when a system failure is detected or there is a power failure is the master cylinder directly connected to the actual brakes.

Bear in mind this is an MG4 answer in the ZS forum.
The 4 is a ground-up EV design which would be expected to have integrated systems.
The ZS is an ICE design with an EV system added.
My observation in the ZS is that the braking and regen systems are not integrated.

 

[johnb80]

JohnB80, well neither of us know how it is done, and it will differ between cars (when reviewing a design, engineers usually comment "well I would not have done it like that"!). Also, what you are told (eg applying the brakes increases the regen) tells you nothing about HOW it is done. If it were me designing, I would start with a signal from the brakelight switch and the rate of change of speed.

That would not be a good way, the brakelight is on or off, ditto the regen, not good for mechanics, passenger comfort, stability or safety.

It might be possible with just the latter, as the computer will already know the level of regen.

You keep going back to change of speed, how would that work, Im slowing down from 60, foot off the accelerator, I now want additional braking and would normally start to use the brakes, the inital part of brake travel being regen in my case, how it work with this change of speed.?

As to brake failure, this can happen at any time due to flexible pipe leaks, rusty pipes etc and any extra items that could fail only increases the possibility.

Agreed hence why I suggest the BMW way is seemingly superior to thers in additional regen from the first part of pedal travel.

Hydraulic failures are rare nowadays, but computer malfunction is not, especially in a low cost system with inadequate backup. (Unlike planes for instance).

Theres always exceptions - Chinook helicopters, Eurocopter, Boeing 737 Max etc

 

[mg4mc]

I don't have access to a ZS system description document, but from the remove and refit manual, from what I can see the ZS appears to have a "conventional" master cylinder with vacuum assist and separate ABS module. No drive-by-wire apparent.

 

[Tim Green]

JohnB80, no it would work. The brakelights are already a known signal to the electronics (eg discontinue ACC) and the easiest way to increase regen would be to introduce positive feedback into the regen control loop. If the feedback is proportional to rate of change of speed that would obviously increase proportional to how hard you push the pedal.

But this is all much too deep. The regen available is very small compared to battery use (maybe 1%?) So its not worth too much effort. And all the manufacturers will do it differently.

 

[johnb80]

JohnB80, no it would work. The brakelights are already a known signal to the electronics (eg discontinue ACC) and the easiest way to increase regen would be to introduce positive feedback into the regen control loop. If the feedback is proportional to rate of change of speed that would obviously increase proportional to how hard you push the pedal.

Hmmm, is there not a danger that as your rate of deceleration increases, so does regen and then in effect it can wind up to max regen very quickly which is not what we want?

But this is all much too deep. The regen available is very small compared to battery use (maybe 1%?) So its not worth too much effort. And all the manufacturers will do it differently.

I agree

 

[RobinH]

Theres always exceptions - Chinook helicopters, Eurocopter, Boeing 737 Max etc

737 Max is a bad example of software malfunction. The crashes were down to a sensor failing without redundancy or fault detection. The software did exactly what it was meant to do, but it was reacting to an erroneous input.

 

[Tim Green]

Exactly right. And mind boggling that the engineers were willing to rely on one sensor for such a safety critical application.