Just can't get regen braking working after all this time

[RooGM]

Hi all!

I've struggled with this for 8 months and thought it was time to ask for help.  Since I got my Magic Pie 4 I've never been able to get regen braking working on 48v.  It worked perfectly on 36v.  Now I have a new 48v battery setup it's time to fix this issue.  I have tried every possible setting I can think of in the controller with no luck.  Here's a table of what's happening with a few different configuration:

Battery: New 14S 5000mAh 35C Lipo (also tried with LifePo4)
Standard 48v settings with the following changes for regen:

Decel level: 65
Max EBS Phase Current: 30
Outcome: Motor actually speeds up momentarily then cuts out when regen brake is activated.  When testing without load the motor speeds up indefinitely.

Decel level: 65
Max Phase current: 70
Outcome: Regen kicks in for about 0.5 seconds, then fades smoothly to completely inactive over the next 1 second

Decel level: 100
Max Phase Current: 70
Outcome: Regen works for as long as it's being depressed but it's way too powerful to be useful/safe. 

Having decel set to 100 is the only condition in which I can get it to work, but as I said, the braking force is too hard and sudden to be useful.

If you can think of something I might be doing wrong please let me know.  Also, I don't know if this could be the reason but I'm using a 36v throttle and the LEDS flicker and are warm when running on 48v due to overload.

[Bikemad]

I am puzzled by your comment "Having decel set to 100 is the only condition in which I can get it to work" as I seem to recall that the Decel level 1 (rpm/s) setting (which will accept values between 15 and 380) only affected the maximum motor speed in reverse (when wired, enabled and selected) not the regenerative braking.

The EBS phase Current (A) parameter (which accepts values between 20A min and 80A max) is the setting that should determine the amount of regenerative braking (when the brakes are applied with regen enabled).

If you try reducing the Max EBS phase Current (A) to 60A, 50A or 40A etc. it should reduce the regenerative braking effect accordingly.

The throttle LEDs should not have any effect on the regenerative braking whatsoever, but if you are not 100% convinced, feel free to unplug the throttle and try the regen again to rule out the LEDs completely.

Alan
 

[RooGM]

Thanks for replying, it really is a puzzle to me too.  Adjusting Decel level 1 is what strengthens or weakens the regen braking power when it does work on the 36v battery.  For instance, setting it to 20 is too weak, setting it to 100 is way too strong, so I settled on 65 and that felt good for me.  I have pasted a screenie of my settings but they look pretty standard.  I also note that the battery I'm using is a simple 14S lipo made up of Hobbyking cells with no BMS or diode.  I must be doing something stupid that I'm not aware of.  Surely a light bulb moment is around the corner.

I also wonder if it's the fact that my battery is more like 58v than 48v as it's 14S.  But I'm not sure how that would cause the regen to be faulty when it works fine every time on 10S Lipo packs.

And then there's the weird Decel level 1 setting.  What should that be set to and how come it affects my regen braking percentage?  I sense we are close to some clues.

[Bikemad]

I have just tried programming a spare controller with the same settings as yours and here are the screenshots taken from two different program versions which have loaded the parameter settings from this controller:



Notice that the Decel level 1 (rpm/s) value of 100 remains the same for both programs, but the description does not.

Try using the later PI-800 program (version 3.1.3) and see if it makes any difference.
(Copy PI-800.exe and text.ini to the same folder/directory on your computer before you run it)

I expect that the regen behaviour will still be exactly the same, but it would at least be sensible to try it.

The battery I'm using is a simple 14S lipo made up of Hobbyking cells with no BMS or diode.

I also wonder if it's the fact that my battery is more like 58v than 48v as it's 14S.  But I'm not sure how that would cause the regen to be faulty when it works fine every time on 10S Lipo packs.

The 14S LiPo pack should not be responsible for your regen problem as its fully charged voltage of 58.8V is virtually the same as the GM 48V LiFePO₄'s 58.4V, so the controller should still be operating well within the expected battery voltage range.

Alan
 

[RooGM]

Hi Alan,

Thanks so much for your time here.  The version you linked to is 3.2.1 which is even later.  I tried it but nothing changed.  I can't test on the road as we have thick snow today but testing indoors with the wheel off the ground gives the same results as before.  With the settings as shown in the screen capture, when I spin the motor, then let go and press the regen brake, the wheel actually speeds up.  This is what it has always done since I bought it when running 48v.  I think the likely explanation is that I received a faulty unit. 

What I find the most strange is that Decel level 1 affects the regen braking when you say it is only related to reverse speed and should have no effect on the regen braking.  What else I find strange is the choice of words for this software.  In the latest version it's still called "Decel level 1."  That doesn't make sense to me if it is supposed to relate to reverse speed only.  "Decel level 1" sounds like it's related to "deceleration" which, if it did relate to regen braking, would make some sense as a title.  On mine it does seem to be related to regen braking as can be shown when adjusting the number and testing the brake. 

I am wondering if I should send it in for repair under warranty or if I should just remove the regen brake level and run without.  I am concerned that even if I did have a working unit, repeatedly using regen braking would reduce the life span of my battery.  I imagine that the motor pumping sudden and inconsistent bursts of unbalanced and varying current back into the cells can't be that great for them, can it?  I always baby my batteries, only ever balance charging them and always at no more than 1C.  For my bike batteries I charge at 3.5A which is a lot less than 1C.  And yet my previous LifePo4 battery failed a lot sooner than I expected on my Magic Pie.  The same type of battery has lasted for hundreds of charges on my electric skateboard and is still going.

[Bikemad]

The version you linked to is 3.2.1 which is even later.  I tried it but nothing changed.

Oops, I should have double checked it was the correct version before posting the link. 
To avoid any further confusion I have now modified my previous post so the previous link should now be correct.



Here are the links to Version 3.2.1 and Version 3.3.0 of the PI800.exe parameter setting program in case anyone else wants to try them.

What I find the most strange is that Decel level 1 affects the regen braking when you say it is only related to reverse speed and should have no effect on the regen braking.  What else I find strange is the choice of words for this software.  In the latest version it's still called "Decel level 1."  That doesn't make sense to me if it is supposed to relate to reverse speed only.  "Decel level 1" sounds like it's related to "deceleration" which, if it did relate to regen braking, would make some sense as a title.  On mine it does seem to be related to regen braking as can be shown when adjusting the number and testing the brake.

The earlier Version 3.1.3 of the PI800.exe program seems to be the only version that has that parameter labelled as Maximum Reverse Speed (rpm) instead of Decel level 1 (rpm/s).

I don't have a reverse switch wired into my motor harness so I am unable to test this out on my MP4, but it would be interesting to hear from other users who are using the reverse function to see what setting affects the reverse speed and what affects the regen.

I am wondering if I should send it in for repair under warranty or if I should just remove the regen brake level and run without.  I am concerned that even if I did have a working unit, repeatedly using regen braking would reduce the life span of my battery.  I imagine that the motor pumping sudden and inconsistent bursts of unbalanced and varying current back into the cells can't be that great for them, can it?  I always baby my batteries, only ever balance charging them and always at no more than 1C.  For my bike batteries I charge at 3.5A which is a lot less than 1C.  And yet my previous LifePo4 battery failed a lot sooner than I expected on my Magic Pie.  The same type of battery has lasted for hundreds of charges on my electric skateboard and is still going.

To ensure the battery is not overloaded during regen, the maximum charge rate for you battery should be higher than the current output from the Magic Pie during high speed regen.

As many of the LiFePo4 packs are only rated at 2C for charging, it might be a problem if you have a small capacity pack (e.g. 5000mAh) as this should only be subjected to 10A or less charging current, and I have already discovered that with a 29.6V 8S LiPo pack, the Magic Pie current output can be over 25A at 20mph.

If your LiPo/LiFePO₄ packs cannot accept the high regen current, your battery might be happier if the regen is disabled.

Alan
 

[RooGM]

Alan, thanks enormously once again. I have learnt something incredibly useful regarding the regen current.  My old LifePo4 battery was designed for a skateboard with regen but my new Hobbyking packs can only accept a charge rate of 2C so in this case I will have to do without regen altogether.  However, I have submitted a support ticket anyway because I would like my MP4 to be fixed should I ever want to use a larger battery that can accept the higher charge rate that regen offers. 

To be honest, apart from the very useful braking ability that regen offers, I am quite skeptical that it provides a real world benefit when it comes to topping the battery back up.  It's normally only used for a few seconds at a time and over the course of say, a 30 minute ride, the total regen engagement time probably doesn't add up to more than 2 or 3 minutes and depending on the terrain, often less. What real difference a 2-3 minute charge is going to make has got to be very slight.  I used to work in the electric skateboard industry and the whole "regenerative braking" thing was just a marketing point.  It's the simplest way to brake on a skateboard and it removes the need for actual brakes and that's why it's used.  However, it's pitched as being able to top up your battery when braking but in real life, at least on a skateboard, it really doesn't add any perceivable range. 

Thanks again for the support and advice; you are basically a one man support system for these products and I think you are a huge asset to GM.  I hope some day you are compensated for all your efforts within the community.  I personally would have been lost without the information you have shared on the forum.

All the best.