Grin Baserunner Regen shutting down bike

[robotron2084]

I've been trying to get help from Grin on this but I'm thinking that there isn't a solution here. Its been a 3-month long saga and I'll try to focus on the current problem. I haven't heard from Grin in a week, which unfortunately seems par for the course. So I'm trying to see if maybe other people might have ideas.

I have their G-Line kit as a base system: https://ebikes.ca/other-kits/specialty-kits/brompton/brompton-gline-front-kit.html
I have the following modifications:

• I have built my own 36v battery, which I have used for months on another bike without issue.
• I don't need the throttle so I've removed it. I have determined that the problem occurs with or without the throttle.

The problem is the bike shuts down during regen, and I have to pull over, turn it off, wait, then reboot, then repeat until it works.

• The bike tapers off power up to 20mph, and about 22mph starts regen. Once it hits a higher speed it will dump > 700w into my battery and 'something' shuts down. No setting changes modify this behavior.
• I am not touching the pedals when this happens. I live in a hilly area and will naturally end up coasting up to 25-30mph on downhills.
• This happens regardless of the charge level of the battery. There was a theory this might be related to the bike being fully charged, that's been debunked.
• I have removed the throttle, but it happens with or without the throttle attached.
• This is essentially the stock brompton g-line kit, with my own 36v battery (which works great on other bikes).

I have tried every potential configuration I can to resolve this, including:

• Max regen phase current set to 10A/1A/0A
• Max Regen Battery Current set to 10a
• Turning off backpedal regen.
• I have tried setting the Max Regen Voltages as mentioned, nope.
• I even resorted to AI, which gave me a bunch of useless ideas.

I have attached the most current settings I have on the bike.

On my other Grin bike, which has a CycleAnalyst, I have the ability in the CA to limit the current and control the speed at which regen occurs, and whether it was on. It seems like what I'm seeing here is that there regen is just on, there are no adjustable settings, no way to turn it off. I don't know, maybe I'm looking in the wrong place? This seems like a huge missing feature if that is the case, and I'd imagine there would be a lot of people complaining. Please help!

Thanks,
Chris

 

[highlandr]

Do you have the ability to try a different battery? What BMS did you use for your home made unit?

 

[robotron2084]

Do you have the ability to try a different battery? What BMS did you use for your home made unit?

I could try a different battery, I do have a 48v battery.

I have this Daly BMS 10s 20A.
https://www.ebay.com/itm/127344674101

And I realize that the problem is very likely that the power is going over the 20A limit ( 36 * 20 = 720 ). Its more that it seems that there is no setting to control the output. That does make me think that maybe the cheapest option here might be to buy a BMS with a higher amperage if that won't cause a fire.

 

[highlandr]

I could try a different battery, I do have a 48v battery.

I have this Daly BMS 10s 20A.
https://www.ebay.com/itm/127344674101

And I realize that the problem is very likely that the power is going over the 20A limit ( 36 * 20 = 720 ). Its more that it seems that there is no setting to control the output. That does make me think that maybe the cheapest option here might be to buy a BMS with a higher amperage if that won't cause a fire.

I would bet that your BMS is the culprit. The more I dig into e-bike problems the more I realize why companies want to lock users into proprietary hardware, a lot to go wrong in the DIY scene. Feel bad for Grin customer service

Try that 48v battery and see what happens, hopefully you’ll isolate the battery as the issue. What cells did you use in your DIY battery?

 

[Alex12345]

Max Regen Battery Current set to 10a

Are you sure this BMS can accept charging at 10A (since the problem is triggered by regen)?

 

[robotron2084]

I used Samsung 35Es.

I don't see this as a battery problem. I see this as a motor controller problem, because there are no settings to control the regen or turn it off. I have a CycleAnalyst that had this feature, and I use that on the bike with the 48v battery, which also shut off once the watts got too high. So I'm likely not going to go through the effort to swap out the battery because the other battery will likely do the same thing ( i didn't make that battery, its from Affordable Ebike and very solid).

I feel bad for Grin's customers, not their customer service. The product seems broken and a step back from their CycleAnalyst based systems. They talk about creating an open system, but aren't hitting the basic minimum. I'm not asking for a lot here, just some basic settings to control regen.

 

[robotron2084]

Are you sure this BMS can accept charging at 10A (since the problem is triggered by regen)?

DALY BMS Li-ion 10S 36V 13S 14S 48V 16S 17S 60V LFP 4S 8S 16S 24V BMS

bmsdaly.com

Yes, you can see that it can take 20A. But again, I don't see this as a battery problem.

 

[robotron2084]

But I did buy a 40A version of the Daly BMS just now to possibly route around the motor controller problem!

 

[Alex12345]

The bike tapers off power up to 20mph, and about 22mph starts regen. Once it hits a higher speed it will dump > 700w into my battery and 'something' shuts down. No setting changes modify this behavior

So it only happens when you going well over the assist speed limit and try to either pedal or throttle? It doesn't happen when you say going 19 mph and engage the regen?

 

[robotron2084]

I don't have a throttle, and I am not engaging regen on purpose. I am coasting downhill ( i live in the foothills of mountains and so any direction I go is downhill or uphill) and as I increase in speed without touching the pedals the regen kicks in and I don't want it to.

Edit: I do have backpedal regen on, and that correctly limits current to 10a if i backpedal. But this isn't related to when it shuts down.

 

[Alex12345]

Right yeah, it's just a very specific situation, that's why I'm wondering. I don't get any uncommanded regen above the speed limit while coasting with my phaserunner or baserunner (at least coasting at 25mph with a limit of 20mph). I wonder if there's some parameter that controls this.

 

[robotron2084]

Right yeah, it's just a very specific situation, that's why I'm wondering. I don't get any uncommanded regen above the speed limit while coasting with my phaserunner or baserunner (at least coasting at 25mph with a limit of 20mph). I wonder if there's some parameter that controls this.

Thank you for confirming on your baserunner that you're not seeing this, that's good to know. Are you using the new superharness system or a CycleAnalyst?

I did try to look into the xml file to try to find settings that might control this, but there's a lot in there and I couldn't see anything that looked like 'beginRegenSpeed' or some such. I also put Claude on it without any luck, I think its too niche for it to give useful info.

 

[99t4]

Please tell me if I have this right:

• Regen engages automatically at higher speeds
• Regen causes the system to turn off
• No way to disable regen

Can you identify which component shuts the system down? This behavior looks suspiciously like when the BMS shuts down when sensing LVC under load, and needs a reset to resume.

I don't see this as a battery problem.

Maybe not, but swapping in the other battery might help us better isolate the problem.

 

[robotron2084]

Please tell me if I have this right:

• Regen engages automatically at higher speeds
• Regen causes the system to turn off
• No way to disable regen

Can you identify which component shuts the system down? This behavior looks suspiciously like when the BMS shuts down when sensing LVC under load, and needs a reset to resume.


Maybe not, but swapping in the other battery might help us better isolate the problem.

You nailed it. I cannot easily tell which part is causing the problem but I was getting an error in the phaserunner software that I sent to Grin way back of:

Warnings[8]: Max Regen Voltage Rollback

I understand that you want to isolate this to see if this is a battery related issue, but it shuts down at around ~700w and this is a 20A bms so its very likely the battery is shutting down. But does that mean its a battery issue? If a motor controller sends too much current in either direction that feels like the root cause to me. I think we're overindexing into the battery because its not a 'blessed' grin battery when its not the real issue, which you've outlined above.

Regardless I will try to pull my 48v battery off my cargo bike and flash the baserunner and see what happens.

 

[Alex12345]

Warnings[8]: Max Regen Voltage Rollback

Hm, yeah interesting, I know with my phaserunner I do get occasional situations in winter, after my bike parked at work with the battery on it, my EM3EV BMS would not allow regen at below freezing. So I can see regen very briefly engage, BMS cuts off, but doesn't fully shut the system down (lights, CA3 etc still work), voltage reaches the max regen 55V threshold as seen on CA3 display, I loose the throttle, but the system doesn't just die (and then I just disable regen and ride home peacefully). So at least my phaserunner can deal with rejected regen and doesn't overvolt or otherwise trigger a fault requiring a restart.

 

[Alex12345]

Also, I ran a diff command on your xml against my baserunner and phaserunner xmls, I don't see anything super different. Motor parameters and current regulator PLL etc settings I assume you got from Grin's presets. And I don't have any backpedal stuff in case that relates in any way to your problem. I see both my controllers have regen torque release ramp at 50ms, but yours is 500ms. Also I have parameter 77 sensor start and sensorless run, which means at high speeds it's sensorless, but you have the 77 set to 0 (sensored). Maybe all of this is unimportant to your issue, idk. But if the battery is suspect at cutting out everything, I would simply try lower the regen battery current (currently set at 10A), not the regen phase current (you also have this at 10A in the shared xml).

 

[lnanek]

A separate port BMS instead of a common port BMS might help. I ran into the same issue. If I go fast downhill my baserunner controller sometimes dies and shows an over voltage fault when I use a battery blender or DC-DC boost converter. I fixed it by always having one battery attached directly and just feeding that battery additional charge from the other one through a diode. That way the controller always has a place to dump back EMF.

Similarly, a separate port BMS won't reject any amount of charge from the controller, only from the charger. A BMS has no way to limit current. It can just turn it on or off. And my batteries' BMS need me to turn a key off and on to turn them back on after tripping them. So having the BMS trip in a situation where the controller is just trying to dump surges isn't ideal.

Following the link to your BMS I just saw a B- and P- pad, not B-, P-, and C-. So I think it is common port, not separate port.

I do have a regen clamp laying around that would likely work as well:

ODrive Regen Clamp

The ODrive Regen Clamp provides a high-performance braking energy management solution for ODrives without an external brake resistor driver. It automatically detects power regeneration from downstream motor drives, and diverts excess energy to an external brake resistor. Compatible and shown...

shop.odriverobotics.com

But haven't tried hooking it up yet. There's also similar devices that can be hooked up in parallel with the battery that are easier to wire:

In Stock MTBR1812 Rheostatic Brake 10S Lipo Voltage 5A Max Current Pro

If you want to use other cheaper but slower shipping method, please contact sales11@maytech.cn Shipping time by DHL or UPS is fast, normally 3-7 days arrive to you. Maytech MTBR1812 Rheostatic Brake 10S Lipo Voltage 5A Max Current Protect ESCs Specification: Model No. MTBR1812 Proudct Name...

maytech.cn

 

[robotron2084]

Just a quick report to say that I'm able to repro with my other 48v battery. I wanted to wait until it was cooler in the morning here to ride. Pretty much what I expected: I go down a hill, accelerate to 25-30 mph and it cuts out. Pretty much the same wattage, maybe 100w higher?

 

[robotron2084]

Also, I ran a diff command on your xml against my baserunner and phaserunner xmls, I don't see anything super different. Motor parameters and current regulator PLL etc settings I assume you got from Grin's presets. And I don't have any backpedal stuff in case that relates in any way to your problem. I see both my controllers have regen torque release ramp at 50ms, but yours is 500ms. Also I have parameter 77 sensor start and sensorless run, which means at high speeds it's sensorless, but you have the 77 set to 0 (sensored). Maybe all of this is unimportant to your issue, idk. But if the battery is suspect at cutting out everything, I would simply try lower the regen battery current (currently set at 10A), not the regen phase current (you also have this at 10A in the shared xml).

I shall try to change parameter 77 and see if that makes a difference, thanks for checking that out, I haven't tried changing that. Who knows? I've already tried changing the regen battery current and phase to 10/1/0a and it makes no difference. Backpedal settings also make no difference. And correct, this file is based off the base settings from grin.

 

[robotron2084]

A separate port BMS instead of a common port BMS might help. I ran into the same issue. If I go fast downhill my baserunner controller sometimes dies and shows an over voltage fault when I use a battery blender or DC-DC boost converter. I fixed it by always having one battery attached directly and just feeding that battery additional charge from the other one through a diode. That way the controller always has a place to dump back EMF.

Similarly, a separate port BMS won't reject any amount of charge from the controller, only from the charger. A BMS has no way to limit current. It can just turn it on or off. And my batteries' BMS need me to turn a key off and on to turn them back on after tripping them. So having the BMS trip in a situation where the controller is just trying to dump surges isn't ideal.

Following the link to your BMS I just saw a B- and P- pad, not B-, P-, and C-. So I think it is common port, not separate port.

I do have a regen clamp laying around that would likely work as well:

ODrive Regen Clamp

The ODrive Regen Clamp provides a high-performance braking energy management solution for ODrives without an external brake resistor driver. It automatically detects power regeneration from downstream motor drives, and diverts excess energy to an external brake resistor. Compatible and shown...

shop.odriverobotics.com

But haven't tried hooking it up yet. There's also similar devices that can be hooked up in parallel with the battery that are easier to wire:

In Stock MTBR1812 Rheostatic Brake 10S Lipo Voltage 5A Max Current Pro

If you want to use other cheaper but slower shipping method, please contact sales11@maytech.cn Shipping time by DHL or UPS is fast, normally 3-7 days arrive to you. Maytech MTBR1812 Rheostatic Brake 10S Lipo Voltage 5A Max Current Protect ESCs Specification: Model No. MTBR1812 Proudct Name...

maytech.cn

The regen clamp comes with 100w of resistance, I'd imagine that I'd need 1KW...at $100 plus resistors plus whatever else I need (new bms with separate port?) it makes me wonder if i'm throwing good many after bad trying to make this work.

I had never heard of a lot of this, lots of cool information, thank you! Yes, my BMS is a common port bms. I may look into these items as a potential option, and weigh them against my current options of a different kit or a higher amperage BMS (which has been shipped and I'll get it soon).

 

[99t4]

I think we're overindexing into the battery because its not a 'blessed' grin battery when its not the real issue, which you've outlined above.

Well, my thinking was that if it was easy enough to swap in another battery, just do that in an effort to troubleshoot at an upper level to hopefully clearly identify where the problem lies. I'm not caring if it is a

'blessed' grin battery

or not. Sometimes we can get so wrapped up in the minutiae (obscure settings parameters) that we miss a more simpler explanation.

But nvm, I see you swapped in an entirely different battery with no change in the behavior. So you were and are correct, the problem is not the battery or its BMS.

Keep calm and carry on.