[SOLVED] Need help please: intermittent problem with controller/display

[squarewheels]

I would be SO GRATEFUL if someone can point me into the right direction, finding a nasty intermittent problem with my Ebike system.

Controller: minshine.cn type OEM controller, FOC type
Display: T24-V2 minshine
battery: home-made LiFePo4 , 41V
BMS: Daly Lifepo4 12S 36V 30A
Motor: OEM, high quality 250W motor.

My first system with cheaper controller/motor worked great for 9 years(!)

But after a upgrade with new components, the problems started: intermittent complete faillure of the system. (would not turn on anymore: no display)

I already bought another controller from a local E-bike shop who branded the minshine controller with their own brand name.

Now the system WILL turn on but the display indicates there is no battery power.
- measuring the battery voltage, when no power is indicated, turns out OK, so no problem there.
- when I change the controller (I have 2 as I said) will often solve the problem until suddenly the system will go dead again (after 10 minutes or so)
- motor is OK, I have the same problem with our without motor connected

DONE SO FAR:
- replaced display wiring
- checked all wiring with DMM
- checked wiring controller at least 1001 times
- exchanged controller
- checked for solder bridges inside the controller

My plan is to buy again, for the 3rd time, a new system but I would be much happier if I could find the problem!!
My wife needs the Ebike so any help would be great!

 

[jetpackjbd]

My first guess is still the battery, even if it reads fine. Check the voltage while it’s in use, it might display full voltage but dip significantly under load. If it’s a 9 year old battery it’s probably dead, but maybe not being LiFePo4. Strange that it would die without the motor connected though, your controllers would have to be very inefficient for them to drain the battery. Have you tried replacing the BMS? It could also be a messed up thermal switch or breaker inside the pack.
It’s possible that you’ve gotten unlucky twice with a controller, unlikely but not unheard of.

 

[docw009]

It's likely the battery. You're going to have to determine if it is an unbalanced battery shutting off because of a low voltage cell, or if the controller had detected a low battery voltage due to voltage sag.

Is it a smart BMS that shows all the cell voltages on a phone App? Are they all looking pretty uniform and not close to triggering LVC? What's the overall battery voltage when it shuts off?

I've had a 10S Li-ion pack read 40.8 volts and shut off because one cell group was just a few tenths of a bolt above the low voltage cutoff of 3V, while the rest were close to full charge, I didn't have a smart BMS, but had to take the battery apart to measure the group voltages.

 

[squarewheels]

My first guess is still the battery, even if it reads fine. Check the voltage while it’s in use, it might display full voltage but dip significantly under load. If it’s a 9 year old battery it’s probably dead, but maybe not being LiFePo4. Strange that it would die without the motor connected though, your controllers would have to be very inefficient for them to drain the battery. Have you tried replacing the BMS? It could also be a messed up thermal switch or breaker inside the pack.
It’s possible that you’ve gotten unlucky twice with a controller, unlikely but not unheard of.

Thank you for your reply!

The battery is from may 2022, sorry that I was not clear about that...

• 12S LiFePO4, max 42 Volt. (48 cells: JGNE 26650 3200mAh - 9.6A LIFEPO4 - 3.2V)

I measure the battery while the motor was running and the voltage dropped 0,5V. The frustrating bit is the intermittent fault, showing itself within minutes so from working to dead and back in 10 minutes or so??

 

[squarewheels]

It's likely the battery. You're going to have to determine if it is an unbalanced battery shutting off because of a low voltage cell, or if the controller had detected a low battery voltage due to voltage sag.

Is it a smart BMS that shows all the cell voltages on a phone App? Are they all looking pretty uniform and not close to triggering LVC? What's the overall battery voltage when it shuts off?

I've had a 10S Li-ion pack read 40.8 volts and shut off because one cell group was just a few tenths of a bolt above the low voltage cutoff of 3V, while the rest were close to full charge, I didn't have a smart BMS, but had to take the battery apart to measure the group voltages.

I though of the battery also, but I have checked and measured under load but I see nothing out of the ordinary? I will attach screenshots I just made...

AFAIK it's all OK??

 

[jetpackjbd]

I though of the battery also, but I have checked and measured under load but I see nothing out of the ordinary? I will attach screenshots I just made...

AFAIK it's all OK??

Hmm, the battery seems alright and the BMS appears to be reading everything fine. Weird. I hate to say it but you might want to buy a used eBike with missing parts and swap your old parts into it if you or someone else on the forum can’t figure it out.

 

[amberwolf]

If this controller is the type that autodetects which battery it is connected to, then it is probably detecting a full 36v battery as an empty 48v battery. It's not uncommon.

The only fix I've seen is to discharge the battery some so it is below the threshold the controller uses to detect a 48v battery, and then it will see it as a 36v.


If this works, then since your BMS is probably programmable you could set the HVC for charging (and the balancing start/end points) full to a lower voltage, so this doesn't happen in the future.


If it's not an autodetecting controller, it could just be whatever circuit in the controller is used to detect the battery voltage has a poor connection to the battery positive line, either in the traces to it, or the resistors used as a voltage divider, or the op-amp used as a buffer to the MCU, etc. (it's more likley to be int eh controller than the display)

If you've had this problem with a previous version of this controller, you might try a different brand/model of controller/display.

 

[docw009]

Your BMS appears to me to be properly be set for 12S LiFEPO4. Cells are balanced. It's only seen 33 cycles?

I like amberwolf's thought. The battery is sitting at 39V right now. That would be cutoff if your controller thinks it is working with a 48V Li-Ion battery where LVC is 40V,

I use dual voltage controllers too, and since 10S and 13S Li-ion packs are 30-42V and 40-54V respectively, the only time they get confused if if I switch from a mostly discharged 48V pack to a fresh 36V pack. Then it thinks I still have 48V, I have to remove the battery and power up the controller so it bleeds residual charge off the input before connecting the 36V.

 

[squarewheels]

If this controller is the type that autodetects which battery it is connected to, then it is probably detecting a full 36v battery as an empty 48v battery. It's not uncommon.

The only fix I've seen is to discharge the battery some so it is below the threshold the controller uses to detect a 48v battery, and then it will see it as a 36v.


If this works, then since your BMS is probably programmable you could set the HVC for charging (and the balancing start/end points) full to a lower voltage, so this doesn't happen in the future.


If it's not an autodetecting controller, it could just be whatever circuit in the controller is used to detect the battery voltage has a poor connection to the battery positive line, either in the traces to it, or the resistors used as a voltage divider, or the op-amp used as a buffer to the MCU, etc. (it's more likley to be int eh controller than the display)

If you've had this problem with a previous version of this controller, you might try a different brand/model of controller/display.

Your BMS appears to me to be properly be set for 12S LiFEPO4. Cells are balanced. It's only seen 33 cycles?

I like amberwolf's thought. The battery is sitting at 39V right now. That would be cutoff if your controller thinks it is working with a 48V Li-Ion battery where LVC is 40V,

I use dual voltage controllers too, and since 10S and 13S Li-ion packs are 30-42V and 40-54V respectively, the only time they get confused if if I switch from a mostly discharged 48V pack to a fresh 36V pack. Then it thinks I still have 48V, I have to remove the battery and power up the controller so it bleeds residual charge off the input before connecting the 36V.

Thank you both very much for a new insight to me. I did ask the vendor if for a controller that would support a LiFePO4 battery but maybe it IS a problem.

The only frustrating thing (again) is: the 2nd replacement controller did better and "only" shut off completely 1-3 times every ride of 6km. Then gradually in a few weeks, the problem disappeared and the system did OK for 5 months. Until lately suddenly the system failed again but now NOT shutting OFF completrely (Blabnk display also) but only displaying an empty battery and also no motor power of course. Sigh...

So the only thing I have left is buy a new controller I think, WITH a new (expensive) display.
Throwing 380 euro's down the drain is sad but <beep> happens.

But I would like a controller over which I have control and not the vendor who has to program it?
I need to dive deep into this forum looking for a suitable controller I guess (European market)

Many thanks, so great to meet people willing to help here!

 

[amberwolf]

Thank you both very much for a new insight to me. I did ask the vendor if for a controller that would support a LiFePO4 battery but maybe it IS a problem.

It doesn't matter what chemistry the battery is--the controller doesn't know, or care.

It's just the voltage range the battery itself covers, vs the controller's HVC and LVC.

The only frustrating thing (again) is: the 2nd replacement controller did better and "only" shut off completely 1-3 times every ride of 6km. Then gradually in a few weeks, the problem disappeared and the system did OK for 5 months. Until lately suddenly the system failed again but now NOT shutting OFF completrely (Blabnk display also) but only displaying an empty battery and also no motor power of course. Sigh...

If a system shuts off completely, no display, no power, no lights, etc, that's almost always a battery issue--too high a current draw from the controller at too low a battery voltage (or a cell in the battery too low, unbalanced battery pack) causing the BMS to protect the cells from damage by turning off the output.

If the battery voltage is still correct ( just measuring it with a multimeter at the controller/battery connection point without unplugging anything as that would reset a battery that shut off for excessive load) but the controller won't turn back on, it's not the battery, but could be a connection fault, anywhere between the battery and controller (including crappy fuseholders, spread contacts in connectors, etc).


If a system shuts off assist but does not actually turn off, it's not usually the battery, unless it simply sags in voltage so much that it drops below the system LVC, under a high enough load (current draw).




We can help you find a new controller/display, if you specify the features you are looking for, the things you still have on the bike it must connect to (throttle, PAS, brake levers, etc), and the voltage the controller must operate within, and the power it must be able to supply to the motor, and the max current your battery is capable of supplying to the controller.

 

[squarewheels]

It doesn't matter what chemistry the battery is--the controller doesn't know, or care.

It's just the voltage range the battery itself covers, vs the controller's HVC and LVC.





If a system shuts off completely, no display, no power, no lights, etc, that's almost always a battery issue--too high a current draw from the controller at too low a battery voltage (or a cell in the battery too low, unbalanced battery pack) causing the BMS to protect the cells from damage by turning off the output.

If the battery voltage is still correct ( just measuring it with a multimeter at the controller/battery connection point without unplugging anything as that would reset a battery that shut off for excessive load) but the controller won't turn back on, it's not the battery, but could be a connection fault, anywhere between the battery and controller (including crappy fuseholders, spread contacts in connectors, etc).


If a system shuts off assist but does not actually turn off, it's not usually the battery, unless it simply sags in voltage so much that it drops below the system LVC, under a high enough load (current draw).




We can help you find a new controller/display, if you specify the features you are looking for, the things you still have on the bike it must connect to (throttle, PAS, brake levers, etc), and the voltage the controller must operate within, and the power it must be able to supply to the motor, and the max current your battery is capable of supplying to the controller.

Thank you so much again for your help!

After endless testing the past year and looking for a fault, I now want to buy another controller and would be so glad if you could point me in the right direction.

Personally, I would like to have control and be able to change settings if needed so I thought of a KT controller with display?
I added some photo's below.

This Ebike is for normal use, no high speed etc. Max speed allowed here is 25km/h (15.53 Mph)
We live below sea level in The Netherlands so it's flat like a pancake here.

What I have and I think could use:

- Motor:
custom branded one, "premium type"
Brushless,
9 pole connector,
250W / 25A,
torque 40-50nm,
Sinewave or FOC,
HALL + SPEED sensors
RPM with 28" rim: 225RPM (16km/h = 9.94 Mph)

- Battery:
48 cells, 4P-12S LiFePO4,
12.8Ah,
38.4 Volts NOMINAL
LVC: 31.2V - HVC: 43.8V (set in BMS)

PAS sensor:
currently 10 magnet type

Brake sensors:
Hall sensor type

Throttle:
Thumb- hall sensor type,
ONLY FOR WALKING SPEED due to regulations!

What I don'k know is if a 36V/48V controller would give trouble with the voltage range my battery is putting out, like one of the comments in this trhead pointed out...

As said, I myself was thinking about a KT controller because if it still is not working then I won't loose too much money.
Another problem in The Netherlands is, that we have very little dealers of such products, so I can't go to a shop for advise.

Well, many thanks for pointing me in the right direction!


------

I made my LiFePO4 battery according to this info:


almost new motor:


brake sensors

 

[docw009]

Discharge the battery using an appliance til it's under 42V. Then ride the bike to test Amberwolf's suggestion. If it now works, you can try adjusting your Daly BMS.

Can you lower the "cell high voltage protect" parameter from 3.65V to 3.50V? Also the "sum high voltage protect" from 43.8 to 42 volts? Check it with a meter to make sure this is working when you recharge the battery,

 

[jetpackjbd]

Thank you so much again for your help!

After endless testing the past year and looking for a fault, I now want to buy another controller and would be so glad if you could point me in the right direction.

Personally, I would like to have control and be able to change settings if needed so I thought of a KT controller with display?
I added some photo's below.

This Ebike is for normal use, no high speed etc. Max speed allowed here is 25km/h (15.53 Mph)
We live below sea level in The Netherlands so it's flat like a pancake here.

What I have and I think could use:

- Motor:
custom branded one, "premium type"
Brushless,
9 pole connector,
250W / 25A,
torque 40-50nm,
Sinewave or FOC,
HALL + SPEED sensors
RPM with 28" rim: 225RPM (16km/h = 9.94 Mph)

- Battery:
48 cells, 4P-12S LiFePO4,
12.8Ah,
38.4 Volts NOMINAL
LVC: 31.2V - HVC: 43.8V (set in BMS)

PAS sensor:
currently 10 magnet type

Brake sensors:
Hall sensor type

Throttle:
Thumb- hall sensor type,
ONLY FOR WALKING SPEED due to regulations!

What I don'k know is if a 36V/48V controller would give trouble with the voltage range my battery is putting out, like one of the comments in this trhead pointed out...

As said, I myself was thinking about a KT controller because if it still is not working then I won't loose too much money.
Another problem in The Netherlands is, that we have very little dealers of such products, so I can't go to a shop for advise.

Well, many thanks for pointing me in the right direction!


------

I made my LiFePO4 battery according to this info:
View attachment 342910

almost new motor:
View attachment 342911View attachment 342912View attachment 342913View attachment 342914

brake sensors
View attachment 342915

Definitely don’t go for the same brand again. It’s weird if your battery/BMS wouldn’t be able to handle the current draw since it’s only a 250w motor, I suggest you go over spec (capable of more amps than necessary) on the next BMS if that is the issue here.

 

[squarewheels]

Definitely don’t go for the same brand again. It’s weird if your battery/BMS wouldn’t be able to handle the current draw since it’s only a 250w motor, I suggest you go over spec (capable of more amps than necessary) on the next BMS if that is the issue here.

No, I won't touch the same controller brand again... The BMS worked fine all the time, is capable of 30A discharge current and when I had the controller/display failures, I also tested the system with a bench PS, to rule out battery issues but to no avail.

 

[squarewheels]

Discharge the battery using an appliance til it's under 42V. Then ride the bike to test Amberwolf's suggestion. If it now works, you can try adjusting your Daly BMS.

Can you lower the "cell high voltage protect" parameter from 3.65V to 3.50V? Also the "sum high voltage protect" from 43.8 to 42 volts? Check it with a meter to make sure this is working when you recharge the battery,

Thanks for your suggestions! But I had the strange intermittent problems with a battery voltage of for example 39V also. Testing with a bench PS set to 36V also gave the same issues. Then suddenly the next morning a working controller but when riding a day or 2 dead again. etc etc etc. So the controller and 150 euro's display will be promoted to a trophy on my wall...

I have never ever had such stupid problems with electronics before!

 

[docw009]

If you believe a new controller is needed, here is a 15A KT model from topbikekit in China. I like the LCD3 display because they're easier to see in sunlight, You will need to add a 9 pin motor extension cable (first option) of the right length, and the IT4 harness that goes to the display. If you can solder, than you can adapt your current brake switches. Strongly suggest the throttle, even if they are illegal in your country. Useful for trouble shooting. Also buy the PAS sensor, because it will fit, and save you the trouble of rewiring the old one,. If you want more power, the 22A model is available. There are lots of other vendors on aliexpress, but I've bought controllers from this seller. His shipping charges are high though.

For example, the above seller doesn't carry this small 25A KT controller which I have on one of my bikes. But then you would have to find the motor cable, display, and other accessories on your own. But for myself, this is a better price.

 

[squarewheels]

If you believe a new controller is needed, here is a 15A KT model from topbikekit in China. I like the LCD3 display because they're easier to see in sunlight, You will need to add a 9 pin motor extension cable (first option) of the right length, and the IT4 harness that goes to the display. If you can solder, than you can adapt your current brake switches. Strongly suggest the throttle, even if they are illegal in your country. Useful for trouble shooting. Also buy the PAS sensor, because it will fit, and save you the trouble of rewiring the old one,. If you want more power, the 22A model is available. There are lots of other vendors on aliexpress, but I've bought controllers from this seller. His shipping charges are high though.

For example, the above seller doesn't carry this small 25A KT controller which I have on one of my bikes. But then you would have to find the motor cable, display, and other accessories on your own. But for myself, this is a better price.

Thank you very much, I have ordered the system and I will report back here if the problems are solved...

 

[squarewheels]

IT WORKS!

First of all, the controller/display I ordered solved the problem. For 90 euro's I have a working system now!

I did not know how to set the P1 setting but on YouTube, I saw a video with the tip on how to count the number of magnets in your motor by connecting 2 phase wires of the motor. By turning the wheel, you feel each magnet, and that's how I could count them.

I ended up with 240 steps, but there is also a gear reduction in the motor, so I'm not sure if that's the exact number of magnets.

Anyway, I set P1 to 240, and now everything works perfectly.

Many thanks to everyone here who took the time to help me; I am very grateful!

 

[docw009]

P1 is (the number of magnets) x (gear ratio). Your count includes the gears turning the motor. If you're getting the right speed, then that confirms P1 is right.

That's a interesting way to count magnets. I will have to try it.

 

[squarewheels]

P1 is (the number of magnets) x (gear ratio). Your count includes the gears turning the motor. If you're getting the right speed, then that confirms P1 is right.

That's a interesting way to count magnets. I will have to try it.

Thank you! so count is including the gears... That makes sense so now I now for sure the setting is OK.

I first tried P1=202 before 'counting' and then the walk assist speed was a bit off. With P1=240, walk assist speed is spot on!

 

[calab]

How often are you using walk assist?

 

[squarewheels]

How often are you using walk assist?

NEVER...

I only used it to check the speed while trying out the settings.