52v setup - KT36/48SVPRL + 1000w DD hub motor cutting out after 5 seconds

[Pug1]

Hi all,

I have built an a stand on scooter using a KT36/48SVPRL controller, a 1000w direct drive hub motor with hall sensors and a 52v 18ah battery with a JK BMS. The battery is all good and the motor works fine. I’ve tested the hall sensors and get clean 0v - 5v - 0v on all 3 lines. I know I’ve got the wiring correct as its quiet, uses hardly any power on the stand and has torque from a stand still. I’ve tried all 36 variations of the wiring combinations and there are 2 that don’t cut out after a few seconds but they have no torque and grumble when under load and the wiring gets hot.

The problem is that it cuts out after a few seconds no matter what amount of throttle I use. If I wire up a false positive the motor will spin indefinitely but uses more power on the stand.

It’s driving me daft now. I’m sure the wiring is correct because it’s the only variation that has torque from a stand still. The motor, battery, BMS and controller are all brand new. I bought all the stuff from EBike Masters in the uk to ensure I got official parts.

Has anyone got any advice/ideas what I can try next please. I know there is some open source fw from stancecoke but I can’t see how flashing that fw would fix my problem.

Thanks in advance everyone

 

[Alex12345]

Can you describe the cutout better? Does the motor stop but the system (the display for example) still works? Does it cutout at a specific RPM or RPM doesn't matter (even at crawling speed on the stand it still cuts out after some time)? What does it take to run after a cutout, do you have to power down the controller or merely reengage the throttle? Are you sure the BMS is not responsible for cutouts?

I can describe a situation I had with KT cutouts, it was due to bad hall sensor connector/pins. It was my first build and I mangled that connector while guessing hall combinations etc, so occasionally on bumps it would cutout but the system would still be on and I would have to let off and reengage the throttle to get power to the motor again. But no hall error was ever displayed, I think because the connection dropout was momentary. Replacing the connector and also straightening / fixing pins helped in that case.

 

[Pug1]

Thanks for your reply Alex
Doesn’t matter what speed/amount of throttle I apply, after about 5 seconds it cuts out. No warning/error on the screen. Screen doesn’t flicker or turn off, the system stays on. The motor just stops receiving power. I watch the volts on screen, they take less then a volt on the stand and nearly 3 I think under load. Regardless, after 5 seconds it cuts the power. The same amount of time every time. I can let go of the throttle and reapply it instantly and the motor will start moving again.

I’ve built a CNC and a 3d printer and managed to fix that no problem every time, but this blinding thing though properly kicking my ass lol

 

[Pug1]

sorry Alex, I forgot to mention I back probed the hall sensor plug to ensure the signals I got were nice and clean and could eliminate a bad connection/funky wire. It’s definitely not the hall wires as the pins are out of the connector at the moment and are individually connected to eliminate that as a problem as well lol the motor wires are all bullet connections as well and they are all nice and snug. However weird it sounds it feels like a software problem

 

[gromike]

A 48v controller powered by a fully charged 52v battery might be cutting out on high voltage.

 

[Alex12345]

A 48v controller powered by a fully charged 52v battery might be cutting out on high voltage.

That's actually a great point I somehow overlooked

KT controller with 52V battery

Hi, I'm looking for a new battery and considering a 52V one. I was using a 48V battery with a 48V sine wave 25A KT controller on 750w gear hub motor. I’ve read that it should work fine, but I want to ask if anyone is using this combo with the LCD8H display. Does it show the correct voltage on...

endless-sphere.com

I bet that's what happening if so precisely timed...

 

[Alex12345]

sorry Alex, I forgot to mention I back probed the hall sensor plug to ensure the signals I got were nice and clean and could eliminate a bad connection/funky wire. It’s definitely not the hall wires as the pins are out of the connector at the moment and are individually connected to eliminate that as a problem as well lol the motor wires are all bullet connections as well and they are all nice and snug. However weird it sounds it feels like a software problem

Seems like sometimes KTs can run with a 52v pack but other times they blow up

I burned my new 48V controller with a 52V battery

Hello ESFMs, I have got a overvoltage protection error on a new build. Do I need to buy another controller? If so, what do I have to look out for? Here are the facts: Bought this kit: Motor: 1000W geared rear hub motor Controller: Rated voltage 48V, Limit current 28A, Nominal current 14A...

endless-sphere.com

How to know how much battery volts the ebike controller can take?

How do I know how much battery volts the ebike controller can take? I mean how much above the voltage I could go on the battery above the controller's stated voltage? When I open the controller those battery kind of looking things has text on them saying something about 63volt. The text...

endless-sphere.com

Probably depends on components choice in each particular controller. Maybe yours somehow least tolerant to 52v. Does it behave weirdly if you discharge the pack to below 54V?

 

[E-HP]

Measure the voltage of the throttle signal wire while the bike is on the stand and confirm whether or not the throttle signal voltage is still present when the motor cuts out.

PS. I've never heard of a KT controller that implements HVC.

 

[99t4]

Moving thread to Stand-up EScooter/ESkateboards.

 

[docw009]

Measure the voltage of the throttle signal wire while the bike is on the stand and confirm whether or not the throttle signal voltage is still present when the motor cuts out.

PS. I've never heard of a KT controller that implements HVC.

I've run the little 20A six FET 36./48 model on 60V (from a DC-Dc booster) for miles.

Could be a weak cell in the battery. I would try a another battery if I had one, or try that battery in another ebike,

 

[Pug1]

A 48v controller powered by a fully charged 52v battery might be cutting out on high voltage.

But wouldn’t that happen on a false positive as well? It only happens on this combination of phase and hall wires and the combination in the opposite direction.

Measure the voltage of the throttle signal wire while the bike is on the stand and confirm whether or not the throttle signal voltage is still present when the motor cuts out.

PS. I've never heard of a KT controller that implements HVC.

Thanks I will test when I get back home shortly and update the thread.

Seems like sometimes KTs can run with a 52v pack but other times they blow up

I burned my new 48V controller with a 52V battery

Hello ESFMs, I have got a overvoltage protection error on a new build. Do I need to buy another controller? If so, what do I have to look out for? Here are the facts: Bought this kit: Motor: 1000W geared rear hub motor Controller: Rated voltage 48V, Limit current 28A, Nominal current 14A...

endless-sphere.com

How to know how much battery volts the ebike controller can take?

How do I know how much battery volts the ebike controller can take? I mean how much above the voltage I could go on the battery above the controller's stated voltage? When I open the controller those battery kind of looking things has text on them saying something about 63volt. The text...

endless-sphere.com

Probably depends on components choice in each particular controller. Maybe yours somehow least tolerant to 52v. Does it behave weirdly if you discharge the pack to below 54V?

is it the caps that are the problem? How would I rectify the problems with this controller?

I've run the little 20A six FET 36./48 model on 60V (from a DC-Dc booster) for miles.

Could be a weak cell in the battery. I would try a another battery if I had one, or try that battery in another ebike,

I have a JK BMS and 21700 cells I think they are. The whole setup is brand new. The BMS has propa active balancing and all the cells were within 0.0007v if I remember correctly. Unfortunately I don’t have a spare battery to try

 

[Pug1]

Seems like sometimes KTs can run with a 52v pack but other times they blow up

I burned my new 48V controller with a 52V battery

Hello ESFMs, I have got a overvoltage protection error on a new build. Do I need to buy another controller? If so, what do I have to look out for? Here are the facts: Bought this kit: Motor: 1000W geared rear hub motor Controller: Rated voltage 48V, Limit current 28A, Nominal current 14A...

endless-sphere.com

How to know how much battery volts the ebike controller can take?

How do I know how much battery volts the ebike controller can take? I mean how much above the voltage I could go on the battery above the controller's stated voltage? When I open the controller those battery kind of looking things has text on them saying something about 63volt. The text...

endless-sphere.com

Probably depends on components choice in each particular controller. Maybe yours somehow least tolerant to 52v. Does it behave weirdly if you discharge the pack to below 54V?

It was doing it the other day when the battery was almost empty, somewhere around 48.5v. I’ve just freshly charged the battery so it’s goin to take a while to discharge down to 54v/53v. Could the software be tweaked to deal with problem or is it more fundamental than the software?

 

[Alex12345]

It was doing it the other day when the battery was almost empty, somewhere around 48.5v. I’ve just freshly charged the battery so it’s goin to take a while to discharge down to 54v/53v. Could the software be tweaked to deal with problem or is it more fundamental than the software?

Honestly idk then. So you are saying it doesn't cut out at a wrong phase combination at all even if you run it for like a minute? I would then suspect the tuning still

https://www.pedelecs.co.uk/topic/44448-fiido-d2-conversion-to-kt-controller-upgrade-not-working-noisy-motor-and-cuts-out-after-5-seconds/

Can you ask the shop that sold you all the components what would be the correct settings? Any changes if you tweak C2 or L3 parameters?

I would hesitate to flash the open firmware, I tried it at some point, but it's really tricky, you have to know the right repo to use, change the appropriate settings from their defaults etc.. if Ebike masters sell the controller and the motor, presumably the two can work together and the motor doesn't have any weird offset angle (setting which would be one of the benefits of the open source firmware).

 

[docw009]

The JK BMS must be bluetooth then. Can't be a battery problem if it will pull the current when the phases are miswired.

Will it spin on the stand for more than 5 seconds? Or is this a problem that only occurs under load.

Which display are you using if it shows volts? Does it show watts too? Did you set the speed limit too low? You can take the display out of the equation. KT's will run without them.

If it's a flat JST display plug, jumper the red wire to the blue. The second jumper isn't needed for testing. If it were permanent, the second jumper would ground the RX data line. I assume this is receive data relative to the dis[play,


If it's a water proof plug, this is how I've done it.



Double check everything, as your passing 52V from the red pin to the blue pin. This powers up the controller's logic and low voltage circuits. The phase circuits already get 52V from the battery, If you connect 52V to the wrong pin, you will fry something,


The controller will store battery voltage on its capacitors for days. SO it's a good thing it's a female plug,

 

[Pug1]

Honestly idk then. So you are saying it doesn't cut out at a wrong phase combination at all even if you run it for like a minute? I would then suspect the tuning still

https://www.pedelecs.co.uk/topic/44448-fiido-d2-conversion-to-kt-controller-upgrade-not-working-noisy-motor-and-cuts-out-after-5-seconds/

Can you ask the shop that sold you all the components what would be the correct settings? Any changes if you tweak C2 or L3 parameters?

I would hesitate to flash the open firmware, I tried it at some point, but it's really tricky, you have to know the right repo to use, change the appropriate settings from their defaults etc.. if Ebike masters sell the controller and the motor, presumably the two can work together and the motor doesn't have any weird offset angle (setting which would be one of the benefits of the open source firmware).

Thanks Alex, after reading that post I have a little more hope. So yesterday while testing I change C5 from 10 to 00 and it acted horribly, then I set it to 01 and it was just as bad. Set it to 2 and it is a lot more responsive but still only staying on for a few seconds and cutting out, I reapply the throttle and it instantly works again. Once the motor is spinning it’s got good torque and speed but for literally 5 seconds. I will play with C2 shortly and update here. I don’t have access to the ‘L’ settings. KT lock them down

Measure the voltage of the throttle signal wire while the bike is on the stand and confirm whether or not the throttle signal voltage is still present when the motor cuts out.

PS. I've never heard of a KT controller that implements HVC.

I tested the throttle and the voltage is still present when the motor cuts out. The reading were something like 0.9v closed and 3.6v full throttle

The JK BMS must be bluetooth then. Can't be a battery problem if it will pull the current when the phases are miswired.

Will it spin on the stand for more than 5 seconds? Or is this a problem that only occurs under load.

Which display are you using if it shows volts? Does it show watts too? Did you set the speed limit too low? You can take the display out of the equation. KT's will run without them.

If it's a flat JST display plug, jumper the red wire to the blue. The second jumper isn't needed for testing. If it were permanent, the second jumper would ground the RX data line. I assume this is receive data relative to the dis[play,
View attachment 388982

Double check everything, as your passing 52V from the red pin to the blue pin. This powers up the controller's logic and low voltage circuits. The phase circuits already get 52V from the battery, If you connect 52V to the wrong pin, you will fry something,

I will try and eliminate the screen shortly when I do some more testing. When I got the controller it came with a jumper for the screen so you can use it without one.

It cuts out on the stand or under load. I am starting to think its software as changing C5 to 02 did make a difference. 02 is better an 00, 01, 03, 04 and 10. This setting says is a voltage divider and should limit the volts from the battery

 

[Pug1]

Ok so I’ve done a bit more testing because I’m still not satisfied with the ‘P + C’ parameters.
I think all my problems are stemming from using a 52v battery on a 48v controller, even though I was told it ‘should’ work by EBike Masters. I realise it’s ultimately my problem for not getting them to hook everything up on the bench and testing it before they sent it out and I’m kicking myself about it what a dick but hay ho

I have just been testing the settings again. I turned the system on and hit the throttle on the stand and it’s like the system was struggling for power, sounded crap and grumbling. I let go of the throttle and instantly reapplied it and got better response from the throttle. Still cutting out after a few seconds. I’m now thinking this is because the system had discharged itself for half an hour and then had to recharge when I started riding again
I would say it’s more like 3 seconds than 5 seconds and it’s exactly the same every time, which has to be either the components or a software problem

I thought that C2 & C5 options were making a difference but I’m not too sure now. If I change the ‘P + C’ parameters and then cycle the system, when I turn the system back on and hit the throttle, it seems like I have to let go of the throttle and then reapply it for the system to register a clean signal and the motor not struggle and make a grumbling sound. So I turn the throttle, let go, turn the throttle again for it to not make a weird noise, like it’s struggling for power, it then has loads of power and torque.

I’m starting to understand after a lot of learning and reading that hub motors are not very happy from a stand still, especially under a heavy load, but once they are moving they are extremely efficient and provide plenty of torque and speed which I am experiencing on my tests. The hill climbing isn’t very good but I’m putting that down to the controller basically resetting itself after 3 seconds because it’s not happy with the power delivery.

I really don’t want to mess about with my 14s battery but can I make it a 13s battery by disconnecting the 14th cell? When I was making the battery I had to solder the 15 BMS wires to the cell groups. My specific BMS is a 24s 80a unit and I had to solder wire 16 - 24 to the same pad as wire 15. But am I not understanding correctly or am I, if I take 1 cell group off the BMS it will make it a 48v battery not a 52 v battery?

Sorry I know I’ve waffled on a bit again

 

[docw009]

If you built the battery, then you can certainly tap it at 13S to test your theory. Running it long term would require you to rig the BMS for 13S so you could charge it.

I think it will do the same thing because you might have a bad controller. No HV cutoffs on KT controllers that I ever seen.

 

[Pug1]

Cheers docw009
I’ve been doin a good amount or reading over the least couple of days… I’ve probably read 80+ pages of the KT OSEC thread
I’m goin to bite the bullet and flash the OSEC fw as I have nothing to loose at this point.
I am after a better detailed guide to changing out the resistor and lm317 though if you or anyone can help with that. It’s been causing problems form some users on a 52v setup. I have a 60v to 12v-8a dc-dc converter I will be using for the lights phone charger etc.
Is it just a case of desoldering the lm317 and cutting the resistor off and feeding 12v to the output pad of the lm317? praying that it is