Take a look at this (back emf) seek knowledge (motor shutdown)

General Discussion about electric bicycles.
Post Reply
Hulken   1 mW

1 mW
Posts: 11
Joined: Nov 24 2022 2:07am

(Check new video)Help please controlleK-15G-48v-35a-yksz-2000w r

Post by Hulken » Nov 24 2022 2:30am

Hello im stuck with this controller, I did some modifikation changed battery wires to 6mm² copperwires, and phase to same with better contacts.
I did solder on copper tread 2,5mm² along plus and minus and along mosfet to phase line.
I tested on a motor and I got a reaktion on trottle for a second and The motor spinned to cut of after 1sek for about 5times then No more response at all tried another motor no response. Everything else works lights etc. I got 0,8v to 4.3v on trottle signal wire and 5v on trottle plus. I got 5v on each hall sensor for each phase. On plus hallsensor i got some lower value around 3.6V.
Before my solder i got 9.3 kohm resistens between phase and ground and now its around 3.5kohm.
Is this the problem? I guess there is some over current protection that hits? I did notice that there's a thinner line by default last sektion on minus i did go all the way.
If you would help me i would be very happy :o
20220831_114700.jpg
20220831_114700.jpg (2.65 MiB) Viewed 251 times
20220831_114741.jpg
20220831_114741.jpg (2.5 MiB) Viewed 251 times
20220825_235115.jpg
20220825_235115.jpg (2.63 MiB) Viewed 250 times
Last edited by Hulken on Nov 26 2022 1:59pm, edited 1 time in total.

Hulken   1 mW

1 mW
Posts: 11
Joined: Nov 24 2022 2:07am

Re: Help please controlleK-15G-48v-35a-yksz-2000w r

Post by Hulken » Nov 24 2022 8:39am

Do i need to desolder all on plus and minus and phases? There are some copper treads from bottom to top. Maybe I offered an easy way to pass these?
If so, would someone be kind to tell/show me the path of the current and basic information. My thought with the mods was to get better flow on circuit and to decrease heat also lead away heat from fets better. But fail

User avatar
E-HP   10 GW

10 GW
Posts: 4616
Joined: Nov 01 2018 9:20pm

Re: Help please controlleK-15G-48v-35a-yksz-2000w r

Post by E-HP » Nov 24 2022 12:34pm

Did it work prior to your mods?

User avatar
amberwolf   100 GW

100 GW
Posts: 35332
Joined: Aug 17 2009 6:43am
Location: Phoenix, AZ, USA, Earth, Sol, Local Bubble, Orion Arm, Milky Way, Local Group
Contact:

Re: Help please controlleK-15G-48v-35a-yksz-2000w r

Post by amberwolf » Nov 24 2022 2:25pm

Hulken wrote:
Nov 24 2022 2:30am

I did solder on copper tread 2,5mm² along plus and minus and along mosfet to phase line.

Looks like you bypassed the shunts with the wire in red. This prevents the controller from detecting current flow and protecting itself from overload, and makes it easy to blow itself up.
20220831_114700.jpg
20220831_114700.jpg (126.71 KiB) Viewed 227 times
The most common failure mode is FETs failing shorted (stuck on). If the motor is harder to turn by hand in either direction than it used to be, then this is likely what happened. If so, you can go here for some testing info.

https://ebikes.ca/learn/troubleshooting.html

If you find FETs have failed, you may also have to repair the gate drivers if replacing the FETs does not fix it. At that point it's usually easier to replace the controller with one that is designed to reliably do the job you wanted it to do rather than trying to fix and improve the broken one to reliably do that job.

Static electricity (ESD) can also damage electronics, so if during the work you weren't using ESD precautions, that can also cause failures that may not happen immediately. So can heat damage, if during the soldering parts on the board (especially FETs) heat up beyond their specification limits (easy to do in this type of mod).
Hulken wrote:
Nov 24 2022 8:39am
My thought with the mods was to get better flow on circuit and to decrease heat also lead away heat from fets better.
Was it overheating before the mod?
If you found this advice helpful, supporting contributions are accepted here.

Beginning of a New Life


Give a listen to my music!

(More here, including experimental)
.

Hulken   1 mW

1 mW
Posts: 11
Joined: Nov 24 2022 2:07am

Re: Help please controlleK-15G-48v-35a-yksz-2000w r

Post by Hulken » Nov 24 2022 6:08pm

amberwolf wrote:
Nov 24 2022 2:25pm
Hulken wrote:
Nov 24 2022 2:30am

I did solder on copper tread 2,5mm² along plus and minus and along mosfet to phase line.

Looks like you bypassed the shunts with the wire in red. This prevents the controller from detecting current flow and protecting itself from overload, and makes it easy to blow itself up.
20220831_114700.jpg
The most common failure mode is FETs failing shorted (stuck on). If the motor is harder to turn by hand in either direction than it used to be, then this is likely what happened. If so, you can go here for some testing info.

https://ebikes.ca/learn/troubleshooting.html

If you find FETs have failed, you may also have to repair the gate drivers if replacing the FETs does not fix it. At that point it's usually easier to replace the controller with one that is designed to reliably do the job you wanted it to do rather than trying to fix and improve the broken one to reliably do that job.

Static electricity (ESD) can also damage electronics, so if during the work you weren't using ESD precautions, that can also cause failures that may not happen immediately. So can heat damage, if during the soldering parts on the board (especially FETs) heat up beyond their specification limits (easy to do in this type of mod).
Hulken wrote:
Nov 24 2022 8:39am
My thought with the mods was to get better flow on circuit and to decrease heat also lead away heat from fets better.
Was it overheating before the mod?
Thanks for the reply.
Acually, the shuntmod was the first thing i did and use for maybe 50km before more mods as above. I did remove the shuntmod you marked and instead put an extra battery wire on other side of shunts to be able to connect/disc. with battery easy if i perfere more power or more range.

Well i guess the fets could been warm, the heatsink got quite hot. I would like too think the fets can make quite high temps than specs when not in operation. I still got the exakt same resistens on each phase but lower value.
There should be some difference in resistens if there is a difference in circuit as an damaged fet.
The motor rotate easy by hand on both tried motors.
Everything did work before.
Remember after mod it did begin to spin the motor but did cut off direcly to no response at all after a few attempts.


Excuse spelling swedish autocorrect .

Hulken   1 mW

1 mW
Posts: 11
Joined: Nov 24 2022 2:07am

Re: Help please controlleK-15G-48v-35a-yksz-2000w r

Post by Hulken » Nov 24 2022 6:08pm

E-HP wrote:
Nov 24 2022 12:34pm
Did it work prior to your mods?
Yes

Hulken   1 mW

1 mW
Posts: 11
Joined: Nov 24 2022 2:07am

Re: Help please controlleK-15G-48v-35a-yksz-2000w r

Post by Hulken » Nov 26 2022 1:58pm

Okej i got the motor to react again. Dont ask how. Just disconnect and meachure then connect and meachure etc. And suddenly first trottle rev up 3sek then second trottle gas like this.
Is there a sign for possible problem?

[youtube]https://youtu.be/1aeWsWa90x8[/youtube]

Now i got it to run. If I take very very slow on trottle at start it wont shut of. Maybe need to ad that copperewire to shunt? To trick it not to feel the current?

Hulken   1 mW

1 mW
Posts: 11
Joined: Nov 24 2022 2:07am

Take a look at this (back emf) seek knowledge (motor shutdown)

Post by Hulken » Nov 28 2022 8:02am

I got problem with motor and controller. I have to be very carefull on trottle to start rev and keep rev.
The motor gets high current and goes bang when starts rotate and shuts down. If you hot trottle very very careful and stops when it starts you can keep rev but to fast trottle twist will shutdown.
When meashure phase it shows infinite ohm when this happens.

Is it correct to assume this small 47uf capacitors aside every phase is for back current and when exeeded capicitance it shorts the motor to shutdown?
So the solution is to change these capacitors to a lager capititance like 100uf.

Any advice on this? We can leave standard protection of higher currents out of discussion, there is no point ot relevant to modifikation to higher current to learn try and error and keep standard circuit.

And yes there is pick on different controllers but this funktion is The same.

Very grateful to be shared knowledge in someones experience
Screenshot_20221128_132959_Gallery.jpg
Screenshot_20221128_132959_Gallery.jpg (147.89 KiB) Viewed 219 times
Screenshot_20221128_133330_Gallery.jpg
Screenshot_20221128_133330_Gallery.jpg (495.71 KiB) Viewed 219 times
20221128_133246.jpg
20221128_133246.jpg (1.34 MiB) Viewed 219 times

User avatar
E-HP   10 GW

10 GW
Posts: 4616
Joined: Nov 01 2018 9:20pm

Re: Take a look at this (back emf) seek knowledge (motor shutdown)

Post by E-HP » Nov 28 2022 9:00am

Hulken wrote:
Nov 28 2022 8:02am

Is it correct to assume this small 47uf capacitors aside every phase is for back current and when exeeded capicitance it shorts the motor to shutdown?
Interesting. Can you explain how you arrived at that conclusion?

Hulken   1 mW

1 mW
Posts: 11
Joined: Nov 24 2022 2:07am

Re: Take a look at this (back emf) seek knowledge (motor shutdown)

Post by Hulken » Nov 28 2022 10:31am

E-HP wrote:
Nov 28 2022 9:00am
Hulken wrote:
Nov 28 2022 8:02am

Is it correct to assume this small 47uf capacitors aside every phase is for back current and when exeeded capicitance it shorts the motor to shutdown?
Interesting. Can you explain how you arrived at that conclusion?
Not shure at all wild guesses based on a little bit knowledge (to low) and some logic. Im in process to understand the circuit and learn. the trace on thin cap goes from battery minus trough cap to phase and leads back to one mosfet on same phase my conclusion was either this is a small capacitor that will short the phase if current back exeeds the phase current. My conclusion is that my mods from the battery plus and minus to the phasewires to motor and shunt mod is creating a very large current att start up and the motor cant match it and the backcurrent exeed the out current because of the peak level is too high vs continious current. So my guess is to upgrade the small 47uf caps or to upgrade the large 470uf (gets deleepted faster than charged) to deliver more current over time to keep things rolling. My guessing is this but I might be totally wrong. And would be very grateful to have knowledge to move on further.

When I meashure phase to ground at zero trottle or when hit the trottle i get a value that changea with speed obvious. When the motor cuts of either at start or att low to high speed with fast twist on trottle it shows infinite resistense. Equal what can be experted from a behaviour with a capacitors function in a circuit. And to expect when make changes on the Units abilty to pass current the capititance Maybe is insufficient.
Or the phase wires 3mm² uknown wire quality to motor vs 6mm² pure copper wires from controller mismatch and current cant pass the wire fast and current is still left in wires of it when not suppose to and creating a failure do to low rotation cant make use of the peaklevels. Either way i need some advice what to look at. I hope there is knowledge out there and a person kind enogh to share some knowledge.

Knowledge is power they said and im after power here but lack of knowledge so I can only make it half way.
Last edited by Hulken on Nov 28 2022 10:55am, edited 1 time in total.

lnanek   10 W

10 W
Posts: 70
Joined: Sep 26 2022 11:32pm

Re: Take a look at this (back emf) seek knowledge (motor shutdown)

Post by lnanek » Nov 28 2022 10:51am

The controller doesn't just dump that sort of stuff back in the battery? I know with my Grin Technologies Baserunner, if I keep regen enabled and hook up a boost converter to go from 48V to 58V, or have an ideal diode in front of each battery in a multiple battery bike, I constantly get controller over current faults in the situations you're talking about.

Hulken   1 mW

1 mW
Posts: 11
Joined: Nov 24 2022 2:07am

Re: Take a look at this (back emf) seek knowledge (motor shutdown)

Post by Hulken » Nov 28 2022 11:01am

lnanek wrote:
Nov 28 2022 10:51am
The controller doesn't just dump that sort of stuff back in the battery? I know with my Grin Technologies Baserunner, if I keep regen enabled and hook up a boost converter to go from 48V to 58V, or have an ideal diode in front of each battery in a multiple battery bike, I constantly get controller over current faults in the situations you're talking about.
This is an cheap ebike controller for a ev-ultra 2000w bike i dont think there is an regen to charge batteries.
Could be wrong? Whats the conclusion? Do I need to provide more information. I can take pics do meachurements with fluke 177.
Obviously there is somerhing i miss i had an exact same controller Who i mod from 30amps to 65amps with no cuttof. I guess the difference here is there less resistense do more mods and the peaklevels is higher and a problem occurs that either is from an protection circuit or somerhing else. There is a large current peak at start and its really hard to twist trottle so carefully that it will continiue to rotate and not cut of. When cutof you have to twist trottle back and carefully pass the moment when it hits. I forgot mention the wheel is in air
Last edited by Hulken on Nov 28 2022 11:18am, edited 1 time in total.

User avatar
fechter   100 GW

100 GW
Posts: 15959
Joined: Dec 31 2006 3:23pm
Location: California Bay Area, USA

Re: Take a look at this (back emf) seek knowledge (motor shutdown)

Post by fechter » Nov 28 2022 11:05am

Your description sounds like the motor hall/phase wires aren't properly matched to the controller. Don't assume the colors should match. If the hall/phase wires are mismatched, it causes overcurrent which can shut down the controller.

Here's a topic that describes finding the right combination:
viewtopic.php?f=16&t=3484
"One test is worth a thousand opinions"

Hulken   1 mW

1 mW
Posts: 11
Joined: Nov 24 2022 2:07am

Re: Take a look at this (back emf) seek knowledge (motor shutdown)

Post by Hulken » Nov 28 2022 11:26am

fechter wrote:
Nov 28 2022 11:05am
Your description sounds like the motor hall/phase wires aren't properly matched to the controller. Don't assume the colors should match. If the hall/phase wires are mismatched, it causes overcurrent which can shut down the controller.

Here's a topic that describes finding the right combination:
viewtopic.php?f=16&t=3484

Yes i understand that but the phase and hallsensor wires are as the bike was set original yellow to yellow etc.. match in colour exept ive change to pure 6mm² copper and named then phase A B C instead as it is in controller. But A,B,C equal to same colour as before.
And would the motor run perfekt at all speeds if not trottle to fast from low to high?

User avatar
fechter   100 GW

100 GW
Posts: 15959
Joined: Dec 31 2006 3:23pm
Location: California Bay Area, USA

Re: Take a look at this (back emf) seek knowledge (motor shutdown)

Post by fechter » Nov 28 2022 1:48pm

If the combination is wrong, the motor could run but it would take much more current than normal. A rapid acceleration could put it over the limit and shut down.

Do you have a way to measure battery current? Measure with the wheel up (no load) at full speed. This will tell you if it's right or not.
"One test is worth a thousand opinions"

User avatar
amberwolf   100 GW

100 GW
Posts: 35332
Joined: Aug 17 2009 6:43am
Location: Phoenix, AZ, USA, Earth, Sol, Local Bubble, Orion Arm, Milky Way, Local Group
Contact:

Re: Take a look at this (back emf) seek knowledge (motor shutdown)

Post by amberwolf » Nov 28 2022 2:13pm

If you're still bypassing the shunt as when you started out, then as noted the controller can't correctly measure current flow, and can't correctly protect itself against overcurrent.

If the FETs have changed resistance as you've noted, then they may all be damaged, and no longer respond the same to controller MCU inputs, no longer driving the motor correctly.(a non-damaged FET shouldn't have a different resistance measured now vs a previous time)

Another issue is that if the current flow is greater than before (because the controller is no longer limiting current as it was designed to do), voltage sag on the battery will be greater. Under higher loads it can be enough to trigger controller LVC, even if it is not enough to cause the battery's BMS to shutdown and completely cut power.
If you found this advice helpful, supporting contributions are accepted here.

Beginning of a New Life


Give a listen to my music!

(More here, including experimental)
.

Hulken   1 mW

1 mW
Posts: 11
Joined: Nov 24 2022 2:07am

Re: Take a look at this (back emf) seek knowledge (motor shutdown)

Post by Hulken » Nov 28 2022 6:35pm

amberwolf wrote:
Nov 28 2022 2:13pm
If you're still bypassing the shunt as when you started out, then as noted the controller can't correctly measure current flow, and can't correctly protect itself against overcurrent.

If the FETs have changed resistance as you've noted, then they may all be damaged, and no longer respond the same to controller MCU inputs, no longer driving the motor correctly.(a non-damaged FET shouldn't have a different resistance measured now vs a previous time)

Another issue is that if the current flow is greater than before (because the controller is no longer limiting current as it was designed to do), voltage sag on the battery will be greater. Under higher loads it can be enough to trigger controller LVC, even if it is not enough to cause the battery's BMS to shutdown and completely cut power.
Its not fets damage. Its lower resistance do to more copper and better quality and thicker wire. The current peak big at startup to shutdown. The peak hits at startup and when twist trottle fast from low rpm. What the phase show when this peak happens and the shutdown occurs is that phases hits infinite resistense and my guess is do to an capacitors behaviour with more current the capacitors either cant filter noise out as disired and or the capititance needed to cope with the new desired current cant supply long enough to get things spinning enough to keep it spinning so the peaklevels counter the contionius levels differ to much and leads to shutdown. Or the backcurrent is to large and shutdown. What would you think of that. Could a lager capititance help.

There is no bms there are 4x 12v=48V lead batteries orginal. Ive got many and tried to add more batteries upp to 12 lead batteries in paralell and serie and also tried 5x12v=64V also to get higher voltage to deliver higher wattage faster with less current. So I think we need to fokus on some fysical shutdown at the controller maybe some resistance needed to be added or some capititance? Maybe the larger cupper wires and controllers power line uppgrades also would need to go all the way to the motor. Its 6mm² quality copper phase wires about 25cm length to 3mm² 100cm length uknown material silver cables to motor.

User avatar
E-HP   10 GW

10 GW
Posts: 4616
Joined: Nov 01 2018 9:20pm

Re: Take a look at this (back emf) seek knowledge (motor shutdown)

Post by E-HP » Nov 28 2022 8:06pm

Hulken wrote:
Nov 28 2022 6:35pm
Its not fets damage. Its lower resistance do to more copper and better quality and thicker wire. The current peak big at startup to shutdown. The peak hits at startup and when twist trottle fast from low rpm. What the phase show when this peak happens and the shutdown occurs is that phases hits infinite resistense and my guess is do to an capacitors behaviour with more current the capacitors either cant filter noise out as disired and or the capititance needed to cope with the new desired current cant supply long enough to get things spinning enough to keep it spinning so the peaklevels counter the contionius levels differ to much and leads to shutdown. Or the backcurrent is to large and shutdown. What would you think of that. Could a lager capititance help.

There is no bms there are 4x 12v=48V lead batteries orginal. Ive got many and tried to add more batteries upp to 12 lead batteries in paralell and serie and also tried 5x12v=64V also to get higher voltage to deliver higher wattage faster with less current. So I think we need to fokus on some fysical shutdown at the controller maybe some resistance needed to be added or some capititance? Maybe the larger cupper wires and controllers power line uppgrades also would need to go all the way to the motor. Its 6mm² quality copper phase wires about 25cm length to 3mm² 100cm length uknown material silver cables to motor.
Not following any of your logic, but interested in finding out how you resolve the issue. :bigthumb:

Hulken   1 mW

1 mW
Posts: 11
Joined: Nov 24 2022 2:07am

Re: Take a look at this (back emf) seek knowledge (motor shutdown)

Post by Hulken » Nov 29 2022 12:02am

E-HP wrote:
Nov 28 2022 8:06pm
Hulken wrote:
Nov 28 2022 6:35pm
Its not fets damage. Its lower resistance do to more copper and better quality and thicker wire. The current peak big at startup to shutdown. The peak hits at startup and when twist trottle fast from low rpm. What the phase show when this peak happens and the shutdown occurs is that phases hits infinite resistense and my guess is do to an capacitors behaviour with more current the capacitors either cant filter noise out as disired and or the capititance needed to cope with the new desired current cant supply long enough to get things spinning enough to keep it spinning so the peaklevels counter the contionius levels differ to much and leads to shutdown.. What would you think of that. Could a lager capititance help.

There is no bms there are 4x 12v=48V lead batteries orginal. Ive got many and tried to add more batteries upp to 12 lead batteries in paralell and serie and also tried 5x12v=64V also to get higher voltage to deliver higher wattage faster with less current. So I think we need to fokus on some fysical shutdown at the controller maybe some resistance needed to be added or some capititance? Maybe the larger cupper wires and controllers power line uppgrades also would need to go all the way to the motor. Its 6mm² quality copper phase wires about 25cm length to 3mm² 100cm length uknown material silver cables to motor.
Not following any of your logic, but interested in finding out how you resolve the issue. :bigthumb:

Haha 😅 then im lost and should maybe not try further.
Im on amberwolf trace. But why would fets be damaged and still make performance good? But only to fail when it comes to fast and high current switch, then they should not short to ground to make an restart easy time after time? Without fail completely.

My logic is maybe fail, but I think there could be an problem with capacitors in circuit as we see mods resistance lowered on the circuit and will allow higher currents wich will be creating higher inductance and in turn demand higher capititance to deal with these condition. It requires more energy to start or stop current in an inductor than it does to keep it flowing.

User avatar
fechter   100 GW

100 GW
Posts: 15959
Joined: Dec 31 2006 3:23pm
Location: California Bay Area, USA

Re: Take a look at this (back emf) seek knowledge (motor shutdown)

Post by fechter » Nov 29 2022 11:42am

I don't think it's the caps. I've seen controllers where the caps exploded off the board and the motor still ran OK. Until the FETs blew.

You really need to see what the battery current is. There are a number of possible things that can cause a controller to shut down, but based on the observations, it would seem to be over current related. If the battery voltage sags too much under load, it could shut down for low voltage.
"One test is worth a thousand opinions"

Post Reply