KT motor controllers -- Flexible OpenSource firmware for BMSBattery S/Kunteng KT motor controllers (0.25kW up to 5kW)

[stancecoke]

how use stm8-stvd software?

See the "how to" in the link of my signature. Read carefully, if needed two or three times.

Based on what output parameter do I adjust the settings of these parameters?

To find the right motor specific angle: Set all speed limits to 100, set Gear Ratio to 100. This makes sure, that no speedlimit cuts the power at high rpms.

Than change the motor specific angle in steps of 5 in both directions starting from zero (0, 5, 10 .... 255, 250, 245) until 20 and 235.
Look at the current draw with wheel in the air at full throttle. The setting with the lowest current is the best. If you found the optimum, you can finetune with steps of 1.

The gear ratio has no effect on the commutation, it only defines the threshold where the motor cuts the power due to the speedlimit. So just check the speed where the motor cuts the power with a normal speedometer, when you set the speedlimits to e.g.25 kmph. Then adjust the gear ration and the wheel circumference.

regards
stancecoke

[endlessolli]

Based on what output parameter do I adjust the settings of these parameters?

To find the right motor specific angle: Set all speed limits to 100, set Gear Ratio to 100. This makes sure, that no speedlimit cuts the power at high rpms.

Than change the motor specific angle in steps of 5 in both directions starting from zero (0, 5, 10 .... 255, 250, 245) until 20 and 235.
Look at the current draw with wheel in the air at full throttle. The setting with the lowest current is the best. If you found the optimum, you can finetune with steps of 1.

The gear ratio has no effect on the commutation, it only defines the threshold where the motor cuts the power due to the speedlimit. So just check the speed where the motor cuts the power with a normal speedometer, when you set the speedlimits to e.g.25 kmph. Then adjust the gear ration and the wheel circumference.

regards
stancecoke

Thank you soo much, @stancecoke!! That really helped me a lot.

@geofft: Funny thing: I ended up with almost the default setting, as you recommended.
'239' brought me down to minimum measured current (1.3A)

P.S. Would that 'how to' be a good addition for the Wiki?
It could probably help more people like me, who can not 'read' / understand the code ....

[geofft]

Thank you soo much, @stancecoke!! That really helped me a lot.

@geofft: Funny thing: I ended up with almost the default setting, as you recommended.
'239' brought me down to minimum measured current (1.3A)

.....thought I'd give stancecoke's wheel-in-the-air-current-measurement method a try, had an interesting result.

Found that values around my existing 237 deg gave lowest current but noticed a sharply defined peak in unloaded top speed at 239 deg. Got 22.6mph at this setting but moving just 1deg either side (to 238,240) caused a 1.3mph reduction in speed to 21.3mph. This was a firm and repeatable result.

Not sure if this will translate to any difference in performance out on the road, but at this setting it seemed to pull very strongly from rest across the length of my garage, so if nothing else an interesting diversion from lockdown boredom.....

[endlessolli]

.....thought I'd give stancecoke's wheel-in-the-air-current-measurement method a try, had an interesting result.

Found that values around my existing 237 deg gave lowest current but noticed a sharply defined peak in unloaded top speed at 239 deg. Got 22.6mph at this setting but moving just 1deg either side (to 238,240) caused a 1.3mph reduction in speed to 21.3mph. This was a firm and repeatable result.

Not sure if this will translate to any difference in performance out on the road, but at this setting it seemed to pull very strongly from rest across the length of my garage, so if nothing else an interesting diversion from lockdown boredom.....

If I optimize for speed I get a different setting from what I stated earlier ('5' instead of '297') - but in that value-area, the current quickly increases drastically and the motor gets pretty hot (wheel-in-the-air setup). So, I stayd at 297.
But I am happy now with my 'lowest current' optimum: Motor starts reliably, runs smooth and is responsive to the throttle.

Now a different question:
I am using the excellent BluOSEC App to do parameter optimization.

@Xnyle, @stancecoke: Is there anywere more detailed explanation for this App than in https://github.com/stancecoke/BMSBatter ... ndroid-App?
I.e:
- What is the meaning of the different 'State Symbols'?
- What is the meaning of the ifferent wave tables? (Mid Clamp, Nip&Tuck, ....?)
- When I look at the Phase B current diagram: I can make the curve flatter by slight adjustments of the 6 Hall-Angle values (in std. riding mode, not in 'DC static Zero'). Is this the right approach to fine-tune the angles?

Thanks again for all the work you guys did here!

[endlessolli]

Based on what output parameter do I adjust the settings of these parameters?

To find the right motor specific angle: Set all speed limits to 100, set Gear Ratio to 100. This makes sure, that no speedlimit cuts the power at high rpms.

Than change the motor specific angle in steps of 5 in both directions starting from zero (0, 5, 10 .... 255, 250, 245) until 20 and 235.
Look at the current draw with wheel in the air at full throttle. The setting with the lowest current is the best. If you found the optimum, you can finetune with steps of 1.

The gear ratio has no effect on the commutation, it only defines the threshold where the motor cuts the power due to the speedlimit. So just check the speed where the motor cuts the power with a normal speedometer, when you set the speedlimits to e.g.25 kmph. Then adjust the gear ration and the wheel circumference.

regards
stancecoke

I added that information in the bottom of 04 The Java Tool chapter in the Wiki.
https://github.com/stancecoke/BMSBatter ... -Java-Tool
I hope, thats ok

[stancecoke]

I hope, thats ok

of course, thank you very much!

If I optimize for speed

with a too large advance angle, you'll get flux weakening. The motor spins faster but with high losses which lead to heat...

regards
stancecoke

[szkuba]

I do have Bafang motor with 6 magnets for speed reading. How to setup this properly so i can get real time speed read? Shall i select "external" speed sensor? How about defining six magnets? Appreciate any response.

[geofft]

I do have Bafang motor with 6 magnets for speed reading. How to setup this properly so i can get real time speed read? Shall i select "external" speed sensor? How about defining six magnets? Appreciate any response.

I think if you select 'External speed sensor' and then divide your wheel circumference figure by 6 (in the configurator) it may work.

I say 'may' work, but there might be a problem. Your speed pulses will be arriving at the controller at 6 times the usual rate and they may get blocked by the firmware low pass filter at a fairly low road speed. If that proves to be the case it may be easiest to buy an external single magnet speed sensor and use that instead.

[szkuba]

I think if you select 'External speed sensor' and then divide your wheel circumference figure by 6 (in the configurator) it may work.

I say 'may' work, but there might be a problem. Your speed pulses will be arriving at the controller at 6 times the usual rate and they may get blocked by the firmware low pass filter at a fairly low road speed. If that proves to be the case it may easiest to buy an external single magnet speed sensor and use that instead.

Thanks
Just had run a test with the wheel in the air. My impression is: i can confirm speed readout overflow with real max speed of 33km/h the readout was 54km/h (should be 33 x 6 => 198 ) . From the other side i heard some poeple remove the excessive (five) speed magnets from the motor and then it works.

[stancecoke]

From the other side i heard some poeple remove the excessive (five) speed magnets from the motor and then it works.

Are you crazy?

You just have to add a "*6" to one line of code!

ACAcommons.c line 98:

Code: Select all

ui32_speed_sensor_rpks_accumulated += (((uint32_t)ui16_pwm_cycles_second)*1000) / ((uint32_t) (ui16_time_ticks_between_speed_interrupt*6));

regards
stancecoke

[szkuba]

From the other side i heard some poeple remove the excessive (five) speed magnets from the motor and then it works.

Are you crazy?

You just have to add a "*6" to one line of code!

ACAcommons.c line 98:

Code: Select all

ui32_speed_sensor_rpks_accumulated += (((uint32_t)ui16_pwm_cycles_second)*1000) / ((uint32_t) (ui16_time_ticks_between_speed_interrupt*6));

regards
stancecoke

Not really crazy to me it is hardware mod. Looking at the MXUS 01C motor we can ser one magnet there but also space for additional 5. Thanks for your reply for now i do not feel confident to modify the code, although would like to learn more in future as this firmware works really great.

[stancecoke]

to me it is hardware mod

Of course it works to remove the additional 5 magnets, but it takes at least 2h to disassemble the motor, remove the magnets and reassemble the motor. It only takes 2s to type "*6"

regards
stancecoke

[szkuba]

Of course it works to remove the additional 5 magnets, but it takes at least 2h to disassemble the motor, remove the magnets and reassemble the motor. It only takes 2s to type "*6"

regards
stancecoke

Then it is really faster, although i do not know where is the main code

[Vbruun]

Hello

I am using the erider torque sensor om my bike.

I recently changed the wiring of it so that it is now conmected in this way

https://ibb.co/Y8gxnc6

After this I blew the power supply for the torque sensor, replaced it and blew it again. Can anyone see something obviously stupid I have done?

The power supply is one of these

https://a.aliexpress.com/_mOOiyBf

[mehmetoz1980]

Thanks all. I was able to install open source software to my controller. It works fine.

[endlessolli]

I am trying to use this firmware for a project, where I only use the thumb drive input. Unfortunately, my incoming signal is rather noisy. (I smoothed it already with filtering & averaging, Still noisy.)

But while I get good motor control results with the std. firmware, this opensource firmware results in erratic / non smooth Motor behavior (under load).

Therefor my questions:
- Does someone have a recommendation for Gain P an Gain I settings, which makes the thumbdrive behavoir as close as possible to stock performance?
(right now, I am at values around 1 for both parameters; when I go down to i.e. 0.2 the system does not react at all anymore on the throttle input, which I find odd as that are the default settings. Can this be triggered by too fast throttle input changes?)

- Can someone explain in more words than the wiki(*) how these 2 parameters impact throttle behavior?

(*) GAIN P: Proportional factor of the PI controller. The higher the value is selected, the higher the risk, that the control starts oscillating.
GAIN I: Integral factor of the PI controller. The smaller the value, the smoother the control runs into the setpoint. Both gain values must be written with a dot as a decimal separator.

[stancecoke]

this opensource firmware results in erratic / non smooth Motor behavior (under load).

What is your setting for max battery current and for max phase current?
Set max phase current much higher than max battery current!

regards
stancecoke

[endlessolli]

this opensource firmware results in erratic / non smooth Motor behavior (under load).

What is your setting for max battery current and for max phase current?
Set max phase current much higher than max battery current!

regards
stancecoke

Thanks for the reply, stancecoke!
These are my current settings:
Battery current = 70
Phase current = 100
Cal a = 50 (I have small a 6 FET controller - so I should put this to 100 accroding to the wiki. (Is that a fudge factor or is there a way to calculate the correct value?)

I also found another potential issue: I digged some more in the forums (also german pedelec forum) and realized that my 'pre-controllers' RC low pass filter for the thumb input was maybe too low. (You used 1kOhm & 5µF for some experiments, the 'Forum-controller' uses 150Ohm and 100µF; I used 10kOhm & 5µF. Still wondering why it worked w/ the std. Firmware and not so good with your open source f/w.
I will also experiment with the current settings.
But the weather forces me to take a break......

[Valopallo]

Hello. I suppose I ordered the so called "high level" brake sensor for my s06s controller, which has 3 wires instead of 2.https://bmsbattery.com/parts/459-hwbs-h ... parts.html The Blueterm shows the following when I activate the brake:

Code: Select all

4350,0, 0, 325, 0, 0
4350,0, 0, 325, 0, 0
4350,0, 0, 325, 0, 0
4350,0, 0, 325, 0, 0
1,0, 0, 324, 0, 0
1,0, 0, 325, 0, 0
1,0, 0, 325, 0, 0
1,0, 0, 324, 0, 0
1,0, 0, 324, 0, 0
1,0, 0, 325, 0, 0
1,0, 0, 325, 0, 0
1,0, 0, 325, 0, 0
1,0, 0, 324, 0, 0
1,0, 0, 324, 0, 0
1,0, 0, 325, 0, 0
1,0, 0, 325, 0, 0
1,0, 0, 324, 0, 0
1,0, 0, 325, 0, 0
1,0, 0, 325, 0, 0
4350,0, 0, 325, 0, 0
4350,0, 0, 324, 0, 0
4350,0, 0, 325, 0, 0
4350,0, 0, 324, 0, 0
4350,0, 0, 325, 0, 0

If I'm correct the brake is ON at default and prevents any operation whatsoever. My question is, can I use these with my s06s controller at all or do I need to buy the "low level" ones? Maybe modify these to make them work with my controller?

[szkuba]

motor spin all time at very low speed

Nice to see the motor spinning, this is a great sucess already! Congratulations!

It seems that the battery-current measuring is not stable fast enough at startup. We already have build in a pause in the autozero procedure in the adc_init(void) function, but it seems to be too short.
You can try to increase the number of loop runs in the delay-loop, or just use a higher offset (e.g. 5 instead of 1) in line 79 of the adc.c

Code: Select all

ui16_current_cal_b -= 1;

regards
stancecoke

I had similar problem with wheel spining after running the motor and tried to modify the code. It worked well with offset "3"

[szkuba]

I wanted to share two small e-bike conversion with Open Firmware recently acomplished to my family and friends:
1. 24inch e-MTB with Q85 motor, and Sempu T4

IMG_20210109_183319.jpg (949.14 KiB) Viewed 246 times

2. 20inch with bafang 020.350 DC and eRider T9

IMG20210123181227-1.jpg (300.31 KiB) Viewed 246 times

Both have reduced battery current to 7A and Phase current to 20A, also to accomodate smaller rider size the cranks are 152mm instead of 170mm
Basically i use stock perimeterers from custom software besides TQ calibration lowered to "1000" to accomadate lower mass of entire setup - rider plus bike
my fellings are:
1. that firmware on Sempu T4 does react faster - due to more PAS magnets perhaps . But the sensor feels like cutting of randomly
2. eRider T9 setup reacts slower but is more stable.
3. Both are very gentle and can safely be used by kids - under adult supervision of course

Thanks to all who contribute to this great software. We can now democratise e-biking among the kids as well

[Valopallo]

This is the effect of the throttle.

Very good! So nothing's wrong with the controller
We have to check the hallsensor-signals next.

Put in GPIO_ReadInputData(HALL_SENSORS__PORT) & (HALL_SENSORS_MASK) instead of ui16_setpoint now. Then turn the motor shaft slowly by hand. The signal sequence has to be 4-6-2-3-1-5. If you have to turn the motor backwards due to a freewheel, the order is vice versa of course. If the sequence is OK, you have to find the right motor specific angle.
The meaning of the first value (ui16_control_state) can be decrypted as described here.

regards
stancecoke

I was in this situation a year ago with another controller and HALL sensors. I managed to fry them and now I'm attempting to build back my system once again. This time the problem seems to be at the HALL sensors, even though I used these same sensors(same lot) to my rear motor without any problems whatsoever. This is the same smaller geared hub at the front wheel as year ago. New sensors though.

When I add this line to the main.c

Code: Select all

GPIO_ReadInputData(HALL_SENSORS__PORT) & (HALL_SENSORS_MASK)

and read the controller data with Blueterm it gives me this:

Screenshot_20210124-215835.png (131.03 KiB) Viewed 243 times

After 10 seconds the controller stops sending any info. Turning the wheel does not give any HALL sensor sequence reading. Also the output voltage of the sensors switch from 5V to -5V(between Vin and HALL signals) and 0 to 8V(between GND and HALL signals) measured with a multimeter. What's happening here? There are no pull up resistors or capacitors installed because I have never needed them with these motors and controllers. Could the pull up resistor help with this?
Without HALL sensors plugged in, the Blueterm seems to read data as it should.

How to troubleshoot?

[stancecoke]

How to troubleshoot?

The system reboots all the time.
I guess this is a problem with the 5V supply of the controller. Disconnect one external device after the other (Throttle, PAS, Motorhalls ...) from the 5V rail and see, if the controller still freezes.
Do you have an extra DC/DC converter for the BT module?

Other possibility is that the watchdog resets the processor. But no idea why...

regards
stancecoke

[Valopallo]

How to troubleshoot?

The system reboots all the time.
I guess this is a problem with the 5V supply of the controller. Disconnect one external device after the other (Throttle, PAS, Motorhalls ...) from the 5V rail and see, if the controller still freezes.
Do you have an extra DC/DC converter for the BT module?

Other possibility is that the watchdog resets the processor. But no idea why...

regards
stancecoke

Yes, I have done that and the rebooting happens only with the HALL sensors plugged in. But, do you think that those voltage values of the HALL sensor I mentioned are an indication of a faulty HALL sensor behavior or do they seem normal? They do switch clearly, but the voltages are not +5V and 0V. Is it even relevant? I tested those values with external 5V PSU so it's not related to the controller 5V rail.

EDIT: Nevermind. It seems these shitty HALL sensors somehow blew the whole 2WD system. 5V rail dropped to 2.6V at both controllers and that includes PAS, Throttle, Brakes and HALLs which are not connected when measuring. Blueterm does not read anything anymore, even though it did just yesterday. The rear motor actually worked like it should with PAS the other night. Any hope left?

[stancecoke]

Any hope left?

If the voltage returns to 4.3V after disconnection the motor halls, the controllers should be OK at least.

You can check if the hall cables are weared somewhere and cause contact to the bike frame e.g. But it's quite easy to replace the hallsensors and the cables. Of course you have to open the motor for repair.

regards
stancecoke