Unusual hall problem,due to controller KU63 :workaround

[beppe000]

I edited this thread, since i've found a workaround for my problem, so if someone will experience something similar, here is my solution.
I must thank people from this forum who helped me find out the which was the problem; i also found a lot of help(not here) and information from another member of the forum, who provides the schematics of the ku63 controller (photos and schematics on his web page: http://www.avdweb.nl/solar-bike/electronics/ku63-motor-controller.html)

I have failure of the 5v power line, which cause the hall sensor and throttle working bad/stop working intermittently; the cause is the 12 to 5v converter 78l05, which overheat and don't work properly (you can find this component near the processor, on the ku63 board).
Possible solution are:
1- Arrange some kind of heat sink on the plastic part of the 78l05.
2- Replace it whit a 7805 converter, or another 78l05 version which does have a the heat sink connection.
3- Use an external 12 to 5v regulator to power the throttle, pas and hall sensors.

I've chosen the 3rd, because i already have a lot of cigarette lighter-to usb andapter, and it was easier to solder 2 wire on the back of the board (on ground and 14v supply), instead of unsolder the 78l05 from board and replace it.

So now i have connected just the signal wires of pas, throttle and hall sensor, while the power wires (black and red) are all connected to my external power adapter.
I tested it on road and it works just fine.

In this way the power consumption of my controller is increased a bit (because i now have two 5v power source, even if one almost unused), but it should not be a problem since the 36 to 12v converter (lm317t) inside the ku63 controller is on the long side of the board where there is the heat sink bar.




Here is my old post:

Hi guys, first of all, this is a great forum, i've found a lot of usefull info i used to build my e bike.

Then let me (try to) explain my problem, hopeing someone can give me some clue: i have a foldable bike (hoptown) i elecrified with a 250w 36v bafang QSWXH rear and a ku63 controller from bmsbattery; some week ago my bike stat to work just for few minutes (or seconds) and then no more until i disconnect from the battery and reconnect, then it work again just for a short time.
Since it seem to be a hall sensor problem (very common on bafang from what i read here) and since my controller can work without sensor i disconnected all the sensors wire, and after this my bike works for a bit longer. Anyway it do not work very well, cause the power is intermittent and often fail (i suspect because the engine's phase are powered with the wrong timing).
So i tested the sensor as described in this forum (led circuit) and found out something strange: hall sensor seems to work (do turn on and off rotating the wheel backward or using the engine) but when they are in their ON position(angle), they start blinking instead of remaining just turned on, as if the 5v power supply from the controller (red and black wire) cant provide enough current to keep the led on.
So i tried the same test, but this time i provide 5v power from an external lipo battery, and this time everything worked perfecly(no blinkin led, they stay turned on and off properly). Then i connected the yellow, green and blue wire to the controller (as usual) but i didn't connect the red and black, which instead i connected to my 5v lipo battery, and now everything works perfectly.

I think that my controller somehow fails to provide the 5v power source, even if my throttle and pas sensor work fine (they both depend on 5v power supply, but maybe they require fewer power than the hall sensor).
So here are my questions:
Any suggestion on how to solve this problem without using an indipendent power source(5v battery)?
I found out (with a multimeter) that on the controller side of the yellow, blue and green there is 5v voltage, is it normal?
Can this configuration cause damage to the external battery i'm using? Should i connect at least the black wire (on sensor side) to both the external battery and the controller black wire, maybe using a diode on the former to prevent reverse current flow?

Sorry for this long ad a bit confusing post... i hope someone can help with this.
Thanks in advance...
Beppe

 

[dnmun]

can you measure the input voltage to the controller? do you have a voltmeter and know how to use?

 

[beppe000]

I'm not sure what you mean by "input voltage to the controller", if you are speaking about the red and black cable on the controller side of the connector, i can measure 5 v (on the red cable +) for one second, then the voltage fall rapidly to 0v, instead between the black (-) and any of the tree other wire i measure a stable 5v voltage.
All this measurement had been made on the controller side, while disconnected from the motor.

The battery is fully charged.

 

[dnmun]

what happens to the voltage where the battery connects to the controller when the 5V on the hall sensor power jumps up and down?

 

[beppe000]

I'm not sure i can make this mesurement, since when i measure the voltage it immediately drop to 0. If instead of measure it i just put a led between + and - it just blink(it should remain on, but seems that the voltage continue to drop from 5v to 0v intermittently).
I dont know how to explain better, the has sensor work well when tested with an external battery, but as soon as i connect it to his proper supply they dont. They do not turn on and off but they all remains off (at ground voltage 0v on the engine side).
I'm afraid i''l have to open the controller and check, but i have no schematics for the ku63 controller and even if i think i have the soldering skill to repair it, i'm not sure which part of the circuit can be defective.
The strange stuff is that i tried to provide power from another source (the throttle one, 5v too), and even if here the voltage remain stable this time (no blinking led) the sensor do not work.

I think that the hall sensor do not work for a not adeguate supply, the controller seem to be able to "sense" the signal from the sensors, but not to power the sensor.

From what i see the position of the sensor (on or off) do not affect any of the voltage i can measure, on both side of the connector, as long as i power it with the default input: i have, for any position of my wheel: between red-black an oscillation from 0 to 5 v, between any other-black, oscillation from +or- 3,8v to 5 v, but with no change with the position of the wheel.

 

[dnmun]

not sure why you wanna open the controller. you need to test the hall sensors first. plugging the voltmeter in to the red and black should show a steady 5V. when you put the led in there it may be turning off when the 5V drops under the led load. who knows?

can you disable the pedelec function first so we can try to help you figure out if the halls are working?

no pictures, can you plug the 5 pin connector back together so the controller is connected to the hub motor? leave the phase wires disconnected.

 

[edavinci]

I had a similar problem with my BPM, Ku93, 11AH LIFEPO from BMS Battery. Power would cut out every minute for no apparent reason. For me, the battery BMS was to blame! The white (+) BMS lead had disconnected from the (+) bus. You may not have a Hall sensor problem at all!

 

[dnmun]

yep, that was the sense wire for #12, so when it loses the connection it would cause the BMS to shut off.

we first need to see if he can determine if the hall sensors are bad as he thinks already.

 

[amberwolf]

IIUC, you are getting a steady reading of 5V on actual hall signal wires (yellow, green, and blue) at the controller, because the pullup resistors are correctly holding the 5V up (at least when the controller is not connected to the halls).

Then when reading the 5V supply to the halls themselves (the red wire) the voltage starts at 5V but drops to 0V.

Is this correct?

Does the same thing happen if you read the throttle's 5V wire?

If not, then this is probably a bad connection at the PCB to the 5V hall power or faulty wire or crimp in the pin.

If the throttle 5V does exactly the same thing, then there might be an issue with the 5V supply inside the controller, but I highly doubt this, since you are correctly getting 5V from the pullup resistors at the signal lines.

 

[beppe000]

Ok, first thank to dnmun for your help.

This is what i've done:
I've Disconnected everithing (pas, throttle, battery-meter, phases etc..) except from the 5 wire connector of the hall sensors; my measurement are not 100% trustable, because after few second voltage always drop so i took the first value i red immediately after i turn the power on.
Wires: red - black: +4,6v, but after few seconds it start to oscillate between 0v and 5v.
Wires: any of yellow,blue,green - black: according to the position of the wheel i measure either 5,2v, which after few second start to oscillate from 5,2 to 3,8v or 0,1v, in which case it does oscillate from 0,1 to 3,8v.
If,in the first few seconds, i move the wheel forward, the voltage doesnt change, while if i move it backward it does switch from 5,2v to 0,1v or viceversa, according to the position.

Then i disconnected everything (hall sensor too) and i checked the voltage on the controller side and i found 5v between every wire and the black one, then i connected the wires one by one and found out that the "oscillation" happened as soon as i connect the red wire (power +). If i do not connect the red wire to his connector, but to an external battery, it takes much more time for the oscillation to happened, but it still happened at last.

So my problem are: the hall sensors seem to work just when rotating backward, is it normal? and second my controller became unstable as soon as the hall sensor power is connected.
According to this data i'd say that or i have both a defective hall sensors (not working forward) and controller, or (much more likely) i just have some bad connection on the hall sensor power which make some shortcut somewhere inside the engine that cause my controller instability.

I'd really appreciate any suggestion before opening my bafang to look for some defective connection.

for amberwolf:
i thought the same at the beginning, so i tried not to use the hall power but the throttle's 5V wire, but it didnt solve my problem. And even if an issue with the 5V supply inside the controller may explain the problem, i do not think, cause it does work, i can read 5v, i use throttle, pas and (sometimes) i also charged my phone frome the 5v supply line. Anyway it is a possibility...

 

[gwhy!1]

without the halls plugged into the controller do you get a steady 5v out of the controller ?. you may have a sort on the 5v line into the hall sensors, this will shut down the 5v line

 

[beppe000]

for edavinci:

thanks for the clue, i'll check, but i dont think that is my case: from your picture seem to be a problem in the battery case connection, anyway i do not experience a power cut, i still have power, but my engine do not work, or better, it does a bit if i keep accelerating and i change the position (angle) of my wheel (and my batt meter still work, it wouldnt if the bms would have shut off.

 

[beppe000]

for gwhy!

yes, i have 5v out on the wires red, blue, green and yellow, but as sson as i connect the red one to the hall the voltage bacame unstable.

I didint understand what you mean by about " having a sort on the 5v line into the hall sensors, this will shut down the 5v line"

 

[gwhy!1]

for gwhy!

yes, i have 5v out on the wires red, blue, green and yellow, but as sson as i connect the red one to the hall the voltage bacame unstable.

without the halls plugged into the controller measure the resistance between the red and black wire going into the motor this should be very high

edit :
Sorry 'Short'

 

[beppe000]

I measured the resistance between black and red and is high, but maybe not enough (at the beginning 1000kohm, then slowly fall to 600khom); but the strange thing is that i also measured the resistance between each wire and between each wire and the motor cover and i always measure almost the same values as above. It should not be so? am i wrong? if thats the case, thanks a lotgwhy! for your suggesion, there may really be some bad connection inside the engine.

 

[dnmun]

do not open the motor. first we have to figure out why your 5V supply is oscillating. do like AW asked and measure the 5V on the throttle to see if it does the same thing. don't check resistance while there is voltage in the controller.

 

[gwhy!1]

600k ohm between red and black is high enough I would have thought, but I would not expect to have the same resistance on all the wires and to the motor cover .

do not open the motor. first we have to figure out why your 5V supply is oscillating. do like AW asked and measure the 5V on the throttle to see if it does the same thing. don't check resistance while there is voltage in the controller.

I second this , is the 5v steady on the throttle without the hall sensors plugged in or unplugged or both ?. if the 5v is steady on the throttle 5v line with/without the halls plugged in then the chances are you have a very high resistance somewhere between controller and the halls 5v line ( connector/ fractured wire , or dry joint ) . if the throttle 5v line drops when the hall sensors are plugged in then you defo have something funky going on with the hall sensors/wiring ( short )

 

[fechter]

It does sound like the 5v supply from the contoller is getting overloaded and dropping out.
It would be normal with most controllers to see 5v on the signal lines with the halls disconnected. There are typically pull up resistors to the 5v supply on each input.

If the 5v still drops out even with the hall sensors disconnected, there is something wrong with the controller.

Most controllers use a 78L05 linear regulator for the 5v supply. If overloaded, they can overheat and shut off. After some cooling, it can come back on. It could be just a bad regulator chip, but usually there is something else that's drawing too much power.

It would be ideal if you could measure the current consumption of the controller when powered up, but not throttled. Likewise, it would be good to measure the current draw of just the hall sensors. This usually involves disconnecting only the red wire and trying to place an ammeter in series with it. Don't try to measure any voltage souces with the meter in current mode or something will blow up (meter will be like a short).

The hall sensors should draw something like 10mA or less (not exactly sure).
The whole controller on standby should draw less than 100mA.

 

[beppe000]

It does sound like the 5v supply from the contoller is getting overloaded and dropping out.

Most controllers use a 78L05 linear regulator for the 5v supply. If overloaded, they can overheat and shut off. After some cooling, it can come back on. It could be just a bad regulator chip, but usually there is something else that's drawing too much power.

The hall sensors should draw something like 10mA or less (not exactly sure).
The whole controller on standby should draw less than 100mA.

I couldnt measure the hall sensor consumption, i'm not sure why, but maibe is too small for my multimeter. Anyway i disconnected everithing in order to measure the consumption from battery (i have no cycle analist, just a multimeter), and.... seems that my sensor are not broken!
I tried to connect just hall sensor, phases and throttle, and seem that my 5v supply is now stable. I still have to test it, but i couldnt reproduce my issue anymore.
Next i'll try to connect PAS,Brakes etc.. to find out which component does overload my 5v power supply; i'll also check every connector, looking for some short.
I'll let you know the results of this test and the power consumption data.

 

[dnmun]

why do all that? why not just focus on testing the halls sensors to see if they work? you will have no way to measure the current consumption of the hall sensors anyway.

what you need to do is see if the hall sensor voltage toggles on and off as you rotate the wheel backwards. this has not been done yet. that is what we need to know. measure the voltage between the black and each of the BGY hall sensor wires by putting the voltmeter probes into the back end of the 5 pin plug while it is plugged into the motor, and the controller powered up.

 

[amberwolf]

I think he has already tested the actual sensor toggling as ok, except for the interference of whatever is causing the power fluctuation:

So i tested the sensor as described in this forum (led circuit) and found out something strange: hall sensor seems to work (do turn on and off rotating the wheel backward or using the engine) but when they are in their ON position(angle), they start blinking instead of remaining just turned on, as if the 5v power supply from the controller (red and black wire) cant provide enough current to keep the led on.

 

[dnmun]

my impression was that he had put an led on the hall sensor wire and it blinked. i was not aware he had ever measured the voltage on the hall sensor wire as the hall sensor was turned on and off by the magnets.

 

[beppe000]

I think he has already tested the actual sensor toggling as ok, except for the interference of whatever is causing the power fluctuation:

If i didnt make any mistake in my measurement this is exactly what happened to me. I first thought it was a hall sensor problem, but they work fine when not connected;
now i'm checking or replacing every connection (cable, some of which were a bit too long, and plughs), but i dont think i'll finish before tomorrow.

Thanks to everybody for your your help... i'll tell you as soon as i've finished.

 

[number1cruncher]

my impression was that he had put an led on the hall sensor wire and it blinked. i was not aware he had ever measured the voltage on the hall sensor wire as the hall sensor was turned on and off by the magnets.

I think led = lead here. As in meter probe.

I'm unclear whether he has checked the 5V signal on the throttle. It does sound like the hall 5V is fine if not plugged into the motor (engines use petrol Beppe :wink: )

If the power oscillates on the throttle, there's something wrong in the controller, probably the 78L15 or 7805 regulator as Fechter mentioned.

and (sometimes) i also charged my phone frome the 5v supply line. Anyway it is a possibility...

Beppe, please confirm that you have tested the throttle 5V. Also, its probably not wise to use the 5V line as a charger for your phone as the regulator is not made for that kind of load.

for gwhy! I didint understand what you mean by about " having a sort on the 5v line into the hall sensors, this will shut down the 5v line"

If your controller is working properly, a short on one of the halls could cause the 5V regulation to stop working, is what I think he's saying here.

Good Luck!

 

[dnmun]

an led in avalanche is essentially a short. that's why they normally put the resistor in series with them. he did say led, not lead. i just never saw anywhere that he actually tested the voltage on the hall sensor leads with a voltmeter.