Motor diagnostics..

[thelarkbox]

I have tested the hub motor hall sensors (pictures attached) Using the diode setting of my meter each signal connection of the sensors all read 466-467 against common ground, sensors B and C read similar between + and -. sensor A Read 780 between +, and - 300 units higher than the others which according to the 2 minute video guide i followed indicates a fault with the suggested fix a replacement board.. Sensor A initially read - nothing (1) until i really dug into the solder-blobs with the meter probe points.

I am in the dark.. I get what a diode is but dont know what metric the meter is showing and ?????

is this enough to raise the hall sensor error err-03 on the kt controller? should i look for replacement sensors and or boards??

Cheers..

 

[saneagle]

I have tested the hub motor hall sensors (pictures attached) Using the diode setting of my meter each signal connection of the sensors all read 466-467 against common ground, sensors B and C read similar between + and -. sensor A Read 780 between +, and - 300 units higher than the others which according to the 2 minute video guide i followed indicates a fault with the suggested fix a replacement board.. Sensor A initially read - nothing (1) until i really dug into the solder-blobs with the meter probe points.

I am in the dark.. I get what a diode is but dont know what metric the meter is showing and ?????

is this enough to raise the hall sensor error err-03 on the kt controller? should i look for replacement sensors and or boards??

Cheers..

Why do you guys make stuff so difficult? You power it up, set your meter to 20v dc, then check whether the 5v switches on and off when you turn the motor backwards. That's what the controller is looking for. It doesn't care about diode values!

 

[Nealh]

D8veh has beaten me to it.
Check the 5v switching , that alone will tell you if one is faulty.

 

[thelarkbox]

That was my initial plan @saneagle until i started to google, and circuits with 1 or 2 resistors that would require an ebay/amazon order to be made before i could find some at the back of a drawer started showing up and the diode mode test 'seemed' like a extra component free option..
My limited/working concept of electronics is simply that of black boxes with expected i/o..

So no 10k resistor necessary as per in the grinn iirc guide? , That i can do this afternoon, cheers.. now to find a charged up powerbank and sacrificial usb lead..

I was expecting the sensors to be ok as the injured cable entry point to the motor is the #1 suspect as it has taken many knocks, and one right befo.

 

[thelarkbox]

.........Before the error first showed up.
I must have cut the last bit of my post above..

 

[saneagle]

That was my initial plan @saneagle until i started to google, and circuits with 1 or 2 resistors that would require an ebay/amazon order to be made before i could find some at the back of a drawer started showing up and the diode mode test 'seemed' like a extra component free option..
My limited/working concept of electronics is simply that of black boxes with expected i/o..

So no 10k resistor necessary as per in the grinn iirc guide? , That i can do this afternoon, cheers.. now to find a charged up powerbank and sacrificial usb lead..

I was expecting the sensors to be ok as the injured cable entry point to the motor is the #1 suspect as it has taken many knocks, and one right befo.

There's too much intellectual masturbation on the internet. We need practical solutions.

You only need resistors if you use a separate power supply to test the halls, and even then, there's often already one there attached to the pcb. I can't see yours because of all the white gunge, but there's probably one hiding under it.

 

[thelarkbox]

DOMINOS and not the pizza one thing lading onto another.. I got the whole bike back to test the wheel connected to power, and upside down the battery is not making a good connection.. and as this battery has no on/off switch, trying to juggle the battery against gravity into place just results in unhealthy sounding pops an crackles.. ;(

So time to replace the battery and sled connectors to the round pin connectors used on my other bike so i can A) use a battery with an off on switch when expedient and cure the dodgy connection issue suffered atm.. and B) swap/use batteries as needed..

However i may have space to fit a low profile switch at the battery top end near the indicator??

what do you think? circa 15-20mm headroom off the slant ?? and about 15mm between the lock and charge port backs. any suggestions for suitable switches to install???

I would gladly sacrifice the battery indicator for a switch if necessary, ironically its the only aspect plugged in with everything else soldered/welded..

 

[harrys]

You need a pretty stout switch to handle the whole current of the battery, Most switched Hailong packs use a cheap low current switch that turns off the BMS, but the BMS needs to be provisioned with the on/off circuitry. And some are like yours with no switches. I own one like it.

 

[saneagle]

You need a pretty stout switch to handle the whole current of the battery, Most switched Hailong packs use a cheap low current switch that turns off the BMS, but the BMS needs to be provisioned with the on/off circuitry. And some are like yours with no switches. I own one like it.

It's easy to add a switch to a BMS that doesn't have one. You should be able to wire a switch into any of the sense wires because if any of them go open circuit (0v), the BMS will switch off if it's working properly.

A better, but slightly more complicated way is to put a switch on the 12v- 14v that switches the MOSFETs. The 12v-14v is switched by a small transistor, which gives a convenient line to add the switch to.

Of course if you have the two pads for soldering an optional switch to, that would be your first choice.

 

[thelarkbox]

Cheers @harrys, thats why im asking any switches i have rated for DC current are both quite beefy (large) and not waterproof at all..

@saneagle.. -- say what??
Perhaps its worth opening up the shrinkwrap for a look at the bms pcb.. Even if i find well labelled solder pads however i think i will post a pic or 2 for confirmation before making any definate alterations.. If i can find a suitable waterproof spst switch either in a box or on ebay i will be back for more help.. Thanks again ..

 

[thelarkbox]

Found a lovely SS waterproof switch but its a momentary switch ;( looks like ebay is where i am off to.

 

[thelarkbox]

The world is out to get me and stop this bike from getting sorted.. 2 x household devices went 'poof' and released the magic smoke.. tefal will pick up the airfryer for a fix (only 18mnths old..) phew, but the TV is dead.. 1/2 the screen is just a gray block.. hmmm thats a subject for youtube browsing later...

Back to the bike.. - its set up for testing with the wheel in the frame upside down and a consistent battery connection, the err-03 issue persists as soon as the motor is started.. but i need a hand holding the probes staedy and turning the wheel and being able to look at the meter as it spins.. IM just too kak handed and lacking in chopstick skills necessary.. thankfully a pal will pop over for a cuppa and natter sooner or later

meanwhile the latching waterproof switches arrived.. so .. watch this space for pics of the bms and requests for aid identifying suitable points for switch connections this evening.. fingers crossed for a helpful silk screen print on the pcb..

Any tips on re-wrapping a 36v battery pack and bms using the original split wrap and tape ???

 

[saneagle]

Any tips on re-wrapping a 36v battery pack and bms using the original split wrap and tape ???

I use duct tape.

 

[thelarkbox]

I use duct tape.

cheers i have some of that and some thinner 2"wide pvc tape but was concerned it may add too much 'girth' and cause issues re-clamping the lid to the base with its already damaged tongue and groove lips and one threaded screwhole.. worry averted cheers

 

[Cisco-man]

It's easy to add a switch to a BMS that doesn't have one. You should be able to wire a switch into any of the sense wires because if any of them go open circuit (0v), the BMS will switch off if it's working properly.

A better, but slightly more complicated way is to put a switch on the 12v- 14v that switches the MOSFETs. The 12v-14v is switched by a small transistor, which gives a convenient line to add the switch to.

Of course if you have the two pads for soldering an optional switch to, that would be your first choice.

I’ve noticed a swytch BMS uses the same connections as the overheat thermistor - putting the switch in series with the thermistor.

 

[saneagle]

I’ve noticed a swytch BMS uses the same connections as the overheat thermistor - putting the switch in series with the thermistor.

You can put it anywhere that causes the BMS logic to switch off discharge. The BMS is already a switch. You just have to tell it what to do.

 

[thelarkbox]

Opening this up was uncomfortable...



Looks like a pair of convenient solder pads sat next to the 'SW' label on the pcb (bottom left side of pictured pcb)

However i still balk at shorting those pads out and testing the battery output terminals..
----hand holding required..


ignore what may look like moisture I went ott misting 99% alcohol to free the glue holding things down.

 

[thelarkbox]

so the 2 x solder pads next to the sw label on the pcb have no effect on battery output voltage if connected to a latching switch or shorted.. I have validated the switch breaks and maintains continuity.. pictures attached with the button depressed and not, output voltage remains constant.. 1 oddish thing with the multi wire plug disconnected the output voltage was 36-37v while when plugged back in i get 39v output.. i assumed that since only ground side is fed into the bms via a wire Positive voltage for running must come from the multi wire plug.. however either way with the bms provided +ve voltage and without get voltage on the output??

??? doh!

 

[saneagle]

The MOSFETs are often leaky, so it's not unusual to get voltage on the ouput when the BMS switches off. What voltage you get, depends on the impedence of your meter. If you want a true reading, you need to put any load on the output. Even a 5k resistor should be enough to collapse the leaked voltage. You therefore need to try that switch again with any load attached to the output.

One other thing to look out for is that the switch might be bridged by default if there's no switch, and if the BMS is controlled by a micro processor on the other side of the pcb, the switch might be a momentary one to change its state.

 

[thelarkbox]

ok correction time, in an attempt to document the 36v output with no +ve voltage suppled to the bms, i pulled the plug with the led meter attached and it died instantly??

I have no clue? I swear i measured 36-37v on the battery output without the bms multiwire plugged in MORE THAN ONCE as it was the 2nd or 3rd time the difference from the 'full' 39v registered??

OH well - still the sw pads are not functioning as I expected?