Bike great for 6 months now LCD claims no charge on charged battery, motor doesn’t work

UphillBattle

Pedelecer
May 7, 2019
50
6
I’ve been running my self build for a few months now with no problems and no sign of loss of battery range.

Its an old mtb with a CST rear hub motor, KT 9 Mosfets controller from TopBikeKit, Hailong battery from Eclipse and a KT LCD3 Display.

A couple of weeks ago the display started going funny (suddenly showing 60mph, then 3 mph kind of thing) but I put it down to the extreme cold (NE Scotland after dark) on that short ride.

Since then the display would never show much left in the battery despite it being freshly charged and now the motor won’t kick in at all (assist or throttle). Display powers up ok but reads zero battery.

The battery’s own level comes up as full and measuring with a multimeter (as suggested by Eclipse) reads 54.0 to 54.1V.

Im kinda at a loss as to how to proceed here and am hoping someone recognises these symptoms!
 

Michael Love

Pedelecer
Aug 26, 2018
155
35
58
Do you know anyone close with another display or controller because it sounds like you will just have to swap parts until it starts working.
 
  • Like
Reactions: UphillBattle

Nealh

Esteemed Pedelecer
Aug 7, 2014
19,994
8,173
60
West Sx RH
The controller has issues and has failed, the obvious to test is Controller Mosfets, though with Julet cabling isn't directly easy to diagnose with out spare Julet cable with tails to apply probes.
My old Oxydrive (Lishui controller) was the same lcd worked but no pas /throttle yet batteries were fine, integrated controller went tits up.
 
  • Like
Reactions: UphillBattle

UphillBattle

Pedelecer
May 7, 2019
50
6
The controller has issues and has failed, the obvious to test is Controller Mosfets, though with Julet cabling isn't directly easy to diagnose with out spare Julet cable with tails to apply probes.
My old Oxydrive (Lishui controller) was the same lcd worked but no pas /throttle yet batteries were fine, integrated controller went tits up.
That’s interesting. I just raked around in the garage and found the screenless LCD I had kicking about - it also powers up but the low power light flashes. I also got the motor to work this time (it has worked a few times when the battery is freshly charged but it soon gives out) and with the usual LCD display it reads 0 watts motor output the whole time the wheel is turning - so that aberrant behaviour fits with what you’re saying about it being the controller.

I do have a green JULET 5 wire cable that I had previously attempted to use to wire in a Bluetooth dongle. I could desolder that, plug the wire in instead of the display and have a go on the multimeter, though I don’t know what to look for?
 

Nealh

Esteemed Pedelecer
Aug 7, 2014
19,994
8,173
60
West Sx RH
For Controller Mosfet testing you need a spare eight or nine pin motor to controller Julet cable with tails or you have to remove the controller, open it and apply probes direct to the PCB points. No battery power is required as it is a resistance test.
 

UphillBattle

Pedelecer
May 7, 2019
50
6
For Controller Mosfet testing you need a spare eight or nine pin motor to controller Julet cable with tails or you have to remove the controller, open it and apply probes direct to the PCB points. No battery power is required as it is a resistance test.
Ah thanks, I don’t have that cable. I wonder if it may be best just to order a new controller rather than mess about trying to diagnose it.

Though I guess if I’m thinking to replace it anyway there’s no harm in opening the controller up to take a look directly.

Its the controller that integrates directly under the battery so it’s going to be a long wait for a new one.
 

UphillBattle

Pedelecer
May 7, 2019
50
6

This is the controller, though I seem to recall the nine chips (they are the mosfets right?) being submersed in glue stuff from when I received it.
 

wheeliepete

Esteemed Pedelecer
Feb 28, 2016
2,047
755
60
Devon
This is the controller, though I seem to recall the nine chips (they are the mosfets right?) being submersed in glue stuff from when I received it.
That's interesting, I assumed those controllers came unpotted, no glue stuff, as in the photo. Could you post a photo of your one when you have a chance. Are there any signs of water ingress on your controller?
 

Nealh

Esteemed Pedelecer
Aug 7, 2014
19,994
8,173
60
West Sx RH
Yes.
Once in you simply need to attach your meter Black probe to the Black power wire and then the Red probe in turn to the three phase wires and then like wise Red probe to Red power wire and Black probe in turn to the Phase wires. Meter set to 200k ohms setting, readings for each three measurements should be the same but both groups will/may be a different value.
Repairing controllers isn't practical as you won't know if other components have failed so is just a practical test to confirm where failure lies and takes away any guessing.
 

vfr400

Esteemed Pedelecer
Jun 12, 2011
9,822
3,985
Basildon
All your symptoms point to water in the LCD has destroyed it. Test by disconnecting it, then short the red to the blue wire and short the two data wires (not red, not black and not blue). If it works OK on the throttle, you need a new LCD.
 

UphillBattle

Pedelecer
May 7, 2019
50
6
Looks to me like water has gotten in through the holes as it inevitably would given that the design leaves the controller exposed underneath the battery.
 

UphillBattle

Pedelecer
May 7, 2019
50
6
All your symptoms point to water in the LCD has destroyed it. Test by disconnecting it, then short the red to the blue wire and short the two data wires (not red, not black and not blue). If it works OK on the throttle, you need a new LCD.
I tried it with an LED I had lying around and it also complained of low power on a full battery...?
 

UphillBattle

Pedelecer
May 7, 2019
50
6
That's interesting, I assumed those controllers came unpotted, no glue stuff, as in the photo. Could you post a photo of your one when you have a chance. Are there any signs of water ingress on your controller?
I must have misremembered - no glue!
 

vfr400

Esteemed Pedelecer
Jun 12, 2011
9,822
3,985
Basildon
The control panels measure the battery voltage directly, so if more than one says the voltage is low, you have to find out why. Measure the voltage between the red and black LCD wires, then the two battery wires going into the controller, then the battery terminals to find out where the drop is.

If you get full voltage at the LCD connector, bridge the red and blue, then see if it drops. If it doesn't, only the LCD is faulty.

Measuring tests and logic will always get you quickly to the cause of a problem. Preconseptions, random disassembly and replacement won't.

Alternatively, you can test the LCD by disconnecting it and applying battery voltage to the red and black wires. It should switch on and light up and stay on showing the correct battery voltage, and it should respond to the buttons in the normal way.
 
Last edited:
  • Like
Reactions: UphillBattle

Andy-Mat

Esteemed Pedelecer
Oct 26, 2018
2,214
561
77
The controller has issues and has failed, the obvious to test is Controller Mosfets, though with Julet cabling isn't directly easy to diagnose with out spare Julet cable with tails to apply probes.
My old Oxydrive (Lishui controller) was the same lcd worked but no pas /throttle yet batteries were fine, integrated controller went tits up.
If anyone is interested in testing MOSFETs, there are plenty of videos here:-
Which should aid anyone wanting to go in that direction.
I myself test in place, as this can still give an indication that one has failed.
I then de-solder it and test it fully.
Hopefully the MOSFET markings are easy to read and the replacements easy to find.
There are also tables available on the web, where one can find an exact replacement, making sure that the pins are in the exct same position, though I have in the past added some stiff wire and re-soldered the pins to a PCB, where that replacement was not exact!
I used to repair high power DC Supplies in this manner, many years ago! Though in a bike controller, space is probably tight, so do insulate carefully/well, even if it is only hot glue to steady and insulate the wires.
I do such things for fun, Scottish blood, but of course a replacement controller/LCD or similar, will be a far easier way to go.
Best wishes
Andy
 

UphillBattle

Pedelecer
May 7, 2019
50
6
I’ve done a little more testing.

I couldn’t see how to get at the wires in the controller with the multimeter without desoldering them so I’ve just left that.

I used a stripped JULET cable in place of the LCD to test the voltage reaching the lcd and found it to be over 52V. Wondering if somehow the LCD was putting a load on the battery that was causing it to dip I plugged a USB lamp into the USB socket of the battery - this has zero effect of the voltage reading, leading me to believe that the battery is ok and supplying full voltage to the LCD correctly.

With both a KT LED and and a KT LCD display producing low power readings when they are being given over 52V, this leads me back again to the controller. I think I may just have a look on eBay for a cheap controller I could try patching in to see if that fixes things.
 

vfr400

Esteemed Pedelecer
Jun 12, 2011
9,822
3,985
Basildon
You must measure the battery voltage before and after you switch the LCD on or bridge the red and the blue wires.
 

UphillBattle

Pedelecer
May 7, 2019
50
6
You must measure the battery voltage before and after you switch the LCD on or bridge the red and the blue wires.
There’s no way for me to measure voltage with an actual LCD in place: the controller integrates with the battery mount so I can’t get in there with the battery plugged in. If I strip the LCD’s wire I won’t be able to reattach the Julet socket, my skills are very rudimentary.

Im measuring the voltage between the black and red wires: do you mean twist the blue and red wires together then measure the black against blue-red?
 

vfr400

Esteemed Pedelecer
Jun 12, 2011
9,822
3,985
Basildon
Im measuring the voltage between the black and red wires: do you mean twist the blue and red wires together then measure the black against blue-red?
That's right. When blue and red are joined, the blue powers the controller. If the voltage drops in that condition, the controller is faulty. if it doesn't, it's the LCD.
 
  • Like
Reactions: UphillBattle