Why is my E-bike unreliable?

[tdibertje]

Hello.
I am new here and I have a question about my DIY E-bike.
About 4 years ago I build my first DIY E-bike.
I used an old full suspension mountainbikeframe and the electrical components are from:
Golden Motor: Magic Pie 2 48v, with external controller BAC-281P, thumb throtle.
Battery and BMS from vpower 48v.
And Cycle Analist 2.23.

Now I have the folowing problem:
In the past couple of weeks my E-bike sometimes don't respond on the throtle or the pedelec when I want to use my E-bike.
i thought this problem was caused by the heat from the son, because the bike had been standing in the sun all day long.
But today was diferend. After my work I used my bike to go to a shop, about 18 km ride. No problem till now.
After less than 5 minuts I wanted to go home, but no respond on the throtle or the pedelec.
But after a 10 minut walk my E-bike was ok again.
The voltage on the Cycle Analist is ok.

Can someone give me a clue about what part can cause this strange problem?

Bye the way, I am from Holland and my English is not very good.

Kind regards tdibertje

 

[Willow]

This is an intermittent fault, which is usually a sign of a poor connection or broken wire. I'd start by checking all the plugs - start with the one that the throttle signal travels along. Check for loose pins, broken wires etc. sometimes worth buying some decent plugs to replace the ones that cam with your kit.

 

[Gregory]

And after all the plugs, then start looking at the battery, measure individual cell voltages, look for physical damage etc.

A 4 year old VPower battery could have a faulty cell interconnect, dead sting of cells or intermittent connection/broken balance wires.

 

[dogman dan]

When it's not the battery, 90% of the time it's the wiring. look hard for problems at the plugs, like connectors that back out of the housing. Or toasty connectors from the battery, or to the motor.

Look also, for nicks and cuts on the wires. Especially at the axle. It took me a month once, to notice a tire had rubbed the motor wire on one of my bikes. Same thing, sometimes worked, sometimes didn't. Depending on the bumps, the cut wire made contact, or didn't.

If you want to try replacing one thing on the bike, see if the problem goes away with a new throttle. Sometimes the magnet inside comes unglued.

 

[tdibertje]

This is an intermittent fault, which is usually a sign of a poor connection or broken wire. I'd start by checking all the plugs - start with the one that the throttle signal travels along. Check for loose pins, broken wires etc. sometimes worth buying some decent plugs to replace the ones that cam with your kit.

When I build the bike, I didn't use the connectors that cam with the kit. I cut the plugs off, soldered and isolate them with shrink sleeve. i only used the conectors directly into the controller.

And after all the plugs, then start looking at the battery, measure individual cell voltages, look for physical damage etc.
A 4 year old VPower battery could have a faulty cell interconnect, dead sting of cells or intermittent connection/broken balance wires.

My battery is very well packed, in an aluminum case (isolated with foam) and fully stuffed with foam blocks. This case is bolted on top of an iron rack above the rear wheel.
Under this rack is a smaller case for the elektronics.
The battery voltage is 57.6, so I think this is ok.
I will check the cell voltages later.
Now I am drawing a new chematic diagram, to make it a bit easier to analyse my problem, I will post it later in this topic.

If you want to try replacing one thing on the bike, see if the problem goes away with a new throttle. Sometimes the magnet inside comes unglued.

I have a new trottle, so I will replace that also.

With kind regards, tdibertje.

 

[beast775]

Did you ever commute in the rain? if so, i would look at your solder joints under the shrinkwrap..

 

[dnmun]

when it will not respond, then immediately measure the voltage at the battery. you need to eliminate the possibility of there being a loose sense wire inside the pack that turns it off. if it is not that then the possibility of intermittent connector failure in the wiring is most likely.

 

[dogman dan]

From the sound of his motor wiring, I'd sure be looking hard at that battery too.

But also look for water in that controller.

 

[tdibertje]

Hallo,
It is a few weeks since the last post.
What have i done?
I checked the wiring and made an electrical drawing:





I also replaced the trottle.
And I also simplified the wiring around the throttle.
I connected the new trottle directly to the throttle, without the relais and the resistor.

And did it solve my problem?

 

[tdibertje]

After some test drives I found the problem is still there.

This problem is temperature related.
The controller dont start up when it is warm, for example after a ride or when the bike is parked in the hot sun.
After that, I opened the controller to check this out:


Bad cleaning.


Bad soldering on the shunt.


A loose tin wire.


More bad cleaning.

View attachment 1

 

[Voltron]

Sometimes solder joints, even under shrink wrap, flex enough that the strands break right where the stiff soldered section has made a stress point. Crimps can actually be better in some high vibration situations.

 

[tdibertje]

I cleaned the controller and repaired the bad soldering.



And made also the modification I found in this forum:
http://goldenmotor.com/SMF/index.php?topic=4072.0



I used not the same value (220 Ohm), used in this tread, but 249 Ohm.
And the to modify resistor is not R54 but R51.

 

[tdibertje]

But how did the modification work on my e-bike?

The first impresion was in one word; wow!
The acceleration is much better and the max speed also increased a little bit.
My E-bike is much more fun to drive with.



And how about my problem?
It is still the same.

 

[tdibertje]

After this it was time for some new modification of my wiring.

The problem is: the controller dont start when it is warm.
Why? I dont know.
Can I get around this problem?
When I take the bike, I switch the bike on and the controller is also switched on by a relais in the powerline.
When the controller is to warm at this moment, it dont start.
If the controller is cold and i switch it on, it works fine. And it stays on, also when it is warm, until I switch it off.
So, I removed the relais in the powerline. In this way the controller is always connected to the power and always on.
This works fine.

I know, the problem is not solved by this, but now I can get around it. For now it is ok for me.
I think the problem is some where in the electronics of the controller.

 

[dnmun]

you did a great job working through this an obviously are competent enuff to have the confidence to fix it without needing the assistance of a greek tv repairman. kudos to you. need more like this.

 

[d8veh]

In your first photo, labelled "bad cleaning", it looks like there's an unsoldered component leg near to R69. It's the middle one of three. Is that just the camera angle?

 

[teklektik]

After less than 5 minuts I wanted to go home, but no respond on the throtle or the pedelec.
But after a 10 minut walk my E-bike was ok again.
The voltage on the Cycle Analist is ok.
...

After this it was time for some new modification of my wiring.

The problem is: the controller dont start when it is warm.
If the controller is cold and i switch it on, it works fine. And it stays on, also when it is warm, until I switch it off.
So, I removed the relais in the powerline. In this way the controller is always connected to the power and always on.
This works fine.

Good descriptions, pics, and congrats for a workable get-it-on-the-road solution.

The fact that the CA registers good voltage, both the throttle and PAS are inoperative, and the controller works without difficulty or interruption once it IS working, suggest that this is unrelated to a battery, throttle-specific, or connection problem. This seems to specifically be a problem with the controller power-up sequence that is disabling the controller. I have no experience with these controllers, but here's some thoughts:

• Some controllers will prevent power application if the throttle is not at zero when powered up. If your throttle sense line is not at a low enough voltage, the controller may be locking out power until it sees you have closed the throttle. Try grounding the controller throttle sense wire on power up and see if this remedies the situation.
  
  
• Some controllers detect a missing throttle or faulted throttle by an out-of-bounds voltage on the throttle sense wire. In your case, there is a relay associated with your throttle which likely takes a bit of time to close compared to the micro processor firing up. It may be that that the two are racing and the controller occasionally sees the throttle voltage as incorrect on power up and so locks out operation. This may or may not be able to be recovered by rectifying the situation - a fresh power up may be required. (This idea doesn't make good sense since the controller should have a pulldown resistor on the input, but the relay contacts are an unusual feature of your design, so...) Here I would try wiring the throttle directly to the controller (jump the relay contacts) so the controller always sees a 'good' throttle voltage.
  
  
• I found a reference to a known issue with BAC-28x controllers and CAs - please see this post. It looks like you are using the proper 1K resistor - but a change to R6 as described may be required if the limiting is not good. The article is dated, so things (parts) may have changed in the controller. Here is a referencing post.
  
  
• I'm not sure I understand the purpose of the s7 k7 relay hookup. Switching the CA throttle override wire serves no purpose. Also, CAs have an internal diode on the throttle override output already, so the external one is unnecessary - but harmless. Killing the power to the throttle is a difficult way to disable the throttle - it's going after the output rather indirectly. I would discard the relay, connect the CA override permanently to the throttle as you have shown, wire the throttle power ON all the time, then wire the 'throttle disable switch' from the controller throttle input to ground. This will mimic the CA throttle limiting strategy by dragging the throttle signal to zero when the switch is closed. This will leave the PAS operational and, of course, will not disable the controller in case in goes bonkers - so not a good kill switch solution.
  
  
• The previous point assumed the 'throttle disable' was to provide a means to disable the throttle while leaving the PAS operational. If you had intended it as a kill switch, I would recommend rewiring things as shown below. Here the controller 'Gate' wire is used to enable/disable the controller. I have jiggled the key switch connection as well so the CA (speed/distance/etc) are still active if the controller is switched off. The Gate input controls power to the processor, so killing power there will make the controller brain-dead and stop it cold since there will be no smarts left to drive the FETs.

Anyhow - just some thoughts.... (BTW - very tidy build - nice reinforcement on the rack, too!)

 

[tdibertje]

At first, i like to thank you for your helpfull replays.
I know, this thread is almost 2 years old and alot off things happend meanwhile.
A few months after my last post in this thread I blow up the BAC-028 controller.
I replaced the controller by the more powerfull BAC-0501 (48V/2000W).
At the same time I also replaced the controller, BMS and connecton box.
And also reorganised all the wiring and connections.
The bike runs realy fine since this rebuild.
The only downside is I have no PAS now. The controller don't support this function.
But my problem is solved now. The controller start now under all conditions.

Last summer, 1 week before my hollydays i blow up my charger.
Luckily I found in Singapore a very nice Electric Bike Dealer, R R Motor Pte Ltd named.
There I bought a new charger and a little 48V horn.