I got my controller/LCD3 from PSW power, a Chinese company but some of their stuff is shipped from Germany.

I also added a throttle which you can't do with the controller that comes with the bike so I have all three options of control, Torque, PAS & Throttle. Remember the controller would need to be the smaller 24/36v version or it won't fit inside the frame. Running cables from the controller and attaching a meter is problematic because of the way the battery connects up. You have to have the bike folded shut to have power so I found it easier to put the bike on a jig I built myself that allowed me to carry out tests with the bike powered up but still unfolded.

That required a lot of very thick cable and some cable clips I fashioned myself from crimp terminals, it's all possible but not easy. I'd still like to know why the original setup is failing though.

Where possible, use Amass XT60 connectors instead of crimp fit bullets etc. I assume that's 1.5 m there and back, total length?

I got my controller/LCD3 from PSW power, a Chinese company but some of their stuff is shipped from Germany.

I also added a throttle which you can't do with the controller that comes with the bike so I have all three options of control, Torque, PAS & Throttle. Remember the controller would need to be the smaller 24/36v version or it won't fit inside the frame. Running cables from the controller and attaching a meter is problematic because of the way the battery connects up. You have to have the bike folded shut to have power so I found it easier to put the bike on a jig I built myself that allowed me to carry out tests with the bike powered up but still unfolded.

That required a lot of very thick cable and some cable clips I fashioned myself from crimp terminals, it's all possible but not easy. I'd still like to know why the original setup is failing though.

 

Okay great, I will check that site out. If they do not ship from Germany and only China I will prob have to pay VAT & Duty to UK ? Was it the Brainpower controller you got as that seems to be the perfect pair for the KT-LCD3 ?

 

Did you have problems with the wiring or did you cut off the wire connectors from the original controller and then paired them up the new one with the same connectors? Did all the color of the wires match up or was it alot of guess work ?

 

I think my money now is on the controller, heating up and then causing the motor to malfunction, there prob is nothing wrong with this motor now that I have learnt alot from all on this thread. However I jumped the gun and now have a MXUS XF15R motor on this bike, which fits well and is 250w/350w motor does def look much more beefy then then the Bafang one that came with the bike.

I used the KT24/36SVPR-12A Controller because it matches the LCD3 Display so no compatibility problems.

You can't get a beefier one because as I said it won't fit in the frame, this one does.

Small packages from China usually get through without having to pay any extra charges I've always found.

 

When it came to connecting mine up I cut off all the plugs & sockets and soldered the whole lot together covering the wiring with heat shrink. It was a bit of a squeeze cramming all the wiring back into the frame space especially as in my case I had extra cables for the throttle plus I fitted a separate lighting circuit that has an LED indicator on the bars to show if the lights accidently get switched on during the day without me noticing.

With all that extra cable in the frame I decided to go the soldering route so everything is covered and no plugs can shake loose.

I used the KT24/36SVPR-12A Controller because it matches the LCD3 Display so no compatibility problems.

You can't get a beefier one because as I said it won't fit in the frame, this one does.

Small packages from China usually get through without having to pay any extra charges I've always found.

 

When it came to connecting mine up I cut off all the plugs & sockets and soldered the whole lot together covering the wiring with heat shrink. It was a bit of a squeeze cramming all the wiring back into the frame space especially as in my case I had extra cables for the throttle plus I fitted a separate lighting circuit that has an LED indicator on the bars to show if the lights accidently get switched on during the day without me noticing.

With all that extra cable in the frame I decided to go the soldering route so everything is covered and no plugs can shake loose.

Thanks for all this information, really useful stuff. So I am thinking of ordering the following components, as the other stuff I have on the bike or do I need to get anything else you think which is not compatible with the bike ?

Thanks for all this information, really useful stuff. So I am thinking of ordering the following components, as the other stuff I have on the bike or do I need to get anything else you think which is not compatible with the bike ?

 

[mention=4799]Ross[/mention] - I have ordered the same controller you have, KT-LCD3, plus the light/horn and throttle. It says it should arrive in 3-5 days using DHL Express.

 

I don't want to damage this new motor using the existing "suspect" speed controller, so will now wait for new controller and hopefully I can start to enjoy the bike.

 

24V/36V 250W 6Mosfets 12A Brushless DC Sine Wave Controller

LCD/LED:KT-LCD3

Lamp and switch:with Lamp and switch

Throttle:with left Throttle

I was wondering, for a stock set up, do you think low speeds/full amps is worse for the controller or for the motor? Which would you expect to break first?

That's where you get maximum current. when the motor gets up to speed, the current goes down independent of anything the controller does.

 

The controllers always blow first. I've never seen a hub-motor of the types we use burn in normal use, even when they've been substantially over-powered. Burning will happen sometimes if a phase wire disconnects with some controllers.

 

As I said, solder on the shunt from the factory is quite normal. they do it to make adjustments to the shunt value.

That's where you get maximum current. when the motor gets up to speed, the current goes down independent of anything the controller does.

 

The controllers always blow first. I've never seen a hub-motor of the types we use burn in normal use, even when they've been substantially over-powered. Burning will happen sometimes if a phase wire disconnects with some controllers.

 

As I said, solder on the shunt from the factory is quite normal. they do it to make adjustments to the shunt value.

 

That's what I expected, although sounds worrying about the phase wire disconnection. Is that from a controller failure, or just a wire coming adrift?

 

I suppose in some factories they have someone sitting there with a soldering iron trying to "tune" the shunt. All the controllers I've taken apart and BMS units with shunts have been clean.

Binlow, out of interest, what is the part number on the MOSFETS? The component that looks like a MOSFET on the end is a voltage regulator, but can't see the codes on the MOSFETS.

That's where you get maximum current. when the motor gets up to speed, the current goes down independent of anything the controller does.

 

The controllers always blow first. I've never seen a hub-motor of the types we use burn in normal use, even when they've been substantially over-powered. Burning will happen sometimes if a phase wire disconnects with some controllers.

 

As I said, solder on the shunt from the factory is quite normal. they do it to make adjustments to the shunt value.

 

They do that a lot in the world of consumer electronic products. In theory every item of the same type coming off of the conveyor belt is the same as the first, but because of component variation in the circuit board you get a slightly different performance so slight adjustments are made to even things up.

Binlow, out of interest, what is the part number on the MOSFETS? The component that looks like a MOSFET on the end is a voltage regulator, but can't see the codes on the MOSFETS.

 

I'd guess they're 2SK4145's

The issue is the controller mosfets and their temperature rating, when they reach this they reach the thermal overload and temporarily short out/cut out. Once they cool a again they are fine until it happens again if they don't fail altogether.

The weight, uphill terrain. riding with max current loading and slow speed doesn't take long for the mosfets to reach the temperature ceiling. One can try extra heat sinks, fan cooling or opt to replace with a bigger controller and more mosfets.

That's what I expected, although sounds worrying about the phase wire disconnection. Is that from a controller failure, or just a wire coming adrift?

 

I suppose in some factories they have someone sitting there with a soldering iron trying to "tune" the shunt. All the controllers I've taken apart and BMS units with shunts have been clean.

If a wire comes adrift, the motor can't turn, but you get maximum power down the remaining complete phase. You must never give full power to an electric motor when it isn't running properly. I've seen photos of the aftermath, though I can't remember a single incidence on this forum in the ten years I've been on it. I would have expected it to be a bit more common because its's easy to stretch the bullet connectors or not fully insert them, so I guess you have to hold full throttle for quite a while before it burns enough to fail.

Binlow, out of interest, what is the part number on the MOSFETS? The component that looks like a MOSFET on the end is a voltage regulator, but can't see the codes on the MOSFETS.

 

This is the clearest picture I was able to get for the numbers stamped on the Mosfet. Maybe you may now what make and type this is from your experience .. I'm still a newby .. lol !

This is the clearest picture I was able to get for the numbers stamped on the Mosfet. Maybe you may now what make and type this is from your experience .. I'm still a newby .. lol !

 

Hey Wheezy, I think I found the Mosfets used in my controller

 

https://uk.farnell.com/stmicroelectronics/lm317t/v-reg-adj-1-2-37v-to-220-3-317/dp/9756027?CMP=KNC-GUK-GEN-KWL-MAN-L5Pages-New-Structure-25May2020-STMICROELECTRONICS&mckv=_dc|pcrid|437370421431|&gclid=CjwKCAjwxLH3BRApEiwAqX9ares4Rj3GZpOv1ZC5DaySDMEUwc9XYixbiOYMAULdAZ3dJ3yqy1ue-RoCyp4QAvD_BwE

 

https://www.st.com/resource/en/datasheet/lm217.pdf

This is the clearest picture I was able to get for the numbers stamped on the Mosfet. Maybe you may now what make and type this is from your experience .. I'm still a newby .. lol !

 

 

I use a magnifying glass to read the numbers on mine. It's helpful if you can find out the number then you can look up the specs for them.

 

The one on the end (LM317) is a voltage regulator, not a MOSFET. The MOSFETS are the 6 components to the right.

 

Did your controller have thermal paste on the small heatsink connecting the MOSFETS to the case?

 

I've found most do not and so heat conduction to the case is poor.

I use a magnifying glass to read the numbers on mine. It's helpful if you can find out the number then you can look up the specs for them.

 

The one on the end (LM317) is a voltage regulator, not a MOSFET. The MOSFETS are the 6 components to the right.

 

Did your controller have thermal paste on the small heatsink connecting the MOSFETS to the case?

 

I've found most do not and so heat conduction to the case is poor.

 

Ak ok, they were all blank as far I can remember, that now makes sense why there was 7 of those , as normally the mosfet's comes in pairs right? A neg and pos pair.

 

It had very little paste, like a greyish dust came off my fingers. Should I buy a tube of thermal paste and try putting some on?

 

I am concerned that the controller does not damage my new MXUS motor or am I pretty much safe to try it out again before new controller comes, the same one Ross recommended as he has the same bike as mine?

Ak ok, they were all blank as far I can remember, that now makes sense why there was 7 of those , as normally the mosfet's comes in pairs right? A neg and pos pair.

 

It had very little paste, like a greyish dust came off my fingers. Should I buy a tube of thermal paste and try putting some on?

 

I am concerned that the controller does not damage my new MXUS motor or am I pretty much safe to try it out again before new controller comes, the same one Ross recommended as he has the same bike as mine?

 

 

Ok, it's annoying when they do that and then you can't look up the specs.

 

A bit of thermal paste would help, but don't overdo it, you don't want it too thick either, a nice thin even layer.

 

I would ride carefully and not push things for now and keep an ear out for strange sounds/vibration.

Ok, it's annoying when they do that and then you can't look up the specs.

 

A bit of thermal paste would help, but don't overdo it, you don't want it too thick either, a nice thin even layer.

 

I would ride carefully and not push things for now and keep an ear out for strange sounds/vibration.

Hey All, A quick update!

 

I managed to take the bike out with the new MXUS motor installed and done nearly 20 miles with not a single problem, the bike just worked when I needed motor assistance. So it has to be the old Bafang motor that has developed the fault, I must have over loaded it with my son on the back. As before I started putting him on the bike it was working with no issues.

 

So perhaps the issue was with the motor and not the controller after all ..

 

Now can I repair this motor is the million dollar question .. lol

Have you done a hall sensor test yet ?

Have you done a hall sensor test yet ?

No, how do you perform that test?

No, how do you perform that test?

 

Put a multimeter on the hall sensor wires, one at a time with the controller switched on and you should see each one switching from 0V to 5V and back again as it senses a magnet going past.

No, how do you perform that test?

 

Red & Black hall should show 5v with meter set to 20v, bike battery connected and on. Probe then on the Black hall and in turn the Red probe to the three coloured halls, one should see the voltage switching as one rotates the wheel slowly backwards.

Red & Black hall should show 5v with meter set to 20v, bike battery connected and on. Probe then on the Black hall and in turn the Red probe to the three coloured halls, one should see the voltage switching as one rotates the wheel slowly backwards.

Hey all, I have now cancelled my order with pswpower as it looks like the controller was not at fault at all. It was actually the motor. BTW, the new mxus motor is good, seems to tackle all the inclines and is upto 350w power too if I want. The bike just works so good now. Although in UK, limited to 250w by original controller.

 

So I know you want me to test the hall sensors on the old Baang motor, however the bit that puzzles me is why does the motor operate without any issues until it has to deal with some load. I would have thought that is the hall sensor is failing that the motor would not work smoothly or be jerky that you cannot ride?

 

Also when testing , are you saying to have the motor opened up and the probe it whilst the motor is connected via the 9 pin wire to controller etc?

jerky that you cannot ride?

 

Also when testing , are you saying to have the motor opened up and the probe it whilst the motor is connected via the 9 pin wire to controller etc?

 

No.

You test via the controller if it has a hall wire block by poking the probes in to the rear.