Problem with KT 20A sinewave controllers?

[studio45uk]

Hi guys, I've recently been having a total nightmare with a customer's Gunai 24" fat tyre folding ebike. Initially, it had a lot of problems with water damage to the wiring, which I thought I would solve for good by replacing the controller, display, all connectors, and moving the controller up out of the frame box (which is, stupidly, below the pedals right in the way of any water that splashes up from the road!) into a box on top of the frame. However, I've now replaced the controller twice, and both times it has blown up after about 10 miles of riding. No error codes were produced. It seems that a single FET goes short, permanently connecting one motor phase to the battery, which senses the overcurrent and promptly cuts power. Model number is KT36/48SVPRD-HRD01. Running at 48v with KT LCD5 display, current limit not modified from stock (anyway you can only turn it down, not up.) Motor inspected, tested on other controllers, no fault found. Has anyone recently replaced a KT controller and experienced the same thing? I suspect a bad batch...
Thanks,
Matt

 

[Nealh]

I have been running KT for several years now without hardly an issue, I upgraded to 9 fet sinewave as found the 6 fet models do suffer from overheat/ thermal cut outs. This happens at slow speed on inclines with max current, not yet seen or had issues using the 9 fet models.

 

[Nealh]

Although the 6 fets are rated for up to 20a , it depends on the use and terrain.
If hilly then opt for a 22 or 25a 9 fet model.

 

[studio45uk]

Right on...so the KT 20A 6-FET models are not, in reality, able to really sustain 20A? That is just bloody annoying!
I had one apart and checked the MOSFET model number - they are TK150E09NE - searching that indicates they should be 150A parts - so why should they fail short at just 20A? Even if the heatsinking is not optimal, that is so far inside their rating all should be fine. I suspect the ones in these two blown units, were not really that part number. 111
Anyway, plan is to fit the next size up controller, in an IP65 rated box on the pannier rack. But I'm not happy about it! Makes me look a right silly sausage, when the repair won't stay repaired...

 

[StuartsProjects]

Right on...so the KT 20A 6-FET models are not, in reality, able to really sustain 20A? That is just bloody annoying!

The label will probably state a rated current, what they can provide\sustain long term, and a maximum current. Doubt the controller could provide the maximum current for long.

 

[saneagle]

Hi guys, I've recently been having a total nightmare with a customer's Gunai 24" fat tyre folding ebike. Initially, it had a lot of problems with water damage to the wiring, which I thought I would solve for good by replacing the controller, display, all connectors, and moving the controller up out of the frame box (which is, stupidly, below the pedals right in the way of any water that splashes up from the road!) into a box on top of the frame. However, I've now replaced the controller twice, and both times it has blown up after about 10 miles of riding. No error codes were produced. It seems that a single FET goes short, permanently connecting one motor phase to the battery, which senses the overcurrent and promptly cuts power. Model number is KT36/48SVPRD-HRD01. Running at 48v with KT LCD5 display, current limit not modified from stock (anyway you can only turn it down, not up.) Motor inspected, tested on other controllers, no fault found. Has anyone recently replaced a KT controller and experienced the same thing? I suspect a bad batch...
Thanks,
Matt

My first guess before thinking about it is that the motor cable is damaged. I've seen that happen when somebody tied a cable tie too tight, also when the cable was nicked where it comes out of the axle.

Thinking logically, the MOSFETS blow when you overload them. Just about the only way that can happen is when the moror is running too slow when high power is provided by the controller. These are the circumstances I can think of how that can happen:
1. Heavy rider and steep hill
2. Lazy rider using the throttle all the time, not pedalling up inclines
3. Motor too fast and small for the power provided to it
4. Anything that causes a synchronisation error:
a. Faulty hall sensor
b. Faulty hall connection
c. Bad phase wire connection
d. Damaged cable

 

[Nealh]

Yes the unknown is how the owner is using the bike to cause the fault/failure so a repair is not really a component fault but could be down to the user causing the component fault.
One needs to ascertain from the user how and when the fault occurs, power level , terrain and speed /cadence .

My 20a 6 fet issues though not terminal were caused by riding in PAS 5 on the South Downs, my cadence isn't fast and uphill on a grassy sfc certainly not helping . My controllers never failed but I did have to wait for them to cool down enough to allow use again. Temporarily I added 2 x 5 v fans and a separate 2p 5v power source, but then eventually went to 9 fet models.

 

[studio45uk]

Thanks for your help guys. We decided to use a big 45A controller and restrict it to 20A, because this is the guy's daily commuter ride, he lives up a big hill and he ain't gonna be doing no more pedalling than he does now!
So the next problem I have is finding a good sized, waterproof box to fit it in - ideally about 270mmx140mmx100mm - the controller is the KT 18 FET model. I'm coming up blank from all the electrical suppliers I can find, the only boxes that are long enough are far too wide and tall. I would use a zip-up bag as is commonly seen, only this guy has a tendency to get the bike very wet (hence the initial problems with water ingress). The commonly sold frame-mounting controller boxes don't seem to be big enough either, nor are they IP rated. Has anyone found a source of such boxes?
Cheers,
Matt

 

[Benjahmin]

Try CPC Farnell.
Their search engine is a bit carp, so you may have to wade through a few pages.

https://cpc.farnell.com/search?st=junction%20box&gs=true

Good for spiral wrap and heat shrink too.

 

[studio45uk]

Hmmmm still not really finding anything in those results, seems like there is a pretty traditional aspect ratio for junction boxes, can't see any that are long and narrow as I need. HOWEVER...I did have a brainwave...those LED "bulkhead" emergency lights are in just the right sized box, and IP rated to boot! And just about a tenner for the cheapest one on CPC which is half the price of a Schneider etc box similar sized! Gonna buy one and see how I get on, but I might have found THE solution.
Also worked out that I can get rid of a lot of the spaghetti wiring along the frame by using a couple of RJ45 breakout boards, small project box on the stem tube and a Cat5 network cable. Here's my wiring plan:
1. GND
2. B+
3. B+ RETURN
4. TX
5. RX
6. BRAKE
7. +5v
8. THROTTLE
All grounds combined of course. Wiring for front light and horn all inside the front box, not going back to controller.
Should end up looking super neat, waterproof and reliable...fingers crossed!

 

[studio45uk]

Update: repair worked great. Can recommend using the emergency light enclosures as controller boxes. Ethernet cable wiring also worked great first time, and looks really neat. Much better than all that spaghetti, and fiddly JST plugs. Gave the machine a couple of really good long test rides, and the new controller didn't miss a beat, hardly gets warm, with parameter C5 set to 3 (max current /2, or a 22.5A limit for this 45A max controller).

Next project is ANOTHER bike that had a KT controller fitted last October. Customer has ridden it twice since, and it's conked out already. Melted the insulation on the phase connectors, and shorted itself out.

[insert massive, massive rant about KT controllers here]

 

[saneagle]

Update: repair worked great. Can recommend using the emergency light enclosures as controller boxes. Ethernet cable wiring also worked great first time, and looks really neat. Much better than all that spaghetti, and fiddly JST plugs. Gave the machine a couple of really good long test rides, and the new controller didn't miss a beat, hardly gets warm, with parameter C5 set to 3 (max current /2, or a 22.5A limit for this 45A max controller).

Next project is ANOTHER bike that had a KT controller fitted last October. Customer has ridden it twice since, and it's conked out already. Melted the insulation on the phase connectors, and shorted itself out.

[insert massive, massive rant about KT controllers here]

When the pasw wires melt, it's normally as a result of another problem, like the motor connector not in properly, going to slow for the speed of the motor or anything that stalls the motor, If you were using the 20A controller with a high speed DD motor, that might be the reason. The 20A controllers are only rated at 10A, but if you use a high speed motor, it'll be running at 20A all the time. A 20 mph motor should be fine with it. What sort of motors are you using?