KTM 200 exc goes electric

j bjork

1 MW
Aug 31, 2018
Linköping, Sweden
I bought this bike without motor:


I was´nt sure what motor to use, I looked a little at different alternatives. But most got expensive fast..
One I looked at was QS 138 3000W, I was afraid it will be too weak.
But it seems like others are pleased with the power they can get out of it, so I will give it a try.

I ordered it yesterday, so it will take some time before I get it.

So I started looking at the batteries:


This is a module from a vw golf gte. It is 2 modules of 12s 1p 25Ah panasonic cells.
They are bolted and glued together on a cooling plate.
I can´t fit it to the bike like this..


I cut off the edges to be able to reach the glue a little better.
Then I started cutting with a nylon cord, normally used to cut out windshields on cars.

After a lot of cutting I ended up with this:



Oops, those metal surfaces on the cells have voltage. There is not a lot of space between the cells..


I took it apart as much as possible and covered the bare metal with kapton tape.

My plan is to be able to use 2 modules like they are, and cut down 2 modules to smaller units.
But this is a problem:


The cells are welded together. It seems to be different metals in the positive and negative poles on the cells. They seem to be welded with different metal to a bar between them. I am not sure what metal the bars are, they look like aluminum and are not magnetic.

I wonder if i can cut it in the middle, then solder wires to them?
I would like to have the motor now, so I could mount that and position everything else around it.
To be able to at least see if I am somewhere it the right area I found something with similar size:


I have started to look at what sprockets to use.


This is the 48 tooth sprocket from the bike, (I think it is 520 or 530. That is with 5/8" pitch) compared to a 71 tooth sprocket with 1/2" pitch. The 48t is 250mm dia, and the 71t is 295mm dia.
This size would work, so I would like to find a 72t 428 sprocket.
I have found 60t 520 sprocket, that would work too. But I don´t know the dia of it.
(the wheel is not from this bike, it is a 16" from an electric bike)

Front sprockets are available for both 520 and 428 from 10t up to about 20t.
520 with 10 front and 60 rear would give 6:1 gearing, the same as 12t front and 72t rear with 428. But with 428 I can gear both up and down from there, if necessary.


Here is a picture of the 48t and a 56t with 1/2" pitch, the 56t is about 231mm dia.
I did some calculations: rpm/reduction x wheel circumference x 60 (from meter per minute to meter per hour)/ 1000 (from meter/h to km/h)

2000 rpm with 6:1 and 2,13m wheel circumference = 42,6km/h
3000 rpm = 63.9km/h
4000 rpm = 85.2 km/h

I`m not sure what rpm I will reach, but I doubt I will reach 4000 without field weakening . Maybe I should take a closer look at the sharts.. Anyway, it looks like I won´t need to go lower than 6:1, so 10 front and 60 rear looks like it would be a good start. Later maybe change to 11 or 12 or something.
Very nice motorcycle! the Qs 138 70h is very powerful, I shoot 17kW with the Nuc24F the acceleration is impressive. The motor and the controller do not heat up.
This duo engine + controller tract 400kg without problem has more than 65km / h with a reduction of 8: 1 . On a motorbike it will work miracles.
larsb said:
kV about 60 without field weakening
24s is about 100V fully charged

You’ll get to 6000 rpm. I’d gear it down more, especially if if wheels are 2.13 around

1:8 would give 33% more starting torque than 1:6 :D

Thanks, maybe I should keep looking for 428 sprockets.
I won´t be using 24s, maybe 22s at 90V.

Pitmix, thanks.
I plan to use my nucular 24fet on this too :wink:
The best thing would be to start with the motor, but I don´t want to wait..


So far I am using this as a substitute.


This is how I plan the batteries to sit on the frame.


Time to start cutting batteries.


It seems like I don´t have all the pictures I thought, but the busbars look a little strange. There are different color welds on the cells, and there are a line on the busbars. They look like aluminum, but when I cut them they look like copper inside.

I tried to solder one:


No problem. However I then decided to cut out batteries in the middle to be able to keep the terminals, so I should have test-soldered a different cell.



Shortening the outer shell.



Putting it back together.


This was the plan for the busbars.


But no, the solder does not stick to the other pole.
I guess the busbars are copper in the middle, then plated with different materials for the positive and negative poles.
So maybe I can get flux for alu, and hope that will work. Or grind down the plating to get to copper.
What C rating are those 25 amp hour batteries? I would think you’d want at least 50 amps for decent range? I’m sure the bike will be very light with a 22s 25 amp hour battery though.

I don´t have much data on the batteries unfortunately, I am sure interested if someone else have.
Panasonic cells, nominal voltage 3,6V. if I remember correctly they are used between 15-85% charged. What that means in voltages would be interesting. I might have seen the fully charged voltage somewhere, but I don´t remember.
They seem to be used at about 10c.
75kw from the motor, my guess down to about 3.3V per cell. 96s 1p, 316.8V. 75000/316.8=236.74A. The system can´t be 100% effective, so if I take that x1.1 it is about 260A.
The problem with that calculation is that the fuse is 250A..
I have data on temperatures too, I think it was -28-50 degrees c. But I have to check to be sure.

I am building 22s 2p, so 50Ah. I plan to use the nucular 24fet, so I think 300-350A peak should be no problem. I only have (limited) air cooling though. The car have water cooling.
It won´t be very light though, it is about 40 kilo batteries..
The progress is slow at the moment, I am waiting on a lot of parts. The mostly wanted is the motor itself. It is a big possibility that I already have put a lot of work in to things that I will change when the motor is in position.
I would like to lower the rear battery packs, but there is the question what space is needed for the motor.


I have made some mounts for solenoids for the hv system, and a relay for the precharge. I plan to only cut the positive connection. The bms for the nucular controller doesn't seem to like when you cut the negative connection too. I don't think it is necessary to cut both to be able to get the bike registered (road legal), but the rules are not very clear on that.

I am putting the solenoids and everything hv in the batterypacks, exept the bms. I am not sure if that is necessary ither, or if it is even ok to have the balance wires (that are also hv) outside of the packs. But it will be difficult to do it any other way, when there are 2 packs and one bms.

I got a timer circuit for the precharge:


I am not sure if that is what I will end up using, but it should be able to do what is needed.


I made some bus bars to sit between the packs, they are only 1mm thick copper. But they are 20mm in the narrowest part, I hope that will be enough. Well, if I move the packs they will not fit of course..

There are a lot of other things to take care or on the bike too, so I started to look in to that.


The rear brake whas not very stable. I took it apart, there are 2 bearings with a lot of play in them.. Just standard parts, so easy to fix. Also the uniball has some play, but also standard part.

This is a little trickier:


The rear shock was leaking and didn't work very well. The oil was a grey mess..
The most critical part is available, but it seems difficult to find other seals except for the piston. I would also want the piston rings (or what they call them), but that seems harder too.
Some important parts have arrived, like the rear sprocket:


Here I'm marking out the holes.

I also made a holder to put in the lathe:


I didn't have the right material for it, so I used a brake disk.


I think it turned out alright in the end :)

I also got the motor:


Enlarging the hole to 15mm, so it fits the swing arm bolt:

The installation of the motor was not as easy as I thought, the chain didn't line up.
I had to modify the motor a little. First I tried to mill off a little in my drill press with the reamer I used:


It didn't work very well :roll:

Well, then I made a simple milling steel:


It worked ok. Material got stuck on it sometimes, and made a rough surface.
I was planning to mill down 12mm, but I ended up with 11mm with ok surface. I didn't want to try to take one more mm, and hope that the surface would be ok again..

As you can see I also had to adjust a little on the surface of the motor to get clearance from the swing.


I have also mounted the rear shock again, with some new seals and wear parts.
I modified the filling port for the nitrogen:


It was just a screw before, so you needed special tools to fill it.
I mounted a schrader valve instead.
It seems like I forgot to take any pictures of the rest of the work on it.
Anyway, I think I will have to take it apart again. The rebound setting doesn't work, it is inside of the piston.
I will try to find a manual of how to take that apart. I think the bottom bracket is threaded on the piston, but I am not sure..

I got alu flux for soldering the battery bus bars, so I soldered some copper on the ones I cut:



I am not entirely happy with the result, it was still very hard to get good solder flow on the positive side.
It kind of stuck to it, more or less..
I made the front brackets for the motor:



There are 4 plates, one is missing in the pictures. Probably very much overkill, but it should not flex anyway..

Making the backplates for the batterypacks:



I welded around the bolt holes to be able to get a smooth surface.

Started with the hv wiring:





It is still a mystery how I am going to fit everything, get it watertight and protected from contact to any metal etc..
Thanks, it will be a tight fit..
I am not sure it was a good idea to go with contactors, it would probably have been good enough to just have a bms with a switch to break the circuit. These contactors are more for voltages in the 400V range, and they sure take up some room.
Yes, there will be rubber grommets, things are just test mounted at this stage.
I have to find orange shrink tube and orange cloth tape too, but no luck so far.

I have made a cad design for one of the battery covers (cardboard aided design :wink: )


I hope to be able to make a fiberglass mold from it..
I have never used something this unstable before, not sure how it will end.
I plan to make some finishing touches to the mold when it is finished, hopefully that will be good enough.

I got my hv to 12v converter today too:


It seems to be better than I thought :)
I looked around, and had a hard time to find what I was looking for.
I finally found this at banggood. It did the voltages I was looking for, but only 10A. I don't remember if it said anything about if it was isolated (input to output), or if it said that it was not.
I bought it anyway, when checking I found that it was :D
In Chinese fashion there was 4 wires, but no diagram or anything. But when I opened it up it was easy to see what was what.
There was also a potentiometer, where I was able to set the output voltage.
I set it to 14,4v unloaded, about 14,2v loaded. That was with about 49v input, I have to check to make sure it doesn't change with 90v input.
I have mostly worked with the battery shields lately.



I glued them with hot glue, it got really stable. Better than I thought :wink:

The plan was to just fill out the worst parts, then fix the rest in the mold and on the finished product.
But this is where I am now:


I hope it is the last trimming now.. Then it is rounding off the edges and put aluminum tape on all of it. With this much finishing work I should paint them instead..
I've used those same cells on several builds, but the version with threaded posts. Be sure to keep them under good compression and they'll treat you right. If you charge to 4.15V/cell or above expect more rapid capacity degradation. Note that these 26ah cells will sage less than 50ah Chevy volt modules, so they have a significantly higher C-rate. 1p works great for high powered zipping around town, but won't get you very far in dirt bike use, so start making plans now for how to carry at least double the capacity.

FWIW, as long as you didn't heat damage some with that soldering or too deeply discharge regularly, no BMS is needed, because they stay so well balanced. That is as long as they were perfectly balanced when you got them, so no weak cells. I did 1.5yrs of daily commuting with a charge to full required each way, and they were still at +/-0.01V balance. I manually checked balance more frequently at first, but got to where I only checked them every 3-4 months.

In addition to ease of assembly at the sacrifice of some space on top where the terminals are, the threaded post version has a nice safety benefit. I had a charger fail and overcharge a 10s string and despite my compression approach the top portion puffed while the steel battery rack contained the bottom dimension. That made the string arch upward, which caused the terminals to sheer off, and that prevented any shorts or fire.

I just wish Panasonic made a version of double the height for double the capacity, because the 6"+ width fits easily on most builds.
Thanks, threaded terminals would have been very nice.. I am building 22s2p, that is about 40kg of batteries. It would be difficult to fit much more if I wanted, but I think it should be enough.I plan to charge to 4,1V per cell, 90V total. What do you recommend for LVC?

This build started with that I got hold of the batteries, then I bought the rest to put them on.
When you are doing fiberglass molds it is a good idea to make them a little longer than what you need.
Like this:


I like to have some kind of mark where it is supposed to end, and I will get that where the cardboard starts under the tape.


Ready to make a mold, I have waxed 4 times, and i will use release film (?) too.


This is how it looks now. It sure didn't go as planned. I found some fiberglass that i thought was alright, but it turned out to have very long fibers. It was nearly impossible to get to follow the lines. I worked with it for a while, then I had to start with the other mold. (whit another fiberglass mat) When I went back it already started to harden. I probably used too much hardener. I probably should have made one at a time.

Well, I think the first one may have some serious air bobbles. The second one may have some gelcout rise, or what it might be called. Because it may have taken too long before I was able to lay the first layer of fiberglass on the second one. I guess I will know tomorrow.
I think the forms turned out alright. A bit more sanding on the lines after the tape than I thought, but other than that I think they are fine:


It better work, I don't want to make new ones:


This what is left of my "plugs" or whatever they are called in English.

Well, a lot of waxing and release film just in case. This is what the first cover looks like:


There will sure be some fitting work, but I hope there are no major flaws in the design..
The other one is also finished, but it is still in the form. Time to break them apart and see what it looks like :)
Time for an update, battery covers mounted:


I have built a battery for the 12v system, 4s2p lifepo4 18650:s. It is potted in windshield glue and covered in plastic.
I don't think I can take it apart if I would want to..
It was supposed to have a bms, but I couldn't get the damn thing to work after I disconnected it from the pack I took the cells from.


Some more stuff mounted:



The nucular 24fet controller, and below you can see the battery.


Dc dc step down converter.


Now I have some wiring to figure out, and if I should make boxes to put it in or not. (for the 12v system)

And what to do with what is left of the old fuel tank..
I think it will be some sort of cover/box for the hv wiring, but I don't know how.
This is looking good. Do you have a picture of the battery assembly that goes under the covers? Or are the moulds 2 boxes that sandwich the frame?