Kawasaki KL 250 A build

[wjr]

Got another 30 kWp of photovoltaic modules on my roof, so it's time to enter the world of electric vehicles.
Tractor seems to large a project to gain first experience :lol:
My pickup just runs 5000 km a year, and it's for farm use, so it has to comply economic rules - so its out
There is a z24 cabrio in the garage, that does not want to start, but's it'my son's, and he refuses to electrify it

So my eyes fell onto an old Kawasaki KL250A in another corner of the shed.
It has been used as a farm dirt bike for some years, but idling for 5 years or so now.

I think an Enduro would be a good point to start EV DYI-ng - a fun project, not to be taken too serious.
expected driving time is low, so battery can remain comparably lightweight and cheap
But considerable challenge in terms of peak power.
I have some electric fork lifts running, so some hundred Amps on a wire won't scare me.

The benchmark is the KTM freeride e-xc


With great thanks to this forum ( @larsb on top of all, with most helpful comments in my eyes till now) I decided for the QS 138 midmotor with the votol-150 controller (SP variant, so I hope) here
https://de.aliexpress.com/item/33058696041.html?spm=a2g0s.9042311.0.0.27424c4dG3JdHG

I have started a thread in the german pedelec forum
https://www.pedelecforum.de/forum/index.php?threads/elektro-kawasaki-enduro-auf-basis-einer-250er-kla.70411/
so I'm not repeating previous photos here.

Mentioned forum is focussed on pedelecs, as the name says, so my motobike build is somewhat off topic there.
There is not too much response to my technical considerations.

At time of posting, I'm still busy planning and building my battery.

 

[wjr]

The plan is a 20S32P battery of 18650 size Sony-Konio-Murata VTC4 high current cells.
I got 700 of them as "refurbished" for a decent deal.

Rated at 2100 mA, 32P yield 67.2 Ah capacity.
This multiplies with 72V to 4,8 kWh - so slightly above my KTM benchmark.

I think I will seperate the batt to 4 packs, 5S32P each.
Reasons:

• larger surface for heat exchange
• more flexibility in mounting
• easier to handle for mounting and repair (I don't hope, but I'm realistic)
• and I don't like the beer cradle optics of large square packs

The image shows the current stage of my first trial pack.
I decided to use 5x0.1 nickel strips as cell level fuse, inspired by Tesla design.
I arranged the pack in 2 times 5S16P.
Thus all serial connections only span one cell size length, and carry only one cell current.

Next step is to parallel bus bars by proper cabling

 

[wjr]

I tried to assess resistance and heat production in the battery pack.
Details are here
https://www.pedelecforum.de/forum/index.php?threads/elektro-kawasaki-enduro-auf-basis-einer-250er-kla.70411/post-1327766
But they are not valid per se.
It's just for assessment of feasibility, design errors and focus priorities.

The essence:
If I get the cabling and connections right, the cell IR is the largest source of losses.
At 150 A, they'l ldeliver 420 W or so, which I think is quite some figure for a large brick.
And still below the rated capacity of the controller at 180 A.

Someone raised the proposal to add fans inside the battery pack to improve heat exchange.
Not expensive, but lot of work and additional space.
Could somebody share experience, whether this makes any sense?

Even the 250 burst Amps that @larsb reports using on the votol 150 just yield 7.8 Amps per cell at 32P, which is well beyond the 30 A the VTC4 are rated for.
I tested a single cell with 20 A discharge (so they got empty in 6 min), it just got hand warm.
For my fuse test, I tortured a cell at > 50A - won't like to touch it anymore then.

Capacity test in a Li-500 at 1000 mA produced hand warm cells at the end of charge.
The pack is supposed to be charged at 25 A, so I'll stay below 1A per cell.
...well, but, hmm.... there is an old fork lift charger delivering 190 A for quickcharge experiments

Anyway, I have some 70 100-W-resistors in my Aliexpress pipeline, so I can test the pack with 200 A before I go ahead.

 

[j bjork]

Sounds like an interesting build
It seems like the bike is an old offroad, late 70:s to early 80:s?

I wonder if the motor is enough for a big bike, but I don't have any experience with it.
To get max performance with as light and small battery as possible I would have used lipo, but now you have those 18650:s so I guess you might as well use them.

 

[wjr]

LiPo was my first consideration, ... sorry, mixed that with LiFePO -

So, LiFePO I was considering because that's what I'd look first if it comes to utility vehicles.
but I think they have at least double the weight per capacity.
And the current carrying capacity ratio was even worse.

I also considered LTO.
I have a bunch of 40 Ah Yinlong in the pipeline for a photovoltaic storage.
So this would have been great synergies to learn the technology.
They have hight current carrying.
They promise cycle times of 10.000 to 20.000 and are priced accordingly, but that's not necessary for a bike.

But for the same capacity, weight would have increased from less than 30 to more than 60 kg.
This was the decision to go for Li-Ion

Didn't actually realize that there was so much difference in performance figures between Li-Ion and LiPO
Considered 18650 as the most mature and robust choice in that family - and the VTC4 in particular.
beginner's mistake?

 

[wjr]

Pulled the Kawa from the shed today

Not nice, better not zoom in :wink:
Do I really want to save it from the scrap pile?
Anyway, let's have it a chance :lol:

Put it on the scale, 119 kg unstripped - just for the records

 

[wjr]

I also finished spot welding my first quarter battery pack.



Its being charged now, checking if it heats up.

Next I plan to test it at 200 A - to make a decision about cooling requirement.

 

[DanGT86]

Nice to see you jump into a build. Ive been watching your info gathering on the q3000 motor thread. I've also been skeptical but optimistic about that motor's ability to move a full size bike. I'm looking forward to your review when its up and running. so far everyone seems happy with it.

Those older frames with the twin shock rear ends are so much less intimidating to convert than modern aluminum monoshock frames. Lots of room in yours for motor and battery. Also the steel is much less scary to drill and weld on.

The battery looks good. Those cells seem like they'll be up to the task. 4-5 kwh should provide lots of range especially for a farm bike. Did you consider splitting the pack into 20s 8p or 20s 16p modules? I do my bike packs like this so i can add range on a long trip or have one charging while i ride the other. By splitting them in 20s modules they are all able to move the bike in any combination depending on weight power or range desired.

Good luck with the build.

 

[wjr]

Did you consider splitting the pack into 20s 8p or 20s 16p modules?

Not my plan yet.
Compared the batteries weight to the drivers weight :lol: and decided that it's not worth the hassle.
I'll make 4 x 5S32P I'd attach quite fixed.

When I want the controller to run at 200 or 250 A boost amps, I think 32P is really good for cell health, although the VTC4 might carry more. But I think had to reconsider the whol pack connection design
250 Amps from 8P means 31.25 Amps per cell.
That's max boost rating vor VTC4 cells, but quite a lot for 7 x 0.15 Nickel.

When I gain street legality (which is not yet top priority) I might add some more capacity so I could try a decent ride out as well.
Detachable side cases would be good place, I think.

 

[wjr]

Still worrying about battery temperature.
Ordered power resistors in China for some package scale tests, but looks like they got stuck in Christmas parcel flood
So I built a small test pack (4S1P) that fits my discharger at hand and pack it into some insulation to get a rough idea.

I think the picture tells ist:


Full discharge and charge cycle within half an hour may sound crazy, but that's what the Votol 150 is supposed to sustain.
So I'd not plan for less.
So - battery heat is an issue.
I'd not like to pack all my 20S32P into one large cradle, then.
Either I separate it into smaller packs with larger surface, or I'd add some fans inside to enhance air circulation.