High powered kickibike build

[born49]

Hello guys,
I am complete newbie with this e-vehicle stuff and I am finishing my first build : electric kickbike with
HS3540 motor.

Some specs:

- Frame (steel) is 'Mibo Dextro' (http://mibospace.com/specs/dextro/dextro.htm)
- Motor: HS3540
- Battery: 18s (3 x Turnigy Multistar 16Ah) mounted on front fork
- ESC is Lyen 12 FET Mark II

- Torque arms (both sides) are 4mm titanium from E4Bike mounted to 5mm steel frame. I am not exactly happy with
the mounting hole being located almost at the end of the fork, but it was already present so I used it.
I am monitoring torque arms before / after each ride

- Rear wheel is 16" moped rim laced with 3mm spokes. The tire I use is 20" bicycle as the one targeted
for mopeds (which I bought) had some defect unfortunately.

- Total weight (ready to run) : around 34 kg

First riding impressions:

At the first, I set up controller very conservatively to 40A battery and 80A phase current. I took it for ride
with my friend and was very disappointed - there was almost no acceleration and I cannot climb
even not so steep hills. Also, at that time, Cycle Analyst was not connected yet, so I didn't knew what
watts are pumped into motor. I hoped, that this is not what direct drive motors can do.

For next ride I connected Cycle Analyst (unfortunately in legacy mode, so I cannot use throttle input remapping yet)
And configured the controller to 'unlocked' mode : battery current 100A a 300A phase current. This changed
everything. When I took it for a spin I was literally scared how wild the acceleration was. This exceeded my expectations
by large margin. I need to tame it somehow because throttle is little bit too touchy form my taste. Top speed
is about 65 km.h-1 on flat road and you get to this speed in about 3 seconds I guess. When doing this speed runs
CA reports about 7kW is pumped into motor.

Negatives (which are going to be solved )

- Acceleration response to throttle is too aggressive for smooth riding. As you go over bumps hand on the throttle
moves slightly and scooter jumps abruptly forward. Not pleasing.

- Major negative (somehow expected): when climbing to hill (even mild one) motor gets very hot too easily (90 degs celsius).
I naively hoped, that the venting holes will reduce this problem, but now I see it is not that easy. I am reading
the threads about fan-inside-motor approach and I am keen to try it. Unfortunately this makes this vehicle little bit
impractical, but it is fun to drive it anyway.

Pictures:

The wiring of the ESC / CA is obviously work in progress

Old version:

Updated version:

 

[Samd]

Awesome, I've been thinking of building up one of these. Nice frame.

 

[born49]

Thanks ! It is very nice to ride

 

[Samd]

Where are you based? Its not in your profile...

 

[born49]

Profile updated

 

[Samd]

Sent from my iPhone using Tapatalk

 

[999zip999]

Maybe turn down the amps or three speed switch. Looks fun the 3540 is getting hot after how long ?

 

[born49]

Yes, I am going to tame it somehow, at least before I get used to it. The three speed switch sounds good,
thanks for bringing it up !

I didn't perform exact temperatures monitoring yet, but as it is now, it is kind of show stopper. I tried
to 'climb' about 2 km long hill of very mild slope with speed around 30 km.h-1 (motor consumed < 2 kW
during this) and at the top temp was 94C. It kind of surprised me, because motor is rated for 2 kW so
I thought, it would be fine.

Once it is around 90C, it really does not want to go down under 80C even after driving slowly on flats. Maybe the
fans inside hubmotor mod can help this. If I go to short steeper hill, it is not problem to get to 90C in 1min or less
(starting with 'cold' motor).

Also, I unmounted the rear wheel and checked the axle + torque arms. Apparently 4mm TA are not
thick enough. It already started to bite into axis and also arms now have more play on axis than I like. I am
going to replace it with something much more thicker, probably welded directly to frame. Any tips
for good approach how to solve TA on this frame are welcome !

 

[Samd]

Yes some testing by justing confirmed that below 6mm the metal smears in the slots and wears out fast.

It'd be easy to waterjet some slots in plate, or just make some slots in some metal bar and screw it on both side. Even epoxy.
I can't see which way the axle flats are under the nuts - any picture with a nut off?

 

[born49]

I already removed back wheel for inspection, but it was mounted
like this:

Next torque plates (probably welded directly frame) will be 10-12 mm thick. Should last for eternity I hope

 

[Samd]

Easy to buy some 10mm flat from the local steel shop and cut a slot with a hacksaw, drill a hole to bolt.

 

[born49]

I already attempted to create torque plate you described. However I had no prior experience with cutting steel, so result was not disaster but not
exactly satisfactory Also, I am not sure if I'll be able to create two exactly same pieces. I am looking around local workshops if they can
create two pieces of T-plates for me.

 

[born49]

Have some progress on cooling. I tried to do fans inside motor mod and it is major success !

I put 5 12V fans in series to motor, oriented in such way that one side sucks air and other exhaust it.
No fancy attempts to seal air going possibly around the fans.

I keep fans running all the time - noise is almost not audible. On the 'benchmark' hill (2 km, quite steep)
where I got 94 C at the top I now get 56 C (same ascending speed - 30 km.h-1). Now I can keep riding
without avoiding hills / speedruns and temperature seems to stay under 60C all the time !

Seems to be nice mod - thanks to all guys who pioneered this approach

 

[born49]

I did some upgrades in last few weeks:

- Bought new (second hand) Marzocchi 888 RC3 fork. This fork has very nice and smooth suspension action. Previous one had
no suspension action at all. It was complete garbage - stay away from it ! (Zoom 650DH)

- New fork was white, so I changed color of frame to white too I like it better now.

- New hydraulic disc brakes Shimano SLX BR-M675 (200 mm disc front / rear). Very good brakes - now it has
really nice stopping power.

- I burnt my 12 FET Lyen controller so I replaced it with Adaptto Midi-E controller. Yes, it took me some time
to decide to spend this amount of $ on controller It supports torque throttle control, which I was after. Now
after riding it for a while I would probably never go with speed based control again. It has very smooth & precise
power delivery - you definitely don't need 'throttle tamers' with this.

- I swapped thin (4mm) titanium torque arms with 12 mm thick steel torque plates (with clamping, which is not presented
on picture). Seems to be much more robust now.

 

[johnnyz]

Hey, I am having a similar problem with my setup...Converted a cool looking ebike..took out the SLA's put a 80 volt Headway pack and swapped out the stock controller and my Crystalyte Controller that would pull 65 amps in my trike. The bike is fast in acceleration but like you, cant cruise above 40 kph..very jerky..even when I hold the throttle as steady as I can..jerks on and off...except when going full throttle...and the cycle analyst doesnt help. Is this what you were getting?...when you cruised did it do this?. By the way I ran that motor and the H4080. The motor your running now I couldnt use as it got to hot too quick. The H4080 however, could take the heat better before getting too hot. ...But you are pumping ALOT of amps through that thing..more than it was designed for...max watts on my trike was 5300 and it was ok with regards to heating...


John

 

[born49]

I'll try to summarize issues I had with my setup:

1) At first, I was using 12 FET Lyen. I was impressed with its ability to accelerate vehicle, but it really was
not controllable smoothly during acceleration and at low speeds. Even slightest increase to throttle resulted
in strong pull from motor. It was however fine when you got up to some speed - above about 20 km.h-1 it
was ok. At higher speeds power delivery was smooth, I was able to cruise without issues. So I would describe
this 'issue' as result of controller using speed throttle mode combined with too aggressively setup battery / phase amps.

2) Setting 12 FET Lyen to about 6 kW bursts however took its price. On one hot day I somehow damaged it. It is
not completely dead, but I experienced similar issues what you describe. Power delivery started to be very jerky,
even when not pushing it. Just cruising at constant (even low speed) resulted in what it seemed like motor lost
power for moment completely and in next instant it recovered again. Also, when accelerating from dead stop
(very lightly), time to time controller stopped to provide enough power to CycleAnalyst which in turn started
rebooting and vehicle lost power completely. Quite dramatic when you drive in traffic ! I briefly checked
controller for blow FETs (according to https://www.ebikes.ca/documents/BlownMosfets.pdf), but everything
seemed to be fine. There is however some issue with controller, which needs somebody who actually understand
this to check out. Are you sure your controller is OK ? Also, I found out, that when driving in this 'broken' mode
with Lipo alarms attached to my batteries controller time to time pulled huge amps from battery, even if I
was not hard on throttle, because Lipo alarms started buzzing. This never happened when controller was OK
even when pushing it hardly. Can you check what happens to your battery pack voltage when it is jerky ?

 

[johnnyz]

I'll try to summarize issues I had with my setup:

1) At first, I was using 12 FET Lyen. I was impressed with its ability to accelerate vehicle, but it really was
not controllable smoothly during acceleration and at low speeds. Even slightest increase to throttle resulted
in strong pull from motor. It was however fine when you got up to some speed - above about 20 km.h-1 it
was ok. At higher speeds power delivery was smooth, I was able to cruise without issues. So I would describe
this 'issue' as result of controller using speed throttle mode combined with too aggressively setup battery / phase amps.

2) Setting 12 FET Lyen to about 6 kW bursts however took its price. On one hot day I somehow damaged it. It is
not completely dead, but I experienced similar issues what you describe. Power delivery started to be very jerky,
even when not pushing it. Just cruising at constant (even low speed) resulted in what it seemed like motor lost
power for moment completely and in next instant it recovered again. Also, when accelerating from dead stop
(very lightly), time to time controller stopped to provide enough power to CycleAnalyst which in turn started
rebooting and vehicle lost power completely. Quite dramatic when you drive in traffic ! I briefly checked
controller for blow FETs (according to https://www.ebikes.ca/documents/BlownMosfets.pdf), but everything
seemed to be fine. There is however some issue with controller, which needs somebody who actually understand
this to check out. Are you sure your controller is OK ? Also, I found out, that when driving in this 'broken' mode
with Lipo alarms attached to my batteries controller time to time pulled huge amps from battery, even if I
was not hard on throttle, because Lipo alarms started buzzing. This never happened when controller was OK
even when pushing it hardly. Can you check what happens to your battery pack voltage when it is jerky ?

Ok..your first description of problems was not what I am experiencing.. The acceleration is fine..strong and linear ( under full throttle)..and my problem is somewhat inconsistent. Sometimes it happens and most times it does at the 3/4 range..under 30 kph..it is fine..smooth and steady..also, at full throttle its fine. The hint for me I think is in the cycle analyst...it will suddenly read -amps..as though I am charging the battery. Because I hooked my Crsystalyte controller into a stock Motorino ebike..i think somehow theres an issue with the wiring..also, on my controller, there is an on/off switch which I located on the handle bar of the bike which was used for killing the power. This is now my on off switch. Obviously the FETS are ok, but a controller is composed of more than FETS...According to the CA the volts remain fairly constant when it begins jerking...its not the throttle cause I replaced with a new one. The other HINT is that it wont pull 65 amps like it did only 55...according to the CA.

John