'lightest.bike' 1.7kg 1000w mid drive

As promised, here are the results of my modifications -

This was sent as Feedback to Bikee by 'bedesign' after 60 Km
------------------------------------------------------------
When the issues described below are managed,
I would recommend the set.

Edit: bending = manual shaping
At the beginning of the shaping of the rear chain guide rod to
~45°, it broke like brittle porclaine.
This could be prevented by heating the piece with a heat gun
to 200°C - and not more.
WARNING! never use a gas burner - it will ignite, which is impossible
to extinguish.
The metal is possibly magnesium with a low content of alloy.

I manufactured a replacement from silicon-aluminium.

Edit:
I now tested a 2 mm piece of my replaced rear chain guide rod for
the risk of ignition while heating it to a bright red color,
Well, it didn't burn at all !
So I lower the warning to no or low risk.


The crank arm outer threads are not
parallel with the square hole
-----------------------------
I corrected the wobble to within 0.5mm by manually grinding the square
hole surfaces on only the offending areas to 50% of their axial depths,
just until the eloxation layer was removed at the inner and outer and
top and bottom opposite sides.
- (but this is not for the average customer)


The pinion is not totally stable
--------------------------------
The depth of the grove on the axle for the segger is not very deep,
and may allow the pinion to force the segger out of its
position.

I'm trying a modification that might eliminate the segger
by adding a total of 4.0mm spacers behind the pinion,
and a 1 to 2mm spacer at the front.
Then I added a thread locker and adjusted the screw until a
slight friction was noticed when the pinion is fully
pushed seated.

After a some tenths of Km runs, it's still totally stable,
but should be tested for longer.

Edit:
I had to change the 42T chain wheel to 36T because the chain kept
swinging down on top of it, despite there was clearence with
the chain stretched under load.
(I plan to later add the installation of another pulley to keep
the chain from swinging low and going back to 42T)

This resulted in the chain being too long now, and combined
with a bad, abrupt high gear change under load, it derailed
from the pinion and pulley and put a brutal sideway force
on them.
This resulted in a hard, unwanted test of stability -
but, there was still not the slightest trace of a wiggle.

So I think this simple modification can be relied on.


I needed your support for updating the display
before I found an odd fix, see further below.
--------------------------------------------
My mobile Android version is 13
and the display is not actively BT-bonded
to the phone, only listed as available -
as advised.

The App version is 94
original motor controller fw rev. 072
original display fw rev. 026

The update begins, but then the Bluetooth indicator on the display
turns off and resets with an error message in the App:
DFU Device Disconnected
Error type: 0
Error: 4096

Edit:
By using the Nordic nRF Connect App for monitoring while updating,
the connection was kept and not reset?
And the update to ver. 053 was sucsessful!
Though let's see for future occasions if that's still needed.

Edit:
Just received a morning surprise, an almost auto update (no QR code required)
of the motor controller took ~14 minutes to rev.127.0!

nRF Connect, 2024-06-09
DfuTarg (E8:19:F3:4E:AE:9D)
D 18:54:29.128 [Broadcast] Action received: android.bluetooth.device.action.ACL_CONNECTED
D 18:54:29.145 gatt.close()
D 18:54:29.148 wait(200)
V 18:54:29.349 Connecting to E8:19:F3:4E:AE:9D...
D 18:54:29.350 gatt = device.connectGatt(autoConnect = false, TRANSPORT_LE, preferred PHY = LE 1M)
D 18:54:29.361 [Callback] Connection state changed with status: 0 and new state: CONNECTED (2)
I 18:54:29.361 Connected to E8:19:F3:4E:AE:9D
V 18:54:29.383 Discovering services...
D 18:54:29.383 gatt.discoverServices()
I 18:54:29.561 Connection parameters updated (interval: 22.5ms, latency: 0, timeout: 4000ms)
I 18:54:29.725 Connection parameters updated (interval: 7.5ms, latency: 0, timeout: 5000ms)
D 18:54:29.925 [Callback] Services discovered with status: 0
I 18:54:29.925 Services discovered
V 18:54:29.931 Generic Access (0x1800)
- Device Name [R W] (0x2A00)
- Appearance [R] (0x2A01)
- Peripheral Preferred Connection Parameters [R] (0x2A04)
Generic Attribute (0x1801)
- Service Changed (0x2A05)
Client Characteristic Configuration (0x2902)
Device Firmware Update Service (00001530-1212-efde-1523-785feabcd123)
- DFU Packet [WNR] (00001532-1212-efde-1523-785feabcd123)
- DFU Control Point [N W] (00001531-1212-efde-1523-785feabcd123)
Client Characteristic Configuration (0x2902)
- DFU Version [R] (00001534-1212-efde-1523-785feabcd123)
D 18:54:29.932 gatt.setCharacteristicNotification(00002a05-0000-1000-8000-00805f9b34fb, true)
I 18:54:29.986 Connection parameters updated (interval: 22.5ms, latency: 0, timeout: 4000ms)
I 18:54:30.391 Connection parameters updated (interval: 22.5ms, latency: 0, timeout: 4000ms)
D 18:55:22.486 [Callback] Connection state changed with status: 0 and new state: DISCONNECTED (0)
I 18:55:22.486 Disconnected
D 18:55:22.503 [Broadcast] Action received: android.bluetooth.device.action.ACL_DISCONNECTED



It's very difficult to moderate the speed by the
throttle, especially with a 6 Km/h limit set
----------------------------------------
I think instead of a current control,
a PID speed algorithm would be better.


The torque sensor pulley position
---------------------------------
May make the chain grind against the sensor bar with
a chain width of 7.45 or more (with gears less than 9),
if no extra spacers are installed.
 
Last edited:
The lower/rear chainring pulley arm should be stainless steel. This turned out to be a more crucial part than I originally thought. When the chain comes off the chainring... it really sucks
Another important part is the plastic pulley holder. Note that it is attached to the pulley arm with a single screw. This means the assembly can pivot forward in the rough and tumble of normal bike operation. The plastic assembly can then run into the chainring with disastrous results. How do I know?
 
And my throttle bug is reproducible from cold start:

Turn on display.
Turn off display.
Wait for shutdown.
Hit throttle.

Results: throttle will activate and spin motor.
Expected results: throttle/motor should never spin with system powered off.
Hi Adam, thanks for the heads up. We fixed that.

You can download the APP rev 96 from the Download section here: Download user manual and APP for the Lightest ebike mid drive motor . After that you'll be prompted to update the Display Firmware. You can follow the procedure of the display update and the bug will be gone.
 
Another important part is the plastic pulley holder. Note that it is attached to the pulley arm with a single screw. This means the assembly can pivot forward in the rough and tumble of normal bike operation. The plastic assembly can then run into the chainring with disastrous results. How do I know?
@rivers the plastic pulley holder has a notch which engages with the chain guide rod to prevent it from pivoting.
 
@rivers the plastic pulley holder has a notch which engages with the chain guide rod to prevent it from pivoting.
Okay, thanks. My mistake. I had the pulley holder parts assembled out of order. Yes, there is an anti-rotation notch that engages the guide rod when the parts are assembled in the correct order. The instructions could be better.
 
There is a newer dyname 4.0 too and I think that one is even better noise wise if I'm not wrong.
These bikes use a standard race face bottom bracket that I think I will use on my bike if the bikee bottom bracket wears out.
 
Rider complains of "pedal bob" from the torque sensor PAS. Interesting because it has the exact same torque arm design, and I have the same complaint with the Bikee.
I thought there was a setting in app to filter out the "pedal bob", or give longer time constant. Does this not work? or does it give consequence of delayed reaction?
 
If that is the case, I will investigate. IMO, bikee should tune it higher to be more fluid at default. Most emtb riders actually want overrun. I have seen lack of overrun as a complaint on countless emtb reviews. The problem is, the higher the overrun, the more you are going to want brake e-cutoffs. If your system doesn't have them, it's a balance... but carrying the PAS through the full stroke won't kill anyone.

@nervagon , you have the APP, the Superhero pack and you can find the manual here: Download user manual and APP for the Lightest ebike mid drive motor . You can customize what you want from your smartphone. You'll find the deceleration ramp in the tuning section, that's the second slider. Please let me know how it goes.
 
Good news, i got mine hooked up and the driveline is straight af thanks to the adjustability of this drive.

1719244784587.png

With the bike upside down, moving the cranks by hand results in an infinite loop of strong acceleration until i turn it off. This is as predicted; the good news is that the motor spins.

Throttle does nothing regardless of the power mode set.

I ran the display code update as instructed
I received a motor update code, ran it, it had an error the first time, then succeeded. ( i now see race mode )

I have another code to run afterwards which disables torque sensing. This code gives me an error saying:
! CODE INCOMPATIBLE !

For versions of firmware, i get:
The app displays:
Display: 55.0
Motor: 01/08/2023 REV132

On the display at bootup it shows: rev 66

On the app when i start it, it says version 96.

Any trick to get the throttle to work? I will contact tech support about the firmware update problem.
 
Do you remember where the throttle toggle is? on the display or the app?
 
I looked through the manual and apparently i don't have the superhero pack applied even though i scanned the code, that's why i can't select the throttle i think.

Strangely enough i can select race mode, but don't have a 'race button'.

Did another round of updates and no dice, still waiting on an email back.
 
Yeah. I think i just updated mine yesterday.

Yeah i'm currently going back and forth with lightest on my current case:
- included PAS sensor with dual chainring conversion kit doesn't work or show as an option in 'modes'
- throttle unresponsive and non-selectable in 'mode'.

On the other hand, the controller with the superhero pack can be set to 35A. I'm assuming this is in line with the strength of the gears. I'd like to see this 35A mode work ( 1820W on my 52v batt ), so i'm hanging tight!
 
Sorry, no idea
 
Back
Top