Modifying an Oxygen Lepton with Leaf Batteries and a Kelly Controller

MrBeeeeee

10 mW
Joined
Nov 8, 2018
Messages
22
Hi guys,

I decided to finally get in on the EV game by purchasing a pretty old Oxygen Lepton just for commuting and tooling around my neighborhood. I loved it, but hated the range that four lead-acid batteries in series gave me.

I endeavored to replace the battery pack with 10 Nissan Leaf battery modules setup in series (84v) and replaced the stock controller (made by Lafert) with a Kelly Controller KLS8415H.

The original controller was rated at 60A RMS and 100A peak with a 48v battery pack supplying. The Kelly is rated at 60A continuous and 160A peak with an 84v battery. The performance gain has been absolutely marginal (except for range). I'm not sure there's been in any improvement in top speed or acceleration. Monitoring the controller via bluetooth it looks like I'm getting expected current values when accelerating, but it drops to 30A as I approach 25 mph or so. If the motor is unloaded, it runs up to 45+ mph in a flash.

I am using the stock motor which is a puny little Lafert 1.8kw inrunner with a hall sensor array.

My question is, if I want to improve my hill climbing and ability to get off the line without any shudder, can I squeeze more performance out of the motor with a bigger controller, or am I just running into the limits of the motor's torque curve?

Here's a photo of motor nameplate.

And an album of the build progress
 
Here they are. I had to take screenshots of the Android App and stitch them together for this.
 

Attachments

  • kelly settings 1.png
    kelly settings 1.png
    298.8 KB · Views: 2,724
  • kelly settings 2.png
    kelly settings 2.png
    60.1 KB · Views: 2,724
  • kelly settings 3.png
    kelly settings 3.png
    191.3 KB · Views: 2,724
  • kelly settings 4.png
    kelly settings 4.png
    45.5 KB · Views: 2,724
Try changing your torque ki and kp settings to the highest rated number. I think it's something like 323678.

Also set your battery current back to 100. I think if you set it too night it will just ignore the value and default back to 60 or so.

That should help your hill climbing ability.

Always change the settings withing the range that the app tells you to.
 
So changing the PID settings for ki and kp to max definitely seemed to help off the line and hill climbing definitely seemed to improve slightly as well, but I was still hoping to be able to hit 30 mph easier.

I've attached a screenshot of the controller readout of it climbing a relatively gentle grade at full throttle. The scooter's speed topped out at 25 mph during the climb. On that same slope, I was able to reach 30 mph relatively easily going downhill.

So, if my ultimate goal is 35 mph under all conditions, do I need to look at upgrading the controller or the motor first?
 
Well with only 60 battery Amps you can't really expect much. If the motor temp doesn't reach some crazy values like 120C with the 60 Amps then I would recommend going to at least 80-100A with a drop-in replacement Kelly like:
http://kellycontroller.com/kls7230s24v-72v300asinusoidal-brushless-motor-controller-p-1343.html
or
http://kellycontroller.com/kls7218s24v-72v200asinusoidal-brushless-motor-controller-p-1401.html

I'm sure the leaf cells can handle hundreds of Amps but make sure you have a good BMS that can handle more. it will shut down or overheat otherwise.

I had a Johnway Lambda with a Kelly 7218S and a 72V Battery.
Oh man, that thing was a blast! It would do 44mph on the straight and most serious slopes with the same speed :D
The only time it slowed down on one was on a seriously steep hill. I might measure the actual angle of it.

I did have to mount a PC fan to it because the Controller would overheat on anything but the straights with no airflow.
 
Wait, do you think I'm running into an issue with the batteries not supplying enough current, or the controller not supplying enough current? When I come off the line I'm definitely seeing the readout go over 140A for "Phase Current" before coming down to something more nominal.
 
It doesn't matter what the Phase current is.

Power is measured by multiplying the drawn Battery Amps and Volts.
Phase Amps help acceleration and at full speed Battery Amps = Phase Amps

So your max output at any given time is 84Vx60A= 5Kw or less if you have a lower volt battery.
Unless you count that 30-second boost that Kellys have but I sure as hell can't rely on that to get me up a hill :lol:
 
So, how should I direct my efforts? Am I battery limited, controller limited, or motor limited? Can I apply a methodology to figuring out what my performance bottleneck is?
 
Like I said, Your battery can handle more than enough. Question is the BMS, what is it rated for?
After climbing the slope I would check the Motor temp to see if I can push it further. If the Motor internal temp is above 100C after climbing the slope I'd say it reached its limit.

The Controller is pushing 60A which is obviously not enough. Amps = torque, so going with any of the two I linked above you should see quite an improvement. Unless you have an 84V Battery then you will need an 84V Controller:
80A http://kellycontroller.com/kls8422h24v-84v220asealed-sinusoidal-wave-bldc-motor-controll-p-1452.html
100A http://kellycontroller.com/kls8430h24v-84v300asealed-sinusoidal-wave-bldc-motor-controll-p-1453.html

Also, check the Controller temp while climbing to make sure that it isn't reducing Battery current in order to stay cool.
 
I ran it up the biggest hill in neighborhood, down again, then back up and the motor never got warmer than 50C. Controller stayed cool to tbr touch as well.

So I guess I'll get the biggest controller I can mount and see what it takes to make that motor hot.
 
So I upgraded to a bigger KLS7250H to see if I could stretch more performance out of the motor. I replaced the old KLS7215H with the exact same wiring and configured the KLS7250H with all the same settings as the old KLS7215H, except I bumped the nominal motor current to 100A.

The "angle identification" procedure kept producing an error (Angle Identification Err) and wouldn't revert back to normal operating mode. I finally bumped the value for "nominal motor current" to 50A and the angle identification procedure completed successfully (I heard the beeping and the value changed back to '85'). When I engaged the throttle though, the motor would not turn in any direction. I didn't hear whining or beeping or anything strange. I could observe the throttle values change in the Android monitor application when I twisted it. I could also see the values for the hall sensors change when I spun the motor by hand. When I applied throttle and tried to spin the motor, I could feel some resistance, but not much. I could also see the value for "phase current" flutter in the range of 1 or 3 amps.

I haven't switched the old KLS7215H back in to confirm nothing is wrong with the scooter, but does this seem like a familiar failure mode for a Kelly controller?
 
Is the forward switch turned off?
Otherwise, the Controller will be in Neutral and you need to short the two pins for forward.
 
The forward switch is reading (0), but when I had the old KLS7215H in there, I didn't have it wired any differently and it went forward when the throttle was twisted. This screencap was from my old controller while the bike was under load and moving. https://endless-sphere.com/forums/download/file.php?id=243462&mode=view
 
Are you sure the three gears switch option isn't enabled?
Could you post the settings of the new Controller?

The beeping after the Identify phase is new to me, maybe they made some changes to the firmware?
It would be worth trying to connect the forward switch.
 
I'm almost positive it's unchecked, but I can post screenshots after I get home and have a chance to hook it back up. I'll also try connecting the forward switch to 12v and report back.
 
MrBeeeeee said:
I'm almost positive it's unchecked, but I can post screenshots after I get home and have a chance to hook it back up. I'll also try connecting the forward switch to 12v and report back.

Why would u hook to 12v man
 
Are you sure you're not using up the Oxygen's hub potential? Stick I'd say maybe you notice it a bit better than a moped, so the new controller adds a bit more but nothing dramatic. Then I think you reach that point only a hub change will improve things.

Had one, good choice to get rid of the controller. I think that was most likely my problem
 
Fastassmotors said:
MrBeeeeee said:
I'm almost positive it's unchecked, but I can post screenshots after I get home and have a chance to hook it back up. I'll also try connecting the forward switch to 12v and report back.

Why would u hook to 12v man

Because that's how the wiring diagram instructs you to wire the Forward switch wire.
 
Having owned a Lepton for almost 10 years, I'd say you're barking up the wrong EV. This is a great little scooter, but the drive motor is geared so that it's screaming at 25MPH (and it isn't a hub motor, folks, it's a chain drive with the motor near the swingarm). It isn't stopping at 25MPH because it's under-powered - it's doing that to protect it from blowing the motor with regular overrevving. To get more speed you'd have to find a way to change the final gearing. The easiest way to do that is with the tallest rear tire that will fit in the space. That might get you 30MPH. Beyond that you'd need to change the front sprocket and make a new chain cover.
 
So Fany hooked it up and looked at my configuration.

She had me change "Bat Current Limit" from 120 to 100. TPS Dead High was changed from 220 to 80. The Torque Speed Kp and Ki were changed back from their max values to 3000 and 200 respectively.

She also informed me that the value for "Nominal Motor Current" is only used during the identification angle operation. I now have a working controller, but there are still some issues.

When I accelerate, I have terrific torque off the line, but then it feels there's a dead spot at the top of the throttle. If I back off max throttle, it actually accelerates again. At full throttle I only get to about 12mph (going down the street), but if I let off the throttle about 80% (TPS value 200) it accelerates to 30 mph no problem. If I use the "Boost" button, I get the same behavior where I only get to 12 mph. Watching the response curve in the monitor application seems to show a linear response, but I'm not sure if my throttle configuration numbers make sense.
 

Attachments

  • Screenshot_20181209-130931.png
    Screenshot_20181209-130931.png
    35.1 KB · Views: 2,307
  • Screenshot_20181209-130947.png
    Screenshot_20181209-130947.png
    29.4 KB · Views: 2,307
  • Screenshot_20181209-131006.png
    Screenshot_20181209-131006.png
    31.4 KB · Views: 2,307
  • Screenshot_20181209-131017.png
    Screenshot_20181209-131017.png
    20.8 KB · Views: 2,307
If you haven't done so already, it would be a good idea to verify what the actual output of your throttle pot is. Wheel off the ground, check at your controller pin #3 to ground (#20) you should be getting a good linear reading of 0 to 5 vdc.(closed to WOT) If not, there is a cable adjustment available so that the throttle handle twist operation correctly moves the throttle potentiometer to achieve this. (or testing could show a defective pot, supply voltage issue, etc...)


Just in case you DON'T have it... :)


https://electricscooterparts.com/manuals/lepton-electric-scooter-service-manual.pdf
 
Back
Top