Electric kart

[nuxland]

Tested acceleration yesterday. A123 voltage sage is huge when battery temps are below 25C
https://youtu.be/OiY16Rs5VCc
Corrected thermistor map also, so Sevcon shows correct motor temps.
But one full power acceleration from 0 to 140km/h raises motor temp 16C 8)
And half power from 0 to 114km/h raises 8C

 

[SlowCo]

That's quick! Must be a hoot to drive.

 

[nuxland]

Tested new motor with another track weekend. Had configured max battery current to 500A for profile 1, but is seems the config persisted also for profile 2 where I wanted 650A max
In red there is Maximum discharge current and sevcon is allways in profile 1 by default and I had configured 650A also in blue, but if I wrote 77 to yellow then sevcon automatically overwrites blue to 500A. I was hoping when I select profile 2 with button then it would rise to 650A but no, it was stuck in 500A and I could not enjoy 55kw when accelerating with full power.

Here is video also and it seems that BMS was having some voltage measurement error with cell nr 12. As it went lower suddenly. But when making second run the voltage was ok again, so I think it was nothing serious. Will go over connection just to be sure.
https://www.youtube.com/watch?v=ba7YpXqrQNU
Here is second run smoking tires As tires were hot after a few laps it was not easy to get them to slip.
https://youtu.be/zADbWQazAMQ

Also positive busbar is somehow 10 degrees higher than negative. Maybe there is some connection issue, will investigate that also.

Heres, how battery temp sensors are mounted, they are between plexiglass and terminals. Busbar sensors are ziptied to busbars.

 

[nuxland]

Wanted to test 0-100km/h acceleration and changed my rear spocket to 45T. So max speed 110km/h
Could not get the tires hot enough to prevent wheelspin at start, but under 3s is good enough for me.
Battery was also already 50% used as we tested motor cooling first (I have added a fan and scoop to the motor) and drive 12 laps.
As outside temp was 10 degrees then motor did not went over 71 degerees. So theoretically if outside is 30degrees then motor should not go over 100degrees 8)
But I will install dry ice compartment under the motor for next summer just to be on the safe side and able to use full power to motor more offten.
View attachment 1


Here is the acceleration video also: https://youtu.be/Y9QzMlLOzNo
The second cone was approx 60m distance from the start

 

[nuxland]

Figured out that I have not included my battery build in here. I have 0,5mm aluminium sheet between every cell to get the heat out.
Here are images for various stages of the battery. They are compressed together but I fear that both end cells get a higher compression than the middle ones.

 

[Doctorbass]

Good work! have you read the A123 design app note related to the chevy spark A123 battery developpment? they have alot of great advices.

If there is no traction on the tab of each cells, the pressure of all that cell pile should be uniform as pressure is equal in serie.. just like current is between a serie of various resistance :wink:


Doc

 

[steveo]

Very Nice build, Congrats!

 

[nuxland]

Found a not used road and tried 1/4 mile on my own. Cycle Analyst was counting distance and I got the time from my canbus logging application.
RaceRender was used to put it all together and also get the timings and speed by going frame by frame.

0-60ft: 2.187s @ 39,7mph or 64km/h
0-330ft: 5.089s @ 77,6mph or 125km/h
1/8mile: 7.809 @ 91.9mph or 148km/h
1/4mile: 12.098@102,5mph or 165km/h

https://youtu.be/HUnqy7w-L4E
It was +7 degrees outside. But my rear spocket is still too big for 1/4mile.
Will try again with 24T rear spocket in future when weather holds

 

[Punx0r]

That's crazy fast!

 

[Hillhater]

Thats a damm good Result for a kart.
Karts dont go well on 1/4 mile runs. Not enough rubber for good traction off the line and poor aero due to the sit up seating position down the the end of the run...Normally , very little speed increase after the 1/8 th mark.
...extension hubs with dual 6" dia rear wheels was a set up sometimes used for more traction , together with a "laid back"seat position.

 

[nuxland]

Thats a damm good Result for a kart.
Karts dont go well on 1/4 mile runs. Not enough rubber for good traction off the line and poor aero due to the sit up seating position down the the end of the run...Normally , very little speed increase after the 1/8 th mark.

I think my problem right now is transmission. I have 16T to 30T and I need 6000rpm to get 167km/h. But the problem is that over 3200rpm controller is exchanging torque to rpm.
Lowering Iq and raising Id in motor.
But what drives the kart forward is Iq (phase current) and it was lowered to 400A around 4000rpm and to 300A around 5000rpm. But 100A phase current with that motor is around 25Nm.
So if I change rear spocket to 24T I will lower torque to the wheel by 20% but have the same speed ~173km/h in 5000rpm already where I have 25Nm more on motor than with 30T.
And I need that torque in that speed because of Air resistance. It would help if I change myself to a smaller driver also, I'm 1,9m tall

Here is graph also for illustrating field weakening in that run.

 

[Doctorbass]

In fact what you need is max average power too..

I have similar goal as you for my Zero drag bike. i had top speed limited by lower torque at higher RPM.. so my 1/8 mile is excellent ( 7.3s) at 90% phase current but my 1/4 is 12.2s... and i only get 8 more mph during the last 1/8 mile..

I now replaced my ratio fofrom 4.4 to 3.6 and raised max phase curent to 100%.

I am running 22s LiPo and size 6 and 75-7 6 turns .

in your case you could use the 4 turn version of the 75-7 and reduce the need of field weakening at higher rpm..

Awsome build btw ! :wink:

Doc

 

[nuxland]

24T sprocket arrived and I did the test, Final speed was bigger but I lost time at beginning, so time was worse.

0-60ft: 2.356s @ 33,5mph or 54km/h
0-330ft: 5.602s @ 73,3mph or 118km/h
1/8mile: 8.307 @ 92.6mph or 149km/h
1/4mile: 12.626@106,8mph or 173km/h

It was actually not that dark outside, my old countour camera records poorly with low light.
https://youtu.be/AJnJBeFGBE8

But I did test one more test with that 2 different sprocket runs. With 24T sprocket I limited sevcon to 4800rpm to get the same 169km/h with full throttle.
I measured how much power I need to constantly drive at 169km/h. And it seems that about 13kw more goes to heat when driving with 6000rpm instead of 4800rpm, 41300w versus 28400w. Here are two graphs taken from canbus logging to illustrate all of this.



Here is also graph for speed vs battery power

 

[Hillhater]

Wow, 13kW extra power wasted at the same speed just due to higher motor rpm !
Now i am wondering if there are advantages for using multi ratio gearboxes on EVs rather than fixed ratio drives.
..Or is your motor at 6000rpm just operating way out of its efficiency zone ?

 

[Doctorbass]

In fact yes.. at about 4kRPM the sevcon goes into Field Weakening witch is less efficient.

Doc

 

[nuxland]

Excellent build and 1/4 mile result A few questions for you please:
Is your battery 28S3P?

Its is actually 32S2P A123 20Ah pouches, I will add 2S2P to the pack this winter to get voltage a little higher.

Are you planning aero mods?

No, 1/4 mile was just for the fun, I intend to use kart in race track where max speed is around 130km/h.
And my goal is to beat retro karts next summer

In your video it looks like your cells are near 50% DOD. Have you been able to try with a full charge?

Cells were 100% charged, just LiFePo4 voltage is 3,3V when resting, 3V when under normal load and 2,6-2,8V when under heavy load.

 

[nuxland]

Thanks for your reply. I've got a 27S3P pack here which will see only 10kW in next build. Previously used in 30kW peak setup. They have an easy life compared to yours.
I typically charge to only 3.45V/cell but they'll rest at 3.55+V if I charge fully. Once it drops below 3.375Voc the voltage drops fairly slowly to ~3.15VoC near the LVC cliff.

Is there an issue such as greater resistance in your endmost cell (far right on vid?), or is that cell facing forward and exposed to cold airflow?

yeah, it seems nr 32 cell pair has some issues. Actually they sit at the end of the pack facing rear wheel. I'm replaceing them if I add 2 more cell pairs this winter.
But I have tried to measure resistance but have not succeeded, even i have this until 4milliohm multimeter. It always shows me 0 ohm.

 

[nuxland]

Managed to upgrade my 32s2p pack to 34s2p and change two cells in the progress
Did a little test with brakes wheel off ground and now difference between top on low cell is only 0,11V with 500A load.
It used to be 0,25V so improvement is over 0,14V. I have to get battery warmed up also to test in real situation.

As I had to make new top and bottom plate, I managed to install neat connectors for CanBus and 12V.
Now I can just disconnect two little connectors and main Anderson and remove battery in 5 minutes.
I also have to remove two M6 bolts from back and two M8 from front that are holding it
I should add 1kg to the weight because it is without cover plexiglass and fixation bolts in the picture. So together 49Kg.
Also installed extra anderson connector, so I could charge it when not mounted to the kart.

 

[Arlo1]

Great work that looks very professional.

 

[riba2233]

Isn't that too heavy for 4 kwh pack? Could you make a pack with samsung 30q cells, it would be around 25 kg with box.

 

[catacon]

nuxland,

Have you been happy with the ZEVA BMS? I'm considering using it for a similar pack.

Thanks!

 

[nuxland]

nuxland,
Have you been happy with the ZEVA BMS? I'm considering using it for a similar pack.
Thanks!

For my need it delivers well. I already bought 3 more unit for my second pack.
But these units need BMS controller as well if you want to control something, they only provide you voltage and temperature over canbus.
I have a OLIMEXINO-STM32 for control charger and log data.

 

[riwi77]

Hi Nuxland,

Regarding the temperature difference of the motor and the controller, I've seen some strange behavior with the sevcon and kty. Did the motor temp change as soon as the motor started to spinn? I think it's a software bug. I saw it in sn0058.04

Was racing this sunday with retro karts again
Big thanks to Craig for selling me Zero 75-7 motor and Biff for helping me out with the config when I was stuck and all hope were lost

Had to configure sevcon "profile 1" to 4400rpm limit, because thats when field weakening starts. And with it the motor will heat up more quickly.
I also had 48% torque demand on that profile. But as I have a little button on wheel from what I could trigger "profile 2" where torque demand was 100% and rpm limit 6000 then it was fun using it 4 times in the race. But as you can see each using of it raised motor temp for 7 degrees instantly.

Here is video with data. I struggled with the motor conf in first 2 races, thats the reason I was easy on first two laps with the throttle in final race.
https://youtu.be/pbdnJMd8Reg

I have two extra temp sensors that are measuring motor external temp and controller extra heatsink temp. I have covered the motor sensor with foam tape so it does not gets cooled.
View attachment 1
Does anybody know why Sevcon's motor temp is 10 degrees higher than controller one when both are not used and should be at ambient temperature?
I have it configured as KTY84 in combobox as it should.

Total weight of the kart including me on board was 234.8kg, so its should be little lighter that zero motorcycle?

 

[LuigiDesing]

Hi nuxland,

How long does you take to charge the batteries? I´m following your progress and I haven´t seen this topic. Nice job!

This is the portable charger station I made to charge my ekart.

 

[nuxland]

Hi nuxland,
How long does you take to charge the batteries?

I'm using 3kw elcon charger and it takes about 50 minutes. Depends how much I use up the battery.
Actally charger is charging at 24A and usually i'm using 21Ah in a race, so yeah 52 minutes.
I added also extra charging port to the battery when redoing it, so now I can charge with two 3kw chargers if needed.