50cc auto scooter conversion to electric (with video)

[vanilla ice]

out accelerate most ( not all ) cars/motorbikes upto 30mph and can eventually reach a speed of about 55-60mph.

And get 80 to 100mpg while doing so. After reviewing the low rpm ends of a few electric motor sim charts I aint too worried about cvt inefficiency on this thing.

 

[swbluto]

out accelerate most ( not all ) cars/motorbikes upto 30mph and can eventually reach a speed of about 55-60mph.

And get 80 to 100mpg while doing so. After reviewing the low rpm ends of a few electric motor sim charts I aint too worried about cvt inefficiency on this thing.

You're thinking that the low-end RPM of motors is inefficient, right? That usually only occurs at "heavy loads" which shouldn't happen in real life if you gear it appropriately. Otherwise, lower speed on the flats would be obtained by partial throttle which the motor is far more efficient at. However, you get controller losses with partial throttle but my last calculations suggest that controller losses are somewhere around 30 watts - No big deal for 1000 watt loads - that's about a 3% loss. It might be as high as 5%.

 

[gwhy!]

out accelerate most ( not all ) cars/motorbikes upto 30mph and can eventually reach a speed of about 55-60mph.

And get 80 to 100mpg while doing so. After reviewing the low rpm ends of a few electric motor sim charts I aint too worried about cvt inefficiency on this thing.

You're thinking that the low-end RPM of motors is inefficient, right? That usually only occurs at "heavy loads" which shouldn't happen in real life if you gear it appropriately. Otherwise, lower speed on the flats would be obtained by partial throttle which the motor is far more efficient at. However, you get controller losses with partial throttle but my last calculations suggest that controller losses are somewhere around 30 watts - No big deal for 1000 watt loads - that's about a 3% loss. It might be as high as 5%.

When (if) I have this up and running what would be some good tests for working out the efficiency, What would people want to know ?.

 

[vanilla ice]

I'd be interested to see power use comparisons with the cvt locked out if that is possible. Low rpm is exactly where the heaviest loads usually happen. And on a scooter in the city, it repeatedly happens, block after block.

 

[gwhy!]

I forgot to put a pic up of the motor mount.... Hear it is
View attachment motor mount and crank.jpg
View attachment motor mount and crank2.jpg

I'd be interested to see power use comparisons with the cvt locked out if that is possible. Low rpm is exactly where the heaviest loads usually happen. And on a scooter in the city, it repeatedly happens, block after block.

I think this will be do'able.

swbluto: In your opinion the motor that I am planning to try this with what do you think the most efficient operation is going to be before the drive should kick in : the motor draws about 6A unloaded on 48v ( I will double check this when I get the chance ). With a cvt/clutch it is possible to spin the motor up long before it starts driving anything or do you think it should start driving the transmission instantly. We got to give it our best shot here so will there be any advantage spinning it up to a certain speed first ?

 

[gwhy!]

Bearings have arrived.
View attachment replacement main bearings.jpg
These bearings are 3 mm wider than the originals But this was not a problem, The main bearings would have to be replaced due to no oil will be reaching them so sealed mains was in order.

I fitted the modded crank in to the crank casing and this was the outcome:
View attachment step7.jpg

So far so good, no binding and runs smooth, tomorrow I need to make the spacers to mount the motor to the stator side of the crank casing.

 

[maydaverave]

Thats really cool. Can't wait to see it in action.

 

[fizzit]

The performance gains with a gear box for an electric motor are less than for an ICE because an electric motor maintains torque and efficiency much better across the RPM range... I'm pretty sure you don't know more than companies that make electric motorcycles (neither do I, obviously).

Well he is not an engineer designing a completely new vehicle he is trying to make a cheap, easy, fast, diy electric vehicle and I for one am very thankful for his efforts.

No I have no problem with the original builder using the gearbox, it seems smart and easy, I was responding to the person that said "I wonder why all of these electric motorcycle builders haven't figured this out yet"

 

[gwhy!]

I had some time to spin the motor up with a rc sensorless esc and no load on the motor it takes 5.96A @ 44v ( I still think this is really poor compared to the other motors that I have ) anyway, it seems from what other people have said about this motor that is about right. I'm hoping with hall sensors fitted this can be improved but Im not holding my breath.

The bad news is with the motor connected and powered up just through the crank the motor draws 15A @44v so there is a tight spot some where and Im hoping that it is only the motor mount needs a bit of adjustment, worst case it could be the crank is not true ( this may have been the result of having to rework the crank to take the additioal 3mm off for the wider main bearings ), if this is the case Im a bit stuck at the moment as of a way to quick fix it. I may have to remove the crank and rework it on the lathe which is something I would like to avoid if I can. But I think I am still on track to have the transmission all up and running by this coming weekend, fingers crossed.

1day goes by
update:
I took the crank out And stuck it back in the lathe in work and got some else to check it and it was slightly out and now its as good as its gonna get unless I start again. The crank runs a lot smother now but I think the motor mount may still need a bit of tweeking also, I havent re-fitted the motor yet I had a few other things I needed to do ( totally unrelated ) tomorrow I will re-fit the motor and adjust the motor mount hopfully I can get the crank current down to a decent level, then we are all systems go for testing

 

[el_walto]

I'm excited to see how this build will turn out.

 

[maydaverave]

Not as easy as it first sounded :lol: Hopefully you get this straightened out and it will be the hardest part.

 

[gwhy!]

Not as easy as it first sounded :lol: Hopefully you get this straightened out and it will be the hardest part.

Your not wrong , If I still cant get it to a decent current level with the motor connected directly to the crank tomorrow I have a plan B which is put a sprocket on the crank and mount the motor where the cylinder head was and run a 1:1 chain to the crank. I know this in it self will be lossey and a bit more noise but it will prob be the easiest way to make it more efficient just to see if it will work.

 

[maydaverave]

Maybe you could just run a fixed pulley upfront and just use the variable pulley on the clutch. It might give you more mounting and shaft options.

 

[maydaverave]

never mind that wouldn't work without the variator to pick up belt slack. Could you just leave the tranny cover off, mount the pulley directly on to the shaft, mount the motor on the outside and run it in reverse. These might be really stupid ideas but just throwing them out there.

 

[Thud]

gwhy,
Great thread & build. you beat me to this(only differance is I am using a manual tranny ala' frankG)

quote:

I had some time to spin the motor up with a rc sensorless esc and no load on the motor it takes 5.96A @ 44v ( I still think this is really poor compared to the other motors that I have ) anyway, it seems from what other people have said about this motor that is about right. I'm hoping with hall sensors fitted this can be improved but Im not holding my breath.

I am dangerously close to setting up a motor with external halls, Just need to design a bomb proof mount for the external halls. (my bikes all get jumped, & slammed occasionaly) I await your report regarding the no-load conditions especialy regarding this motor. (I have one & another from bubba to re-wind soon).
It seems this particular motor is showing a higher fail rate than others(?) I wonder if it is some dimentional harmonic flaw? (cart in front of horse there)

I like what you have acomplished so far, Having set up a few CVT clutches for snowmobile drag racing I need to ask, does your smaller clutch have replaceable weights & springs to "tune" the launch rpm?
Awaiting next installment.

 

[gwhy!]

Hi Thud,
I have already fitted external halls to this motor when I first got it did improve the no-load current slightly but I shelft the motor uptill now because I was not to happy about the 6A no-load current (this was more like 7A with the skirt bearing installed ), Im not totally convinced about the reliability of this (my) motor either. I have a coulpe 3.25kw motors fitted with sensors and no load current on these babys is only 2A @44v.

The clutches on these scoots do have removable/replaceable springs for tuning the RPM engagment point and I have always found them to work well if setup well.

I would be intrested to see how you get on setting this up though a manual tranny, good luck.

Cheers,

 

[gwhy!]

update:
With a bit of tweeking, pushing and pulling I have managed to get the current down to a reasonable level when connected to the crank. No load current with stand alone motor is still 5.96A@44v and now when connected to the crank its down to 8.5A @44v instead of 15A. Hopefully this may come down even more as the new bearings start to loosen up and there is the slightest binding in one spot there is also a very slight vibration. This is still rather a lot but it will be low enough to test the rest of the transmission tomorrow. I still think this will be the best way to mount the motor ( directly to the crank ) but it will have to be made to a lot higher tolerances that I can manage with the lathe that I have access to. After some tests on the scoot I should get enough efficency data and performace data to know if it will be worth making a better motor mount and the crank turned down by someone with a lot more patience than me :wink: or find a way of connecting motor to the crank via some sort of flexible coupling.

Edit: I have been think about mounting the motor above the crank shaft and running a 1:1 chain or gear this would be so much easer to get rid of any losses generated by the direct connection method and lot more do'able as a diy project, the only real dis-advantage with this method would be space taken up by the motor ( less space for batterys and a additional motor).

 

[Grinhill]

Hey great build thread Gwhy, I like seeing all the photos.

With the drilling for your motor mount plate, did you just rely on the paper template and hand-marking of the hole centres? I'm just wondering if it's possible that the motor shaft is a fraction off-axis with the crank, it would be hard to measure this though.

How is the motor shaft fixed into the crank hole, a couple of grub screws?

I had similar alignment issues on my second build with a 1/2" shaft extention, and I also considered the possibility of flexible couplings. Most seemed rated only up to a few thousand rpm.

 

[gwhy!]

Hey great build thread Gwhy, I like seeing all the photos.

With the drilling for your motor mount plate, did you just rely on the paper template and hand-marking of the hole centres? I'm just wondering if it's possible that the motor shaft is a fraction off-axis with the crank, it would be hard to measure this though.

How is the motor shaft fixed into the crank hole, a couple of grub screws?

I had similar alignment issues on my second build with a 1/2" shaft extention, and I also considered the possibility of flexible couplings. Most seemed rated only up to a few thousand rpm.

Hi Grinhill,
To get the hole positions I used a non-permanent marker pen and drew ( colored the flat edge of the crank casting ) then turn the casing over and pressed it onto a piece of paper. Yes the plate may be slightly off and like you say it would be very difficut to get it 100% correct, this is what I think will make this method to tricky to get correct. I have 4 grub screws on 4 flats on the motor shaft, I had 2 grub screws originally but I think even the slightest misfit in the hole can push the shaft off center from 2 grubs. I was thinking it may be possible to make a semi-flexible coupling using a 15mm bolt head and a 15mm socket ( primitive cv joint )but this would have made the motor stick out further from the casing.
If this works I will revisit the motor mounting method.

Here are some pics of the motor mounted to the crank casing.
View attachment step8 fitted motor.jpg
View attachment step8a fitted moto.jpg

 

[gwhy!]

Its ALIVE!!
View attachment step9.jpg


So far it is as expected, I didnt have any direct replacement rollers and this effected the gearing, all the testing I done today was on the stand with the back wheel off the ground, it was sticking in low gear most of the time and when it did eventually shift up it was not smooth and only ramped up to about halve way ( maybe due to the flat spots on the rollers )( wheel speed 52mph ) . When loaded in low gear it felt like it would pull a train and only was drawing 35A @44v . The current never seemed to rise above 40A no matter what I did. From a dead stop to WOT the current would peak around 60A then soon come straight back down to 16A'ish this was on the stand and No load. I am unable to do a proper road test until I get some replacement rollers, The clutch springs will need softening for a better launch rpm and I think the rollers will need to be a lot heavier than the stock ones ( I think the motor could take it ) . Anyway thats it for now till I get some rollers. Hopefully the next instalment will include some video :wink: .

 

[maydaverave]

Get some lighter weights and video.

 

[gwhy!]

Get some lighter weights and video.

I will stick with standard rollers for now just to see how it goes ( it maybe fine without the flat spots ). The motor wasnt being taxed very much and I think it could pull a heavy roller but I do like the idea of a lighter roller just maybe it will wheelie 8) . Its pretty easy to increase or decrease the weight of rollers without buying new ones , I think the clutch will need some adjustment though it grabs at about 6.5k rpm so the throttle is not very responsive there is only about 2k rpm of throttle range. Im thinking if the clutch engaged at around 4k rpm that will give me a lot more range on the throttle and also put the motor into non-taxing rpm for when it starts getting loaded up. The only issue that may be a problem is the motor temp It gets a bit warm, it may be possible to rig up some sort of blower off the crank ( sucking through carb opening and blowing out through the piston opening and directed to the motor ). Spose I should not really be thinking this far ahead until I have actually ridden it . Video will come :wink:

Edit: I will also try locking the clutch on to see what ( If any ) difference this will make in the real world with the current and gearing.

 

[gwhy!]

just a little update:
While I am waiting for some new rollers I thought I would do the clutch, I had some standard blue springs from a old yam50 auto ( soft ) Fitted them and now the rpm engagement now is a more acceptable 3.5k rpm, just got to wait for the rollers now .

 

[captainslug]

This project is very interesting.

 

[vanilla ice]

I have a couple of Yamaha scooters and some spares.. I'd send some rollerz to get this rollinz but I'm so far its not worth the freight! I assume that backlog of flights thing is over by now?