50cc -> 6kW conversion

[gwhy!]

I just edited my previous post as you were posting, not sure if you seen it.

Just read your edit and the linked post. I presume that all of those current measurements were taken are at full speed?

To clarify, the "skirt bearing" = oil seals?

I'm not going to try fiddling around inside the bearing/linkage unless it throws up any problems - it's been the hardest part to get right, and I'm a bit nervous that anything I do to it from now might end up breaking it further - if it runs acceptably as-is, I won't make any changes. I'm seeing 10A vs your 11A at the crank/shaft only, we'll see what happens when I put the belt on mine - I might even end up popping a multimeter fuse or two... (I can only measure up to ~10A until I sort out the Hall sensor - probably won't have time tonight). I don't think that no-load is going to be particularly representative of how it will perform, to be honest - especially because I stupidly forgot to measure current pre-mounting, and so don't have anything to compare it to.

I'll post back tomorrow (hopefully with some current readings, if not with a smoking multimeter )

The skirt bearing on the motor ( its the big bearing that screws to the inside on the outer of the motor can ). the no load without the belt ( just running the variator ) I would have thought should be around 6A @48v max at WOT ( if the motor on the bench @48v WOT was 4.5A ) , With the belt fitted as long as the belt is a good quality one and the pulley faces are in tip top condition I would have thought this would take around 1 or 2 A and the final reduction and clutch engaged driving the rear wheel upto full speed should be around another 1 or 2 A ( but this will depend on how heavy the wheel is, I never got round to tweeking mine to try and lower the no-load current at WOT through the cvt so these values are what I would be aiming for ( total 10A no-load @48v WOT ). The no load current was huge on mine the belt was not the best and the pulley faces were also worn and stepped ( around 30-35A ) but this would not change even with the read wheel locked stationary, the only time it hit the current limit of the controller was when accelerating , but onece a given speed was achived the current would drop to the 30-35A and be maintained even going up hills as the gearing would adjust itself .

[charliebruce]

Re current readings (SamTexas?) - I put a spacer ring under the motor to fix the mechanical issues/rubbing I'd been having (the base of my motor has some lovely deep gauges in it! All cleaned up now though, shouldn't recur). Certainly sounds smoother now. With motor and splined shaft in place (no variator or belt) I'm seeing current going from 0.7A at "idle"(120rpm?) up to a little over 10A (WOT). Will post updated readings with the variator in place, but probably not until I've got the Hall sensor set up instead of a multimeter (10A fuse).
...

That's video from testing earlier today. Sorry for the angle, and I realise that the meter's not ideal, but I didn't want to risk a good one since I didn't know how high the current'd go.

Thanks for the info. Let's make sure I understand what I saw in the video: You were doing a no-load test (driving wheel off the ground), correct? There are only 3 components in your drive train: Motor ==> CVT ==> Rear wheel, correct?

If so, the WOT 10A current at 48V is quite disappointing in my view. I know the CVT is going to consume some power but I didn't expect that much. I did, however, noticed that the test was very short. The 10.xA reading only lasted about a second, so it might just be showing a transient state. I hope/think that the current will be much lower (5A?) if you could keep the throttle at WOT for a few continuous seconds.

BTW, what is the capacity (Ah or Wh) of your 48V battery used in the test? And what type of Lithium battery is it? RC Lipo, Laptop Lico or what?

SamTexas: The engine components I have (alternator, flywheel, cylinder, spark plug, regulator, coil) come to 3.8kg. Add another 500g of piston/crank/bearing/engine casing I milled out, and a couple of kilos for the exhaust and that's the weight saving. Then remember that Li cells are heavy... (It's probably going to end up about 10-15kg heavier, overall?)

Ok, thanks for the numbers.

Nope, that's just Motor -> splined shaft that accommodates variator. I tested it again, it was actually nearer 8A when the value settled (took about 5s to settle from the time it reached WOT) - it has other problems though, this linkage is the hardest part.

@Gwhy! I left that bearing in, wasn't aware that it could be removed safely.

So, after testing it again, I attached the variator and spun it up - after a few seconds, there was a small but noticeable wobble on the shaft which led to the motor "stalling" (controller losing sync and resetting the motor?). The vibrations from that at WOT were small in amplitude but pretty intense (centering tool from a lathe in contact with the shaft - uncomfortable to hold because of the vibration). The grub screw linking the motor shaft to the splined shaft had also somehow slipped through. Since it was clearly not going to work like that I popped the motor out - the bearing was in contact with the base of the motor and was rubbing, and the process of removing the motor had damaged the protruding grub screw (head damage - should be removable with pliers). WOT with the bearing still attached was also resting around 8A so the power is either the skirt bearing. heating of coils. magnetic losses or the motor bearings/main bearing.

Stuff I don't know (just putting my thoughts in writing, helps me think about it - comments/suggestions welcome though):
Is 8A reasonable for WOT, or does this suggest that the bearing (which I left on) is eating too much power in the grease?
How the motor base managed to come into contact with the bearing - I suspect that might have been when I tapped it from the other side to remove the motor, and not happening when the bearing was in place.
How the grub screw managed to work its way loose.
How to fix it all and prevent vibration/slippage in future.
What I did differently last time, when it all worked.

I'm not going to be able to work on it now until the 22nd, so I've got a fairly long while to think it all through.

[datahar]

Laptop cells will be around 2.2Ah new, LiIon is only good for 2C, so 4.5A max current - that's not necessarily enough to spin your motor up, and is quite possibly your issue (mine spins up temporarily at a few amps before settling at 0.7A at the lowest throttle - I suspect the voltage is dropping too much as the motor starts, which resets the ESC). If I stick a resistor in series with my cells (simulating the issue you might be having with the cells), the ESC "twitches" the motor - it starts moving but fails to commutate properly, but each brand/firmware ESC might behave differently. Get a better set of cells or a bench PSU, or put a few cells in parallel (carefully!) and you'll probably be able to turn the motor.

Fantastic! That was indeed the problem, I attached the cells as 1(3s + 3s1p)s and had perfect motor function!

my new cells have been dispatched but what with the jubilee who knows when they will arrive, so my weekend will be spent on the motor mount now

Legal wise I think I am first get it MOT'd then register a change in propulsion as electric and then maybe just pay the £15 pound a year tax and keep it classified as a moped? Or do we reckon that sounds too easy?

[Jeremy Harris]

Legal wise I think I am first get it MOT'd then register a change in propulsion as electric and then maybe just pay the £15 pound a year tax and keep it classified as a moped? Or do we reckon that sounds too easy?

Sounds reasonable, although as soon as you have it registered as electric it will automatically be zero tax I believe. The helpful local DVLA chap (not the numpties at Swansea) told me that it would remain a moped in legal terms, but be free from road tax.

[SamTexas]

Sounds reasonable, although as soon as you have it registered as electric it will automatically be zero tax I believe.

Just curious: How much is the annual road tax for a 50cc motorcycle in England? How about annual insurance for the same motorcycle? Here in Texas, annual registration (road tax and what not) is $60 for my 500cc Kawasaki Ninja, annual liability insurance is $100.

[Jeremy Harris]

Sounds reasonable, although as soon as you have it registered as electric it will automatically be zero tax I believe.

Just curious: How much is the annual road tax for a 50cc motorcycle in England? How about annual insurance for the same motorcycle? Here in Texas, annual registration (road tax and what not) is $60 for my 500cc Kawasaki Ninja, annual liability insurance is $100.

Road tax is £15, so around $23 a year. Insurance depends very much on age, a youngster would pay a fair bit more than $100 a year here, whereas someone over 25 would probably pay a bit less than that, if it was just for third party cover (here we tend to have fully comprehensive insurance, though, which costs a bit more but covers accidental damage, theft, etc as well as third party liability). On top of this you also need a licence here to ride a moped, although if you're a car licence holder and passed your drivers test a few years ago you automatically have a moped licence.

[SamTexas]

Road tax is £15, so around $23 a year.

Thanks. I thought the saving would be a lot more. Sounds like you guys convert to electric more for fun/green than economic reasons, right?

[Jeremy Harris]

Road tax is £15, so around $23 a year.

Thanks. I thought the saving would be a lot more. Sounds like you guys convert to electric more for fun/green than economic reasons, right?

Yes, in part, but remember we also pay over $8 per US gallon for fuel here, more than that if it's a two stroke, because of the high oil cost................

[SamTexas]

True. But even at a low 100mpg ($0.08/mile) it's about the same as the wear and tear cost of battery, in my estimate.

[datahar]

True. But even at a low 100mpg ($0.08/mile) it's about the same as the wear and tear cost of battery, in my estimate.

Well with fuel going up with no end in sight I'm hoping that the battery wear and tear will evetually win, but personally I'm converting for green and to have a (hopefully) reliable machine which will save money on constant repairs

[Arlo1]

you may even have to try to lighten the clutch shoes ( so they fling out at even a slower speed )

gwhy you are a smart mo fo and a great person on ES but the clutch shoes swing out based on their weight and rpm so there for if they are lighter they will fling out at HIGHER speed!

You have this backwards buddy.

[gwhy!]

you may even have to try to lighten the clutch shoes ( so they fling out at even a slower speed )

gwhy you are a smart mo fo and a great person on ES but the clutch shoes swing out based on their weight and rpm so there for if they are lighter they will fling out at HIGHER speed!

You have this backwards buddy.

The lighter something is the lower the speed it will need to fling it outwards, its true , honest

edit : I think , But it will have slower acceleration outwards

[Arlo1]

you may even have to try to lighten the clutch shoes ( so they fling out at even a slower speed )

gwhy you are a smart mo fo and a great person on ES but the clutch shoes swing out based on their weight and rpm so there for if they are lighter they will fling out at HIGHER speed!

You have this backwards buddy.

The lighter something is the lower the speed it will need to fling it outwards, its true , honest

edit : I think , But it will have slower acceleration outwards

No dude. If something weighed 0 it would never have any force to make it fling outwards no mater how fast you spun it. You put something heavier in and it will fling outwards at a lower rpm.
The way to look at this is a weight in motion wants to stay in motion but it wants to go strait when it is spining in a circle it still wants to travel strait this is how centrifugal force is created.

Another way to look at it is you had to put very heavy rollers in the front clutch on your scooter... Why do you think that is? It is so the added weight will make the rollers slide out words at a lower rpm causing your clutch to shift up at a earlier rpm which works better for the electric motor because of its added torque!

[gwhy!]

No dude. If something weight 0 it would never have any force to make it fling outwards no mater how fast you spun it. You put something heavier in and it will fling outwards at a lower rpm.
The way to look at this is a weight in motion wants to stay in motion but it wants to go strait when it is spining in a circle it still wants to travel strait this is how centrifugal force is created.

Another way to look at it is you had to put very heavy rollers in the front clutch on your scooter... Why do you think that is? It is so the added weight will make the rollers slide out words at a lower rpm causing your clutch to shift up at a earlier rpm which works better for the electric motor because of its added torque!

Arlo1 of course you are correct ( thanks for correcting it ( me ) ) and I really dont know why I had this in my head , I think I must be going a big crazy or something ( thats a few things recently I have been ass about faced with ) , I keep getting mixed up with why you do things with a gas engine of course making the clutch lighter puts the engagement point at a higher RPM which is defo not what you want with a elecy setup, so heavy clutch shoes with no ( or very weak clutch springs ) are the order of the day as there is no real power band for a electric motors as there is with a gas motor.

[mauimart]

Another way to look at it (through physics):

F_c=(mv^2)/r

Given fixed tangential velocity and radius the centripetal force is proportional to mass i.e., more mass will overcome the springs restoring force sooner for a given speed and radius.

[charliebruce]

I'd read "lighter" as "weaker clutch springs" so no issue there. The mechanical/linkage issues are the real deal-breaker at the moment not the clutch engagement, and they'll have to wait until after exams now.

[charliebruce]

Dragging this thread back up again with an update.

To stop the vibration/play in the shaft, I removed the kick-start and modified a live centre to fit in the place where the kickstart "contact" was:





I re-assembled, spun it up and the motor/controller didn't seem to "stall" or have any issues like it was before, and it might have been my imagination/the lower RPMs but it sounded quieter with different vibrations. It didn't spin up as high as it previously did, so I measured the pack voltage and it was right down at about 38v (12s) so I took it home for charging. I first checked each cell voltage before slow-charging and found that one of them had dropped to 1.3v without recovery, another at 2.7v but recovered fine when slow-charged, so it looks like I'm going to have to remove the dead cell and rebuild this pack before I can properly test it

Hopefully, the live centre will do its job and it's all now mechanically sound.


Considerations now:
I want to design a 3ph controller for e-bike/other projects, but I also want to get this vehicle road-legal. What I'm considering is getting it tested and road-legal using the current pack (dead cell removed, 11s), then when I've designed and built a decent controller, replace the cheap chinese controller with my version and a LiFe pack (Victpower or sponsored, running at a higher voltage so I can hit 10k motor RPM). The other option is to get a 48v pack and continue with the current controller.

Questions for you guys:
1. How high a voltage can I safely (=>5k miles before dying) run a C80100 both with and without cooling? This is with a modded variator, now 14g rollers up from about 8g. I've heard anywhere between 72v and >100v as an upper limit, no definitive answer.

2.Do you think I'd be better off just sticking at 48v and calling the project "finished", and hope that the increased weight in the variator means that it'll reach 30mph, or will the increased voltage/speed be worth it?

Controller thoughts so far:
dsPIC base, 3ph, torque-based control, 24khz, IGBT.
Synchronous model >100V limit, 160A, sensorless possible, regen-capable, air-cooled.
Asynchronous model 600V limit, 600A, sensored only, regen-capable, water-cooled.
Both with thermal + emergency cut-off, potentiometer or serial control, current-limiting and voltage cut-off.

[Arlo1]

If you up the voltage it will spin faster for better cooling. But it may run hotter if you run the same amps or more. As well there will be more eddy current losses.
As for the controller??? Is this a kit or diy?? Dont use igbts below 200v use mosfets!

[charliebruce]

The aim is for higher voltage, slightly higher amps (so that I can hit the higher speeds). It currently runs cool enough (58c on coils after a decent run, no fan), but if I do run at a higher voltage I'll be keeping a close eye on temperatures. The way it's set up makes fan cooling easy enough if necessary. I'm just wondering if the efficiency quickly drops off at a higher voltage, or if the insulation will fail at higher voltage, and if so at what kind of voltage this is.

I'm thinking of designing and building a controller myself - 2 versions: a high voltage, high current, asynchronous IGBT one for induction motors, and a lower-voltage, lower-current synchronous version for e-bikes and lower-powered vehicles with permanent magnets. I've not decided on the switches yet, but I'll definitely look into MOSFETs for the lower-power controller. The aim is to keep power and logic almost completely separate so I can keep code and PCBs similar