Converted alternator on ebike controller

[Ekart]

Hi Stan, thanks for your thoughts and ideas, yes varying the field current on the alternator rotor would be good and I have worked with Arduino quite a bit with multiple sensor types. An inductive sensor could work well to monitor Alternator rpm and adjust current through windings accordingly. I had thought to add a motor cycle starter motor for initial takeoff power also... I will add some permanent magnets also and I will look for the article on automotive alternator conversion / tests you mentioned.

 

[stan.distortion]

The magnets in the alternator looked more trouble than they where worth when I looked into it, getting them between each pole on the rotor seems pretty common on those "energy assist" alternator/motor things some cars have, tricky to machine though, an engineering company might charge more than the cost of the controller. Extra circuit protection looked like the better option to me but I've no idea what to do there, I was just going to stick fuses everywhere initially.

Have you plans to do more with it? If so I'm looking forward to seeing how it evolves, automotive alternators have loads of potential and a kart is ideal.

EDIT: You'd really need a dyno to get the field current right and I'm not sure if it would have any major gains over a simple vary with rpm. Any flywheel would work for an inertia dyno, it's going to have some sort of rotation sensing anyway so just need to read in time between pulses.

 

[markz]

That smaller bolt/nut is to hold the shaft.
Must be common on all alternators then.
Interesting

 

[Ekart]

Hey markz, here’s my first project you may like .... an ebike using a printed circuit motor (pancake motor) running off two 12v 18Ah batteries in series. The motor is from a mainframe computer reel to reel tape drive ( from about 1976) I had kept for a rainy day. I used a DC controller for speed adjustment (under the seat) and potentiometer on the handle bars. :lol:

 

[markz]

When I was at the junk yard, a lot of older trucks had a narrow pancake like alternator.
I think I will still make a junky bicycle into an alternator ebike. It will be a fun project! and I can lock it up without worrying about it, like today I went to a big mall to watch The Joker, then across the street I went to Marks clothing store (chain) then across the other street a yupee natural foods grocery chain (I wanted to find low carb bread or wraps), then hop back on the train, then to a normal grocery store, stocked up on 4 bags of groceries. But since I hoofed it, and bussed it, and took the train my feet were tired! I could have gotten a lot more done if I was on an ebike.

I have to ask, how much power are you getting out of the motor?
I dont know how you'd explain that, but perhaps going up a steep hill with what weight.

 

[Ekart]

Hi markz
Yes I saw Joker as well.. very “dark” movie, the first of a few I think!
The pancake motor is rated at 36v dc and draws about 11.6Amps, so thats 417.6 watts which is just above 1/2 hp roughly, but I had it running on 24v only 2x12v in series, so maybe 1/3 hp realistically. It was ok once moving as a “helping hand” but not much more. I should fit another converted alternator? I have a motorcycle starter motor which I could use??

 

[stan.distortion]

Starter motors are no good, they're designed to work for short bursts only and even at that they're not particularly powerful. Car starters typically put out between 1kw and 2kw and will overheat if run for long, it would be possible to re-configure them for constant duty operation but it wouldn't be worth the cost and effort.

They're a relic and as strange as it may sound the auto industry is full of them, up to and including piston engines (and I'm a petrol head saying this). If the whole starting system was replaced with the same kind of drive train used on e-bikes it would weigh less than a typical starter motor and that's for everything, battery, motor, controller, wiring and would likely combine the alternator and starter in a single unit.

The reason an auto maker doesn't make those kind of changes is simple, the rest of the industry would pull out their patent portfolios and sue the crap out of them because that would cost less than re-tooling to keep up with the tech. "Tucker, the man and his dream" is a film worth watching, he tried to make all kinds of advances and the rest of the US auto industry came down hard on him for it.

 

[Ekart]

Many thanks Stan...that's great info and saves me wasting my time. I had thought to use it to "kick start" the ekart from standstill but probably not worth the effort and extra weight/cost involved. I will check out "Tucker, the man and his dream"..

I appreciate all the great info you guys are willing to share and help that a newbie is still learning about... cheers Guys!

 

[stan.distortion]

I can't find the thread I'd mentioned earlier on using an alternator as a motor :/ It might be on a different forum but all my searches are coming up blank so far, hope it's not gone for good as there was some really good info in it. Here's some of the dyno charts from it, iirc he was running a standard 12v alternator at 60v for them:

Over 3kw continuous, not bad at all for something that can be bought for less than $100 new and little or nothing used and iirc he was confident around 6kw continuous would be possible with a modified alternator. If you're looking for more power then your batteries are by far the weakest link imo, for mobile applications lead acid batteries are only useful for keeping wheelchairs upright. The main reason is their capacity is nowhere near its calculated value, Ah x volts:
https://en.wikipedia.org/wiki/Peukert%27s_law

 

[Ekart]

Wow thanks Stan, those charts really show the efficiency changes with those field current inputs.
Yes the lead acid batteries were a cheap starting point. $50 for the four. But I’m looking at around $3-400 for lithium packs so thats the next step. Even making your own is not cheap .. 18650’s are about $3.50 ea I think?

 

[stan.distortion]

Yes, there's way more variance from changing the field current than you'd expect. The first time I hooked an alternator up with a variable power supply on the field (LM317) I was almost afraid to run it up to full speed, they'll often run at over 20,000rpm under the bonnet but it's scary to see one doing half that on the bench!

Just as a note, it's not necessary to remove the diodes to run an alternator as a motor and it can be useful to keep them in place as they can be used as for regenerative braking. I'm not sure if you've pulled them out or not, I did and it was useful for changing between star and delta but I should have at least kept the diode plate in a usable state (I broke it up to get at the field connections).

For the batteries, it's a catch 22. You need to know how much power the motor can draw before you can spec the pack but you need a bigger pack before you can test how much power it can draw. I went with RC lipo packs for testing, not much capacity but masses of current. If you go that route be sure to read up on them first, they're much higher maintenance than 18650s and under the wrong conditions they can burst into flames! The warnings about them can't be understated, at least one forum member has had their house burn down but for power/cost and power/weight they're pretty much unbeatable.

The controller will be the limiting factor after the batteries. I'd done most of my testing with a cheap RC ESC, 80A 6S (24v), about $15 and good for about 1.5kw. I've abused the crap out of it and it's still going strong in another project that's had 100s of hours of use but the general opinion here seems to be RC ESCs aren't all that dependable. Considering that's coming from folks who've done tens of thousands of miles on ebikes it's probably advice worth taking, I'll be sticking with them for off road projects but would get something more tried and tested for projects that can leave me stranded miles from home.

 

[Ekart]

Hi Stan thanks for all that great information!
Running an alternator at 20,000 rpm you would hear it coming down the street!
The RC 24v /80Amp batteries sound good! Where can I get those please?
I bought my 1500w 48v controller for the ekart from Aliexpress for AUD$39. Which I thought was pretty good. A friend of mine from Mens shed, (Aussiejester) built a rippa ebike complete with gearbox, fibreglass covering panels and electric - lifting stabilising wheels (whilst he gets onto it from his wheelchair)!!! See here the motor he rewired and tested. Turnigy CA120 https://m.youtube.com/watch?v=6ioKGpAgMBo

 

[stan.distortion]

I got the batteries on ebay, not going to link it though because I'm pretty sure the spec was overstated. There are plenty to choose from on ebay but it would be worth hunting out reputable suppliers as batteries on ebay are a minefield, $200 should get enough power for at least a few minutes of full power testing.

RC LiPo info can be a little confusing, typically something like "3S 30C 5000ma/hr". They're pouch type cells and that means 3 pouches in Series, max discharge 30x Capacity and 5000 milliamp hour or 5A/hr capacity.
The pouches are rectangular with a nominal voltage of about 3.85v so 3 in series would be 11.5v.
30C would be 30x5A/hr so 150A maximum but that figure can be tricky, it could be for just 30 seconds or even just 1 or 2. The packs continuous rating is the main thing to look for and it should be well above the current you need, 2 or 3 times or more if possible.

5000ma/hr is probably about the minimum to get any kind of run time, probably only be a few minutes but it's no problem to add more packs in parallel later, all going well it would probably be better to move to 18650 cells at that point. I went with a couple of 6S 6000ma/hr packs, about 24v in series or 48v in parallel and that worked out well but in hindsight 4x 3S (12v) would have been better, I've never needed 12v or 36v but it would be nice to have the option.

Those packs need a special balance charger, I got an Imax B6AC, cheap and cheerful, maybe a little slow for high capacity packs so it might have been worth getting 2 or a more upmarket charger. Mine is a clone (most available are clones) and it's been 100% so far but some folks have balance voltage issues with them. BMSs are worth looking into early on and may be good to use instead of a balance charger, they'd be needed for 18650s anyway so nothing lost.

It sounds like you got a good deal on the controller, you should be able to push it up to full current and might find it's enough. If it holds up ok and you want more power then it would probably be better to go with a higher current model from the same supplier than with an RC ESC. I've been impressed with the RC ESCs so far, loaded them up as hard as I can to see what they'll hold up to but every application is different and they just don't have the bulk of a purpose built controller, get them in a state where they're heating up and they couldn't take it for long.

I tried to find the vid that got me interested in alternators, an eastern European loony with a Kart but it looks like it's gone from youtube. Impressively quick and it was very hard to believe a couple of batteries no bigger than a bag of sugar could put out so much power and the whole system could be built for less than $300, turned out to be true

 

[Ekart]

Thanks again Stan for your great info and knowledge on the subject, I really appreciate your help!
I will look into those cell packs and do plenty of research before buying. I’m also interested in building my own “power wall battery pack” for my solar panels which will be about 10KW as the Tesla ones here in Aus are about $10-15k!
So maybe 18650’s are still the way to go with BMS’s, I’ve seen a few home made one’s for 1/3 the cost and built in blocks to allow for expansion. In Perth we get credit of $0.07c/KWh for power we upload to the grid and are charged $0.24c/KWh for power we use from the grid. Cost of a decent quality 6KW solar system with 24 panels and inverter is about $3,800 installed, but the battery pack is where they get you on top of that!

 

[stan.distortion]

If you're planning to build solar storage then 18650s could be a better option than RC lipos for the kart, get some practise on a small pack for the kart before moving to the big pack and it doesn't matter if you go over the top with the spec on the kart pack as any extra can go into the house pack

I know what you mean about the Tesla power walls, maybe I'm missing something but it doesn't look like they could ever give enough return to pay back the purchase price. I'd go with lead acid for that job, new or refurbished forklift packs but that's just personal preference, li-ion is acknowledged as being ahead of lead acid in just about every application including simple price/capacity/lifespan with no regard to size and weight.