First post, electric motorcycle build.

General Discussion about large electric scooters and motorcycles and other things with no pedals.
phate   10 W

10 W
Posts: 70
Joined: May 28 2018 2:17am

First post, electric motorcycle build.

Post by phate » May 28 2018 6:36am

I've been lurking for a while, and finally got a start on my electric motorcycle project. I figure I will document my (slow) progress here.

I'm a mechanical engineer working in the controls industry, and lately I've felt the need for a new project that will stimulate my need to design/build things. I screw around with 3d printing, RC cars, and real cars (hondas and subarus mostly).

Why an electric motorcycle?
- I don't have room to get another car to mess around with, so I want to get a motorcycle.
- I've been amazed by the performance offered by the lithium/brushless combos that are becoming ubiquitous in the RC world
- It only needs to be practical enough to commute 20 miles per day, otherwise it's a toy.
- I like the idea of being able to hoon around on it at low-ish speeds without annoying people.
- I can get some decent performance out of it without spending a ton of money (budget around $5000).
- I helped with a much more tame conversion a few years ago at school (sub-40mph Etek/Alltrax/batteries I don't remember on an old GS550 frame)
- I can feel less-bad about the environmental impact of my modified subaru.

Basic goals before I dive into the details:
- Easy to register for road use. If the donor bike has clean paperwork, I basically just tell the DMV what I started with and check "electric" instead of "gasoline" as the fuel type.
- Small: Designed around a Ninja 250, but open to use whatever entry-level full-fairing sport-ish bike I can scoop for cheap on craigslist.
- Lightweight, low center of gravity
- 25 miles will get me to work and back, but I can charge at work if I completely screwed up my range calculations.
- Highway-capable, geared for 70-75mph top speed: I'm not planning to ride this thing on the highway, but I'd like to not die if I have to do it for an exit or two.
- Strong acceleration at low speeds. I'd love it if wheelies were possible, but it at least needs to be faster than most things in the 250cc range up to 40-50mph.

I'm building the bike in a few steps to allow me to work out issues before I spend too much money on things that aren't going to work:
Phase 1 - Battery and Charging system.
To get the bike going, I'm building a smaller (2.6 kWh fresh off the charger) pack that will be expanded once I have it working well enough to ride around and decide I want more range. Also I can't really do any motor/controller testing without the battery pack, so this only makes sense as a starting point.

I'll be running a 20s4p pack of SPIM08HP lipo's housed in a case made of 3d printed ABS. If the specs aren't complete BS, the cells are 8Ah rated at 25C discharge/8C charge. All of the cells I've cycled seemed right around that 8Ah mark, so our total capacity seems pretty believable at least. I'm pretty sure these were salvaged from a chinese hybrid electric bus. If the 25C rating is accurate, the pack should be capable of kicking out 800 amps, which is more than I'll ever really need to be honest.

Messing around in Onshape, I've more or less figured out the overall shape and size of the pack. The cells will be held together with 3M tape, and the crimped area around the pouch will locate the cells in the enclosure. I've got plenty of room for BMS wiring, temp probes, pretty much whatever. With the enclosure it's a bit over 14in long, 11.5in wide, and about 9in tall. Should end up right around 60lbs assembled pack weight.
Image

I'm doing experiments with smaller 4s and 2s2p packs to work out any issues I might run into with the battery case construction and securing the cells. They'll most likely end up running my daughter's power wheels for a chunk of the summer.
Image

Currently, I'm just using my hobbymate elite to charge/cycle a few cells at a time, but I'm planning to bulk charge with either an S1600 or S2500 from BMSbattery. At home I will be charging from a 120v outlet, but at work I have access to 220v for faster charging. J1772 capability falls firmly in the "I want it but we'll worry about that later" category.

Phase 1.5 - Battery Management and monitoring.
For battery management, I'm going to take a swing at using a one of the several arduino mega's I've got kicking around and a few multiplexers to monitor individual cell voltages. Cell balancing will be handled using a relay (solid state or otherwise) to connect the 4 cell module to a beefy resistor. If I run into problems making it work on the smaller packs, I'll go with an off the shelf solution.

Once the cell voltage data is in the arduino, I can either send drive a simple display right off the arduino, or I can pass it to a raspberry pi if I want a bigger/nicer screen.

Phase 2 - Powertrain
Once I have the battery pack together I can start playing with the motors and controllers.

I'm going to attempt to run a pair of water cooled 50kv sensored MP12090's (the same ones you'll find from Alien Power Systems, Ehobbypower, FRCHobby, etc) with a Kelly KLS7230S for each of them. I'll be running a primary chain with a slight reduction, with the goal of locating the countershaft/drive sprocket in roughly the original location and using more "normal" rear sprocket sizes. The motors will share a single, longer shaft. I'll attempt to use the original radiator from the donor bike, but the cooling system really is not that challenging to figure out.

The motor mounts and primary are going to be made out of some aluminum L-channel.

At the maximum rated current, these guys should kick out around 30-35lb-ft of torque each if the specs are not complete BS, so geared to hit 75mph we're looking at more than 250lb-ft to the tire, which should be plenty of fun to rip around with.

Phase 3 - Donor Bike
I love the styling of the Ninja 250, and was pretty comfortable on the last one I sat on, so that makes the choice obvious, but really I'm open to any full-fairing entry-level sporty bike I can get for cheap on craigslist. I'd really like to stay full fairing for both aero as well as the bike looking much cleaner with all of my electrical nonsense covered up.

This is where the general maintenance, teardown, and actual conversion will happen.

Phase 4 - Nice-to-Have's
So once it's working and sorted out, we've reached the "nice to haves". These are mostly things to make it more useful like the addition of J1772 charging and a big pile of additional batteries to bring the total pack size up to 5-6kWh, which should be enough to bump the range up to 60 miles or so.

User avatar
Dauntless   100 GW

100 GW
Posts: 6965
Joined: May 29 2010 1:49am
Location: Coordinates: 33°52′48″N 117°55′43″W

Re: First post, electric motorcycle build.

Post by Dauntless » May 28 2018 2:57pm

How about a donor bike you can get cheap new? Do away with the Chinese engine; new brakes, new seat, new tires, etc. $1,549 before you sell off the unused drivetrain.

https://helcher.com/collections/motorcy ... motorcycle

Image
Any sufficiently advanced technology is INDISTINGUISHABLE FROM MAGIC!
- Arthur C. Clarke

phate   10 W

10 W
Posts: 70
Joined: May 28 2018 2:17am

Re: First post, electric motorcycle build.

Post by phate » May 29 2018 12:31am

That's awesome, I had seen the Grom clones, but I hadn't looked at any of the "Ninja" clones. That's certainly in the running, but I'll have to look into what kind of parts availability I can get.

The Boom 2kw electric Grom clones seem interesting as well for $1200. Crap out of the box, but grab one of the 8kw QS hub motor/controller combos for $700 and some lipo's and it'll be a blast. If my own EV conversion is as fun as I think it's gonna be, I'll probably try to pick one of them up in the future. https://e-grom.com/products/e-grom-fly

I've already got enough chinesium in this build, I might as well go all-in.

User avatar
amberwolf   100 GW

100 GW
Posts: 25477
Joined: Aug 17 2009 6:43am
Location: Phoenix, AZ, USA, Earth, Sol, Local Bubble, Orion Arm, Milky Way, Local Group
Contact:

Re: First post, electric motorcycle build.

Post by amberwolf » May 29 2018 2:16am

May I suggest using some Meanwell HLG type LED PSUs as chargers? They are waterproof fanless units that you can bolt to the bike without worry about vibration breaking them, and you can series and parallel them as needed to get the charging voltage and current you need. They work on 120v or 240v ac input.

I use an HLG-600H-54A built into the SB Cruiser trike for typical charging on the go or at home (600w+), and a parallel pair sit at home for the times I need rapid charging (1200w+). These plus another unit were used by their previous owner in 2 series 2 parallel for on-bike charging of a motorcycle, IIRC.

User avatar
flippy   10 kW

10 kW
Posts: 851
Joined: Aug 12 2015 3:07pm

Re: First post, electric motorcycle build.

Post by flippy » May 29 2018 2:30am

if you go with only a 20S solution i recommend just getting a off the self BMS with a decent current rating. check the bluetooth bms topic for tips. if a pack is balanced yo dont need much to keep it there. and most smartbms solutions can balance manually, just hit the button and wait. costs 130ish bucks with free shipping for a 300A model.
Lithium beats liquid dinosaurs.

phate   10 W

10 W
Posts: 70
Joined: May 28 2018 2:17am

Re: First post, electric motorcycle build.

Post by phate » May 29 2018 4:35am

Thanks for the feedback so far! I feel like I was mostly-there as far as figuring out how to make this thing go, but was definitely weak on a few details. I've never had a need to charge anything larger than 4S or so, and aside from occasionally recovering of over-discharged lithium packs I really haven't done much aside from selecting voltage/charge current and letting my RC charger do its thing.
amberwolf wrote:
May 29 2018 2:16am
May I suggest using some Meanwell HLG type LED PSUs as chargers? They are waterproof fanless units that you can bolt to the bike without worry about vibration breaking them, and you can series and parallel them as needed to get the charging voltage and current you need. They work on 120v or 240v ac input.

I use an HLG-600H-54A built into the SB Cruiser trike for typical charging on the go or at home (600w+), and a parallel pair sit at home for the times I need rapid charging (1200w+). These plus another unit were used by their previous owner in 2 series 2 parallel for on-bike charging of a motorcycle, IIRC.
Those look like a water resistant and less-janky version of the power supplies everybody uses on their 3d printers (which I think are also supposed to be for driving LED's). If I understand Meanwell's naming scheme, the middle number is the power output, and the last number is the output voltage? So if I wanted to charge my 20S pack to the full 84 volts I would need to grab two -42A units to run in series. Looking at the adjustment range of the HLG-480H-42a that would let me set charge voltages anywhere from ~72-88V (so I could bump it to 21S if I really wanted to) with current settings anywhere between 6-11A (~500-950W). I do like that they are nice and small, water-resistant, around 13lbs for two of them, and that they only cost $130/ea on Digikey. If it charges in 3 hours off a normal 120V outlet I can't complain too much. In the future for fast charging I guess I would just grab two more of them to run in parallel whenever it's plugged into an outlet that won't melt if I try to pull 1800W from it for an hour or so.

I'll have to do a bit more reading, but I would imagine that the BMS would be able to disconnect the charger(s) so I wouldn't need to babysit this thing when I plug it into the charging station out in the parking lot. Is that even 100% necessary? If the charger only reaches 84V and I leave the thing plugged in for a few hours, I would think it wouldn't be putting much current into the pack anymore (and the BMS can do its thing top-balancing the cells).
flippy wrote:
May 29 2018 2:30am
if you go with only a 20S solution i recommend just getting a off the self BMS with a decent current rating. check the bluetooth bms topic for tips. if a pack is balanced yo dont need much to keep it there. and most smartbms solutions can balance manually, just hit the button and wait. costs 130ish bucks with free shipping for a 300A model.
I just found the 8s-24s 300A smart BMS on aliexpress, it's pretty much exactly what I'm looking for. Yeah, that's better than whatever I was going to put together.

https://www.aliexpress.com/item/Smart-D ... 94534.html

I would guess if I'm going to be pulling more than 300 amps I'll have to bypass the BMS when I connect the controllers, unless there's a way to only read from one of the two controllers and just double the value to get useful data (doubt it). If the specs on these cells are not BS this pack is rated for 800A discharge, and the controllers aren't even capable of drawing that (they also have built in current limiting). I'm not worried about overcurrent protection, but it would be cool to not lose out on being able to measure the current flow.

I also found the Chargery BMS24 screen/BMS combo viewtopic.php?t=85357, which is expensive but pretty cool in that it's one of the only ones I've seen that I don't have to bypass if I start beating on the motors (potentially pulling 5-600A from the pack). Unless I can find more options that's looking like the ticket.

I'm mainly worried about balancing/management because I'm using salvaged cells and I want to keep them happy and not puffed/on fire.

User avatar
flippy   10 kW

10 kW
Posts: 851
Joined: Aug 12 2015 3:07pm

Re: First post, electric motorcycle build.

Post by flippy » May 29 2018 5:03am

i recommend bypassing the BMS and feed the controllers directly and put a regular 400A fuse on each controller. use the BMS only for charging and balacing. you can put a contactor on the output that you can put between the rest of the bike (that includes the igintion) so the bms does not see the real power but can shut down the bike (no control voltage into the kelly and it dies instantly) but you dont have to worry about burning up the BMS because you are pulling 600A into the controllers. lot less heat as well.

+1 on the mean wells. get them. but put in a AC relay that cuts the connection to the battery when you disconnect the AC, otherwise it will very slowly trickle drain the battery. yo can leave them always connected and leave them floating. but i seriously recommend only charging to 4.05V that is massivly increase the lifespan of the battery.
Lithium beats liquid dinosaurs.

phate   10 W

10 W
Posts: 70
Joined: May 28 2018 2:17am

Re: First post, electric motorcycle build.

Post by phate » May 29 2018 5:42am

flippy wrote:
May 29 2018 5:03am
i recommend bypassing the BMS and feed the controllers directly. use the BMS only for charging and balacing. you can put a contactor on the output that you can put between the rest of the bike (that includes the igintion) so the bms does not see the real power but can shut down the bike (no control voltage into the kelly and it dies instantly) but you dont have to worry about burning up the BMS because you are pulling 600A into the controllers. lot less heat as well.

+1 on the mean wells. get them. but put in a AC relay that cuts the connection to the battery when you disconnect the AC, otherwise it will very slowly trickle drain the battery. yo can leave them always connected and leave them floating. but i seriously recommend only charging to 4.05V that is massivly increase the lifespan of the battery.
Sounds like a plan. If I really want to monitor the total pack charge status that badly I can just install a standalone coulomb meter.

Was already planning to use a couple of relays to connect/disconnect the chargers depending on what kind of outlet I'm plugged into (2 series 480H-42a's for a 120v outlet, 2s2p or 2s/3p for fast charging if I plug into 220, although at that point I'm coming up on 40lbs of chargers).

The Kelly controllers will take up to 90 volts, so if I dropped the charge cutoff on the cells to 4.05-4.1v I have the option to bump it up to 21 or 22s. I'll worry about that later though, since I'm pretty sure this thing is gonna scare me if I don't turn the current limiting way down.

User avatar
flippy   10 kW

10 kW
Posts: 851
Joined: Aug 12 2015 3:07pm

Re: First post, electric motorcycle build.

Post by flippy » May 29 2018 5:56am

i plan on rebuilding my pack from 16S to 21S with the same controller and i will make a 61Ah 75V nominal battery (max charge 4.05, 85V total) running on that exact kelly. it should get the maximum from the controller without crapping itself.

for acceleration you need to make a balance between phase amps and battery amps. phase amps give lots of low end and battery amps give you the "Go" on the high end.

for fast charging you should just get a high power supply just for incidential charging. that is lighter but bigger. you can find those on the oriental shops.
Lithium beats liquid dinosaurs.

phate   10 W

10 W
Posts: 70
Joined: May 28 2018 2:17am

Re: First post, electric motorcycle build.

Post by phate » May 29 2018 6:24am

flippy wrote:
May 29 2018 5:56am
i plan on rebuilding my pack from 16S to 21S with the same controller and i will make a 61Ah 75V nominal battery (max charge 4.05, 85V total) running on that exact kelly. it should get the maximum from the controller without crapping itself.

for acceleration you need to make a balance between phase amps and battery amps. phase amps give lots of low end and battery amps give you the "Go" on the high end.

for fast charging you should just get a high power supply just for incidential charging. that is lighter but bigger. you can find those on the oriental shops.
That's about the size of the pack I eventually want to put in this thing, but I'm going to see what kind of range I can get out of the 20s 32Ah pack first. In theory it's like 35-40 miles if I don't do rippers with it the entire time. Probably a whole lot less when I do.

That Kelly seems to be the beefiest controller you can get in the smaller form factor. Coming from a car guy perspective, it's really cool how much control I can have over the powerband of this thing just by messing around with the controller settings. I'll turn up the wick more and more once I get used to it.

It seems backwards, but I actually don't have an available 220v outlet to charge with at home, so unless I plug into multiple circuits I'll end up slow-charging. The EV charging stations we have at work (which are free, and entirely vacant for my 10 hour overnight shift, lol) have at the very least a few beefy 120v outlets as well as 220v J1772. That's getting ahead of myself though.

User avatar
flippy   10 kW

10 kW
Posts: 851
Joined: Aug 12 2015 3:07pm

Re: First post, electric motorcycle build.

Post by flippy » May 29 2018 6:30am

the mean wells run on 110 just fine, just reduce the current. it tells you on the tin how much you should lower the current by if you have them not in free air i recommend lowering the current to 2/3's current rating.

if you dont need to charge fast: dont. you will kill the battery for no reason.

ps: 0-30 can probably be done in less then 2~3 seconds depeding on how haw far "balls out" you intend to go...
Lithium beats liquid dinosaurs.

phate   10 W

10 W
Posts: 70
Joined: May 28 2018 2:17am

Re: First post, electric motorcycle build.

Post by phate » May 29 2018 6:53am

flippy wrote:
May 29 2018 6:30am
the mean wells run on 110 just fine, just reduce the current. it tells you on the tin how much you should lower the current by if you have them not in free air i recommend lowering the current to 2/3's current rating.

if you dont need to charge fast: dont. you will kill the battery for no reason.

ps: 0-30 can probably be done in less then 2~3 seconds depeding on how haw far "balls out" you intend to go...
That's pretty in-line with what I was able to come up with on my spreadsheet performance-wise.

My planned use is to slow charge at home and at work, but fast charging would allow me to run the pack down riding somewhere interesting, park it at a 220v source, do things/eat lunch/whatever to kill an hour, then have enough juice to get home. By "fast charging" I really just mean "more than 1000W". I doubt I'd actually end up pushing more than 1C into them.

User avatar
amberwolf   100 GW

100 GW
Posts: 25477
Joined: Aug 17 2009 6:43am
Location: Phoenix, AZ, USA, Earth, Sol, Local Bubble, Orion Arm, Milky Way, Local Group
Contact:

Re: First post, electric motorcycle build.

Post by amberwolf » May 29 2018 1:39pm

phate wrote:
May 29 2018 4:35am
If I understand Meanwell's naming scheme, the middle number is the power output, and the last number is the output voltage? So if I wanted to charge my 20S pack to the full 84 volts I would need to grab two -42A units to run in series. Looking at the adjustment range of the HLG-480H-42a that would let me set charge voltages anywhere from ~72-88V (so I could bump it to 21S if I really wanted to) with current settings anywhere between 6-11A (~500-950W).
Basically, yes to all that. Just make sure the cells you're using can handle that charging current safely without warming up during charge.

two more of them to run in parallel whenever it's plugged into an outlet that won't melt if I try to pull 1800W from it for an hour or so.
Since they are isolated, you can run them off separate outlets if you like, so that if the outlets run off separate breakers you can draw up to twice as much wall current as if only on one outlet/breaker.

Similarly, at home you can use the 240v outlet like for stove or dryer, etc. (though these are hard to find out in the wild for guerilla charging).



I'll have to do a bit more reading, but I would imagine that the BMS would be able to disconnect the charger(s) so I wouldn't need to babysit this thing when I plug it into the charging station out in the parking lot. Is that even 100% necessary? If the charger only reaches 84V and I leave the thing plugged in for a few hours, I would think it wouldn't be putting much current into the pack anymore (and the BMS can do its thing top-balancing the cells).
BMS are generally setup so that as soon as any cell reaches HVC, it will shut off input current to the pack. Before that point, any cells above the BMS's balancing voltage point would begin to be drained down by the BMS, and it would continue to do that whether current is flowing in or not, until they drop below that point.


As long as the cells are good quality and of the same batch they ougth to all have the same characteristics and stay balanced on their own until they get old. RC type cells aren't really known for these properties, but perhaps the ex-bus cells are better. :)


If it's a programmable BMS you can set the balance point and HVC lower (say, 4.0v or 4.1v), and set the LVC higher (3.0-3.4v? depends on teh cells' discharge curve) which while it will give a little less capacity, will be safer and extend the life of the cells and they'll get unbalanced less. (you'd also set the charger voltage lower to match the lower total final voltage)


I'm not worried about overcurrent protection, but it would be cool to not lose out on being able to measure the current flow.
You can always use an external shunt wired just at the output of teh battery/BMS, to measure current from multiple controllers at the same time, rather than trying to merge data from each separate controller or measure just one controller.

phate   10 W

10 W
Posts: 70
Joined: May 28 2018 2:17am

Re: First post, electric motorcycle build.

Post by phate » May 30 2018 11:19am

Might have found my donor bike! If all goes well I'll be picking it up within a week.

Because I want to keep this thing small and light, I've mainly been focusing on small-displacement bikes to convert (i.e. the Ninja 250). I did a bunch of reading, and the chinesium 125's seem like a decent option to start new, but without having ridden one I'm not sure how comfortable I would be putting 4 times the original power into it. Also, no clue what kind of parts availability there would be to keep it running.

Just doing a quick search through the sub-$1500 options, I found a surprisingly clean looking 1993 Ninja 600 listed for $500. The ad mentions having taken apart the cylinder head and not getting it back together, and includes phrases along the lines of "get this worthless piece of crap off my lawn". Clean title, full plastics. It's heavier than what I wanted originally, but otherwise it's pretty much ideal.
Image

Doing a bit more research, 1993 was before they switched over to the twin spar frame, so I've got bottom tubes that will make it a whole lot easier to mount everything and keep it as low as possible.

Running the numbers back through my spreadsheet now, it's obviously going to be heavier than starting with a Ninja 250, but I should still be able to keep things around 300lbs, at least with the 20S4P pack.

User avatar
Bernel   100 W

100 W
Posts: 148
Joined: Feb 03 2013 3:33pm
Location: Peterborough, England

Re: First post, electric motorcycle build.

Post by Bernel » May 30 2018 12:03pm

That is a good choice for your first ev build. At 20S4P what capacity of battery cell will you be using? When you source your controller get one that will allow you to expand the performance of your bike as and when you can afford to. My mistake was buying the wrong Kelly Controller and limiting myself to 90V. I have a Gen4 Size 4 in my YZF600R at the moment which will soon be replaced by a Size 6.

More power!

phate   10 W

10 W
Posts: 70
Joined: May 28 2018 2:17am

Re: First post, electric motorcycle build.

Post by phate » May 30 2018 2:45pm

Bernel wrote:
May 30 2018 12:03pm
That is a good choice for your first ev build. At 20S4P what capacity of battery cell will you be using? When you source your controller get one that will allow you to expand the performance of your bike as and when you can afford to. My mistake was buying the wrong Kelly Controller and limiting myself to 90V. I have a Gen4 Size 4 in my YZF600R at the moment which will soon be replaced by a Size 6.

More power!
Currently I have 20S4P worth of salvaged SPIM08HP "bus lipo's", which have pretty nutty specs (25C continuous discharge, 10 or 15C charge). They're 8Ah each, so the pack will be 74v nominal and 32Ah. I haven't assembled the pack yet, since I'm still test cycling all of them to ~3.8v so I can work on figuring out the best way to connect them all.

That should be enough to get a bit under 40 miles out of it if I ride it reasonably, which is enough for my day-to-day usage. If/when I need to go farther I'll be adding an additional pack (probably made of different cells that have more energy density than power density).

I'm running dual motors, each one fed by a 300A Kelly (KLS7230S, aka the biggest small controller they sell). As far as room to grow, the motors are rated for 250A each, but they're water cooled and I figure as long as they don't overheat I can probably max out the controllers and force feed them 300A each in short bursts. I've also got room to add some more voltage if I really want to, but yeah I'm gonna be limited to 90v. I can get a good amount of punch back if I decide that it doesn't actually need to go 75-80mph and gear it lower.

phate   10 W

10 W
Posts: 70
Joined: May 28 2018 2:17am

Re: First post, electric motorcycle build.

Post by phate » Jun 03 2018 4:38am

Really disappointed that I haven't heard back from the guy with the Ninja 600 yet, I really want to use it as my donor bike.

Played around with how I'm actually going to construct the battery pack some more. Haven't 100% decided whether I'm going to just bolt the cells together (using the bolts themselves as busbar) or if I'm going to solder things to the battery tabs.

I also put in price quotes for my motors. If the cost difference isn't huge, I'm going to eat the extra 2 kilos and grab a pair of 120100 outrunners instead of 12090's. The larger stator should allow for me to max out the controllers without too much drama, and when I inevitably end up wanting more I've got a bit more on the table after I upgrade to a more powerful pair of controllers.

Edit: It's a difference of $20 per motor. Yeah I'm going with the watercooled/sensored 120100's. They're rated for 350A peak, so I can pretty much just yolo the controllers without hurting them.

User avatar
Dauntless   100 GW

100 GW
Posts: 6965
Joined: May 29 2010 1:49am
Location: Coordinates: 33°52′48″N 117°55′43″W

Re: First post, electric motorcycle build.

Post by Dauntless » Jun 03 2018 4:04pm

phate wrote:
Jun 03 2018 4:38am
Really disappointed that I haven't heard back from the guy with the Ninja 600 yet, I really want to use it as my donor bike.
I've a long history of just not hearing back, not getting that perfect bike when I find it. Sometimes it never was for sale, just some goofball having fun listing this stuff
Any sufficiently advanced technology is INDISTINGUISHABLE FROM MAGIC!
- Arthur C. Clarke

phate   10 W

10 W
Posts: 70
Joined: May 28 2018 2:17am

Re: First post, electric motorcycle build.

Post by phate » Jun 04 2018 12:53am

Yeah, hoping this is just craigslist flakiness.

The ad mentions stuff about calling/texting, but he only linked it to his email, so I'm guessing he just hasn't checked yet.

Finalizing the details on my motor order, which is pretty exciting. I haven't seen too many people play around with these, and it's entirely possible that there's a good reason for that. They're the same as the Alien Power Systems/Epowerhobby 120mm outrunners.

Specs Below:
MP120100 Outrunner
50kV
12mm keyed shaft (I will very likely end up replacing this with a longer shaft running through both motors.
28 Pole (limited to 5000rpm by the eRPM capabilities of the Kelly controllers
Internal Hall sensor board
Water Cooled 60mm stator
350A Peak Phase current
230A Continuous Phase current
~4000rpm straight off the charger with my 20S pack

Also there's two of them, geared 4:1 which should be enough to go 75-80mph.

We'll see how well a "big small" controller like the KLS7230S (well, one controller per motor) will do running these. It should run fine, and I've got plenty of room to grow with these. All I would need to do is buy some beefier controllers when I want more speed than the Kelly's can deliver, and either sell those on here or keep them for other projects (like an e grom, for example).

I figure absolute worst case I can look into either a different controller or learning how to rewind them to be more suitable for a bike.

User avatar
macribs   1 GW

1 GW
Posts: 3481
Joined: Jul 22 2014 5:59pm

Re: First post, electric motorcycle build.

Post by macribs » Jun 04 2018 2:28am

phate wrote:
May 28 2018 6:36am

- I can get some decent performance out of it without spending a ton of money (budget around $5000).

Basic goals before I dive into the details:
- Easy to register for road use. If the donor bike has clean paperwork, I basically just tell the DMV what I started with and check "electric" instead of "gasoline" as the fuel type.
- Small: Designed around a Ninja 250, but open to use whatever entry-level full-fairing sport-ish bike I can scoop for cheap on craigslist.
- Lightweight, low center of gravity
- 25 miles will get me to work and back, but I can charge at work if I completely screwed up my range calculations.
- Highway-capable, geared for 70-75mph top speed: I'm not planning to ride this thing on the highway, but I'd like to not die if I have to do it for an exit or two.
- Strong acceleration at low speeds. I'd love it if wheelies were possible, but it at least needs to be faster than most things in the 250cc range up to 40-50mph.


I'm planning to bulk charge with either an S1600 or S2500 from BMSbattery. At home I will be charging from a 120v outlet, but at work I have access to 220v for faster charging. J1772 capability falls firmly in the "I want it but we'll worry about that later" category.


Looks like this will be a very cool build. However do you really wanna spend 50% of what you sink into the complete bike on a charger? Build budget around 5.000$. Then 2.500 $ on a charger? If you can stretch your total budget to 7500$ you could do an epic build, the less you sink on a charger the more you can get radical performance parts that in return give you the happy e-grin.

If it was me I would just be happy with 110v charging and be done with it. You are building a full fairing bike, with less wind resistance then naked bikes. Less punching holes in the air, easier on the battery. When building the final battery I would strongly consider to build battery as large as you can possible fit within your budget and bikes real estate. Bigger pack less stress on each cell, and longer lifespan. Less heat and less Ohm. More efficient. And plenty of battery amps to really get amazing acceleration.

I don't know the Alien power motors you plan on using.Those are RC motors? Will they need some sort of cooling fans to tap into the peak power? I assume in a RC flight scenario the prop would feed the motor cool air constantly. How well will the cooling be tucked inside the fairing of a Ninja 250?

I do know that it seems the Sur Ron motor is capable well beyond its ratings as people start pushing those motors ie with the ASI 8000 controller. From memory I think I saw claims of 25-30 kw peak for such a combo. Now imagine you run a dual setup of that....https://evnerds.com/electric-vehicles/e ... companies/ sur ron is now selling a ready made dual motor 8)

A little less money on the charger and you can get that dual motor and two ASI 8000 controllers.Seems the dual sur ron motor will be about the same as the dual Alien motors. The dual ASI 8000 controllers will cost more then the Kelly's but in return the ASI's will get you peak current of 700 or 800 amps from each controller. Should really help you off the line when the light changes. Btw, size wise the ASI 8000 controllers are really close to the Kelly. Look at this forum for first hand reviews of the ASI controller.
http://accelerated-systems.com/wp-conte ... -Sheet.pdf

There are other controllers as well, like the Sevcon controllers and others as well to look at. Controllers are important as they pretty much will dictate the width of your grin while cracking the throttle.

Kelly controllers have in the passed had a known ramp up issue, I guess it is firmware related to safety of something. When cracking the throttle you don't get max amps right away, which will kind of limit the acceleration a little from standstill. Don't know if there is a workaround. Maybe worth looking into before you sink money into parts.

As for a donor I would spend some time looking for the perfect donor. Every now and then there are great donors around. Like try to find a bike that is in great condition but with a failed engine. Or a trashed gearbox. If you find a donor with great fork, shock and brakes you will get an easier build because your baseline bike is very capable. Sometimes you can find donors with upgraded parts that lacks ie a working engine so they sell for cheap. Depending on your time frame for this build don't rush into buying a crappy used up donor. You will spend much money and time on changing trivial parts, bearings, stripped bolts, rusted or damaged this and that.

phate   10 W

10 W
Posts: 70
Joined: May 28 2018 2:17am

Re: First post, electric motorcycle build.

Post by phate » Jun 04 2018 6:42am

macribs wrote:
Jun 04 2018 2:28am
Looks like this will be a very cool build. However do you really wanna spend 50% of what you sink into the complete bike on a charger? Build budget around 5.000$. Then 2.500 $ on a charger? If you can stretch your total budget to 7500$ you could do an epic build, the less you sink on a charger the more you can get radical performance parts that in return give you the happy e-grin.
The S1600 and S2500 are the model numbers for the chargers. The S1600 is a 1500W charger that costs $190, and the S2500 is a 2000W charger that costs $240. I've pretty much decided I'm gonna screw around with a few Meanwells instead.
macribs wrote:
Jun 04 2018 2:28am
If it was me I would just be happy with 110v charging and be done with it. You are building a full fairing bike, with less wind resistance then naked bikes. Less punching holes in the air, easier on the battery. When building the final battery I would strongly consider to build battery as large as you can possible fit within your budget and bikes real estate. Bigger pack less stress on each cell, and longer lifespan. Less heat and less Ohm. More efficient. And plenty of battery amps to really get amazing acceleration.
110V charging will be 100% fine if I'm only riding to work (10hr shifts) and back or around town, but I would eventually like to be able to take longer rides, hang out at my destination for an hour or two, and pick up enough range to make it home, which is where the 220v charging would come in handy.

The 20s4p pack was sized mostly to be able to meet whatever current demands I could ask of it in order to get the bike running properly. It's enough to commute with, but it's biased much far more towards power density than energy density. While I'd love to just use some more of these, at 290 grams per 8000mah, once you already have "enough" current capability, adding more of these things is pretty wasteful.

I will probably want to add another 2.5 - 3kWh (bringing me up to 5 and change). If I rebuilt the pack of bus lipo's, I would need another 20s4p (to reach 5.3kWh at 84v) or 20s5p (to reach 6kWh at 84v). At that point we're looking at total battery weights of ~110lb for 5.3kWh or ~125lb for a 6kWh pack. Yuck.

Comparison of energy/power density, ignoring the additional hardware that goes into connecting everything into a pack
- Bus Lipo: 290g/ea, 8Ah, 200A cont. 8.6g/Wh, .34g/W.
- LG HG2: ~50g/ea, 3Ah, 20A cont. ~4g/Wh, .59g/W.
- LG HB6: ~50g/ea, 1.5Ah, 30A cont. ~7.9g/Wh, ~.4g/W.
- LG MJ1: ~50g/ea, 3.5Ah, 10A cont. 3.4g/Wh, ~1.2g/W

It's actually pretty surprising how closely that balance trades off one for another. If we assume that since our current requirements are more or less satisfied, the possibility of using high energy density low power cells becomes apparent. If we decide that our add-on pack only needs to be able to feed a highway cruise on relatively flat ground, that's only 50 - 70A. The connection to the add-on pack would need to be disconnected when current draw is too high, and reconnected when load decreases, which should be fairly easy to accomplish. Because high power acceleration would not happen for long (because it gets up to speed quickly), voltage should not go drastically out of balance between the bus lipo's and the add on pack. If things do get too far out of whack, a resistor can be used to limit the current as the add-on pack "charges" the bus lipos.

With HG2's I would need at least 20s4p (12Ah, 4kg, 1kWh), and with a low-drain cell like the MJ1 I would need at least 20s7p (24.5Ah, 7kg, 2kWh). With a couple of creatively hacked up Turnigy Multistar 20Ah packs, I would be somewhere in the 20lb range if I wanted to add another 1.6kWh, but I'd have more current delivery available.

macribs wrote:
Jun 04 2018 2:28am
I don't know the Alien power motors you plan on using, not the price. But I do know that it seems the Sur Ron motor is capable well beyond its ratings as people start pushing those motors ie with the ASI 8000 controller. From memory I think I saw claims of 25-30 kw peak for such a combo. Now imagine you run a dual setup of that....https://evnerds.com/electric-vehicles/e ... companies/ sur ron is now selling a ready made dual motor 8)

A little less money on the charger and you can get that dual motor and two ASI 8000 controllers.Seems the dual sur ron motor will be about the same as the dual Alien motors. The dual ASI 8000 controllers will cost more then the Kelly's but in return the ASI's will get you peak current of 700 or 800 amps from each controller. Should really help you off the line when the light changes. Btw, size wise the ASI 8000 controllers are really close to the Kelly. Look at this forum for first hand reviews of the ASI controller.


And then you can then benefit from others running the same motor/controller combo in terms of getting the best possible settings dialed in. Kelly controllers have in the passed had a known ramp up issue, I guess it is firmware related to safety of something. When cracking the throttle you don't get max amps right away, which will kind of limit the acceleration a little from standstill. Don't know if there is a workaround. Maybe worth looking into before you sink money into parts.
For whatever reason I can't find anywhere selling the Sur Ron setup to compare, but the twin 120100's are under $800 shipped and will do 25kW each on the Kelly's. I'm not married to the big KLS-S, but honestly 50kW on this thing scares me a bit already. As long as those controllers will work I can worry about adding more power later (after I've added a bunch of range).

User avatar
macribs   1 GW

1 GW
Posts: 3481
Joined: Jul 22 2014 5:59pm

Re: First post, electric motorcycle build.

Post by macribs » Jun 04 2018 7:22am

Hehe ok, I read it as $ not the letter S thinking you should get a high end charger of some sort. I guess thats what I get for not being a good boy and using my reading glasses. :lol:
For whatever reason I can't find anywhere selling the Sur Ron setup to compare, but the twin 120100's are under $800 shipped and will do 25kW each on the Kelly's. I'm not married to the big KLS-S, but honestly 50kW on this thing scares me a bit already. As long as those controllers will work I can worry about adding more power later (after I've added a bunch of range).
I have not used kelly myself so I can't say how big impact that slower throttle ramp up plays. But from reading here on the forum it seems many people can't get the Kelly's dialed in to give best possible acceleration. When I read your first post it seems to me you fancy acceleration way more then highest possible top speed.

I don't think Kelly got problem pushing peak power, giving you fill amp out to motor. But if the controller limit the output say from standstill, it does not matter what peak output you get. Your acceleration will suffer. From stand still that is. Seems if you roll on the throttle while already cruising the pick up is faster and the Kelly don't hold back.

50 kw seems like a lot. But when you factor in the weight of the completed bike and rider that number is not outrages. Have you ridden any of the Zero motorcycles? They got a mode setting, can't remember if it was rain mode or eco mode but one setting puts serious strains on acceleration, from being boring in limited mode to being strong even from standstill in sport mode. My thinking about the Kelly is that you get that kind of "eco mode" feeling when you crack the throttle. Would be a shame really to loose potential grin producing take offs by the choice of controller.

I guess it all up to what you fancy. If you like to be quick of the line when the light changes, I think there are other controllers out there that will give you the most rapid acceleration.

Maybe post a question about the Kelly controller to check how noticeable the hold back is? If you can live with it, then sure no problem.

phate   10 W

10 W
Posts: 70
Joined: May 28 2018 2:17am

Re: First post, electric motorcycle build.

Post by phate » Jun 04 2018 9:02am

macribs wrote:
Jun 04 2018 7:22am
I have not used kelly myself so I can't say how big impact that slower throttle ramp up plays. But from reading here on the forum it seems many people can't get the Kelly's dialed in to give best possible acceleration. When I read your first post it seems to me you fancy acceleration way more then highest possible top speed.

I don't think Kelly got problem pushing peak power, giving you fill amp out to motor. But if the controller limit the output say from standstill, it does not matter what peak output you get. Your acceleration will suffer. From stand still that is. Seems if you roll on the throttle while already cruising the pick up is faster and the Kelly don't hold back.

50 kw seems like a lot. But when you factor in the weight of the completed bike and rider that number is not outrages. Have you ridden any of the Zero motorcycles? They got a mode setting, can't remember if it was rain mode or eco mode but one setting puts serious strains on acceleration, from being boring in limited mode to being strong even from standstill in sport mode. My thinking about the Kelly is that you get that kind of "eco mode" feeling when you crack the throttle. Would be a shame really to loose potential grin producing take offs by the choice of controller.
I have not ridden a Zero at all, although at least on paper this should be fairly close in performance. The only electric bikes I've been on had brushed alltrax setups on them, so I can't really compare.

Everything I could find about laggy acceleration of Kellys resulted from incomplete/incorrect configuration, and were mostly for different models. I think it listed the minimum throttle ramp time as .1 seconds? Either way I'm totally open to suggestions for sinewave controllers that are capable of 3-400A with regen that won't break the bank. Those ASI's are quite pricey.

A pair of Sabvoton 72v200's might work, depending on erpm (need at least 70k) and how their maximum phase current vs battery current settings work. Edit: Yeah if I can't find something that can do 70k erpm and 150-200A cont/300A+ phase that costs less than $400 per controller, I'm going to just grab the Kelly KLS7230S and figure out whatever sweet nothings I have to whisper to the throttle config settings in order to get it off the line.

I'm not at all opposed to just loading the motor up against the brakes. Depends if the "lag" is indeed just a ramp value applied to the throttle position, or if it's dependent on getting "X" amount of rotation out of the motor


Yay for 28 pole motors.

phate   10 W

10 W
Posts: 70
Joined: May 28 2018 2:17am

Re: First post, electric motorcycle build.

Post by phate » Jun 08 2018 1:07pm

Bought a bunch more toys! I'm somewhat striking out on finding a donor bike I really like in the price range I want (not 100% true, there's a halfway decent Suzuki Katana nearby for $900, and a beautiful EX500 as well for $1300), so I'm just gonna build most of the electrical system on the bench and get that all sorted out. Changing up my pack design a little bit. Going to 3d print the actual cell holders, but the cell holders will live inside an aluminum box for more protection and easier assembly/mounting. Still planning to bolt the cells together, using the bolts themselves as busbar and with nuts as spacers so I don't have to fold the tabs over too hard. Probably going to just use aluminum hardware for that.

Grabbed a 300A smart BMS with display for pack balancing and general battery status monitoring, and a coulomb meter that can read up to 750A in short bursts to keep an eye on power draw and how much battery I've used. Between those two and donor bike's existing dash, I think I should be pretty much covered on the instrumentation side of things.

Also grabbed an EV200 contactor, DC-DC converter, a couple of random 12V automotive switches to simulate things like the key switch on the bench, and one of the eventual two controllers (KLS7230S). Pretty sure I have a suitable precharge resistor kicking around somewhere, if not they're cheap and I'll just grab another one.

Only bought one controller so I can evaluate how well it can run these motors. That way if it doesn't work I only have to unload one of them if I can't find another use case for it (like putting it on one of those $1200 Boom eGroms along with an 8kw hub motor).

spectastic   1 mW

1 mW
Posts: 16
Joined: Jul 24 2017 10:21pm

Re: First post, electric motorcycle build.

Post by spectastic » Jun 14 2018 5:39pm

what if the two motors out put aren't exactly the same? wouldn't the motor doing less work slow down the motor doing more work?

what's the reasoning to have 2 motors instead of one?

Post Reply