Custom 80100 RC Drive on Genesis Swing-arm

[DanGT86]

This bike has had an identity crisis since I started working on it. 3 motors later it’s an almost road ready RC mid-drive. I ended up with the genesis bike during my brief adventure into selling my variation of the famous Farfle swing arm. Turns out having a steel swing arm that is easy to weld to and tons of room in front of the rear wheel makes for a great test platform for all kinds of ebike drives. I’m expecting it to be capable of 5kw bursts when geared for around 40 mph. From everything I have read on ES that sounds reasonable to me but it’s my first RC motor as well as my first non-hub build so I really don’t know what to expect. I installed a temp sensor so I plan to sneak up on it and see what I can get away with.

It spins and doesn’t throw the chain or the belt off so I’m feeling pretty good about it so far. I still have a few more parts to machine as well as full voltage testing on 20s. Hopefully there aren’t any issues with the Lyen controller and the high rpm of this motor. I'm also a bit nervous about the motor windings since the re-termination from Delta to WYE was a quite invasive procedure.

I machined all of the parts myself but the inspiration I gathered from years of admiring other peoples’ work is probably obvious.

Here are the current specs:
• Luna cycle’s version of the classic turnigy 80100
• Re-terminated in WYE for 80kv from the original Delta 130k
• Sensored using an external hall board from e0designs.com
• Custom machined reduction drive using Gates belt 1st stage 1:2 and 13:94 tooth final 219 chain drive
• 24” standard mtb wheel with LH final drive sprocket mounted to the brake flange. I’m hoping a 26” wheel fits as well so I can cover a wide range of speeds and riding styles.
• Farfle type rear swing-arm
• 20s lipo
• Lyen controller I have a 9FET 12FET and 18FET at the moment but I’m not sure which will permanently live on this build.
• 3D printed radial flow cooling fan to draw air through the motor.

Here is what it looks like at the moment.

 

[DanGT86]

Did my first test ride last weekend and it moves under its own power! First observations are that the throttle response is very snappy with the lightweight rc motor. I have always had the massive rotational inertia of a hub hindering the spin up time. Also, the lack of weight in the rear wheel makes it do burnouts when I twist the throttle standing next to the bike. Not that impressive but pretty funny looking.

Here is some performance data:
It used 1100watts at 29mph which was full throttle on my speed based 9Fet Lyen controller

Motor temp got to 40deg C after about 20 min of riding with peaks of 2500-3000 watts. Ambient air temp was about 15deg.

No load current is around 5amps at 6000 RPM at 71v

There is some bad stutter under about 10mph. It seems to go away if I increase throttle but once its rolling pretty good everything smooths out electronically speaking. I need to do more testing to find the optimal hall sensor timing. I adjusted it for the smoothest quietest and least current draw per rpm when the wheel was free spinning on the stand. I suspect that is not the optimal hall position under load.
I suspect the stutter could have been my CA3 hitting the low voltage cut as the pack I was testing with was pretty close to 70v on 20s. It is also very possible that the controller was reaching a current limit internally due to the spikes from the RC motor being hard to drive. I will play with the controller settings and try again with a fully charged less saggy pack. Also need to set the CA for direct pass through throttle control to eliminate variables. Once the controller is happy then I will tune the throttle for preference. The controller is also an EB249 board and I have some EB3xx controllers here that may work better. Not sure if the chip is any different or better on the newer generation board. Now I want a sin wave Kelly so if these are too much trouble I might just move on a save the LYENs for hub bikes.

The winding wires have all been soldered to 4mm bullet connectors so I could switch from Wye back to Delta using Y adapters to recombine the windings. Its probably not ideal to have the extra wiring length to bring the 3wire WYE connection out of the motor but soldering this junk varnished wire was such a pain I could not have possibly done it inside of the motor. I don't ever want to solder this varnished magnet wire ever again. My hours invested in the re-termination and soldering fight have far exceeded the value of an 80100 and I should have just gotten an Astro the way I wanted in the first place. I guess we could all say that about any build though. It's my own fault I take on these projects.

One detail I am really disappointed with is the poor machining tolerance of the 219 front sprockets. The sprocket is misaligned with the hub such that the the chain gets tighter and looser within 1 revolution of the small sprocket. This makes it feel really sloppy and will probably lead to a mechanical failure since it manifests as a bad vibration at speed. I really don't want to resort to machining sprockets myself. These small ones have to be hardened due to the stress per tooth at such a small tooth count. That adds lots of extra cost since most heat treat companies have a minimum cost of a few hundred dollars just to turn the ovens on. I'm really worried about the chain derailing with no rear freewheel. This heavy 219 chain would not hesitate to lock up the wheel or rip my foot off. That is the main reason I went for the drive mounted to the swingarm. I really didn't want to mess with tensioners and chain growth during suspension travel. On a related note, the 5mm pitch 15mm wide Gates belt didnt skip any teeth on the primary drive. I was not sure it was going to be up to the task.

So I have a bit of R&D before I can really beat on it but so far it runs and acts like my other ebikes so I think I’m on the right track. Also the regen works great and feels very similar to my hub bikes. Downside is the sliding chain tensioners don’t keep the wheel from scooting backwards when I regen so I need to deal with that. They pull towards the rear for chain tension but thats the way the wheel wants to slide when stopping.

Here are some more pics
Sliding motor plate to tension belt now has a bearing in it so the shaft is not pulled sideways at the motor.

Hall sensor bracket is pinned to the sliding motor plate so once adjusted it moves with the motor. I can now remove the belt without having to re-time the hall sensors.

This is the final drive setup. The sprocket adapter is just a quick plate to test everything. I have plans for a better version later.

Bearing tube sits in those clamps mounted on the swing arm. This allows me to adjust it from side to side for chainline or pedal clearance. The main plate that holds the primary drive is also a split clamp design that clamps to the bearing tube. This gives adjustment of the angle of the drive and chainline as well. Keeping it modular like this makes it easier to move it around if I want to try it on another bike as a crank drive.

 

[litespeed]

That looks super trick. Great job.

I definitely want to try a mid drive one day and yours looks like a great option.

Watching with interest.

Tom

 

[DanGT86]

Thanks for the kind words but i wont be selling them. Might sell this one when Im done with it but I stole way too many ideas from other systems on the market to feel right selling this design.

I will let you ride it for sure.

 

[Byte]

Nice build!

 

[skeetab5780]

Awesome Progress Dan! That is a beautiful looking jackshaft/drive. what is the weight of the motor and jackshaft and mounts total? Curious how much different it is from the LR direct driven

Thanks for linking this thread, any video of the noise?

 

[Ecyclist]

Nice build Dan!
You might want to try another small sprocket and get better results. Mass produced Chinese components are hit and miss. One can be perfect, next total junk. You could make one out of A-1 tool steel and harden it with torch but that would be a project and half.
Have fun. I'm looking forward to your next post.

 

[DanGT86]

Wow thats crazy timing on your post. I was litterally typing this update when I noticed a new post. I've been thinking the same thing about making my own sprockets. These purchased ones use a really big coarse thread setscrew that I don't really like either. I'm thinking 4140 pre-hard material would be fine. Maybe 40 rockwell. I don't want to torch harden something and have the inconsistency cause more problems.

Got my motor control sorted out. I think the issues with the stumble were related to the EB2xx board in my 9FET controller. The EB312 board in my 12 FET seems to work fine so far. I had Lyen beef it up for me when I ordered it. Because of the upgrades It requires that I program it as an EB318 board. I think this has the positive effect of changing the internal current limiting such that the RC motor doesnt trip it. Maybe its just more tolerant of higher current. Either way it works now. Settings are presently:
20s battery
50A battery
100A phase
Block time 0
Motor speed is 6krpm at which is 30mph with my present gearing.

Testing so far at 3kw max on a 95deg F day I couldn't get this little motor hot. After about 20 min of beating on it really hard it was never above 70degC winding temp!! I'm running a centrifugal fan on the end of the motor. I couldn't be happier with this little motor. Hopefully I'll be turning it up to 6kw soon. Thats about where I'm expecting to run into the limits of this motor as well as my wiring size and tired 10ah lipo pack.

Here is the inside of the 12FET with upgraded traces. I hope it can handle 100amps batter current. I think he charged me $10-20 for the labor. Saved me tons of time since I'm not too handy with the solder.


Here are some vids of it operating on the road. It was really hard to hold the phone with one hand while operating the throttle with the other. I need to build a camera mount stick for stuff like this.
[youtube]GQPdBIskAcI[/youtube]
[youtube]MYRJRBDkgAo[/youtube]
[youtube]966RRt2pKoI[/youtube]

 

[skeetab5780]

Thing looks pretty snappy! Controller looks sick too well worth the money

 

[DanGT86]

It has the typical "infineon" style ebike controller throttle response. It was made more apparent in the video because I was holding the phone with my other hand shooting the vid. I only had a few minutes to ride it that day and set the CA3 throttle control in pass through because it needs some tuning to be smooth with this little motor.

As far as the controller I really don't know what it will take without blowing any FETs. I have no idea how Lyen can work that cheap. That much soldering would have taken me days and I would have ruined something for sure. Other than the goofy speed based throttle logic I think these controllers are great.

 

[Matt_92]

Hi DanGT86 ! Great job :wink:

I have a question : do you know where can I find specs of a C80100-130kv ?

Table or chart for : Efficiency vs RPM vs Amp vs Torque vs Volt ?

Thanks in advance for your help :wink:

 

[DanGT86]

It was a super common motor a few years back so there are tons of threads with various detailed specs. The problem is you just have to find them. Google searching this site seems to work better than using the ES search feature. I got my motor from Luna. They usually have them in stock but they don't specialize in RC motors. I was less than impressed with the winding quality of the motor. Some of the wire strands were sticking out of the stator such that they would have rubbed on the magnets and shorted out. If you get the motor from them, be ready to take it apart and look it over before using it. I was able to twist the wires tighter to take up the slack in the loose strands and added some heat shrink in areas I thought would be a problem in the future. Also the wires were bent around a lot of sharp corners. I used a file to take off lots of sharp edges that may have eventually shorted on the front cover.

Not trying to trash Luna here. Just keep in mind that its a bulk produced $175 6000kw small motor and they are just a a re-seller. They only have one winding option so they probably don't take them apart and inspect them. Ordering a custom wind from a place like Alien might get you a higher quality product.

Hope these help a bit.

Lots of good test data in this thread.
https://endless-sphere.com/forums/viewtopic.php?f=30&t=44287

here is the ES wiki page on it.
https://endless-sphere.com/w/index.php/C80100

Motor comparison spreadsheet thread
https://endless-sphere.com/forums/viewtopic.php?t=65757

 

[Matt_92]

Thanks for your help but I had already seen these thread and datas, maybe you already seen this Excel sheet, comparativ with differents 80100-50, 80, 130 and 180kv...

I repeat myself, but good job for your project :wink:

 

[DanGT86]

If you are just comparing different motor winds you might be concerning yourself with the wrong data. As far as I know, you should aim for the highest rpm you can practically make a reduction for. If you want a reliable 6kw out of an 80100 you need to get it up in the 6-8000 RPM range. Somewhere around there bearings become an issue and sensored controller choices become narrow. The number of reduction stages has been said to make a huge difference in system efficiency as well. So don't choose a specific motor for its efficiency if it causes you to add an additional reduction stage.

The turn count of the windings mostly cancels out between different kv if you have the voltage to spin it the same rpm. Turn count doesnt change the air gap, magnet strength, stator lamination shape, tooth count etc. There are small differences in total copper fill between different winding schemes because of the geometry based on the cross section of the wires and that will make a difference. But for the most part all winding counts are going to behave the same at similar RPM if you have enough current to feed the high KV winds at low RPM.

 

[DanGT86]

I'm trying to get back on this project before spring and with any luck I want to race with it in May. I have a 3 month old daughter so I've had zero time to mess with it since November.

Anyway, I'm still trying to figure out how to find the optimal hall sensor position. As mentioned above somewhere, I set this up for the lowest possible no load current draw with the wheel spinning in the air. There is an occasional stutter at really high load and low RPM which I think is the hall signal being washed out by the field strength from the high current or just bad positioning.

So I brainstormed a testing solution and I'd like somebody to tell me its either a good or bad idea.

I have an extra 9c hubmotor laying around. Its a crazy low kv motor that goes 25mph on 80v in a 26" wheel. The regen from it is pretty stiff. So my thought is to mount the 9c in my frame with my rear 94 tooth sprocket on it. Then I would use a second controller to control the regen from the 9c to create some load for this 80100 drive. I have 2 cycle analysts so I can monitor everything while I tune the hall position of the 80100 setup all on my repair stand. Like a poor mans dyno. Such a poor mans dyno that I can even use the same battery pack and just send the energy on a useless trip through 2 motors. Its like a perpetual motion machine, humanity is saved!!! :wink:

I would like to directly compare the hall position behavior and efficiency of the mid-drive at no load, max load and everywhere in between. I suspect the lowest no load current draw is not the magic spot. What's been weird is that the further I move the halls the faster the motor seems to go. I even machined more adjustment into the bracket and once the motor is spinning I can just keep getting more rpm out of the wheel until I hit the physical limit of the hall bracket. I would have liked to see it pass a point and begin slowing back down so I could be confident I was passed the point of highest efficiency. Im thinking with a bit of load the point of diminishing returns would become very obvious.

 

[Thud]

Dan,
In my experience, my best overall setting for hall timing was between 3 & 5 degree's advanced from lowest amp draw unloaded.
you will get some stutters at the extream load (starting from a dead stop) but this should wake the motor up from 1/3 speed to the top.

Idealy these motors are producing their best power @7500 rpms. Gear accordingly.
(in all my trials, I could never load these to 6kw for any reasonable time....they allways showed some heating issues.)

 

[DanGT86]

Great. I'll try that first. You mean physically 3-5 deg around the full 360 of the can or 3-5deg electrical timing?

Does it matter which stator teeth the halls are positioned relative to? For example the windings on each 2 teeth are the same phase. So if a hall was in between two teeth of the same phase would that act differently than if it were between two different phases?

Seems like it wouldnt matter since the magnets are a regular patern about the whole can.

As for power id be thrilled with 6kw bursts but realistcally dont expect to go charging up hills like that. My 12fet, Andersons, 20s 10ah lipo etc. are all suited to a max of 5-6kw anyway. Geared for 30mph at 5krpm this motor stayed super cool with 20min of repeated wot runs at 3kw. Its pretty fun like that but 6kw bursts would be nice to win over my gasoline motorcycle friends.

 

[Thud]

yes, physically (into the direction your spinning the rotor).
Easyest done with external hall set up.

For internal placement, I started using "between the phase prs" for positioning & payed attention to rotation to have the advance going in the correct/desired direction. (if my math was right, this equates to about 15ish deg of advance or retard depending on rotation)

I have a bunch of projects I need to document & update. Been doing other stuff lately...ahh Life.

 

[DanGT86]

Seems like a big difference in position 15 vs 3-5deg.

Id like to see your math on that. I really dont understand hall timing and positioning.

 

[Rikobonito]

Amazing build Dan !!
Love it soooo Look like a killer
Sorry to dont understund everything well in your last post, I will google-trans in detail. Also Im totaly noobie in ebike and not familiary with all bike part or controler, wattmeter etc. but maybe an inside sensor will help you with a angular fixed position ? (Alien Power System)

Or an encodeur on your shat ?.. not sure in ebike ...

I was planing doing my diy middle drive too but light/soft version (63mm motor with pedal assist ) ... and YOU make me want to do an emotorbike just like you ! You don't pedal ? Never ?
sorry if already say, whats your wh/km ?

I think I will have to make a topic too to anoyed you guys.

Good continuation Dan,

:v:

 

[DanGT86]

I think a 63mm motor would be fun if you plan to pedal along most of the time. Small motors like these and the 63mm can be a bit tricky to gear down enough for normal pedal cadence rpm.

I don't know my wh/km yet. I just don't have much time riding this setup yet.

I can't pedal this bike at the moment because the chainline is obstructed by the drive unit. I need to run the chain under it and back up with use of an idler sprocket. The pedals will just be there for starts and very low speed control similar to trials riding. Also the group I race with mandates that there must be a functional pedal drivetrain to compete.

I looked at the motors from Alien. I like the internal sensor boards but I was worried they would block airflow through the motor. I like the ease of adjustment of the external sensor. This was my first RC motor experience so I wasn't sure how easy it would be to take apart to adjust the internal sensor board. Turns out these motors are pretty easy to mess with. I love how easy they are to fan cool. What a huge difference that makes for a motor!

An encoder puts me in an entirely new price bracket with controller options. I'll figure that out when I have the money to do a full on motorcycle conversion some day. Affording a Sevcon for $750 using an encoder doesn't sound too bad compared to a box of burned up Castle HV160s.

If I did it again I would buy an 8072 from alien already sensored.
http://alienpowersystem.com/shop/brushless-motors/aps8072s-sensored-bldc-motor-165kv-6000w/

This 108mmx50mm looks really interesting too. Nice and narrow to fit between the pedals. Looks like the pole count is pretty high though and there just aren't a lot of proven sensored controller options for these larger RC motors. Hopefully that changes soon.
http://alienpowersystem.com/shop/brushless-motors/aps-10850s-sensored-outrunner-brushless-motor-50kv-5000w/

Thanks for the kind words.

 

[Rikobonito]

Thanks for your feedback Dan !!

Yes it was the plan, plus keep it lighter possible ... it will be a firt try on a old bike I got, then if all work good and want more power for less effort I will go the next size

Ok, yes, I see and understand better now, its the time to make the prototype works :wink:

Yes I see your point. I got bldc/FOC controlers with adc command control and sensor or encoder capcable and lot of 50,63 and 80mm motors so I play with some.
Got 2 like the 8072 and a 8085 sensored. Theyre for other project now, but I could swap it if needed.
The last one are scary ... for a real e-motorbike this one

Thanks for all, I will start soon I guess :v:

 

[DanGT86]

Had some time to mess around with this project this past week. I am currently having the best results running the motor with a lyen 12FET EB3xx generation controller.

It seems to work really well accelerating but has a weird stumble under load that is very apparent when trying to maintain a consistent smooth speed. I was guessing the hall signal was weak and under high amps the field noise from the stator is confusing the sensors. Just a guess really. I also had been very curious if the optimal hall sensor position would be different under load compared to unloaded. I had an old 9c hub motor laying around so I mounted that and used it under regen as a controlled load so I could fiddle with the hall sensor board while the motor was pushing against a load.

As far as I can tell the Hall position that yields the highest RPM using the lowest current draw seems to be the same while free spinning and under load. There was no clear difference. That makes me feel better about future builds where I wont have a convenient "dyno" setup like this.

It took me quite a few tries to figure out how to keep the CA from ramping up the output to compensate for the load. I ended up setting the CA3 to output in speed mode and then set the max throttle output voltage to 2.5volts. I then set it in cruise mode and adjusted the hall board to yield the lowest amp draw for that commanded wheel speed. I may try again with it in pass-thru mode with the throttle voltage clamped at a specific value.

It was pretty fascinating to watch the power flow from one CA display into the other. This loop from one motor to the other is about 50% efficient.

As a bonus, now that I know this 26" hub wheel fits with the mid-drive Im looking forward to giving it a test ride with both motors

Are there any USA based retailers offering internal hall sensor boards for these motors? I would like to try one but the shipping from Alien in the UK brings me to over $50 for just the tiny hall board.

Also before anyone glances at the pic without reading I would just like to say NO, I AM NOT TRYING TO BUILD A PERPETUAL MOTION MACHINE!!!
View attachment 1View attachment 2

 

[DanGT86]

Just checking in with a little update. I got a killer deal on an alien power systems Ebike ESC. Its a 120Amp 20s model. I figured that should be good for 30-40amps if I watch the throttle ramp. Its my first experience with a senseless RC esc. It's pretty hard to program it and it makes lots of fun beep noises that I don't understand. After lots of messing around I got it to operate relatively smoothly. This was a good troubleshooting measure to make sure the motor was not damaged internally and that the belt and chain will all stay in place. Trying to run sensored was causing such rough operation that I was beginning to wonder if the motor was damaged. It appears to be all good. Belt stays in place. Chain stays in place. Bearings and motor are all nice and cool. After about 30 min of repeated 3kw burst up to 30mph the motor was about 50deg C. Still lots of tuning to do if I want to keep this setup but its nice to finally get the satisfaction of riding it around.

[youtube]S_8qxKV6TYI[/youtube]

[youtube]05Dvvg5daR8[/youtube]

[youtube]UdkIiPTbDFA[/youtube]

[youtube]8bWDiU46K40[/youtube]