A2B Metro Drivetrain Conversion

[altonk]

Hello all,
I am new to this forum and this is my first electric bicycle project. I've built many internal combustion trikes, racing karts, etc but this is my first electric project. I bought this Metro to use around town and have left it mostly stock up to this point. I'm on motor number three in the four years I had it and this one is starting to act a little flakey. So, I decided that a preemptive strike was in order, and proceded to look at a drive train replacement. I've been reading a lot of the info here and have looked into several alternatives, including middrive and lefthand drive.

I thought that I would document the project here and hope for some advise.

So here is the bike as it stands now.



https://www.dropbox.com/s/azk0o6uh79xa9oz/IMG_20130227_184830_3.jpg

I upgraded the brakes to avid bb7 from the stock five.

https://www.dropbox.com/s/rblv9sq9nls0prm/IMG_20130227_184918_1.jpg


also put on good high grip wide Crank Bros Pedals


https://www.dropbox.com/s/1tnsmys674zh89f/IMG_20130227_185004_3.jpg


I also replaced that toy shock that came with it with a lower end Rock Shox air shock. This made a world of difference with all that unsprung weight on the rear end.



https://www.dropbox.com/s/cq15a2382pqytth/IMG_20130227_185031_1.jpg



I added a topeak seatpost mounted rack and bags as the a2b was not standard. I also now have a protected area to put the controller, on the stock rack and below the topeak.

https://www.dropbox.com/s/7cu4vgvu7...m/s/teqq825rhd8ggnp/IMG_20130227_185128_1.jpg

I think that this is a great frame for the street fighter commute bike. The Steering geometry is good and the suspension works with the aftermarket shock.
The Battery storage and wire routing issues are all dealt with also. Plus I think it looks great.

Now as to the drivetrain, cheap junk.


So here are my goals with this project.

1. Reliability.
2. Moving the top end from the speed limited 20 mph to maybe 25, 30 tops.
3. Being able to deliver as much torque as possible in short bursts in the 15 mph and below range.

I basically want a street electric trails bike.

So I looked into two main drivetrain options.

1. Mid Drive
The mid drive option looked promising, I liked the ideal of a gearbox, but it was too complicated and expensive. I got a quote from ecospeed. too pricey. I looked at the GNG product but I don't have a machine shop.

2. Hubs
This turned out to be a better solution for me. This upgrade is simple and I can solve my power problems by throwing electrons at it like the electrical engineer that I am.
A hub based solution really leaves me just having to worry about breaking spokes and vaporizing copper. The thermal management can be done in the controller.

I'm not trying to build a fire breathing 8 Kw monster. I just want to have somewhere around 3 Kw around anytime I need it for short bursts.

From what I have read here I believe that's an achievable goal and can be done reliably. If I'm wrong please let me know.



So next the design decisions and what I chose as the replacement drivetrain



I wanted the cruise speed to be around 20 mph and still have a decent range.


So I went to http://www.ebikes.ca/simulator/ ebikes great simulator and chose the following:

HS3548 motor
ebikes 40a infineon controller
A 20" wheel
mountainbike aerodynamics and a 330 lb gross weight.
I set the battery to 72v 15 ah and a low enough internal resistance to make the controller the current limiter in the system.

When set at 42 % throttle on a 2% grade the numbers look like this at 18.7 mph


Mtr Torque 20.3N-m
Mtr Power 663W
Load 556W
Efficiency 78.6%

Electrical

Mtr Amps 29.4A
Batt Power 844W
Batt Amps 11.8A
Batt Volts 71.4V
Performance

Acceleration 0.281ft/s2
Consumption 45.1Wh/mi
Range 23 mi
Overheat In never

This means I can go up at 18.7 mph for 23mi on 15 ah @ 72v.

I'm going to try to stuff 30 ah in the frame and secondary battery so range is good and it shouldn't overheat.

looks like it will generate 95 lbs of thrust at takeoff and max out at 37 mph all on a 2% grade.

 

[richmpdx]

Are you sure that the motors have actually failed rather than the controller or simply some part of the wiring. The A2B hubmotor has a lot more capability than can be obtained using the stock controller. I suspect that the HS3540 offers little or nothing over the stock A2B hubmotor. The following thread describes a big upgrade in performance using a non-stock controller and lipo batteries:

http://endless-sphere.com/forums/viewtopic.php?f=6&t=43482

Rich

 

[altonk]

I'm sure it was the controller, The first one lost a phase then died, but I was in the middle of a 47cc two stroke kmx trike at the time and just wanted my ride back. I'm in a hotel in College Park, MD now and i don't have the machine tools to play with the stock motors here but I will have two to play with when I get back to north carolina. This is corrosion related too I suspect. I'm a Electrical Engineer of sorts, a signal processing guy, and i checked all the wiring related stuff. from looking at both motors I believe that i can push more current through the HS3540 and I like the screwed on side covers as opposed to pressing the stock one in and out every time I want to get into it. I can do this conversion in my room with hand tools and a soldering iron.

I am starting with a 72 volt 40 amp controller and am going to series connect the stock batteries till i figure out a replacement. I saw in another post by ypedal that he had connected a CA to his and that the pack was limited at 35 amps.

The plan is to find the highest capacity/discharge rate rc lipo battery packs that I can stuff into the existing battery storage space. I have a secondary battery and I'll start here.

this 10s1p 20c Constant 50c Peak 5000mah cell looks good.

http://www.hobbyking.com/hobbyking/...y_nano_tech_5000mah_10S_25_50C_Lipo_Pack.html

I'll run it in 20s pairs and parallel as many as I can stuff in the bike.


I'm going to epoxy a thermistor onto the windings and I have ordered the Cycle Analyst V3 which has simple thermal management/ current limiting.

Ideally I would like to have the CA control the max current such that the controller can predict the temperature of the windings, using a kalman filter, given the outside temp is relatively constant and then limit the current profile to prevent overheating. Then it's just up to me when i want to spend my heat budget with the throttle.

The more cooling i add the more current i can push through it. but i think it starts to saturate around 8 kw. but with the batteries I'm selecting my peak current budget for constant current is 100 amps or 7.2 kw.

 

[Ypedal]

the stock motor is more than capable of taking lots more power with an external controller.. could save yourself a bunch of cash.. yes, pain the butt to un-lace and re-lace the hub but radial is an easy build.

Th, the stock controller peaks out at about 25 amps, and that's about all this pack ( 2C ) can muster, if the bms does not trip at 40 amps, expect serious voltage sag and pack death..

Be carefull with lipo inside the frame, if they puff, it could become a big mess trying to get it out...

My a2b is about to get the treatment......

 

[altonk]

I was planning to current limit the controller to about 25 amps till I have a battery replacement worked out. But at 72v thats twice the power.
I have a tidal power project planned for those other two motors but that's a different forum.

On the battery front I plan to have full HVS and LVS cell monitoring and High enough C ratings that I won't push them that hard. I am going to have the balancing leads available so I can bulk charge them daily and trim them up as needed.

 

[altonk]

Ok the conversion is done and i've figured out how to display pictures.

Before:



After:


Every thing went well with one exception. I assumed that the phase wire connector was configured for the motor as i got them from the same vendor. But alas no .., I twisted the throttle and a few poles later it stalled actively with about 10 amps. My first thought was the hall sensors but i checked them with my handy $89 dollar android oscope and they were ok. after checking every thing else and having visions of unspoking the rim RMA ..... I decided to look at the Phase wire connector and they were wrong. But I also discovered that the individual poles of the connector slide apart and i realigned them to match colors from the online docs. After buttoning the shrink wrap back up it spun up perfectly.


Here are some more pics.








In general I like everything about it. It's more responsive, maxs out at around 31 mph @ the 20 amp current limit of the battery and seems efficient. I am in the process of getting log data over the weekend and will have some plots next week. More on setup and configuration although it was pretty much straight forward out of the manual. I've tried all three controller modes and I find the feedback loops hard to tune across the whole band. I'm assuming it's the delay in the loop. I also think I will get a good linear resistive throttle. If anyone can recommend one a half throttle with a button, I would appreciate it.

So next post I'll go thru how I configured everything, I don't have a controller programing cable yet so that's all stock

 

[AussieJester]

how much are you willing to spend will ultimately decide what upgrades you will l get.. I would recommend a mid drive, Recumpences da Vinci drive coupled with a 3220 astro motor would.fit nicely under the swing.arm.I.think? either way best of luck, it's n nice looking biker imo

KiM

 

[malylemon]

Hi. Can you give me more information about rear air shock? Thanks