AWD fatbike (first build, first draft..)

I really like what you are doing with this ride.
I am dreaming up a dual geared hub Fatbike myself.

Can you share an update/opinions of the Roboteq HBL2350?
How well does it run the two different types of motors? Any sync issues?
Can you give more detail/pics of your throttle solution?
I have long desired a snomobile/jetski-style pull-lever throttle, I think you found a nice solution.

Are you now running one C.A. from the battery for monitoring?

thanks for sharing.

Enjoy
 
REALLY Really curious about how you managed to get your crystalyte motor to fit into a 170mm dropout, any enlightening words of wisdom? :D
 
I STILL want to know how this bike fairs on the sand with the fat tires. THAT is the major advantage of this particular bike. You can't take your ducati to the beach and stomp around.

I guess if you don't have access to a beach it makes it my point null. But I'm interested since I have access to the beach and wouldn't mind being able to pass all the stupid rollerbladers on the bike path at the boardwalk.
 
I haven't been attentive to my own thread-- I plead ignorance. Let me try to answer what I can:

- Thanks for the kind words about the throttle I hacked up. It works really well... if you like thumb throttles. (I'm discovering that I don't.) I just stuck a linear displacement sensor into the throttle body, and when the outside lever is pressed, an inside lever squeezes the sensor, altering the voltage. It is the usual +/-/signal wiring, and I had to play around with resistors to keep the signal strictly inside +0v,+5v. Pictures might make this a little clearer:

outsideThrottle3.jpg insideThrottle9.jpg


- I am super happy with the Roboteq controller. I keep discovering new ways to use the fairly primitive programming language to add reasonably sophisticated functionality: speed sensitive acceleration, continuously variable regeneration, launch control, et al..
In fact, back to my previous point, I'm thinking of replacing the throttle with a two-pushbutton throttle. It will be easy to use two digital inputs and the controller's "microbasic" to implement a system like: double-click 'up' button for launch control, double-click 'down' disengages motors, triple-click goes into regeneration. Except if speed is zero, down button reverses motor.

The two channels is also a big win, especially since I can still treat each motor differently. For example, I have a geared motor on channel 1 and a direct drive on channel 2. So I have deceleration set to 50,000 RPMS/sec for ch1 (since that motor freewheels), and 200 RPMs/sec for ch 2.

- I wish I could tell you how I managed to get that motor on a 170mm dropout. I didn't even realize that was unusual, to be honest. I got the motor from EBIKES SF (Ilia) and as far as I know it was stock.


Hope that cleared a couple things up,
s.
 
Ah-- I'm back in love with my bike, thanks to LiPo. I had been using a 50v 10ah LiMnCo battery from batteryspace which seemed solid, but only put out 30 amp continuous/40amp burst.

I didn't realize how much that was holding me back until I bought two 7s 11,000 mAh LiPos from MaxAmps, supposedly rated at 40c. I'm not sure if these batteries are actually capable of putting out 200+ amps, but my bike's performance says it's close. The bike now pulls faster than I can hang on, I've had to modify my controller code to moderate the acceleration right off the line.

I stuffed the batteries into a battery box I had lying around from some other kind of chemistry, and then put a whole bunch of n53 magnets down the spine to keep it attached to the bike. (The magnets make a very satisfying "ker-chunk" noise when putting the battery box on, but it does rattle a bit on the road so I need to work on that.) I also need to modify the box so I can get to the battery balance taps without taking the batteries out of the box etc.

The bike is so much more fun now that I can race ahead of cars from a stop...




s.
 
Awesome build. The only worries me that I do not see any torque arms at all..
 
I agree if you don't have TA'S you NEED to get some!
 
Awesome build!! Have you ridden on sand yet?
What is the top speed now with the Lipo batteries and what is the current configuration of them?!
How is top speed handling with this balloon tires?

Also this part:
So if the handlebar torque sensor signals intent-to-turn, the Arduino will give more throttle to the front motor and less to the rear, to help pull the bike though the turn.

How did this work out for you? It goes against the rules of high speed cornering on motorcycles where you should be gassing through a turn to hold higher cornering speeds. Seeing that you also ride a duc, I assume you know that and may likely know something else that I don't.
 
I agree you definitely need torque arms. It is only a matter of time before the axles spin out and the wires get ripped out by the roots.

Having a direct drive hub on the rear retains the future option of using regen as a brake (which keeps your stock brakes cool so they work at maximum effectiveness when you really need them). However, regen would make the lack of torque arms even worse...

Cool build though, thanks for posting...
 
I have to admit that I've been primarily enjoying the bike more than working on it. (A distinction without a difference to some, I realize.)

Right now I'm running off two maxamp 7s 8ah, which in maxamp's inflated world are sold as '150c'. (While that seems high, they have
lasted a lot longer than other lipo's I've tried.)

I'm constrained to 14s by the Roboteq controller's max voltage of 60v. The exciting news is that roboteq is supposed to release
a new dual-channel controller Any Day Now that will support FOC, sinusoidal, etc. However, it will still be 60v, 60amps per channel.
I've spoken to the designer and he claims it would be fairly trivial for them to do a 100v version, but they want a minimum order of 500
(at about $700/ea). If anyone has money burning a hole in their pocket..

s.
 
Nice build...but i have concerns about ur front fork setup...unless u have a good dental plan....if not no less than a cromoly fork, bmx ones are sturdy and cheap....carbon blings but also breaks..just google images carbon fiber failure...
 
I'm Happy to see this thread relived!

I was looking at doing a Fatty 2wd build. Nice work here!

Tommy L sends....
 
Hello.

Is that a MAC motor in the front?
I am currently making a almost identical rear-wheel for a fatbike, what kine of spoke lenghts did you use in the front wheel?

Regards
Magnus
 
motor: similar to a mac, but it's a bmc v2t (rear version in front thanks to the 130mm dropout)

torque arms: yes, I do have torque arms on front and back. They are pretty crappy though-- generic two-piece kind, and using hose clamps on the rear wheel. I'm now putting together a v2.0 of this bike, this time using a lynskey stratus frame which has slider (paragon ti) dropouts. Has anyone made (or sell) bolt-on replacement dropouts that integrate with/include torque arms?

speed: right now I'm using 14s LiPos so bike only gets up to 40mph (with peak 75 amps to front, 150 amps to back, acceleration is reasonably snappy tho). It's probably a terrible idea, but I've hacked the roboteq firmware so it won't give an overvoltage error above 60v, so I hope to try out an 18s battery later this week...

fork: I'm actually pretty happy with my carbon fiber fork, and I haven't heard of anyone cracking this particular product. But I do worry about hitting an unseen pothole/curb at 40mph and losing control of bike. Suspension forks for fatbikes are pretty limited... Rockshox just came with the bluto, but it's kinda crappy. So I'm currently rebuilding a used Marzocchi shiver fork from the 90s that provides just enough clearance for fat tires.


s.
 
Why don't you use a motorcycle frame? You are already well within motorcycle territory. A motorcycle frame will have components that will be able to handle the torque and tire width you need. I seriously doubt you are going miss the pedals at 40+mph and 10,000w.
 
Moto frames are usually built for a different set of forces than a bicycle or hub motor e-bike (e.g. reaction force on the motor, drive chain tension, torsion from the motor's tendency to gyro-precess). But it would be a very good idea to build a frame for the purpose rather than subjecting a bicycle frame to a very un-bicycle-like set of conditions. If desired, it could even look just like a normal bicycle frame, just using larger diameter and/or thicker-walled tubing.

If you built a chromoly frame to the same dimensional specs as a typical aluminum bicycle frame, it would weigh three times as much. But it would be about three times as stiff and strong, too.
 
I've built a new version of my fatbike -- slightly different motors, slightly different frame, a lot less spaghetti wiring. But the biggest improvement so far is the software.

I stole Roboteq's clever method of using speed (rather than current) sensing to, essentially, calculate BEMF in order to provide a torque throttle and variable regen. (Their explanation is at http://roboteq.com/index.php/applications/how-to/156-controlled-regen-braking). I then modified it to support variable voltage, a (geared/nonregen) front motor, and a reverse mode).

The result is significantly superior to my old voltage-based throttle in every way... recommended!

s.
 
Going Backwards

I continue to improve my bike, albeit at a much slower pace since I prefer to actually ride.

Hardware:
On the hardware side, probably the biggest improvement has been the wheels. I found 36H, 65mm carbon fiber (I know, I know..) rims from Nexie. After some struggles I've been able to run them tubeless with essentially no problems, using the Speedster 26 x 3.5 tire from Vee. Since I was previously using the 65mm steel Weinman's (?) with two tubes in each wheel (for fail over) and Surly's Black Floyd I've saved a ton of (unsprung) weight. Although the main thing I've noticed is superior rollability.

Admittedly, the new wheels may end up costing my life when the CF goes out, but... did I mention the weight savings?


Software:

I'm most interested in playing to the advantages of electrical motors, and in particular, how a direct drive (in back) and a geared motor in front can be combined to give you the best of both.

To those ends I've wanted to add a reverse mode for some time: on my controller, I just have to provide a negative power value to make the rear motor go backwards. Now seemed like a good time to do this since I had replaced my throttle with a single axis thumb switch/joystick that defaulted in the middle.

Remember, I'm using a speed-sensing (rather than current sensing) torque throttle, implemented in software. That means when the throttle switch is in the middle ("neutral"), the controller gives just enough power to overcome cogging as I pedal. (It feels to me as if my back motor is freewheeling just like the front BMC.) Obviously, pushing the switch to the right causes acceleration with increased torque, while pushing the switch to the left is engine braking/proportional regen.

I wanted to overload (in the CS sense of the word) the throttle semantics so that if I was going slow enough, and pressed the throttle all the way to the left, the back motor would start turning backwards. It took some experimenting to get the constants right but I ended up with:

if ( ThrottleCommand < -800 AND MeasuredSpeed < 12 ) # MeasuredSpeed is from back brushless motor, unit is RPMs
SetCommand( _G, 1, 0 ) # Turn off front motor so it freewheels
SetCommand( _G, 2, -200) # Reasonable reverse speed
end if


This makes the delay between coming to a (near) halt and then reversing tolerable. Of course, you have to remember the pedals are gonna start revolving against you, and, no, I can't maintain balance going backwards for longer than a few seconds, but it makes for a pretty neat trick.

s.
 
Newsflash: If you make enough dumb decisions in a row, something bad will likely result.

In my case, that day's dumb decisions began with deciding to race a motorcycle from a stop light, on a damp evening, without any lights on my bike, and culminated with me deciding that the car at an adjacent stop sign saw me.

It did not.

Going 33mph into, and then over, a car hood is hardly the stuff of evil knievel, but it was sufficient to cleanly crack my front fork in two, throw me quite a few feet, and break all the major bones in my ankle. (I figure, though, that my carbon fiber fork saved me from worse injury, by sucking up all that kinetic energy.)

Live, learn, etc.. Once I was off the knee-walker, I rebuilt the front end of my AWD ebike and was... off to the races, nach.

Still, it sparked plans for a new version of my ebike. I want to mess with the newer, smaller, FOC-supporting Roboteq controllers. They have improved CAN support, so they can be networked together pretty easily. Although the two channel version can only handle relatively few amps per channel, I figure by using two of the one-channels, I can get much quicker acceleration, shorter wire runs, and the FOC/field weakening will make up for the slightly lower voltage, in terms of top speed.

That's the plan. But, for now, after ordering a custom frame and the major components, I'm out of cash. So I've built up the bike to work as a pedal bike until I can afford to go further. It looks like:

newBike2.jpg


Noteworthy aspects: the 24" x 80mm CF rims, the Pinion XR gearbox, and the Onyx hubs.

s.
 
Yikes! Sorry to hear that. Did the car's insurance not pay to replace your bike since they ran the stop sign?!
 
Thanks to some bystanders who saw what happened, the police report put the blame squarely on the driver. But given the shabbiness of his vehicle, driver's lack of english, and -- most importantly -- my ultimate culpability in driving like an idiot, I chose not to pursue...

S.
 
Back
Top