First E-Bike build - Full suspension, geared & DD motor

[electraflyit]

Good looking build.
I use a 6 fet controller with the updated fets up to 35amps in short burst with no problems.
I have a 9c motor and 14s lipos.
I built up the traces with solder and partily bridged the shunt with solder.

Eddie

 

[shock]

SUCCESS!

I finally had some time to temp wire and bench test all this hardware and we are go for launch! I got it first try (although a lot of research) and to my amusement a half twist of the throttle and she put some serious torque to the ground, in the correct direction! I think I understand why guys get addicted to this stuff, seeing the wheel spin quietly by itself is awesome! Honestly I am amazed how much torque this thing has from a standstill, and for testing I am only running 12c1p @ 2.6 amp hours.

For reference I think I'll be using a relay to energize the controller, there's a nice solid connection spark when plugging in the batt, I assume a lot of current is flowing and filling caps.

So here is the connection diagram, phase first, halls second.

LYEN 6 FET > EBIKEKIT GEARED HUB

Yellow > Yellow
Green > Blue
Blue > Green

---------------------------------

Yellow > Yellow
Green > Blue
Blue > Green
Black > Black
Red > Brown


And here is the throttle connection:

LYEN 6 FET > EBIKEKIT HALF TWIST THROTTLE

Green > White
Red > Red
Black > Black

**I'm Excited!!**

 

[kfong]

Problem with relays are the contacts can be fused together at these power levels if you don't get the correct ones. Using a precharge resistor is a better approach or you can do what I did and build a solid state switch. Do a search on precharge resistor or follow my link on the smart switch in my sig.

 

[kfong]

FYI Hyperion 1420i is on sale on eBay for $129. I just picked up a spare. That and a server supply would get you a great charging setup for $150 total.

 

[shock]

Mocking up the shape for the triangle container with some stryofoam

 

[kfong]

Looks like I won't have any spare lipos after all. I found out I can get 12 packs in the triangle, so I ordered 3 more. Tested 3 packs so far with a total of 2.4 usable ahrs average.

 

[shock]

More component layout

 

[shock]

Wow, how are you fitting 12? My triangle case width comes to 2"

 

[joe tomten]

I put that hub on a fs mongoose like dogman has done, and I took off the 13t freewheel that came with it and went with the 11-28 (32?)t freewheel that you can get from ebike.ca and elsewhere. At 48v that is the most straight forward way to get the gearing low enough to still effectivly pedal at wot.

 

[kfong]

The layout is posted. It all fits nicely. Only half are shown. Width is double wide, but still slim enough not to interfer with peddling.
http://www.endless-sphere.com/forums/viewtopic.php?f=6&t=12654&start=150

Wow, how are you fitting 12? My triangle case width comes to 2"

 

[shock]

2 things.

As seen in the picture, this setup has no problem laying down some rubber!

Next, the wheel after bolting up is slightly off center. I thought of trying a smaller freewheel, but really the issue is the ebikekit hub axle. Even a smaller freewheel would not relocate the wheel. So is this common? I assume it will affect handling somewhat, ecmven though its barley off. The other cause maybe the frame is not setup for a freewheel? Thought??

 

[Alan B]

Dishing needs adjustment.

 

[Russell]

2 things.

As seen in the picture, this setup has no problem laying down some rubber!

Next, the wheel after bolting up is slightly off center. I thought of trying a smaller freewheel, but really the issue is the ebikekit hub axle. Even a smaller freewheel would not relocate the wheel. So is this common? I assume it will affect handling somewhat, ecmven though its barley off. The other cause maybe the frame is not setup for a freewheel? Thought??

Yes unfortunately having the wheel offset to the left is all too common a problem with ebike kits. The problem is that more space is required on the right side for the freewheel than on the left side. On a regular bike this is also the case however regular bike builders know enough to dish the wheel when they lace it. It actually took the ebike builders a while to catch on to what was common knowledgein the bike world. I thought Jason at E-BikeKit was clued in a long time ago (he claims his wheels are hand-built here).

Still sometimes when a motor is built symmetrically it takes dishing the wheel to the extreme, where the right side spokes are completely vertical, to get the rim centered. This makes for a rather weak wheel so perhaps in this case the wheel was dished as much as it could be. The only way to really fix the problem is to redesign the motor with offset flanges to compensate, and this is what Crystalyte has done for a few years and it's what 9C did with their last redesign.

That said you can try to fix the problem yourself. Loosen all of the spokes on the left (non-freewheel) side the same amount then tighten all of the spokes on the right side all that same amount (try 1/2 turn at a time). If the wheel isn't already dished to the max this will move the rim to the right. If you aren't gonna use disc brakes and your rear dropouts can be spread you can insert a couple of washers on the left side to make the left axle about the same width as the right.

-R

 

[shock]

Thanks for the heads up guys. I'm going to have my local bike mechanic take a look at it while he does the derailer recabling.

 

[kfong]

I had the same issue, had the local REI redo it. I think it was under $30 to have it done, but that was before I had any bike tools or background in it.

 

[shock]

Got the bike back from the shop, cabling is done, seat post has been acquired, and I spent a good portion of the evening doing more assembly and component layout. Also finished up the chain and pedal drive system. I have a lot of pics, going to download them and post them up here in a bit. She's almost ready for a maiden testing run. :-D

 

[shock]

Here's some pics of where I'm at. Things are set up for testing, while I fabricate the triangle case. For starters I am going to run 12c1p and see how far 2.5 amp hours gets me Used some zip ties and cable wrap velcro to temp. attach the controller.

 

[kfong]

I would take it easy on those packs at first. I know they are 20C but they did get slightly warm on me when pulling 7 amps continuously. Be sure to monitor the voltage and how much they end up sagging. You bike at full acceleration from start can easily pull over 40. If you have a CA then set it lower.

 

[shock]

Im running the lyen 6 FET controller, which is limited to 25amps current @48v (at least I think it's limited based on the description: http://endless-sphere.com/forums/viewtopic.php?f=31&t=18744 )

I don't have a CA (yet!) but I do have the USB interface to program the controller, and right now my cutoff is 44v, or 3.66 volts per cell.

Probably safer for the batts. to run a 12c2p config, but this is a maiden voyage around the block a couple times, half throttle. My intent is not to max out the system, rather conclude the LVC works, and give my parts a little "burn-in" :-D

 

[kfong]

I would then pedal first, once it's rolling it takes about 13-15amps to keep it moving. It's the initial torque load that really draws out the current. If these were nanotechs I would not be worried. Would be interesting to know how warm they get under load. 7 amps was all I could push on the bench without having to build a bigger load double that would of been a better test.

Im running the lyen 6 FET controller, which is limited to 25amps current @48v (at least I think it's limited based on the description: http://endless-sphere.com/forums/viewtopic.php?f=31&t=18744 )

I don't have a CA (yet!) but I do have the USB interface to program the controller, and right now my cutoff is 44v, or 3.66 volts per cell.

Probably safer for the batts. to run a 12c2p config, but this is a maiden voyage around the block a couple times, half throttle. My intent is not to max out the system, rather conclude the LVC works, and give my parts a little "burn-in" :-D

 

[Scottyf]

I don't have a CA (yet!) but I do have the USB interface to program the controller, and right now my cutoff is 44v, or 3.66 volts per cell.

Probably safer for the batts.

Very safe, since the voltage sag will be a volt or so lower under load. If anything the LVC is a little high to get the full capacity out of the Cells. Either way it should be a safe setup while the cells bed in but its worth noting alot of people have the LVC to about 3.2v per cell to 3.4v.

 

[shock]

I'll set it up for 40v LVC when she's ready for production

I really need to find myself some good connectors, I use Deans on my RC stuff, but looking into the anderson connectors I see everywhere on this site.

Also, I have some old 3 cell blue lipo's from about 3 years ago that I used on my first Trex 450, and I've abused them pretty good. Ran them to 3.1 volt per cell, crashed a couple times nose first, put them through hundreds of charge cycles, and although puffed out today, they actually still work pretty good. Like I said, I'm a fan of the blue LiPo. For reference, I am not going to treat this high voltage setup in the same manner, I've learned a lot since then, but just wanted to attest to their quality and ruggedness in extremes.

 

[kfong]

I was just getting into helis when the ebike bug bit me. Haven't done much with RC's since. Still have a lot of planes to build and fly and piles of old lipos collecting dust. Glad the blue lipos hold up, I've had my share of crap ones.

 

[shock]

LOL I'm kind of in the same situation. I've been into Heli's and planes for awhile and still fly them regularly, but just started getting into EDF jets when the ebike bug hit. I'm currently about half done with my 3rd A-10, and have a F-35 kit still in the box for a rainy day.

Today I ordered a 50 pack of red and black 30 amp anderson connectors from powerwerx.com. I'm going to convert all the lipo's over to this style connector and make some parallel/series harnesses with them.

I also had the CNC shop I use for fabricating, order me a 4' x 8' sheet of .063" thick white epoxy finished aluminum. Should be plenty for the triangle case and any bracketry needed. I am a mad scientist with this material Love aluminum for applications like this where lightweight, strength, and thermal properties are important. The entire case will basically be a huge heat sink.

 

[kfong]

I look forward to what you come up with. From my RC days, composites worked out well for me. I went that route for the enclosure.

I still have a basement full of RC crap and some helis that I was starting out on, even looked into uav and multi rotor stuff. i let my membership lapse, that's how much I lost interest in it. I will get back to it one day, it just can't compete with mountainbiking, kiteboarding, windsufing and waterskiing. I even built a CNC table just to cut balsa. Where do you fly at? Not too many parks I know of on the east side? My last plane was a wild wing with my own custom vector thrust, that still needed to be covered. I was hoping to at least get that one done since it would have really crazy flight characteristics. Never even had time for it.

I've always been keen on the andersons mainly from my electronics background, another overriding hobby. But the RC guys always bulked at the size. They can be cut down though.

LOL I'm kind of in the same situation. I've been into Heli's and planes for awhile and still fly them regularly, but just started getting into EDF jets when the ebike bug hit. I'm currently about half done with my 3rd A-10, and have a F-35 kit still in the box for a rainy day.

Today I ordered a 50 pack of red and black 30 amp anderson connectors from powerwerx.com. I'm going to convert all the lipo's over to this style connector and make some parallel/series harnesses with them.

I also had the CNC shop I use for fabricating, order me a 4' x 8' sheet of .063" thick white epoxy finished aluminum. Should be plenty for the triangle case and any bracketry needed. I am a mad scientist with this material Love aluminum for applications like this where lightweight, strength, and thermal properties are important. The entire case will basically be a huge heat sink.