Recumbent Friction Outrunner Prototype

[evblazer]

I just finished lowering the frame of my cruzbike sofrider. I really liked the friction drive outrunner setup
http://endless-sphere.com/forums/viewtopic.php?f=28&t=9652
and the advanced friction drive setup
http://endless-sphere.com/forums/viewtopic.php?f=28&t=16550
but both mount on the top of the wheel and I didn't have the option or alot of space as you can see below. I really wanted something lightweight to just give me a little help once and a while and otherwise leave my bike as is. Undo a nut and a small bolt and it's off. Not super quick release but it isn't bad for V1.

I can't put it up front without risking hitting my leg on turns so in the back triangle it goes.

I went ahead a purchased:
HP-Z4045-14 HYPERION Z4045 14 TURN 236 KV BRUSHLESS MOTOR
PHX-ICE60HV CASTLE CREATIONS PHOENIX ICE 60HV (set immediate cutoff limit at 60amps extra caps on order)
RC Throttelizer from EVLogix.com with Dual Current Limit on order (Both amp limits will be less than 30amps)
Thundersky 20ah 24v pack shared with B&D Lawnmower with CellLog8 Monitor for testing.. lighter batteries purchased if I can get it to work.

I planned on mounting the motor alone on the below bracket and put the controller right next to the batteries under the seat. Here is a little mockup I make using sketchup from google.

Ended up using
1/8"x2" aluminum angle
Bearing to support the non-mount side of the outrunner intended for a router bit (22,000 rpm rated)
Aluminum Rivets
1lb pull spring

Top View

Side View

I ran a little test using my R/C Receiver it's control radio and the setup worked pretty well attached with alligator clips and up on the stand. Near the end I applied the brakes a bit which caused the motor to whine a bit and then slip which is where the graph below shows it spiked the amps a bit. Not having much luck converting the 3g2 video to something useful but I did download the Castle Creation output while I had it linked up to my computer to set the amp limit.
The wheel would keep spinning a bit after I took the throttle off and it would even spin pretty easily by hand. Not completely freewheeling but the wheel spins easily by hand with minimal pressure.

I first tried it on 12v then 24v after I knew it wouldn't fly apart. I have thought about putting a little strip of old tube on the contact area but I might wait and see.

 

[coln72]

Are you planning to increase the spring pressure holding the motor against the tyre?????

Just asking because I experienced slip when using a 31cc petrol engine in a friction drive set up with 4 to 6 springs that look a bit more "manly" than the one you are using. It was unuseable in the wet

 

[evblazer]

I don't think I'd want to use my setup in the rain. Were all the springs to try and fix it in the rain or just overall? If rain is the only time I can't use the bike i'll be very happy. If it rains I'll take my other recumbent or just use this one and disconnect the battery.
Hopefully UPS delivers my andersons and I can try it around the block to see if it'll give me any assist with the current setup or if I need adjustments. One I'm already planning is a cable pull to create a true freewheel.
The end goals is to add a bit of power to my ride and flatten some hills. I'll be pedaling 100% of the time it is on.

 

[coln72]

The springs were in use all the time as in previous years it has been found that even in the dry the roller - made from a skate board wheel - would spin on the tyre (20" slick) and wear it out. More pressure seems to stop that happening in the dry but in the wet it was unuseable. At least everyone else with friction drives suffered as well :wink:

 

[Kepler]

Good selection of gear for the job. I presume the bike has 20" wheels. Is there any way you can set the drive up to climb the tire under load? Its important that you dont have any slippage but the drive pressure then required is not good for coasting. I believe having the drive climb the tire under load then back off from tire when not under power is paramount to a good friction drive setup. Good luck with the build, I will be watching with interest.

 

[paultrafalgar]

With respect to slip in the wet, have a look at:
http://endless-sphere.com/forums/viewtopic.php?f=28&t=9652&p=183987&hilit=capstan#p183987

 

[evblazer]

The original pictured recumbent actually uses 26" (559) tires. My main recumbent has 20" (406) tires is going to be a real challenge to even find anywhere to mount the motor given the layout. Can't use the front due to leg interference and/or the front idler pulley. Very hard to use the rear because under the chainstay would be too low I think and the suspension and single pair of chainstays dont' give a really good mounting area. I had thought of a really solid rear fender and slicing a slot and mounting it there but once I sit on it and compress the suspension a little there is even less room then pictured. If I did electric this one up it would probably have to be a hub motor since I dont' think there is even a place for a gear reduction unit.

I'm going to do some measuring today to see if there is enough room to get it to climb the tire under load in any way. Also need to make it a bit adjustable if I change tire size so I don't have to redo the whole thing.

 

[Kepler]

My suggestion would be to mount the motor below dropout triangle rather then above it. There looks to be room to sandwich a nice solid bracket between the pivot and the tire which could easily be made quick relaease. You can the fabricate a post off this backet that gives you enough room for the swing action on the drive. You should be able to keep it compact enough to still give you plenty of ground clearance.

 

[evblazer]

Where was that suggestion an hour ago when I was in my workshop :lol:
It took me a bit to realize but since I'm using two seperate bushings on 1" bolts for the pivot I can put the pivot beside the tube in front. It was a PITA to get it setup but I think it works alot better now.
No Power

Power - Crawls up the tire and pushes into it.

Springs were there in a pitiful attempt to soften the drop when the system turns off. I will wrap a old inner tube around each side and remove the springs. The motor is stopped from rotating higher by the non rotating portion of the bell hitting the pivot mount. I am going to modify it to use a wire held down to the mounting screws and kickstand plate to avoid banging up the motor.

Below if the castle export where I was finding out the LVC of 24 volts really works even if I forgot that I put it on. 12v would kick it on really briefly and the 2 12v sla's which are too weak for my mower sag under 24 volts if I ask for any more than a few amps. Hit the brakes a little or tried to apply full throttle and off it went.

Still using the rc car/radio setup to run it but hopefully next weekend I'll be good for a test run.

I'm going to take a look at under the rear triangle to see how it would mount there. Getting the wheel off in the forward facing horizontal dropouts looks to be a bear the way it is now. There is a beefy kickstand plate there to bridge the chainstays in front of the wheel I could easily connect to.

 

[coln72]

With respect to slip in the wet, have a look at:
viewtopic.php?f=28&t=9652&p=183987&hilit=capstan#p183987

Thats the shape mine ended up after the approx 18hours of racing

 

[Kepler]

Thats more like it. Should give you a nice positive drive now. What stops the motor pivoting too far?

 

[evblazer]

Temporarily the non rotating portion of the motor striking part of the mount stops it. I am going to add a brake cable or picture hanging wire type setup to stop it just before the motor strikes.

Just a few things left:
Build new seat based on the blue recumbent seat.
Mount batteries under seat
Wire in additional caps I just got in the mail
Fix cellog8 to use all 8 cells
Get some parts in the mail so I don't need my rc controls/reciever.
?

You might also notice in my previous post I'm down to three screws on the mount. One of my threaded inserts fell out while I was pivoting the motor to stick the wires out a different side so I gotta use some tweezers and try to get it back in there before I do any riding too. Don't push on the screw when it is not threaded into the insert or anytime really I kind slipped and pushed it while backing it out.

 

[evblazer]

Cellphone cam quality
[youtube]N-kCkcXix5s[/youtube]
Wired up my thunderskys and no more lvc. Wheel spins quite a while after throttle off.
Also found I can use the forward facing horizontal dropouts to adjust tire/outrunner pressure without fiddling with the mount.

 

[EVTodd]

Looks nice!

Is it applying more pressure as the motor engages? It's hard to tell from the video but it almost looks like it's further off the tire when it pivots all the way up.

 

[evblazer]

Now that you mention it. It does look like it is going up just a smidge too much when in video. I paused it and it doesn't look to bad but I will measure closely before I setup the wire similar to kepler's. I did test out applying the brake a bit and getting the amps to spike and the motor groaned but the wheel didn't look to be slowing. I'll have to wire up a speedo to the rear wheel to verify it and maybe get a max unloaded speed. The motor is fully off the tire when it is at the bottom of the pivot.

Edit: Max unloaded speed ~33mph pulling 5 amps. I put move the tire further into the motor and same speed but 5.8 amps. Used brakes for a load and got it up to 15amps before I started get lvc cutout.

I'm not sure how it is going to work with the RC Throttleizer if I try and limit amps alot or rampup so I may need to rework for that and move the whole thing under so I can get more adjustability. On the bench it is having a hard time reengaging if the tire is moving I may have to adjust put stops in both directions to is doesnt' get too far away from the tire. Fighting gravity is hard

 

[Kepler]

15amp current limit might be a tad too conservative. I like 50A on 24V setup. This allows me to get up most hills and maintain a good speed.

In regards to getting the drive to contact the tire, yes gravity is against you here. That’s why mine hangs vertically and as such takes only a tiny torque kick to get it to contact the tire. Unfortunately you don’t have the same luxury and as such you will either need to have it under very light contact in the throttle off position or come up with some way of moving the drive to make contact. My original design used an RC servo with a cam arrangement to push the drive in contact with the wheel. I used the same signal that was being sent to the ESC but powered up the servo with a separate BEC as the BEC built into the Throttleizer doesn’t have the capacity to drive a loaded servo.

Another thought I had was to couple up the rear brake action to push the drive close enough to the wheel to have it pickup. In this config you would apply a bit of throttle then just give the rear brake a light squeeze to engage the drive. During normal braking the throttle would be off anyway and as such have no effect on the drive.

 

[evblazer]

Will I harm the controller if the throttle isn't on and the motor is getting turned by the tire?

I may end up putting both springs on and configuring it to lightly hold it just against the tire and use the friction shifter to pull it off the wheel when I want true free wheel. Another thing I have to try is to tru the wheel since there is a little bump on each rotation which I think is causing a slight issue as it puts alot of pressure on the drive then lets go. This also might be batting away the motor when the wheel is already in motion or may just be my imagination/false hope.Right now I just don't know if I hit a bump hard whether or not the motor will stay engaged. Not a problem if I was able to mount it on the top of the tire but since it is sort of under the front hm.. Well tommorow I took the day off to try and strap on the batteries and duct tape the radio to the handlebars and I'm taking it for a ride

15amps is when the low voltage cutoff on the controller was kicking off which is 24v on 8s thundersky pack. I was just running the demo of the castlelink and noticed the custom option so I could lower it a couple volts now and not have to wait for the throtlleizer. I'm also going to adjust it to a soft rather then a hard limit so it'll try and throttle the amps rather then just cut them off.
I did get the dual current limit of the throtteizer and I planned on putting one with low amps for cruising with just a little assist and on as a higher limit for hills. Even that 15 amps at 24 volts would be 360 watts and say 50%(?)efficient would mean 180 to the ground which is alot of power compare to what most casual riders put down. That may not work in practice but that would ideally what would be what I want to be able to do.

Just to maybe help see what I'm trying to tackle here is my one way commute. As always there is a stop I must stop at right at the bottom of all the hills. There are only a couple of steep spots and they aren't too long. I've been thinking I could just use the assist on some more level areas I need to sprint to survive and chug up the hills with little or no help or figure out whatever helps me not feel like crap the next day after riding my commute 2x in a row. Post surgery I'll be using the bike to do mostly more level rides but want some help if there are any inclines (cardiologist approval etc etc) and working up to and continuing to do my commute.

 

[Kepler]

Having the motor turn when not powered shouldn't be a problem. I have done a heap of testing with mine like this just to gauge if the drag of the motor could be felt when riding. In reality, you can hardy feel the motor if only touching lightly. It just sounds like there is a little bird chirping behind you which can be a little anoying.

Using a control to decouple the drive is one way of solving the issue. The only problem is that it adds another control to the handle bars.

Looks like you in quite a hilly spot. The drive is going to make your ride heaps more enjoyable and still give you plenty exersize which is main reason most people have a bike in the first place.

 

[evblazer]

Final pre-road test version with altered pivot and temporary cable stops so it doesn't hit the frame going down and the motor doesn't hit the mount. The original version was sort of climbing up along the tire this version climb up and into the tire a little more aggressively.
[youtube]N7yGzeBVYjA[/youtube]
I need to get some more robust connectors to hold the cable but I think it is ready for a road test. It even grabs after the tire is spinning which is a nice improvement.

 

[Kepler]

That looks like a nice solid drive now. I think that should work really well. I would also recomend the bike tube sleeve as this definitely improves grip without adding more pressure to the tire. I have pushed mine up to 2.8kW during testing without slip so its well worth the effort.

Do you have a servo tester you could use for testing? If not, they are a snack to build with a few easy to get parts. Here is a diagam of one I built before I got the throttleizer.

 

[evblazer]

Found a local hobby store and picked up a few things including a servo tester. I tried radio shack for some parts but they have crap and there is no where else that isn't mail order so now I have a bar mounted servo tester electrical taped onto the stem till my throttleizer and thumb throttle to be converted to a trigger throttle come in.
Test setup:
-Topeak rack with Topeak MDX bag carrying 8S thundersky 20ah cells (Will build under seat rack before official use)
-7 470uh 50v Panasonic Caps in parallel between battery and Controller
-1" of inner tube silconed to motor can
-Old Cateye cyclocomputer with broken mount taped to handlebar
-Astroflight servo tester taped to handlebar with cat5 cable extension cable.

~1-2 mile short test ride without speedometer went well. Checked all and the only problem was the inner tube had rolled down the motor can so it wasn't in contact with the tire
Went home and strapped on the speedometer and added 20psi to the tire and went for a real ride.
Longer Ride stats:
-8 miles
-2.58ah (did not reset watt meter from 1-2 mile test ride)
-41amp peak
~17mph avg
The drive worked really well as did the servo tester. I got used to flicking it on high to grab the tire in motion and backing it down. Going really fast down a hill would drop the motor off the tire and I could just dial it up a bit and it would grab again. If I slowed down going up a hill I'd dial it down a bit. I did get a workout trying to keep up high speeds although I did let it coast places I'd normally keep pedaling. The bike didnt' feel different but it sure sounded different with all the unsuspended weight on the back now especially over the cobblestone. I could hear the motor and did turn it off when I saw pedestrians and such to kind of sneak by but it wasn't too loud.

 

[lawsonuw]

With the "tire" that you put on the motor walking over to one side, it sounds like the motor and wheel are slightly misaligned.

My other suggestion would be to use a high quality superglue to glue the tire on. I used super glue to fix RC car tires to rims and it holds quite well. (especially when everything is perfectly clean.)

Lawson

 

[Kepler]

I don't use any glue at all the tire tube sleeve stays put without an issue. My alignment is only eyeballed square but that seems to be good enough. Are you supporting the motor on both sides? Your pictures dont show how you support the prop driver side of the motor, only the motor mounting side. If the motor isn't supported evenly on both sides, this would cause it twist under load which certainly would cause the tire tube sleeve to roll off.

 

[evblazer]

The prop side is supported with a router bearing held into a 3/8" aluminum block.

Kepler how big is your tube sleeve you have been using I couldn't find it in the post you make reference to adding one. Is it covering the whole can or just a piece? I checked mine and it was at about 1/2" width with another 1/2" torn up it was from a schwalbe tube I had gotten in teh wrong size 1-1.25" for a 451 tire. I may try to cover more of the can and superglue a new thicker slightly larger tube I'll pickup tonight.

I still have to true the wheel though. I think the tire is swinging back and forth as it taps the v-brakes a little each time around. That and there is a little bump to it both things may be contributing to the sleave rolling off. When I pull it off I may try a different tire on it too and if I end up keeping the bike I'll need to get a tougher rear wheel since the one on it is more of a racing 26" front wheel with a long axle and spacers.

 

[EVTodd]

Kepler how big is your tube sleeve

That's kinda personal isn't it? :lol: Sorry... I couldn't resist the 3rd grade humor.