Pusher trailer build

[izpirkt]

The 24-60V controller seemed like it would solve my LVC problems, so I ordered one. I set it up on the workbench, and hooked up all 4 batteries.

high res link: http://www.flickr.com/photos/59120715@N02/5489106620/

I turned the throttle, and all I got was a blip at about mid twist. I tried again with a slow twist, and got a "groan" or "hum" after about 20 degrees of twist, and then nothing. The Cycle Analyst screen went blank. I let off the throttle, and the CA came back on. It seems like the BMS on the batteries is cutting out. I tried again, just barely twisting past the first hum, and with very careful increases, got the motor to spin all the way up. The active portion of the throttle is only about 20 percent, near the middle of the twist. It hums for about the first half of the active portion, and then quiets. The humming seems to be coming from the hub motor itself, and the inductor that is in between the batteries and controller.

The peak amperage on the CA reads just under 18 amps. This is seems to come right at the first power delivery to the motor. I should be able to get 80 amps out of the 4 batteries in parallel, so I don't know why they are cutting out.

Any ideas? bigger inductor?

[kfong]

It sounds like you are getting voltage spikes back into the Dewalts. They shut down when they see any abnormal spikes or current draw and need to be reset. Try winding more wires through that inductor to increase it's value. You can also add another one on the negative wire. The photo is hard to see. I'm not able to verify your wiring setup due to the resolution.

Looking forward to the pusher build, I've thought of building something similar to exend my range. I hate the idea of having to drive to the trails just to ride my ebike. A pusher would solve that problem for me, just lock it up once I get there. It would also be a great way to carry extra gear or groceries.

[izpirkt]

That photo was supposed to be a link to a high res version. Let's try this:

http://www.flickr.com/photos/59120715@N02/5489106620/

[izpirkt]

I think I figured it out. I had been using the BMS interface boards without the spacer that kfong sent along with them. This spacer puts force onto the PCB to make sure it stays against the contacts in the battery. The batteries would work without the spacer, sending a readable voltage. I thought the spacers were just there to prevent vibration loosening during road use, and would not be needed on the bench. However it seems that it is not just vibration that the spacers help with. I must have been pulling voltage out of only one activated BMS, (hence 20 amps) leaving the others to cut off on their own.

So, with the spacers, I am not getting any blips, and managed to pull 50 amps! I can twist the throttle without having to nurse it along. Yeah! Now I've got some torque to play with!

[izpirkt]

Shoulda done it like kfong said in the first place!

[kfong]

Yeah, you definitely need the spacers. Without them they don't push up against the contacts. Very important to have.

[izpirkt]

So, in addition to the electronics portion of this build, I have also been working on hardware. I wanted to make a very short trailer, basically just a wheel. There was a caution from Rassy about short trailers being twitchy. I welded up an ugly prototype so that I could pull it around and feel some of the dynamics.

I started with a plate that clamps under the rear quick release on both sides of the bike. I needed an additional screw into the frame to prevent rotation of this plate. There is a weld nut on the plate to receive a bolt that will retain the trailer.


The trailer itself is a mash up of random donor bike parts. I've got the rear frame off of a cheap suspension bike, a head tube, and a set of forks. I stuck an old rear wheel in it so I would not have to risk my hub motor in case of major failure. It can move up and down by rotating around the bolts on the plate on the bike. It can rotate side to side on the head tube axis (between the green and gray colored parts)


The trailer followed smoothly and was unnoticeable most of the time. It came right behind me through tight turns and up and down curbs. Originally the tire was at full pressure, and I could feel it bouncing over every sidewalk joint. I couldn't feel much through the bike, but there was an annoying jolt I could feel through the handlebars. I stopped and let a significant amount of air out of the tire, such that I could deform it easily with my thumb. This really improved the bouncing, such that you could forget it was there. I think the whole setup might weigh 10 pounds.



Boy did I get some strange looks pulling up to a stoplight with this contraption hanging off the back of the bike!

[izpirkt]

I do have a couple of concerns about this setup. One is the fact that the trailer will probably behave differently with a significant load on it. The other is that at high speed when the trailer hit a bump it would fishtail a bit. When it landed it would snap back into place with a pretty good thump. When you watch the videos about the extrawheel, it is bouncing and twisting the whole time.

http://www.youtube.com/watch?feature=player_detailpage&v=6ijnEr1kn98

It seems that these trailers have been pretty well thrash tested, and come through all right. I need to test how they will react when powered. Is the bouncing going to really diminish their usefulness as a pusher?

I really like the idea of a short trailer because it will store and park much more easily. I hope to be able to get the wheel to fold up next to the rear wheel of the bike so that it can easily be chained. I'm wondering if an extendable frame would be the answer to this. It can be extended for use, and more capacity, and then compressed for parking and storage.

[Rassy]

Nice job on the test trailer. Be sure to make the adjustable frame pretty ridgid.

The old BOB trailer pictured below made a good pusher. This model with the 2 foot long Action Packer tote only got a little squirelly at speeds over 20 MPH or if heavy items were attached on top of the tote. (this was an actual BOB model with the tote, but I never could find any info on it. I cut the original fork off and replaced it with one from an old bike to accomodate a motor)

[izpirkt]

I did some welding this weekend, and attached two platforms to the sides of the trailer fork. These will support the batteries, to be bungied on top for now. Eventually they will hang underneath the platform, and it will be a shelf for carrying whatever load I need. I decided to invert the fork (mount it on the low side) to keep the mass as low as possible.

My first test run was today. I got the controller strapped on, batteries wired, and bolted the contraption onto the bike. I started with the 26" brushed geared Heinzman wheel, and took a couple of laps of the driveway. I found that over 10mph, the trailer would wag back and forth pretty severely. I could hear the trailer-bike connection rattling as the side to side motion translated to front to back motion on the connecting bolts. I added a couple of spacers and cinched it down tight on both sides. This removed the front to back play, but the joint would still turn to allow for pitch changes. There was a vast improvement in stability. However I could stand next to the bike and when I grabbed the seat and shook side to side, the wheel itself had play that contributes to the wobble.



I took it out on the road, and there was instant smilage! there is nothing like the feeling of a light pedal speeding you up a hill. The bike was steady at 15mph on flats, and would slow to 12 up hill, but I was pushing 750W if I did not pedal. The motor did not heat up to more than warm. There was more noise than I had hoped. It was like a big model airplane following me around.




I then swapped out the wheel for the 20" brushless.


This motor is dead silent. It does not have as much torque, but seems to be able to hit similar speeds. The silence is amazing. The smaller wheel seems to be more stable. Possibly since the weight is 3" closer to the ground, and the wheel is a lot stiffer side to side.

I'm considering whether to make the effort to make a trailer that is 6" shorter for the small wheel. I want something as compact as possible, but still stable.

[spinningmagnets]

I believe a shorter trailer assembly will be a bit 'twitchier'?, and the longer arrangement would be more stable. If you shorten the length from the bike to the pivot and bring the pivot closer to the rear tire, the power of the trailer will have less leverage to try and push you over sideways when in a sharper turn. I would put a cargo box inbetween the trailer tire and the pivot so the cargo weight would be as low as possible (instead of on top of the trailer tire).

I really like the 20" wheeled version, and easy to increase the power if thats whats wanted later...

[Alan B]

You could use a trapezoidal bearing and effectively project the point of rotation forward to increase stability. This is how the Hensley Arrow trailer hitch works. See their website for info on this.

[izpirkt]

You could use a trapezoidal bearing and effectively project the point of rotation forward to increase stability. This is how the Hensley Arrow trailer hitch works. See their website for info on this.

Great idea! I have thought of some linkage joints, but have never seen this trailer hitch. Went to the Hensley website, looked at some competing hitches, and saw a Lego demonstration of the joint. I decided to make my own to try to see what would happen with a single wheel. I broke out the kid's Kin-ex and here is the result:


It seems to track pretty well. Rotation of the green "bike" keeps the trailer wheel (white) aimed at a point well forward of the two orange pins on the "bike". This should be aimed at the contact patch of the rear tire, or slightly forward of that. I still am not sure about the push vs. pull and forward/backward dynamics, but I think I'll need a full size demo to get a feel for that.

I'll have to break out the SolidWorks and model up a joint with the right virtual pivot, that fits in the profile of the trailer, and is strong enough to take these forces. Then we'll see what Home Depot or McMaster has to contribute.

I went looking through the forum for "pivot" and "hitch", but did not come up with anything. Lots of the picts on older posts are not working, so I might have missed something.

Has anyone else tried this before?

[Rassy]

Your model looks interesting, and would probably work okay on a trailer being pulled. But it looks to me like a pusher trailer would force the trapezoid into an unbalanced configuration that would result in the power wheel trying to push to one side. What happens on the model when you hold the green "bike" back with a little pressure and try to move forward by rotating the trailer wheel? If you set the wheel at an angle to the "bike" does the trailer try to straighten out?

It also seems a little like overkill. I have pushed a bike with both one wheel and two wheel (both wheels powered) trailers and they never tried to high side the bike, even though there was some twitching on the one wheel trailers at higher speeds.

Anyway, I'm just curious. Thanks

[izpirkt]

It is amazingly stable when holding the green "bike" and pushing the wheel toward the "bike". The trapezoid does not fold up like I thought it would. The force from the wheel is always pushing directly toward what would be the contact patch of the rear wheel of the bike. I'll have to get either more complex with the model or set up a full size demo to test the dynamics with a two wheeled vehicle in front.

I realize that this does seem complex. But if it offers a compact trailer with no twitching/oscillation/ at high speed or high load, then it will be worth it. The further advantage would be the ability to remove one of the linkages and be able to fold the trailer up beside the bicycle for easier chaining at a bike rack.

Again I will ask: has anyone tried this type of hitch before?

[Alan B]

It works amazingly well on my 10,000 pound trailer. Makes towing as calm as not towing, can't feel the wind or semi's passing. But the hitch I have weighs more than 100 pounds so putting it on the bike doesn't seem very practical. If you scale it down it should be fine.

If it works as well on the bike as it does on the RV you almost won't be able to tell there is a trailer back there. In the name of science it certainly seems worth a try.

[bobc]

Loving the trapezoidal hitch idea - should be relatively easy straghtforward to engineer too.
Still hoping I'll be able to catch you up some time this week izpirkt ;^)

[izpirkt]

So, with the nice weather this weekend, I wanted to take the bike out for a nice trail ride. With the roots and rocks, the trailer would have be impractical, and I wanted some exercise anyway. The trailer came off by removing two bolts, and detaching the throttle and CA connection. These wires were bundled and secured to the bike frame with some velcro straps. So the bike was basically back to normal, with only a few extra ounces.



The ride was a lot of fun, and then this morning it took less than 2 minutes to put the trailer back on, and ride to work. This is the furthest I have taken the trailer from home, and it did great. It averaged about 20 Whr/mi, with some long hills and light pedaling. Just got home from the 2mi/70 degree commute, and no sweat or hard breathing! It would do 18mph on the flats, and still about 12mph on the hills.

Riding on sidewalks emphasized an issue that I'm not sure how to deal with. The trailer would bounce over the seams and irregularities and lose power. I have very low pressure in the tire, so it absorbs a lot of bumps, but when it regularly hits the sidewalk seams, it does start to hop. Adding a suspension could eliminate some of this, but It would need damping of some sort that can be tuned. It has to work with a loaded or unloaded trailer. A 100% potential increase in weight means the damping/ spring force needs a lot of variability.

In addition, it sounds terrible with the whine of the gears fluctuating so much on the bumps. The geared motor is probably not going to be the final solution for this trailer, since the brushless is so much quieter. The brushless may not really be more effective as a device, but will appear so without the noise.

[izpirkt]

Yesterday I swapped out the 26" geared hub motor on the trailer for the 20" brushless. Today I rode to work and amazingly the sway at high speed is nearly gone. I think it has something to do with the stiffness of the wheel itself. The spokes on the 20" are only about 4" long. I think that this reduces the side to side springiness that gets the larger wheel bouncing from side to side. I'll have to see what happens when I put my backpack back there and have more mass bouncing around.

I also enjoyed the silence. You can barely hear the motor, at high speed it sounds like a bee buzzing away back there, but the gear whine is gone.

I rode 11 miles today, using up the 4 batteries right as I turned into my driveway. I didn't really pedal all that much, and had some pretty big hills. At 17 Wh/mi, it is more efficient than the brushed motor, and is equivalent to 2058 mpg. My average speed was 13mph, and I peaked at 25.

[amberwolf]

The sway is probably partly from the longer spokes (especially if they're not fully tensioned), but is also probably from the different geometry of things putting more mass at a greater height with the larger wheel.

[izpirkt]

On a side note, has anyone been able to try out one of these? http://www.ridekick.com/