Matt Shumaker electric recumbent bicycle
An Electric assist recumbent bike project by Matt Shumaker
Page1       Page 2      Matt's Edrive Kit    e-Typhoon project   Velokraft VK3   e-Catrike 700

This project began as a simple exercise in efficient transportation and has evolved into a "Hot-Rod" E-Bike.
I have always wondered why so few recumbents are converted to E-bikes. There are many trikes that have been converted, but precious few recumbent bicycles. This seems strange to me considering the efficient layout of recumbents versus upright bikes. So, to that end, I decided to convert a recumbent.

I had two possible candidates sitting in my garage, a Burley Limbo high recumbent and a brand new Actionbent Midracer. After some consideration, the Midracer got the nod to be hit with some E-bike horsepower.

Now, as I researched E-bike power systems, I quickly became frustrated with the low efficiency, large size, high weight, low power, and poor quality of many E-bike components. There are a few good companies out there, to be sure. But, only a few and their products are CRAZY expensive! So, I began looking into another hobby arena for equipment. I am an avid RC helicopter and RC car buff (I manufacture my own complete helicopters in a small machine shop I have).
I started researching high power equipment that
could push my 190 pound rump around town. The choice of components was easy once I looked
in the RC industry! I chose a large AXI brushless outrunner motor. This motor is 93% efficient at 2000 watts and still over 80% efficient at 4000 watts! So, that was my motor of choice. Also, the AXI is only the size of a shortie pop can, much smaller than a
typical E-bike motor, far lighter, and MUCH high power!
Next up was the controller (ESC in RC lingo). That stands for Electronic Speed Control. In RC, the controller must be VERY small, light weight, and efficient. The controller I chose (Castle Creations HV110) weighs only 5.9 ounces and is the size of a couple match boxes stacked on top of each other. VERY small indeed! This controller is rated at 110 amps at 48 volts continuous. I have two of them in RC cars and have pushed them up to 127 amps at 48 volts without any heat buildup problems, shutdowns, nothing. They just run. Also, the controller is fully programmable via USB. Perfect!
Now, the problem with using RC equipment is the fact that these controllers are designed for use with an RC receiver. The receiver puts out a pulsed signal that RC servos and controllers recognize. This presents a problem. How do I simulate the pulse width of a receiver output to drive the controller? A little internet research netted a perfect solution, a "Servo Tester". This is a simple device that produces a variable pulse via a simple potentiometer. So, bingo, I can drive the controller without an RC receiver.

Next is another component referred to in RC terms as a BEC. This is a battery eliminator circuit designed to eliminate the need for a receiver battery in an RC vehicle. The RC controller I am using assumes there is a receiver and/or small pack to power it up. Without a receiver, I need a way to power up the controller's internal circuitry. So, I picked up (also from Castle Creations) a high voltage BEC that will do the job just fine.
Now, I have the brushless motor, controller, servo tester to drive the controller, and the BEC to power up the controller. Whew, kind of complicated, huh? But, in the end, it is all worth it. All these components are common RC components that communicate with each other very well. Now I have a very tiny, super light weight, efficient, and  reliable controller/motor combo that is good for 4000 watts continuous! Next I looked for a battery. I have decided to go with True RC lithium polymer cells. I will be running 12 cells in series and two parallel for a total of 48 volts 10 amp hours (probably about 40 volts under load). All this from a pack that weighs in at 70 ounces.

So, that is it for the electronics. Now, on to the mechanicals!

I went through quite a bit of research regarding the mechanical portion of the project as well. One item I put a lot of research into was the rear wheel/sprocket situation. An E-bike wheel can be driven in a number of ways;

#1 It can be driven with a friction drive against the rear tire. This was out of the question for efficiency and tire wear reasons.

#2 The motor can be linked to the existing chain using the stock cassette at the rear. The benefit of that design is multiple ratios available in the stock 9 speed cassette. However, I have over 4,000 watts available. So, one speed is fine. Also, that type of system has a multitude of complexities I did not want to deal with. One of which was the drag imposed if I pedal without the motor running. It also is hard to allow the cranks to stay stationary while running the motor if a right side drive is used.

#3 A second drive sprocket can be added to the left side of the rear wheel hub. Early on, I decided this would be the method I would use primarily because the stock right side chain and sprockets would not be affected by the E-bike additional components. Oh, I also wanted the E-bike components to be easy to remove in the future if needed (no welding or drilling allowed).

Since I had decided on a left side drive, I had to figure out how to get the power to the rear wheel. Hmm, the stock rear hub has nothing to mount to on the left side. A disc brake hub would be nice to bolt a sprocket to. But, that would require replacing the rear wheel. Not impossible, but not something I wanted to do. I was hoping to retain the stock rear wheel and save some cash besides. I also considered a CF rear wheel with disc brake hub. Cost was a factor there too. I also want to keep it stock looking if I can.

I started thinking about the hub and how it would be possible to mount the sprocket to it. This would not be easy because the bearing protrusion and the spoke flange are both at a strange angle, not easy to mount a sprocket directly to. So, I decided to use my mill to make a sprocket adaptor and bore the sprocket with the stock taper found on the rear hub. A trip to my shop (small machine shop) and, four hours later, I had the rear sprocket mounted. It was not an easy task! It required the use of a 5 axis mill to make it happen. The adaptor had to be bored with a tapered bore at a very strange 3.5 degree taper. The face of the adaptor also had to be machined with a 3.5 degree taper to match the spoke flange.
The sprocket is screwed to the adaptor with 6 screws, then the whole assembly is mounted to the rear hub with 6 more screws. So, all 12 screws are visible on the face of the sprocket. It looks beefy, indeed.
But that was the easy part. The tough part was making the adaptor match the shape of the spokes and spoke nipples that are on the face of that flange! I spent a lot of time taking the adaptor off the mill and checking clearance, then re-chucking it into the mill for more machining.
It all paid off in the end. The sprocket is mounted to the rear wheel and looks stock. It is straight and true and matches the rear wheel very well.
With that part of the project finished, I turned my attention to the "Power unit". This is a system comprised of the motor and a two stage reduction unit. Two stages are required to drop the estimated 10,000 RPM of the motor to 580 RPM at the rear wheel. This 580 RPM translates into 50 mph top speed! The motor drives a jackshaft by way of an XL pitch toothed belt to a 1/2 inch jackshaft that, in turn, spins a secondary jackshaft which is merely a BMX bike rear hub with a freewheel attached. This freewheel allows the power unit to sit idle while I pedal. That way I will have no parasitic drag if I pedal the bike as a normal bike without any electric assist.

The power unit is made of numerous 1/4 thick 6061 aluminum plates cut on my CNC. All parts are machined to be as light as possible, while handling the roughly 4,000 watts the system is capable of putting out. As a side note, I have a data acquisition system with real time monitoring and black box recording to keep tabs on everything from motor and controller heat, to motor RPM, speed over a given range (GPS), voltage, amperage, watts used, and watt hours.
It is also capable of reading other things such as 2 axis G-force and other pieces of data I have no interest in. It will be an invaluable tool for analyzing the performance and efficiency of the bike.
I have made a lot of progress on the E-Cumbent project. Much time was spent on the CNC and the manual mill as well. The pulleys for the bike's reduction assembly started as basic industrial pulleys. After 7 hours on the 5 axis mill, I had two large pulleys lightened, adapted to the drive system, and prettied up. The largest pulley started at 5.4 ounces and ended up 1.9 ounces.
You can see the primary reduction and secondary reduction. The primary reduction is running an XL pitch belt. XL pitch is good for high RPM and relatively decent torque rating (in case you are wondering, there will be idlers mounted for primary and secondary belt drives to achieve better belt wrap around the small pulleys).
The motor spins the first stage jackshaft at a 4.8 to 1 reduction. The jackshaft (1/2 inch diameter grade 5 titanium shaft) spins the secondary drive at 2.5 to 1 ratio. Finally, the front sprocket is 14 tooth driving a 25 tooth rear sprocket.
So, with 10,000 motor RPM, the bike should be good for about 50 mph. Plus the ratios for each reduction unit are easy to alter. I may find it better to gear the bike a lot lower. But, for now, I will try this.
 The secondary jackshaft is a BMX bike rear hub with a machined pulley mounted on one side and 14 tooth freewheel on the other side. The freewheel allows the power system to sit idle while pedalling. This way, I can run the motor alone, or pedal alone, or use both at the same time.
The secondary belt is a HTD 5mm pitch belt. This belt is not as good for high RPM, but can handle huge amounts of torque. Perfect for the high load taken by the secondary drive.   


Though I have pushed for some "Bling" factor, I may make black side panels to hide all the shiny drive components. With all the drive system hidden under the seat, the only thing that speaks of E-bike is the second chain on the left side. That is part of the reason I went with a black sprocket and chain. I want it to blend in.  
From the pictures, it may look like the seat is jacked up to allow room for the power unit. Actually, that is the stock height. The Midracer comes stock with a lot of room between the seat and frame rail. That is part of what started this project to begin with! Oh, you will notice the space under the seat just behind the power unit. That space will be occupied by the lithium packs I am ordering for it. This entire system will be very well hidden under the seat.
I am a little concerned about the torque load on the chain. I know the chain will be fine, but I am concerned about the torque taken by the secondary shaft (BMX hub) and its axle. The BMX hub I am using has a 3/8 inch aluminum axle with steel bolts threaded inside this axle.
That brings me to the chain itself; I went to the local bike shop to pick up a BMX chain and master link. No problem there. But, when I installed it, I noticed that it was perfect length, not requiring an idler to take up slop except for one thing---------- It was 1/2 link off! Frustrating! So, I did a web search and found an on-line source for BMX chain 1/2 links. Beautiful!  
I may go up to a 35 tooth rear sprocket and gear up the primary reduction unit to compensate. This will reduce the pulling load on the BMX axle at the secondary output. But, we shall see. This may work just fine. 
At this point, I have to machine a few internal standoff braces for inside the power unit walls as well as make the torque brace that will run from the jackshaft mount to the rear axle to reduce the torque on the frame and power unit. After that, I need to mount idlers for the belt drives, two idlers for the right side pedaling chain to get it out of the way of the motor, make a mount for the throttle potentiometer (servo tester), run wiring, and begin testing of the system. I think it will be about 2 weeks to get it running. 
Well, I made some more progress today. It may not look much different from the previous pictures, but I really have made a lot of headway.

One item that I was concerned about what the axle on the BMX hub (secondary jack-shaft). That axle is a 3/8 inch diameter aluminum item. On the chain side, there is over 1/2 inch of that aluminum shaft protruding from the inner support bearing running to the mounting frame. So, all the torque of the chain will be putting a bending load on that 3/8 shaft. Not good.
 So, I machined a shaft support for it.
You can see from the here that the shaft support screws to the side plate with 7 small (4-40) screws and recesses into the freewheel.
This allows the chain torque to be passed as a sheer load on the 3/8 shaft, rather than a bending load. Perfect! That should eliminate any chance of bending the shaft.  
 Next I made the idler for the secondary belt. This is a simple aluminum idler made on the lathe. It spins on two bearings and simply screws to the side plate.
I have to make a similar idler for the other belt as well.

What you cannot see are the various spacers and what-not I machined and installed today. Boy, it is amazing how much time it takes to make something like this when literally every part is either made from scratch or very heavily modified! I have 25 to 30 hours in the project so far.

The drive unit as you see it in the pictures weighs in at 4 pounds 6.8 ounces (70 ounces). Very light indeed for an E-bike power unit.

Next I will turn my attention to a few stiffening spacers within the power unit casing. Then it is on to the rear axle torque rod. After that, I will mount the throttle potentiometer and all that will be left is wiring it up and test running it!

I finished the power unit. First I completely disassembled the entire unit. Next I made an inside brace (center wall) and a couple standoffs. Then I drilled a number of holes in strategic locations to reduce weight. I also took care of a few detail bits like shimming and what-not. Kind of boring stuff, but necessary.
I also fabricated the rear axle mount for the torque rod. The rod will run from the axle mount to the power unit just behind the front sprocket (freewheel).
Lastly, I rough cut two Delrin idlers for better chain management (pedal side). I am hoping for a relatively straight shot on the power side of the chain. The idlers are for the slack/return side.
You can see the interior of the power unit frame. I did my best to make sure everything is 100% accurately aligned.
That is one of the fun parts of machining! You can see it is very stout and strong. Yet it is quite light. My goal is high power without any more weight than is needed.
It seems like most E-bikes are close to 100 pounds. I am pushing for 45 to 47 pounds. Although anything under 50 would be fine with me.  
Oh, my MaxAmps 10AH cells arrived today! They sent me some black heat shrink for me as well. It is looking like I do not need to buy anything else for the bike. It is all a matter of labor at this point! 

All in all, I am having a lot of fun with this. It has been a great deal of work, but I see light at the end of the tunnel.


I have had a number of emails from people wanting to know some other specifics about the bike. One of the primary questions I receive is about the power (expected power) output of the bike. My philosophy on this project is, I would rather have more power than I can use rather than less than I need. Honestly, I doubt the motor will run more than 30% of the time when I am around town. But, back roads are another story!


I fabricated the torque rod. As you can see, the torque rod runs from the front sprocket to the rear axle. This rod absorbs the chain pull. I felt this was needed to eliminate undue stress on the frame due to chain torque. I felt that the power I had available would put alot of stress on the frame and the seat mount that the motor is bolted to. This design should cure that.

The rod looks heavy, but it is very light! It is thin wall (about .050 inch wall) 6061 aluminum tubing, light but very strong. I wasn't too sure about the look of it. But, it is growing on me.

I have a few holes to drill and tap, then I am on to the motor pod cradle. Currently, the power unit is screwed to the frame at the front seat mount. However, the rear of the power unit has nothing to screw down to. I do not want to weld anything to this beautiful 7005 aluminum frame. So, I am making a cradle for the rear that will rest gently over the frame. This cradle will transfer the rotational torque produced by the motor to the frame while the torque rod absorbs the pulling load of the chain. Boy, it is difficult designing a system with 6 HP to be used on a frame designed for 1/2 HP!

Beyond that, I need to make a simple chain guide for the motor chain and deal with pedal chain management. The it is onto the electric wiring!

I have the lithium cells charged individually. Next I will do pack building and mounting. Fun, fun, fun!

I fabricated the torque rod. As you can see, the torque rod runs from the front sprocket to the rear axle. This rod absorbs the chain pull. I felt this was needed to eliminate undue stress on the frame due to chain torque. I felt that the power I had available would put a lot of stress on the frame and the seat mount that the motor is bolted to. This design should cure that.

The rod looks heavy, but it is very light! It is thin wall (about .050 inch wall) 6061 aluminum tubing, light but very strong. I wasn't too sure about the look of it. But, it is growing on me.
I have a few holes to drill and tap, then I am on to the motor pod cradle. Currently, the power unit is screwed to the frame at the front seat mount. However, the rear of the power unit has nothing to screw down to. I do not want to weld anything to this beautiful 7005 aluminum frame. So, I am making a cradle for the rear that will rest gently over the frame.
This cradle will transfer the rotational torque produced by the motor to the frame while the torque rod absorbs the pulling load of the chain. Boy, it is difficult designing a system with 6 HP to be used on a frame designed for 1/2 HP!

Beyond that, I need to make a simple chain guide for the motor chain and deal with pedal chain management. The it is onto the electric wiring!

I have the lithium cells charged individually. Next I will do pack building and mounting. Fun, fun, fun!

I spent more time at the shop today.

I fabricated and installed the power unit cradle. The cradle wraps around the frame at the rear of the power unit. It transfers the rotational torque of the motor to the frame. I drilled it to make it lighter. I think it looks cool.
I have been thinking a lot about the huge amount of torque on the power unit and the chain. I decided to replace the 25 tooth rear sprocket with a 36 tooth. That will reduce the load on the chain and the power unit substantially. I will go up in pulley size on the motor to compensate for the drop in ratio. The only drawback of this new layout is a higher belt speed.
The sprocket started as a very thick and heavy FSA BMX crank sprocket.  
I machined about 1/2 the weight off and bored the center hole to the 3.5 degree taper that the hub flange uses as well as drilled and countersunk all the mounting holes. 
I took this opportunity to show the sprocket adaptor I made earlier in the project. You can see the spoke profile machined into the face of the adaptor compared to the face of the hub. That part took me 4 hours to make!
This gearing change will drop the efficiency a touch, though I doubt it will be substantial. I also plan to keep the ratio a bit lower than my original design of 55mph top speed.
I will re-ratio the bike for a top speed of 45mph. That should be a better starting point. This should make up for any efficiency I lose with the new layout. Plus, I can always gear it up for higher top speed later. Increasing the ratio is super easy by changing the motor pulley.
Anyway, at this point, I am happy with the mechanicals. I need to make a spring loaded tensioner/guide for the drive chain. Next I will deal with the idlers for chain rerouting on the pedal side.

Lastly I will make a throttle mount for the servo tester/potentiometer. After that, I will assemble the pack and do all wiring.

Oh, I did some measuring. It looks like the entire pack will fit under the seat behind the power unit! This is very good news. It will really help keep the package neat and relatively stealth (if you could call this bike stealth). It has kind of developed that "Bling" look.
I spent some more time at the shop over the last few days.

I made all idlers (4 total) and mounts. I have one simple chain guide block to make and all chain management will be done. After that, I will make the throttle mount. I may have that done this weekend. Then I am on to the electrical components.
Each idler was machined from solid 1/2 inch Delrin plastic. I used all my machines to make these idlers. I think I have about an hour in each of them. All idlers spin on two bearings each.
On the tension side of the pedal chain, I can achieve a nearly straight shot from the crank chainrings to the rear sprockets if I squeeze the chain between the seat and the power unit.

All I need is a simple chain guide block to direct the chain through that tight area. Should be easy enough.

Electric drive with idlers.
Electric drive idlers.
I am enjoying making all the parts for this project. I am getting a little nervous about the success of this project, though. I always get a bit concerned right before I finish a project, though. I think it is just part of the fabrication/ R&D process.
I made the upper pedal side chain guide. This guide allows the tension side of the chain to make a nearly straight shot from the chain rings to the rear sprocket. I used the milling machine and a lot of handwork to make this simple part. I also made a few minor adjustments to various parts of the bike.  
Since I do not have the electricals done yet, I figured I would at least pedal the bike around and see if there is any discernable drag related to the second chain. Here is what I found;
The bike rides (pedaling) the same as before. I could not discern any difference with the power unit installed than without. This is a very good thing. That means this is still a perfectly good bike, not just an E-bike.
Second, there was a lot of chain noise at first. When I took the seat back off I found that I neede to make a few adjustments to the guide and the seat for better chain clearance. The sound difference is very good. It sounds pretty normal now. Though I do have some sound coming from the power side (left side) chain idlers. I think I made the idler wheels too small. The chain makes some noise as the links go over the idlers. Either that or I do not have enough wrap around each idler causing a link bumping sound. I will, first, lengthen the chain for more idler wrap to see if that quiets it down. If not, I will make new larger idler wheels. No big deal. All I have left to do is make a mount for the throttle and the mechanical work is finished!
I want to take a side track here and go a bit deeper into the machining aspect of this, or more specifically, my milling machines themselves. I have received a number of emails asking for a closer view of my milling machines. So, I cleaned them off a bit and took some pictures.
My manual mill is a Taig 3 axis mill with two Sherline rotary tables added along with a High Tech systems trunion table. These components make this mill a 5 axis mill. You can see by the pictures that the part can be rotated and the entire table can be tilted. This is the single most versatile piece of equipment I own. The CNC is nice, yes. But, this machine can do things the CNC cannot. However, it takes time to make parts on a manual mill. So, the CNC makes parts alot quicker and repeatable. I have a 4th axis rotary table for the CNC as well, but I have not installed it yet. I do not use any CAD software. I do all programming in G-code.
My CNC is a modified Taig CNC with my own tall fixture and a brushless outrunner motor and speed controller similar to what I am running on my bike. The CNC has about 2000 hours of run time on it and it just keeps running!
I bought these machines new about 4 years ago. I have $5,000 invested in them and they have earned me about $40,000 since I bought them.
Anyway, I have many other tools in my shop. I have a MIG welder, band saw, lathe, bench top belt sander, drill press, etc, etc, etc.

I am 100% self taught. Anyone can learn anything if they are determined enough.

Well, that is it for now. I will get back with more pictures soon. 
I machined the last few mechanical bits for the bike yesterday.

You can see the throttle here. The "Servo Tester" came with a simple plastic knob. That just wouldn't do! So I machined a new one on the 5 axis mill. It is prettier and the diameter is better for this application.  
I decided to go with a simple thumb knob rather than a spring returned throttle lever so I could set the throttle and leave it there for cruising. The controller even has a setting for governor mode (cruise control). I will mount a safety for it to prevent any issues having a throttle without a return spring.
I decided to make a simple fan for the motor. I do not think the motor will ever overheat. But, the efficiency is better if the motor stays cooler. So, I machined a fan blade for it (also made on the 5 axis manual mill). The blade took about 45 minutes to make.
Here's the fan mounted on the motor

I had one major disaster today. While working on the bike, I accidentally damaged two of the motor windings with a pair of pliers (don't ask). That was a $289 mistake! The motor can be repaired, but that will take months. So, I bit the bullet and ordered a new one from Tower Hobbies. I will have it in two days.

Live and learn, I guess.


I am doing all the wiring over the next few days. I expect to have it running in 4 days.

Oh, as a side story, I spoke with the guys at Castle Creations (the manufacturer of my controller) and they told me my motor is WAAYYY overkill. The head engineer told me his concern is the bike flipping over backward or some part of the driveline breaking. So, I am taking a few precautions;

#1 I am programming a "Soft Start" into the speed control. That adds a ramp up feature to the throttle so it does not bang the driveline so hard, but ramps up slowly instead. That should help.

#2 I am making the motor belt a touch loose so it will skip if too much torque hits the driveline.

#3 I picked up a stronger rear wheel. The current wheel is 24 spokes with a very small flanged hub. My new rear wheel is 32 spokes that are larger gauge. Plus the hub is super beefy with very large flanges. Lastly it has mounting for a disc brake on the left side. That will make sprocket mounting a breeze and will tolerate A LOT more torque.

I will run the original wheel as seen in the pics first to see what it can take. But, I will mount this new wheel if I have to.

I will post more pics is the build continues.

Oh, even after the build is done and the bike is running, I will continue the project. I have a data logger to mount, a simple console/dashboard to make out of carbon fiber, and 2 or 3 digital displays to mount in that console. I also want to make a few more bits and pieces for the bike, and I still haven't worked out any bugs in its operation yet.

So, there is much more to come!
Well, it is raining today. So, I am in the house building battery packs and doing other wiring. I am building two 6 cell lithium packs that will be wired in parallel for 48 volts total at 10 amp-hours.

Building lithium packs is an artform and I do not claim to be an artist in this area. However, I have built many lithium packs and am familiar with the ins and outs of this process as I have built many smaller lithium packs over the years.
These cells use copper tabs. This is fantastic because older cells use one aluminum tab on one pole. Soldering aluminum is an absolute nightmare without a special type of flux. I do have that flux, but it also stinks when heated. So, these cells are nice to solder with the copper tabs. 

First I tape the cells together. Then I roll the alternating tabs together for series wiring and solder them.

 After that, I solder the main wires and the balancing tabs. It looks a little busy, but I check everything with my meter before I heat shrink it all together.  
After soldering, I tape the ends of the pack with multiple layers of electrical tape for protection. Then I pull the heat shrink over and shrink it up.

Looks pretty good when finished.  

Word of caution, building lithium packs can be very dangerous!

It is also time consuming and easy to get confused and frustrated by. If you are not thoroughly familiar with soldering, lithium batteries in general, and have a good grasp of safety related to this process, do not attempt this on your own! It is better to buy a completely finished pack ready to go. I cannot stress this enough!
At this point I thought I would show a picture of the controller in my hand to give you an idea of how small it really is. I know the first thought many of you may have is "Nothing that small can work well pushing a bike." I know because I have received a number of comments such as this. However, Castle Creations is the industry leader worldwide in brushless motor controller technology. Typical E-bike controllers are essentially very old technology built in small quantities. Therefore they are large and inefficient. This controller is not only small, they run very cool and are USB programmable. There are many advanced parameters that can be adjusted this way.

Next I will build the second battery pack. After that I will begin wiring and programming the controller.

Oh, the motor disaster I had may have a silver lining; I know someone who rewinds motors as a hobby. He is familiar with rewinding AXI motors. We discussed it and he may rewind my motor for me with a far lower KV (RPM per volt). This way I can run 90 volts through it. Cool stuff! I also have a bid on e-Bay for a shorter AXI 5320 motor. That motor should run more efficient at lower current levels than my 5345. So, if I can get it cheap enough, that motor would be good to experiment with. These motors are very easy to pull out and replace from my bike. I love the experimentation process! 

It lives, it lives, it lives! Read on!

Continued on Page 2

If you have any questions, feel free to email me at


Back to the recumbent projects page