Sunday, November 17, 2019

Converting a regular bike to an ebike workshops

During the last couple of years I've helped several friends add an electric motor to their old bike.  While you do have to pay attention to the details, it's not that hard.  Except for the person who is a complete klutz with tools, almost anyone can do it as long as they are careful.  I have a friend across the river (in New Hampshire) who has also been offering the same help to people (we both are used to doing outreach on town energy committees), and back in 2018 he decided to hold a community workshop for people who were interested in converting their bikes.  About 3 dozen people attended this first workshop and we've done more coaching since.  At least half a dozen bikes have been electrified, and all of the bike owners learned more about working on their own bike. The first half of this post is about those workshops, and the second half is a 2 page handout on tips for converting your bike that I wrote for the workshops.

Both Bill and I had already built a few ebikes for ourselves, and his idea was simply to convert another bike while people watched.  If you decide to hold a workshop like this yourself, be warned that we converted the demonstration bikes ahead of time, and then converted them back to original condition before the workshop.  This is to prevent a show stopper, such as:
-finding out that parts are missing or won't fit (for example odd diameter handlebars, or the chainstays are weird so the motor won't go all the way in and needs shims)
-the battery and motor plugs don't match and have to be changed
-you need a special size socket to fit your bottom bracket
-something is wicked rusty and you spend an hour trying to get it apart
-after breaking the rusty part you need a new bolt and have to make a trip to the hardware store
Converting and unconverting ahead of time gives you a chance to fix any problems like this, and then have a good presentation.

Bill (in green) converting an old mountain bike
at the Cornish Town Hall in June, 2018

Bill chose an old mountain anthropobike to give a new life.  Many people have a bike like this in the attic, it could be nice around town but a little assist would make it much more useful.  Bill added a Tong Sheng TSDZ2 bottom bracket motor to it, which took an hour and 45 minutes.  It was an easy pace, taking time to explain the steps to everyone while he was working, and when he was concentrating on assembling a part of the bike I'd speak for a few minutes to fill in other details.  Afterwards we took the bike outside and enjoyed test rides.  We both also brought extra motors, batteries, and bikes to put on display around the room, to show participants the other choices available.

While the test rides were taking place, I stayed inside to answer questions about the displays, and then had a bit of fun doing an interview.  Cait May and Carlos Guarnizo, who run a Science Cafe, were up visiting family, and they filmed a short interview to introduce ebikes to their viewers:


Tour de sostenibilidad y energías renovables: Construyendo sus propias bicicletas eléctricas!
from the Ciencia Café Pa ́Sumercé channel

Both Bill and I have continued to help people with ebikes (I'm starting on the 9th and 10th projects this year), and we've worked on a few more events since then.  (For example we just held the 5th Upper Valley Electric Vehicle Expo and had 18 ebikes on display for it.)  But along with better bikes I've been spending a little time speaking up for better roads too.  One of the organizations I participate in is the Upper Valley Transportation Management Association, which is a group of businesses, transit providers, planners, town staff, and bicycle and pedestrian advocates.  We have traffic problems here in the Upper Valley, and although they aren't as bad as further south closer to Boston or Springfield (Massachusetts), problems do exist and we are trying to make non car choices available for people.

Lebanon (New Hampshire) is a small Upper Valley city that has done some very good bike/ped work on problems with their roads, and they have a fantastic Greenway and Rail Trail (which now extends all the way down to Concord) that are a delight to ride on.  Usually Vermont is ahead of New Hampshire on issues, but VT is behind on this one.  I'd like to see the Lebanon rail trail extended across the river and then run west at least to Woodstock, if not all the way to Killington.  (Unfortunately during the car centered mindlessness of the 1930's Vermont paved over much of the rail line to create Route 4 so a bike trail will require extra effort.)  Because of the work Lebanon has done, I was very happy to present another ebike conversion workshop this May 17, on National Bike to Work Day, at their Kilton library to help get a few more riders on their roads.
Converting a Hasa folding bike at the Kilton library, May 2019

I added a Bafang BBS02 bottom bracket motor to a Hasa F2 folding bike for this demonstration.  It took an hour and 30 minutes, while taking the time to explain the steps to the audience of about 30 people.  I'll add that I had to spend a couple of hours during the preparatory conversion adapting the full size rear rack to fit on a bike with 20 inch wheels, but otherwise it went smoothly, (for example the original chain guards even screwed onto the new chain ring with only a tiny amount of filing).  Bill brought one of his bikes and helped, and 2 other local ebikes plus some parts were on display too.  Afterwards I actually removed the BBS02 and installed a TSDZ2 without the thumb throttle (so that there is only the little keypad and the brake levers with automatic cutoff switches) to make a very simple ebike that is very easy to use because of the torque sensor built into the pedals.  Although I think the Tong Sheng has less power than other motors I've been using, this combination still rode up my steep hill very nicely, and I have to say that I think it is a sweet little bike set up this way.

I haven't done a workshop installing a hub motor yet because all the projects at the moment are using bottom bracket motors, but I'd expect it to be similar in time and format.  (I'll be comparing the different motors in the Minicargo bikes project that I wrote about in the last post, and they'll get an extended review when they are loaned out for testing.)  I'm finding that while it often costs close to $1000 in parts to convert a bike to an ebike using the specifications that I prefer, it's still much less expensive than most factory ebikes, and when done well will give you a very nice bike.

Several dozen people have attended these workshops, and then the information from them spilled over into the UV EV Expo for many more.  I wrote a 2 page handout for the workshops that has some basic conversion tips, which I'll copy here:

Converting a regular bike to electric drive

Why an eBike?
-Seniors, out of shape people, and hills
-carrying groceries, recycling, or other large loads
-a transportation system designed for cars with destinations spread miles apart
-you would like increased health by doing a light, regular exercise
-reduce air pollution, GHG emissions, and stop sending so much money to fossil fuel companies
-bypass traffic congestion and parking problems
-save money compared to how much your car is costing you
-eBikes have almost exactly the same health and emissions benefits as regular bikes, but are more usable in everyday tasks

What is legal?
-The short version is 750 watts maximum power, and less than 20 mph under motor power only.
Here are more details for the curious:
-The US Federal Consumer Products Safety Commission Guideline (2002) suggests a power limit for on road use of 750 watts (1 horsepower) with a top speed of 20 mph under motor power alone on a level paved road, with no wind and a 170 pound rider.  This applies to the first sale from manufacturer to consumer, and states may adopt their own rules.  (There are also many US Federal and European bicycle Standards, such as 16 CFR 1512, ISO 4210, EN 15194, DIN 79010, and JIS D9301 if you wish to get more serious, but the short version above is good enough for our everyday ebike projects.)
-NH and VT have based their rules on the CPSC guideline, however VT has chosen to raise the power limit to 1000 watts for better hill climbing.  NH recently adopted a system with 3 classes based on 750 watts, which came from California and is being promoted by industry.
-Over 20 mph or above 750/1000 watts (NH/VT) the bike becomes a moped and is subject to other regulations.  However it is legal to use a faster or more powerful bike off road on private property with permission.  It is also legal to pedal or coast on an ebike at higher speeds (that are within the speed limit).  The NH 3 class system does designate a faster eBike category that allows 28 mph, but there are restrictions on use (over 16, helmet, road use only unless local law allows use on paths or trails).
-Some states require that the motor turn off above 20 mph while others do not, and also some states require that the pedals be rotating to turn on the motor, (called Pedelec in Europe), while other states allow use of a hand throttle (which is very useful for starting on a hill in traffic when you have a load of groceries).
-In general eBikes have the same rights, access, and responsibilities as regular bikes, however there are some restrictions on their use on natural surface trails in certain wilderness areas (see state or New England mountain bike organizations for more information).
-Practically speaking, it is good to remember that standard bicycle tires, bearings, and brakes are not designed for motorcycle warp speed.

General guidelines for selecting components
-Use 48 volts, only use 36 volts if you have older equipment that must be repaired.  If you have 24 volt components it's time to move on.  Above 48 volts efficiency still improves but cost rapidly increases, and above 60 volts you start to run into electrical safety practice issues.
-A one piece Copenhagen/MIT wheel is very easy to install, however since all components are packed into one hub it is lighter duty and it is best suited for shorter, easier commuter runs.
-A 750 watt motor is noticeably better for climbing our steep hills than a 250 watt Europe legal motor.

Selection of the motor
There are two types of motors- Bottom Bracket and Hub.  (There is a type that has roller that rubs on the tire, but they are so inefficient that they aren't included here.)  Basically all motors are from the huge market of China, and they build all quality levels- there are both cheap generic motors and excellent motors.
-Bottom Bracket motor: can take advantage of the bicycle gears for climbing hills.  Almost all BB motors use only one front chain ring and have to be customized to have a second one, however with motor power added to pedaling this is OK.  Be sure to pick the right size chain ring (such as 52T for 20” wheels and 44T for 27”)  Some well known brands are Bafang (with 3 models and a proven design on it's 3rd version), and Tong Sheng (with a torque sensor and easy to ride).
-Hub motor: comes in two flavors- 1. Direct Drive (which is better for faster speed to cover miles-16 to 18 mph average speeds) and 2. Geared (better for climbing hills-12 to 15 mph average speeds).  Since both motors are linked one-to-one to the wheel, they can not be shifted down for hills or shifted up for flat stretches, and it is best to limit direct drives to moderate hills (under 10 to 12% grade), and limit geared hubs to covering distance at a lower speed.  Some popular brands are MAC/BMC, MXUS, Leaf, Crystalyte, Golden, 9 Continent, Bafang.

Selection of the battery
-Use lithium cells, which can last up to 5 years, do not use lead acid (which are heavy, corrosive, and you will kill within a year by draining them too far and sulfating the plates.)
-Break in lithium cells gently for the first few cycles while the solid electrolyte interface is forming inside the cells.  Be sure to charge the battery fully the first few times so that the cells are balanced.
-Lithium cells last longest when stored around half charge.
-Storage capacity is rated by Watt hours, not the Amp hours that most battery sellers advertise.  For example:
48 Volts times 20 Amp hours equals 960 Watt hours, or roughly one kWh, (kilo Watt hour, the unit on your electric bill).  A 1 kWh (1000 Wh) battery is pretty big for most riders, but I wouldn't recommend smaller than a 500 Wh battery for trips around our rural area.
-There are several styles of battery cases.  An on/off switch, a Battery Management System, and a meter on the case for checking the charge while in storage are preferred.  Large to small the case styles are:
1. Brick- largest, often shrink wrapped (with hard protective side boards inside), 200 to well over 1000 Wh   (You will need to provide a pack or container on your bike for this style case.) 
2. Silver fish- medium to large, aluminum case with plastic ends, often used on rear rack, 400 to 1000 Wh   (Mounted horizontally in a double decker rear rack, this is the best way to get 1 kWh.  However the vertical style case for mounting near a seat tube is heavy, clunky, and out moded.)
3. Dolphin- medium to large, mounts on downtube bottle mounts, molded plastic case, 300 to 900 Wh   (Lately I've been preferring this case style, as it is a molded plastic and corrosion resistant, as well as naturally insulating.  It is possible -although expensive- to pack in almost 1 kWh and 52 volts in the larger cases.  I look for 4 tab mount cases, not 2 tab.)
4. Bottle- small, mounts in bottle rack, looks like a fake water bottle, 150 to 400 Wh  (I don't recommend this size, unless you really just want to ride up and down your driveway.)

Preliminary size calculations for motor and battery
-While a 500 watt motor is ok for many rides around the Upper Valley, I recommend the maximum 750 watt size for better hill climbing ability, and use a system built for 48 volts.  750 W is a common size and has a pretty good price point.
-The battery size will depend mostly on trip length, with some consideration for what will fit on your frame, and cost.  I find that I use between 10 to 20 Wh per mile depending on the load, how much I feel like pedaling that day, headwinds, etc, and you can use 15 Wh per mile for a rough calculation.  Estimate your desired trip length in miles and multiply it by 15 Wh to get the bare minimum battery size, then round up between half again as large (150%), to twice as large (200%), for both a safety margin when you are out on a trip, and so that the battery will be less stressed and last longer.

Cost estimate
-Motor- A good quality motor in a kit that has all the little pieces that you need cost around $460 (Bafang BBS02B) to $550 (MAC geared hub) including shipping last winter.  There are more expensive heavy duty kits with higher power motors, and there are also less expensive kits on eBay with a weaker motor and battery for the person who only needs to go down the street for donuts.
-Battery- A 600 Wh dolphin battery with charger cost $350 last winter, and a 1 kWh silver fish style battery with charger and a rear rack to fit it cost $460.  Several years ago the quality of the cells were sometimes questionable, but now many battery packs offer ISO, CE, FCC and ROHS certification.
-For comparison, it is possible to buy a decent but not super quality completely assembled new electric bike in several frame styles from RAD bikes, delivered to your door for $1500.  It is also possible to buy a high performance eBike for $3000 to $6000 from many major bike manufacturers.

Sources for parts and kits
-Decent kits can be found on eBay, Amazon, and Aliexpress, but quality will vary by production lot.  I have also found several companies that do a good job- EM3EV has high quality products for a slightly higher price, and Leaf motor has good products, but pay attention when ordering and be careful with them on communication.  Bafang kits are the same from any vendor.  Tong Sheng retailers have been a real mixed bag so far.  Dillenger has OK lower price/lower watt kits.  Lunacycle has OK quality but too high prices.  Ebikes.ca (Justin at Griffin Technologies) has sometimes high and sometimes good prices, but they deserve a lot of support because they constantly experiment with new products and bring them to market, as well as supporting the Endless Sphere ebike forum.

Final tips
-Don't expect much regen, a bike loses most of it's energy to air drag which can't be gotten back. I get 2% to 11% regen on my longtail bike.  Also the direct drive hub motor is the only motor capable of regen.
-If using a bottom bracket motor be sure to shift down for hills.  Don't lug it or it will overheat and burn out.  If you shift so that you are pedaling at a regular bicycling cadence of 60 to 90 cycles per minute the motor will be happy.  Also, ease up on the power when shifting so that the chain and sprockets live longer.
-All eBike motors are most efficient when they are spinning faster.
-If using a hub motor I recommend only rear motors, as front forks are weaker and there are steering effects.  Also use a torque arm.
-Expect to buy a few special tools to fit your bike, but they won't be expensive.




Tuesday, October 1, 2019

The Minicargo Bikes

While I have several new ideas that I'd like to try out on a new bike for myself, I've been spending time on bike/ped/EV advocacy and basic projects instead.  (This past summer I helped five friends put electric motors on their bikes.)  Because I'd really like to build myself another bike, I've decided to combine some of the other projects into one minicargo (short tail) bike project to save time.  This project isn't done, but there's enough that has happened to write an introduction.

To start, here are the other projects:
-First, I've been wondering for a few years now which is a better motor: a direct drive hub, a geared hub, a pedelec bottom bracket, or a torque sensor bottom bracket.  No one has measured the performance, energy consumption, and ease of use on identical frames while running them in typical day to day utilitarian use over an extended period of time with several different riders.  I'd like to run one of each of the motors in a side by side comparison.
-I could also use a spare bike or two, for friends to use when they visit, or for loaning to people that I talk with who would like to try out an electric bike for a few days.  (For that matter, groups near me could really use some sturdy bikes for creating small, local bike share programs.)
-There are several people I meet with who think an electric bike could replace a car for people who don't need to travel far during the day.  We've been talking about DIY conversions as a way to make the cost of an ebike accessible to more people and could use actual data on the choices and cost
-I've also thought a lot about the "First mile / Last mile" commuter problem and what the best bike for this would look like.  A bike powerful and adaptable enough for carrying large loads and climbing hills, but that is still easily used by novice riders.  A bike that isn't too valuable, and doesn't take up much space at home.
-Finally, to inspire planners and transportation budget people.  I think that most planners and policy makers are not able to imagine bicycles being useful for anything more than one mile trips on a sunny summer day for a niche segment of the population.  The reality that I see instead (as I live without a car) is very shortsighted and poor transportation planning that is the main reason why bikes and walking aren't used more often.  After spending almost all of our highway budgets for the last 90 years on cars, there isn't any real choice when you go out - the roads are all about cars.  Bikes and pedestrians are given the nonexistent shoulder.  It's a self perpetuating cycle: "We have to spend all the budget on roads built for cars because people drive cars" but then "People drive cars because that's what they feel comfortable using because of the way the roads are built".  I honestly am not sure that many planners and legislators are able to figure out that you can't fix our transportation problems with the same thinking that created the problems.

So I hope to build at least 4 bikes with different drive systems, and see which one works the best.  Converting regular frames should result in a much better performing bike than the standard commercial bikeshare bike that has been built like a bulldozer,  but they will also be more susceptible to damage by less skilled riders so an important part of this project will be the system for loaning out the bikes.  Several people here in Vermont keep working on bikeshare and carshare programs with moderate success, so I've decided to leave that part alone for now.  Cost is also a very important part of this project, which means keeping the frame modifications simple.  I'm planning on welding on a simple heavy duty rear rack and extending the rear wheel to match it, but the rest of the bike has to stay untouched.

I've collected some frames over the last couple of years for this project.  They are all women's 17 inch (medium size) mountain bike frames, because I've found that a step through is important for many people, and also MTB frames have been better for my cargo bike purposes than other frames.

In my area rust is a problem, and 2 of the frames have splits in the bottom of the chain stays from water collecting inside them, rusting, and freezing.  The tubing has to be cut back to where it is normal thickness and replaced, and this will increase project cost.  (I'd prefer to find another frame or two and avoid the rust.)  The rest of the frame tubes are OK- although I did have to use some extreme finesse (all my weight on a 3 foot breaker bar) to remove 3 of the bottom bracket assemblies.

Simply adding an electric motor and a standard bolt on rear rack to a regular bike can result in a very nice bike, but the goal here is a little more multi use cargo ability (think of a sports utility vehicle).  It looks like a rear rack welded on is the best choice, because it would add extra capacity while making the frame stronger.  An extra feature would be shaping the back of the rack so that the bike could be stored standing up on it's rear wheel to save floor space.  The tail would have to be shorter than medium length though, because about 20% of the people who have ridden my longtail say that it is too big for them, (either it's too awkward to ride, or it won't fit in their storage spot).  There are several companies already building bikes similar to this, for example the RAD Runner, Yuba Boda Boda, Yuba Sweet Curry, Bike Friday Haul a Day, and Tern.  The goal here is to recycle old bikes into small electric cargo bikes with better specs at a comparable price.

The bike rack at one of my local Park and Rides.  Would you leave
an expensive eBike here while you took the bus to work everyday?
The problem is the surrounding roads have steep hills, and 
no one is going to ride here on a clunker bike that's hard to pedal.
(Photo credit Ariel Arwen)

The rear wheel should be extended backwards to better support the extra capacity rack or it will be harder to balance.  I've extended the chain and seat stays on my projects before and don't think it's too hard, but can it be done without adding hundreds of dollars of work time?  A jig instead of strings and squares would save time.  Also splicing the tubing mid length instead of scribing it to fit the bottom bracket or seat post, and using sleeves instead of flush welds on the tubing would help too (but be ugly).

The main question to answer was how far to extend the rear wheel.  There were 3 considerations:
-fitting in a battery pack behind the seat post (or somewhere nearby)
-size of the rear wheel
-how big is too big?

The easiest battery pack to install would be a silver fish (large aluminum case) style, placed over the rear tire and under the rack.  But this tends to be a bit top heavy in the tail and I'd like a novice level rider to feel comfortable.  A dolphin style (black plastic with angled top end) pack won't fit on the down tube of these step through frames, but it might be possible to squeeze one in behind the seat tube after modifying the frame, with the tubing for the seat stays being the main interference.  I'd like to use a dolphin style because it's plastic (corrosion resistant), commonly available (modular replacement), and 700 to 800 watt hours can be fit in without having to use premium cells (for extended running time).  It would be easier to fit a battery in this area the further the wheel is moved back.

I already know that my MXUS 750 watt direct drive hub motor in a 26" wheel will not provide acceptable power up the hills around here for someone who is out of shape.  However the same motor in a 20" wheel could possibly work.  Will a small wheel cause bad handling?  After getting in about 50 miles of test rides on bikes with 20" wheels (front and back), I've decided that a 20" front isn't a good idea on my roads, but that the back wheel has much less effect on the handling and would be worth trying.  This may turn out to be a big mistake because of the rear wheel getting caught in ruts and making the bike too squirrely, or it may turn out to be just a small problem because the tire wears out quickly, and only some riding time will tell.  I'll leave the 26" front wheel that the frames came with in place because that size is less sensitive to poor road quality.  The Yuba Sweet Curry has similar tire sizes, but with a larger rack than I'm thinking of adding.  On the plus side a 20" wheel would also give a lower rear rack height (easier to use and more stable cargo weight distribution), and add a few inches of space for the battery.  On the minus side the gearing won't be standard and I'll have to take into account motor speed.

The two hub motors arrived built up with rims, but I needed two more regular 20" wheels for the bottom bracket motors.  I wanted to use freehubs for strength and gearing reasons, a quick release (most 20" wheels have axle nuts), and 10 mm dropouts, so I got a couple Shimano FH-M525 hubs and laced up two wheels.
For the bike geeks reading this, I decided to push the builds and see how far I could take the tolerances so I got out the dial gauges.  (I know it was obsessive but it was the middle of winter.)  The best runouts that I could achieve were 0.08 mm in both directions, after that the rim surface finish and distorted bead roll around the seam messed things up.  Also I've read that lacing the leading spokes outside or the trailing spokes outside can make a difference, and since I was building two otherwise identical wheels I laced one each way to find out.  As far as I can tell it doesn't matter, but I plan on putting some torque on them and then measuring the width near the derailleur lower sprocket for another blog post.  Also the freehubs are meant for 10 speeds and I'll have to shim them down to 6 to 8 speeds (to be able to use heavier chain, however this might be useful for chain line adjustment), but the important part is that they are built for 135 mm wide MTB drop outs.

In order to find out how big is too big, I made a bike mock up and took it up to the bus bike rack in the front yard of my transit bus company for a test.
The Advanced Transit bus rack training rack

This criteria is directly related to the local poor quality bike parking, not cargo capacity.  As long as only a few bus riders use their bikes for the first mile/ last mile, they can just take their bikes on the bus with them to work (where hopefully there is better bike storage).  Then when more riders start using their bike to get to the bus, the state transportation agencies can put in more serious bike storage, and a bike this size would be more likely to fit in a future bin or rack than a larger cargo bike would.  Also, I hope to live to see commuter rail here (we have the tracks but not the service), and this size would fit in the storage area of the Budd RDC cars that are in inventory.

I took a couple of scrap kitchen shelf brackets and cut them into extensions for the drop outs drilled with many axle holes for testing.  With the largest tires that I had mounted on the rims and the wheels all the way down in the rack, I found that the axle could be moved back 3 3/8".  This is not a large amount and may not be worth the cost, but it also looks like (in combination with the 20" tire) it will give enough room for a battery pack behind the seat post so it's worth a try.


I drew up a sketch using nanoCAD to check for obvious conflicts, and it looks OK enough.  Although the seat stays will be a straight extension (sleeves would work), you can see the angle of the chainstays will have to be significantly changed to put the bottom bracket height at 10 3/4" (the lowest height that I've found OK when riding my other bikes), and it could be a challenge to keep this low cost as it's probably necessary to fit the tubing to the bottom bracket.  I'm not concerned about the small head angle change from slightly dropping the bottom bracket because the frames are all above 70 degrees and there is room within accepted practice for them to be more slack, and along with the increase in the wheel base and trail should help counter the quickness of the 20" rear wheel.  The next step is a full size drawing to double check dimensions before cutting metal.

The bikes have to meet the 750 watt On Road ebike legal limit, but because of the hills around here and rider expectations I have no interest in using smaller 250 or 350 watt motors.  Here are brief descriptions of the four motors to give you some background, you can look them up for more details.  I've listed them in order of what I think is strongest (i.e. will last the longest) to weakest, and it will be interesting to see what actually happens:

1. MAC geared hub motor with a 10T winding (25 mph with a 26" wheel) in a 20" rim, this is a deluxe motor and has been improved over many years to the point that it can take a lot more than 750 watts and survive.  This is my 3rd MAC and I've been very impressed with the first two, I expect this motor in a 20" rim will climb walls.  BMC motors are very similar, though I believe the gears are different material.  The motor controller is large and must be mounted separately in a water protected location.  Price around $550 in a kit with the rim attached, all small handlebar parts (but no display that shows speed), and includes shipping.  Weight of the motor alone is 9.4 lb, controller is 1.1 lb, small parts are 1.5 lb = total weight of 12.0 pounds for the kit.

2. Leaf Motor 1000 W direct drive hub with a 25 mph winding (with 26" wheel) in a 20" rim.  This is sold as a 1000 watt rated motor but it can be programmed to a legal 750 watts and the extra 250 watts of rating becomes a safety margin against overheating.  This motor and it's bigger 1500 W sibling can handle bursts of a couple thousand watts, and have been favorites of ebike hot rodders.  They are strong motors but very heavy compared to the other choices.  This is the only motor of the four that can regen when braking, (depending on the controller it can be set to be automatic at a certain speed), but it also is the only one with a noticeable slight drag when pedaling with the motor off (i.e. when riders have run the battery dead).  Again the motor controller is external and must be mounted in a water protected location.  Price around $425 with rim, all the little parts, and shipping.  Weight of motor alone is 13.3 lb, controller 1.3 lb, small parts 1.7 lb = total weight of 16.3 pounds for kit.

3. Bafang BBS02 with a 52 tooth chainring (for a 20" wheel).  A 3rd generation motor with improved gears and a better internal controller, probably over 8 million have been produced making it a world wide standard, It's very robust, though the MAC will most likely outlive it in high torque usage.  Being able to shift down for hills makes up for a motor that is smaller than the hub motors, but will riders power shift and break the derailleur?  Price around $425 with a nice color display and shipping.  Weight of motor with chain ring is 10.3 lb, crank arms 0.9 lb, controller is inside motor, small parts 1.4 lb = total weight of 12.6 pounds for kit.

4. Tong Sheng TSDZ2 with a 52 tooth chainring.  This is the smallest and lightest motor of the 4, and to me it feels more like a 500 watt motor than 750 watt.  The Bafang is noticeably stronger.  Being able to shift down for hills is a plus.  However this motor has a torque sensor inside the pedals instead of a cadence sensor (pedelec), and it is very smooth and easy to use.  I think this is a great motor for a novice, and the Bafang is a better choice for more experienced riders.  There have been improvements in the gears and heat resistance, but I've had a problem with the speed sensor gap adjustment being far too fussy on 2 bikes causing the motor to not run.  Price around $325 for last year's 3.6 model with a simple LCD display, the small parts, and shipping.  Weight of motor with chain ring is 9.0 lb, crank arms 0.9 lb, controller is inside motor, small parts 1.4 lb = total weight of 11.3 pounds for kit.

While I'd prefer to be trying out some of my ideas on a new bike for myself, putting these 4 bikes together will answer several basic questions that should have been answered a while ago.  I have the meters for tracking the performance of each bike, and the ergonomic outcomes will be interesting.  I've left out a bunch of details in this overview, (such as I think the chains for the bottom bracket motors could be narrower (9 or 10 speed) and still have a decent lifespan because they are running faster than normal because of the 20" wheels), so please feel free to send questions.  My plan is simply to keep working away at this as time and budget permits, so I can't guarantee when the next blog post about them will be, hopefully they will be on the road by next summer.