Kevin lives in a very hilly part of Georgia, and he just completed this conversion of a Suzuki DR650 from gas to electric. That by itself is really cool, but the thing that stands out about this build is that he put a lot of effort into being able to retain the stock 5-speed gearbox, which is amazing. So…let’s take a look!
I’ll start by posting a transcript here, of Kevin’s details from his youtube…
“What’s up Youtube. I’ve been building this bike for about two years. Started with an RC outrunner motor, inside the case of the stock bikes transmission…and, uh…I could never really get that working right with the ASI BAC8000 controller, so I decided to switch the motor to a QS 138/90H. It’s also inside the case.
Later in the video, I’m going to show you how I got all that to fit, with some pictures and stuff I took along the way. But, just spent all weekend getting this thing buttoned up. Like I said for the last two years, it’s [now] out of my garage. So, trying to program the BAC8000 [controller], I have a love/hate with, but…it’s working now so…I guess I’m loving it now. I’m actually finding the controller quite fun to program, now that I’ve made some headway with it and started to understand it a little better.
It runs two Kia Soul EV modules, which are 10S each (36V), so it’s a 20S pack (72V), so two of those modules in series. I haven’t wired in the BMS yet, so this charge cable will go, but it’s hanging on there so I can charge it a little bit. [You] might be wondering what this gauge is for. This is actually the oil pressure gauge, because there’s a lot of modifications to the oil system on this…because the whole top of the engine isn’t there anymore.
So, I wanted to make sure I was getting proper oil pressure. It’s also the cooling system for the clutch, the electric motor (the QS 138 is now oil-cooled), and it flows through the transmission as well, so…gotta make sure that oil’s flowing.
So you can see a little bit about how it works here. I re-plumbed the stock oil system, so these oil lines go back to an oil cooler that’s up in the back of the bike. [I’ll] probably add a fan to it…it’s picking up a lot of heat, so…put a fan on that later.
And anyway, so the oil circulates back through the oil-cooler to the front to this manifold block where it’s distributed out to things like the pressure gauge, to the transmission, and in the future, to a couple more oil-sprayers I’m going to install around the motor. There’s also a large bypass valve in-between the two batteries that you can’t see with the bike assembled, so I’ll show you a slide of that later on.
Just got the lighting rigged-up today. It’s got turn signals, headlight, tail-light…so that’s exciting. It makes it more enjoyable on the street, for sure. It’s got a brake light too. Yeah, that’s about it. I’m going to show you some clips of it ripping and probably put some pictures in of the parts I made.
Feel free to ask any questions in the comments below. Thank you.
[Kevin followed up his build-and-ride video with a “Frequently Asked Questions” Q&A session]
“How fast does this thing go?”
60-MPH (96 km/h)…in the future I’m going to be able to increase the motor RPM’s with field weakening up to around 5,000-RPM’s (from 4,000-RPM’s), and I’m also going to drop the rear sprocket down some teeth to probably around 35T (from 42T). This is going to require a custom billet adapter to the cush-drive, which I’m gonna make in a future video. But, you can see [in the graph], this really spreads out my gear ratios, and makes everything much more useful.
In the last week of riding, or so, I find myself in fifth-gear all the time…which will become the new third-gear, and that will give me an increased top-speed in the fourth and fifth gears, as well as the capability of getting out on the highway for an exit or two between stops. So, I hope that helps on the top-speed question.
I look forward to seeing what it will actually achieve when it’s geared properly.
“What is the range, how far will this bike go?”
It will go about 40 to 50 miles of range, based on my miles-per-kilowatt-hour that I’m recording right now.
“Why did I need a gearbox?”
I live in Northern Georgia, and it’s super hilly here. I’ve done a couple single-speed builds before, and one thing I’ve found with single-speeds is I’m always wanting a different gear ratio. In the hilly mountainous areas, I want something shorter, and on the road where it’s flat and open, I want something taller. And I just could never settle on a [single-speed] gear ratio.
So, I figured with a gearbox, I could quickly click into the gear that I want for the situation I’m in, and have a lot of fun with it. It also adds like…a ton of fun factor to riding the bike. I’m a manual transmission guy in general, I buy only manual transmission cars and vehicles.
And, It gives me that solid feel with a modern bike with the electric drive. So overall I like the fun factor of driving a manual on an electric motor, or a gas motor, box truck, sports car, whatever…as long as it’s a manual, I like it.
So, this is specifically cosmetic. So, a lot of people commented on the looks of the bike. Some people loved it, some people hated it. Either way, I think everyone was right in this case, there’s some things that look really good and, somethings that look yeah, really bad. So, I do plan on doing a whole cosmetic restoration of the bike, as well. I wanted to get the mechanical fundamentals set-up first. Next will be a resto on the cosmetics.
Scrambler style is what I’m going for. So, round headlight, cut the tail down a little bit (pic at 3:38), and then I’m going to make some carbon fiber trim pieces for the side panels, and the gas tank. I’ve done carbon fiber body stuff in the past. So…I’m looking forward to a new carbon-fiber project.
And, uh…we’ll do a video about how you make carbon-fiber parts, and maybe I’ll do a 3D-printed mould even, so…look for that in the future.
“What’s next for the series on the DR650?”
I’m planning on doing a few more videos on the DR650. I want to do the next video on the design and the build of the bike, so let’s go over the CAD-design (Computer Aided Design), programming, and CNC set-up, fabrication of the parts, that sort of thing. We’ll do that one next, probably.
Then I want to do a video on the total cost analysis, about costs to build a project like this, and how much of the money gets diverted into the machine itself, verses how much of it gets diverted into tooling that you can use for future projects. I think that’ll be fun.
And then I want to do a video on tuning the ASI [BAC8000] controller, which…is…actually really fun to get to know it a little bit. But in the beginning, it’s very daunting, and…ASI refuses to open up their support to the public, which…you know…is, I get it with this model, but…I built my entire career building sales and marketing programs for technology companies, and…I…can’t really understand what they are doing. So, um, if you want any advice, ASI…I’m happy to help.
And then I think I want to do a video on the restoration of the bike itself, but that would be in the future, because it’s going to be a longer process. Very detail oriented. Painting, powder-coating, etc, so…
That would be in the future, only enjoy the bike for the summer…probably do that in the winter. So, yeah, little bit in the future. And the last thing I want to talk about in this video is this channel and what we can do on this channel, so…
You might have already seen I sell programs. That’s what I’ve done professionally my entire career. I work in advanced manufacturing. Taking…taking…products like CAD software, or 3D-printers, analysis CT scanning machines, and software for CT scanning machines to market…and, um…I want to mix some of that flair into this channel, so…I know I want to build stuff, but I think I also want to cover the businesses that supply the machines and the components that go into these builds.
I’d like to know if you think that’s interesting, if you think that will make this channel unique. I’d love to hear your thought on anything else you’d like to see, I mean, I think it would be pretty reasonable to interview engineers who design machines and parts that go into these builds.
I think it’s also feasible to visit some vendors that supply, you know, manufacturing machines…or, the parts that go into the builds, so…I think all that’s pretty reasonable. I have a relatively broad network in that manufacturing space, and I think we can come up with some places to go visit. That might be interesting, so…Let me hear your thoughts on where you’d like to see the channel go, and I might be able to try to implement as much of that as I possibly can.
Again, thank you so much for watching, and see you next time.
Here is the 7-minute build and ride video…enjoy!
Our take on this
I am happy to see more builds using hybrid cells from a hybrid EV wreck. Pure EV’s have a huge battery, so they can easily provide huge amps, simply due to the physical size of the pack. this means the designers are free to spec cells that are optimized for long range.
A hybrid car is a little different. Since it has an on-board engine, it is not concerned with range (A hybrid Chevy Volt can typically drive only 80 miles in electric-only mode, while the Tesla Model-S can easily do over 300 miles). This means the expensive battery can be much smaller. HOWEVER!….this small battery pack must be able to propel the heavy car by itself on occasion, so the cells are the “high amp” cells we love…
Motorcycles need high amps to provide great performance without the cells getting hot, and they must do that with a small pack that can fit in a tight motorcycle frame.
Transmissions on an EV?
If you are designing an electric motorcycle from scratch, in order to sell them to the public…there are several good reasons to choose a large motor with no transmission, vs a smaller motor with some gears. The most successful electric motorcycle company is the Zero, by far. However…they use 28 cells in series (28S) which is roughly 104V nominal.
Kevin is using a more available and affordable 20S / 72V battery, so he is starting off with less power. On top of that, his user-profile is for a dual-use enduro. It’s main job is crawling up steep terrain of a widely varying grade, but once in a while…it would be useful for him to take a quick shot down a paved road at highway speeds. Accomplishing that with only 72V is a tall order.
If he had no transmission, the steep portion of the ride would suck down very high amps, which causes lots of heat in the motor, controller, and battery…and also shortens his range. Then, his top-speed on the streets would be compromised if he chose a sprocket-set that was lower-enough to run off-road well.
One of the benefits of an electric drivetrain is the incredible torque that is available down low at the first RPM. But…there are still limits. If you want to build an electric motorcycle conversion that is either “street only” or “off-road only”, there are existing configurations that will do that with no transmission, making the resulting motorcycle simpler to convert, and also leaving more frame-room for a larger battery pack.
However, if you want to have an enduro that can work well for a dual use (dirt and street), then I would humbly suggest that dozens of builds have proven time and again that having two widely-spaced gears is the absolute minimum to pull that off (and I believe that three gears is close to ideal).
Kevin didn’t “choose” 5-speeds. He already owned the Suzuki for a long time and then the engine died. He decided to convert it, and it just came with the 5-speed. In fact, he specifically mentioned that he spends most of his time in 5th gear. It was just lucky that this large 650cc single cylinder had an enormous crankshaft and counter-weight, and as a result he was able to remove them and fit the QS 138/90H inside the case (after shaving all the parts!).
Here is our recent article on the Denzel “motor and 4-speed gearbox” (click here). And here’s our article on bicycle multi-speed IGH transmissions (click here). Once you understand how the vintage 3-speed internally-geared Hubs work, you can see that there is no reason they couldn’t be scaled-up to motorcycle size. I know I will sound nuts for saying this, but the Ford Model-T from the 1910’s had a “planetary gear” transmission that was a very compact and robust unit. It had two forward speeds, a neutral, and one speed in reverse (sometimes listed as a 3-speed). Something like that would be a great option for an electric enduro…
For those who are curious about the QS cylindrical motors, here is our article on them (click here).
Written by Ron/spinningmagnets, July 2021
A direct drive hubmotor is affordable, simple, and robust. However, they turn at the same low speed as the drive-wheel. Electric motors perform best at higher RPM’s and that is why the most common hubmotors are geared, usually allowing the integrated motor to spin 5 times faster than the wheel. A common 26-inch bicycle wheel is only spinning 336-RPM’s at 26-mph (42 km/h).
Once you mount a hubmotor into the frame and drive the rear wheel with a chain, a whole new world of performance opens up. First, you can easily gear he motor with sprockets to run 2:1 or 3:1 to the rear wheel, with the motor spinning faster than the wheel. Even when only using my preferred 52V, a “high kV” motor (low turn-count) can spin at the new RPM’s to give you the same wheel-speed with higher torque and lower waste-heat. If you don’t have a preference already, I would suggest experiments start with the “Leafbike” 1500W DD hubmotor, because it has the thinner 0.35mm laminations in it’s stator, which results in less eddy-current waste-heat…
The second way in which using a DD hubmotor as a non-hub is a major benefit, is that it removes the weight of a large DD hubmotor from the unsprung wheel-weight, and re-mounts it into the frame. This allows the expensive rear suspension of a sophisticated off-road ebike to have a very light and responsive feel to tackling a rough downhill run. A large and powerful rear hubmotor has remained a useful option for street hot rods, but it will always be a compromise for difficult downhill runs.
#1. Custom build from Adam, in France
This builder is well-known to us (ES member bzhwindtalker), and he has started a company selling the LMX-64 (Click here). “LMX” means Light Moto Cross.
#2 Rassy’s 6×10 TerraTrike, from Oregon
When I first saw this build back in 2010, It blew me away. The 6×10 name means that it is a “six turn” winding, using ten strands of copper wire per turn. That’s a reasonably fast winding since he gears it down between the motor and drive-wheel, plus he is driving a smallish 20-inch rear wheel. This results in the trike having a ton of wheel-torque and the motor never seems to get hot, even on steep hills.
“…I easily maintained 8 MPH up the 15% grade…”
The innovative thing about this build is that he runs the motor in reverse (easily done by swapping two of the three motor phase wires), and he drives the rear wheel off of the bottom of the freewheel that is mounted on the hubmotor. Look closely at the pic below. As the rear wheel is spinning in the forward direction, the hubmotor is spinning in reverse (counter-clockwise, as seem from the left side).
This allowed him to mount a conventional freewheel in the standard position on the threaded hubmotor freewheel mount. In the pic below we can see that he is using a 16T heavy-duty White-Industries ENO freewheel. Of course he had to fabricate some custom bracketry to achieve this, but his published pics show a fairly easy way to accomplish that.
There are a variety of ways to achieve the same results, but this arrangement avoids the need for a separate jackshaft between the motor/pedals, and the rear wheel.
For more details, the build-log can be found by clicking here.
#3 Tench’s Specialized Supercharged Racer
We wrote about this custom build nine years ago in 2012! (for that article, click here). It remains a high water-mark that other builds can aspire to. The builder is endless-sphere forum member “Tench”, and the “Supercharged” name is in response to the weak Specialized “Turbo Levo”. At the time, the Crystalyte Direct Drive (DD) hubmotor with a 35mm wide stator was the most common upgrade to more common “1,000W” hubmotors with a 28mm wide stator.
When mounted in the drop-outs, the Crystalyte was known for breaking the axle from builders who raised the watts of power by using extra volts and amps, and sometimes overheating. Simon is from the UK, and his build used 66V and 40A equalling 2600W, but since it used a high-speed winding and geared it down to a lower RPM, the wheel-torque was off-the-charts.
#4 Mystery Custom Build
#5 Paul Brodie’s eBee from 2013 NAHBS
You can find more details about this build by checking out our article by (clicking here). This build is as custom as it can get. Paul used a CNC to make unique parts that he had drawn up on some paper just a few moments before he started the machine. He didn’t use the largest motor, and he built the battery pack to be as small as possible, then mounted the pack as low as he could get it, seen below as the dark block just in front of the pedal-sprocket.
#6 The M55 Terminus
This ebike is no longer available, but for a short time in 2012 it was sold to the public as a VERY expensive option.
#7 The “Super V” from John in Costa Rica
Endless-sphere forum member “John in CR” has been a very innovative builder from a long time ago, with this build being posted in 2012.
“…The 28T at the motor 48T at the wheel reduction makes the 9C [motor] effectively running a 12″ wheel so it will be able to climb anything…(ratio 1.7:1)”
#8 Fume’s 9C custom build
There are crank-sets that have a sprocket on both sides of the “bottom bracket”, and they are for the “stoker” back-seat position on a tandem 2-seat bicycle. Endless-sphere forum member “Fume” used one of these to allow him to mount a DD hubmotor with the sprocket on the left side. Doing this allowed him to use a freewheel in the usual position.
When doing this, the motor can drive the system through the freewheel, but when unpowered…you can pedal without driving the motor…which could have a small amount of drag from the rim-magnets being attracted to the steel in the stator teeth, which is called “cogging”.
Back in 2013, the “9 continents” / 9C motor was a common DD hubmotor to use. For more details, the build discussion can be found by clicking here.
#9 Hannes’ custom mid-drive
This custom build is one of the most professional I have ever seen. The builder is from Austria, and he decided to use a geared hubmotor as a mid-drive. You can find more details in our article by clicking here.
#10 the Appel Cruiser
ES member “APL” built this hub-as-a-non-hub cruiser in 2018.
#11 The V3 Illegalbike, from Russia
The details of this custom build from Russia can be found by clicking here.
Here is a video of this custom build
#12 The APL V4 Dreadnaught
As the name suggests, this is “Version 4” of a street ebike cruiser made by ES member “APL”. A previous version by him can be seem listed above. He named this build the “Dreadnaught” after a battleship from many decades ago, since the high-amp and long-range battery pack weighs 35-lbs! The word “Dreadnaught” means “I fear nothing”…
Written by Ron/spinningmagnets, July 2021
Agnius is from Lithuania, and he is best known for developing the N.E.S.E. battery pack building system. He recently build a longtail cargobike frame from scratch, and it had a few features that caught my eye. Lets take a look.
N.E.S.E Battery Modules
We wrote about the N.E.S.E. battery system in 2017, and that article can be found if you click here. The idea is that anyone from around the world could use a 3D-printer (or hire a local 3D service) to make the plastic housings in any size or color that you want. There is no soldering or welding required, the cells are held onto the buses by compression.
The parts that N.E.S.E supplies are the nickel-plated copper bus-strips, the threaded battery-posts, and the closed-cell rubber compression-foam backing that is made from Poron. The black Poron foam acts as a spring to maintain compression and to take up any heat-expansion in the cells. If the foam was “open cell”, it would be like a kitchen sponge that water could pass through, which would lose its compression-return ability rapidly. Poron is a closed-cell material, so each one of the millions of tiny air bubbles that are trapped in it are sealed away from each other, and this type of foam will never lose it springy-ness. Link provided at the bottom of this article.
The Longtail Cargobike
No matter what material you want to use, a long cargobike will be heavy, especially if you are going to add an electric drive system to it. So, most builders usually concede that its worthwhile to use steel frames, for the ease of welding together the parts from several donor-frames (in the pic below, Agnius also used steel tubes from bedframes). Once that decision is made, steel frames can often be sourced for free by simply keeping an eye out for several months, to grab the old bike frames that are sitting on the curb for trash-day. And that is just what Agnius did.
If you read through the build thread (see link at the end), Agnius tried a 26-inch wheel on the front at first, with a 20-inch wheel in the back. Almost immediately, he changed to using 20-inch wheels on both the front and back. Doing that made the bike slightly lighter and shorter. In fact, if you do not intend to ride on roads with big potholes, the smaller 20-inch wheels are stronger than the larger 26/27.5/29’r wheels.
Agnius wanted a large multi-use tub behind the seat, similar to the Madsen cargobike (click here). The pic below shows the pinching torque-arms for the direct-drive hubmotor axle in to the rear drop-outs. As the powered rear wheel tries to turn forward, the axle is trying to turn backwards. The axle has as much force as the wheel, so 1500W peak is trying to break or spread-out the frame drop-outs. These will be welded over the drop-outs to add enough strength.
As far as which power system to use, my “go to” suggestion for a cargobike is a BBSHD mid-drive using 52V and a small chainring (1500W stock, and 2500W with the Luna Ludicrous upgrade). Alternatives come and go, but that system continues to impress me, year after year. However, part of the appeal of Agnius’ build for me is that he is using a rarely-used configuration that I like a lot. Its a 1500W direct drive hubmotor in a small-diameter wheel.
What’s so great about that? Give me a minute to explain. The most common wheels (in inches) are the 26/27.5/29’r diameter. The larger-diameter wheels can roll over unexpected potholes with less of a jolt, so they can smooth-out a rough ride. However…the larger the wheel-diameter, the slower a hubmotor will spin when traveling at average speeds. Low-RPM’s in a hubmotor which happens to be drawing high amps (from a heavy load) will cause the hubmotor to overheat.
If you try to use a geared hubmotor, it allows the internal motor to spin five times per hub rotation, so its efficiency is better than a direct-drive hubmotor. Geared hubmotors are the most popular drive for “general purpose” ebikes, since they cost less than most mid-drives, and are more efficient than direct-drive hubmotors. However…unfortunately for a heavy cargobike, a geared hubmotor has a poor heat-shedding path from the central hot stator to the outer aluminum shell…
And this brings us back to direct-drive (DD) hubmotors. If they are under-sized for the load, they can also overheat, but now…we have a much better selection of hubmotor sizes to choose from (click here for our article on hot-rod hubmotors). With a direct-drive, the motor spins at exactly the same RPM’s as the wheel, and a common 26-inch wheel will be spinning at only 260-RPM’s at 20-MPH. However…if you swap that same DD hubmotor into a smaller 20-inch wheel, the RPM’s go up to 336-RPM’s at 20-MPH (76 more RPM’s), which is a 30% improvement in wheel-torque and heat-reduction.
Plus, the most common and affordable DD hubmotors use a 28mm wide lamination stack (forming the core of the stator), and they are typically advertised as a 1000W hubmotor. For a “1500W” labeled DD hubmotor, most Chinese manufacturers use a 35mm wide stator. The added width of these motors will provide more torque, so their heat is reduced because it doesn’t “bog down” as much. The 1500W rating is for continuous power, and these have taken 2600W as a temporary peak (52V x 50A). Swapping your 1500W motor from a common 26-inch wheel into a smaller 20-inch will improve the hill-climbing torque and cooling by a full 30%…
The large red block on the top right is the battery pack 20S / 6P (72V). The red box on the bottom left is the controller. The small clear box on the bottom right is the main fuse, and the small gray box at the top left is the 72V-to-12V DC/DC converter.
Big DD Hubmotors in Small Wheels
The most common and affordable DD hubmotors use a 28mm wide lamination stack (forming the core of the stator), and they are typically advertised as a 1000W hubmotor. Most Chinese hubmotor manufacturers use a 35mm wide stator for their “1500W” DD-hubmotors. The added width of these motors will provide more torque, so heat is reduced because it doesn’t “bog down” as much. The 1500W rating is for continuous power, and these have taken 2600W as a temporary peak (52V x 50A). Swapping a 1500W motor from a common 26-inch wheel into a smaller 20-inch improves the hill-climbing torque and cooling a full 30%…
If a 35mm-wide stator in a 1500W DD hubmotor interests you, but you don’t already have a preference…the Leafbike brand (click here) seems to be popular. It uses the thinner 0.35mm laminations, instead of the common 0.50mm lams. That means that it will have less eddy-current “waste heat”, which is where battery watts are converted to heat, instead of work.
Back in 2015, we wrote about using a moped rim on a DD hubmotor (click here), and that is what I recommend for projects similar to this. The reason is because the moped rims use angled nipple holes, and a regular bicycle rim will force the spokes to bend sharply, right at the nipple. If you use a “one cross” spoke pattern (seen in the pic above), instead of a radial pattern (please don’t use a radial pattern, I’m asking as a friend), you will have fewer spoke breakages. Skim the article link I provided to see all the details why.
Moped/motorcycle rims and tires are measured differently than bicycle rims and tires. A 16-inch moped rim is 16-inches at the rim OD, which is the way they should be measured. A 20-inch bicycle tire is roughly 20-inches in diameter at the OD of the tire, but only when using the common average tire from 40 years ago (20-inch fat tires now exist). The majority of the so-called “20-inch” bicycle tires will seat properly on the rim of a 16-inch moped wheel (and vice versa). A 16-inch moped rim is roughly the same price as a heavy-duty BMX 20-inch rim, but the moped tires are surprisingly affordable, along with being very puncture resistant to avoid flats, and also very long lasting (please read the article in the link above!)
I really like the idea of having a 20-inch wheel on the back of a cargobike, since it puts the weight of cargo three inches lower than a common 26-inch rear wheel. If you live where you’d be forced to roll over any unexpected potholes in the street, I’d recommend adding a suspension seatpost (click here), like a Cane Creek Thudbuster or a Suntour NCX. I’d also add a front suspension fork plus a 26-inch wheel on the front if your roads are bad, but…for well-maintained streets that are smooth, a solid fork is simpler and lighter, along with a 20-inch front wheel being slightly shorter and lighter.
Either way, for any serious street commuter hauling heavy cargo, I’d definitely recommend a large hydraulic disc brake on the front (click here to read why). Using a common rim V-brake can be adequate on the back (with upgraded pads like Kool-Stop), but…if there is a mount for a disc brake caliper, I’d prefer using that, even if it is just for an average diameter disc, with a cable actuated caliper, like the popular Avid BB7’s (common brake disc diameters are 160, 180, and 200mm).
The Final Result
As you can see, Agnius is a capable guy who doing some creative and fun stuff. Thanks for reading this far, and go out and have some fun with ebikes…you wont regret it.
His longtail cargobike build discussion can be found here.
His website address is “18650.lt” (the suffix is “dot L-T”, for Lithuania), and you can check it out by clicking here.
Written by Ron/spinningmagnets, June 2021
Back in March of 2017, we wrote about Denzel, and they were retailing electric bikes that are made by the Eastgem company in China. You might want to take a quick scan of that article before you move on to these new products (click here). Denzel is an unusual company in that they are fairly large, big enough that they have survived several years of ups and downs, but they do not have a large presence in North America.
This is the first time I have taken a look at them since the 2017 article, and the reason is their new “GB200” motor/gearbox/and clutch combination, shown in the header pic above. They do now have a US distributor, along with distributors in Australia, China, and Russia (Click here). I’ve contacted their Connecticut USA office to get more information, and I will report back when they reply.
Why a Transmission?
To lower the amps you need for a given speed.
This seems to be a perennial question that never goes away. When the Tesla EV car company was designing their first roadster prototype, it started it’s development as a single motor with a conventional 2-speed transmission. They’ve never openly stated why, but halfway through the process they changed the design to a larger motor with no transmission because a breakthrough in high-efficiency IGBT’s meant that their controllers could now flow more amps without overheating.
An IGBT is an “Insulated-Gate Bipolar Transistor”. It performs the same function as the FET/Field Effect Transistors in electric bike controllers. Transistors were named for “transfer” and “resistor”. It is a variable resistor, and as a variable electro-magnetic field is applied to the bridge material (like a capacitor), that substance becomes more conductive, allowing current to pass. FET’s and IGBT’s can be used in a variety of ways in electronic circuits, and here they are used as an on/off switch to send current to the coils in the motor. FET’s are preferred in smaller devices and lower voltages and currents (along with very high switching cycles). There is some overlaps where one would be preferred over the other, but IGBT’s are for higher currents, higher voltages, and lower switching frequencies (click here for more babble like this). Be aware, I don’t really understand this, and I probably explained some of it wrong.
I am fortunate to have been able to discuss EV systems with Luke Workman (called “Live For Physics” on the web), and when he was a battery engineer for Zero electric motorcycles, he persuaded me that leaning towards lower volts and higher amps to arrive at a given power level is a viable option with distinct benefits. As an example, the current Zero motorcycles use 104V nominal (28S), and over 400A peaks. The Tesla Roadster uses about 375V and puts out over 900A, so they use a similar V/A ratio.
The “now defunct” Alta motorcycle company went the other direction. They used a ratio of high-volts to low-amps, with 350V nominal. “In theory”, you can get the same power either way, and as long as the “volts-times-amps” equals the same number of kilowatts, the results should be similar. Alta felt that there was a small benefit to their range and charging-speed due to the limited selection of available cells at the time, and also their lack of a battery cooling system. Tesla has a battery pack cooling system, and Zero uses the newest high-amp cells that don’t generate much heat in the first place.
I mention all these things because amps and “amp heat” are the main reasons mentioned in the past as to why someone would use a transmission on an electric motor. One example of using a transmission is Jack Knopf’s 1964 Chevelle (click here). Believe it or not, the EV dragrace governing body (NEDRA) has an EV class that is limited to using only 72V (among other classes). Back in 2006, Jack noticed that the times were slower than he expected, and wondered if he could capture a trophy. He found an old Chevelle, and stripped as much weight off of it as he could. It already came with a Powerglide 2-speed automatic transmission, which was part of the reason he chose this particular car.
When using 72V and roughly 4,000-Amps (equaling 192-kW) his best 1/8th mile time is 13 seconds to reach 50-MPH. That may not sound very impressive, but the Chevelle weighed 3,600-lbs, and he was limited to 72V! He credited the transmission with a major portion of the improved performance in this limited class. For years I have pondered a 2-speed transmission to provide an off-road/street range for an electric dual-purpose electric motorcycle, and I still think it may have some merit by using a compact planetary gear-set, like the transmission for the Ford Model-T.
The Denzel Rush
I think that…rather than develop the drive unit and then build an electric dirt bike around it, Denzel took an existing 200cc Chinese dirt bike and custom-modified the left side of the engine-case so that it held a 3000W motor. Having a mould made to die-cast half an engine-case can be expensive, but it’s actually not too difficult. Part of the reason I believe this is because the transmission has 4-speeds, which is common on a gasoline dirt bike like this, and that is a lot more gears than you would ever really “need” if it was an electric motorcycle designed from scratch.
Some electric “trials” riders have stated that they would actually prefer having a standard clutch, but none of them wanted any gears, opting instead for a speed-reduction on the motor, along with only one speed to the rear wheel.
Here is an article we wrote about Rich Benoits EV hot rod (click here), and he used a manual 3-speed transmission. The trans was dirt cheap from the salvage yards, but as he expected, he ended up only using two of the three gears. First gear could provide smoky burnouts, but…for the fastest acceleration, he would start in 2nd gear rather than lose time by needing to shift twice instead of just once.
In fact, the 2020 AWD electric Porsche Taycan uses a 2-speed automatic transmission on the rear axle. So…at least somebody thinks there are still situations where the expense, bulk, and complexity are worthwhile.
Alex is the owner and chief designer for Denzel ebikes. Back in 2017 the pleasure of having an email-chat with him, in order to find out more information. He grew up in the beautiful region around the city of Kislovodsk, near the southern border of Russia, between the Black sea and the Caspian sea, and on the north slope of the Caucasus mountains. Once he got into designing and selling electric bikes, he moved to where the action was, in China.
So, Alex is expanding his company into motorcycles, and lets take a look at this integrated 4-speed module.
The pic above is the right side of the drive case assembly, and it appears to be the same as you would expect from any mass-produced gasoline dirt-bike, and that further suggests that only the left side of the engine case had to have an optional electric version to produce the Denzel Rush E-dirtbike.
In the pic above, this is the left side of the engine case, with the whole assembly flipped upside down. The motor drives an oil-bath reduction gear and the wet-clutch first, then the clutch drives the 4-speed gear-box and output sprocket.
In the pic above, we can see that the motor is an inrunner. Eight magnets on the rotor, and 12 stator-teeth. The magnets are “surface mount” instead of being inset slightly into the lamination stack, which is what the Zero motorcycles have done when they moved to “Interior Permanent Magnet / IPM” a few years ago. Moving to a more modern IPM rotor allows the magnets to run cooler, by moving them a bit farther away from the magnetic eddy-currents around the air-gap. This in turn allows the motor to use higher peak amps when accelerating, without fear that the temp sensor would not limit the amps right when you want them the most.
This suggests that the stator and rotor are off-the-shelf mass-produced units, rather than using a brand new design. Doing this saves a LOT of development and production costs, allowing the product price to remain as low as possible. I can imagine that this also helps Denzel expand into motorcycles with much less risk.
Back in 2011, Brammo tried the same thing (click here) when they took a existing gasoline engine case and adapted an electric motor onto it with the stock clutch and 6-speed transmission. The designer felt that there were many motorcyclists that would feel more comfortable transitioning to an electric version if it had a familiar clutch and transmission “experience”. Brammo closed down due to low sales, but the transmssion did allow Brammo to use a smaller motor than the large Motenergy 75-7…which the successful Zero company is using with it’s one-speed drivetrain.
This is just my opinion, but…if the price is cheap enough, the Denzel Rush dirt bike will sell anywhere that other competitors can’t get product shipped to that location. That being said, Denzel is also selling the “engine case” as a separate unit, which has been a common theme in Asia for converting gasoline vehicles to electric.
The one place where I would actually like this combination of a motor and 4-speed trans is…for a conversion of a mid-sized gasoline quad/ATV. There are millions of old ones in the USA from Honda/Yamaha/Kawasaki/etc. I live in Kansas, and as much as some teens use quads as a fun toy, the ranchers here use them as serious farm equipment. There are two ways in which an EV-quad can be better. First is that gasoline engines (especially old and worn ones) can be difficult to start and run when they are super cold in the winter. Electric vehicles in Scandinavia has shown that owners there appreciate this feature (although the battery pack must be insulated and have a small heating pad installed).
Ranchers tow stuff, like trailers with cargo and getting fallen trees out of the road. Yeah, I want a 4-speed for that.
The second way in which an electric Quad can be better, is that a 48V / 52V ebike battery can run a 110V/220V AC-inverter, plus an ATV conversion would have a fairly large battery pack. This transforms the ATV into a rolling mobile power station. There are dozens of youtubes about converting an ATV to electric, so this is not a new idea, and it is well-proven. I haven’t converted a motorcycle or ATV yet, but the ATV looks to be MUCH easier to adapt to an electric drivetrain.
Zero and Sur-Ron “Crate Engines”
If the whole conversion scene interests you, but you want a larger single-speed motor, you’re in luck. The Zero motorcycle company expanded a while back into selling matched motor/controller/battery kits for the DIY crowd, and I recently saw that Sur-Ron has made a single-motor kit along with a dual-motor unit to increase copper mass, rather than retailing a larger motor.
For info on the Zero “powertrain systems”, click here.
If the Denzel GB200 “kit” with it’s motor, clutch, and 4-speed gearbox still interests you, Denzel recommends their 72V / 275A controller to run it. That equals 19,000W / 19-kW, but there’s no third-party testing data available yet as to how this combo handles 275A peaks…It might handle more, and it might handle less. I guess we’ll see.
The Denzel Cafe Racer
I’m just throwing this in for fun. There’s not enough info for it’s own article, but Denzel has made at least one prototype of a cool retro electric cafe racer, to test customer response.
Written by Ron/spinningmagnets, June 2021
So this is a multi-part build report on my latest ebike build. I’ll start with a quick history, hit on the highlights of the build and my impressions, and then jump into more build details than most will care about. That’s how I like to see such articles arranged so I can see if there is any need to invest in the details based on the highlights and let’s face it, in this ultra compressed, hashtag, YouTube, tic toc world…most people don’t process detailed information very well. Kids these days.
I come from a bike background in a previous life. Some wrenching, some racing, some sales repping and even some chain-store management back in the 80’s and 90’s. I got into building bikes after suffering a severe burn injury back in 2016. I had a lot of time and needed something to engage my brain. I jumped online and eventually found there was a wealth of information from sites lie this. I am mechanocal by nature and started a build and to learn and keep my mind active. Like I said I had a lot of time to think, particularly between the hours of 1:00 AM and 6:00 AM as sleep was scarce during my recovery.
My first build was converting my Ti custom mountain bike with a Tangent Ascent motor and a 17-Ah triangle battery. I made a custom battery box and had some leather fabricated to house the electronics. That build was a beast and it got Eric Hicks’ (founder of Luna Cycle) attention, and we started talking.
From there Eric and I kicked around a bunch of ideas and I told him about some thoughts I had for a battery design. Frame mounted and attached with magnets. I offered to do a mock up of what I was thinking about and that eventually lead to the Wolf pack form factor. Eric and his team did the hard part but it was fun to be a small part of the design process. Look Familiar?
So, on a trip to Los Angeles in May of 2019, I met up with Eric, Ashley and the team, and we did a podcast thing. You can catch it here: THAT was fun…
So the build is basically this: 2001 Schwinn Limited Production Panther frame and fork. This is a pretty cool bike that was made for two years. It uses 7005 Aluminum and a reasonable geometry (71.5/73)
Luna BBSHD mid drive
Luna Ludicrous V2 controller
Luna BBSHD Ludicrous steel gear
Eggrider display from Luna
Luna 52V 17-Ah GA-cell battery pack
Luna Wolf Pack 52V 12-Ah
Sturmey Archer 3×9 3-speed hub
Gates Centertrack belt-drive (55T/22T)
Origin8 Transit handle bars
Brooks Flyer Saddle
Light and Motion Vis-E Combo light set
Vintage Schwinn B6 Frame Tank
Custom Made Schwinn Jaguar replica rear rack
Custom made steel and carbon-fiber battery box
Custom made wood and carbon fiber fenders
BOB Ibex trailer
I have built a couple of these Schwinn Panthers and I like them a lot. I’ve got a nostalgia thing for Schwinn, and I think their cantilever cruiser frame is gorgeous. Heck, I flew to Chicago in 1988 to interview with them, and almost went to work for them. So, the frame makes for a comfy, stable and roomy ride. No surprise there for me. Gates belt drive…ditto, I have built a couple of bikes with them and I like them a lot.
Where I was surprised was with the Luna BBSHD/Steel Gear/Ludicrous V2 Controller/Eggriderpackage. Un-friggin-believeable! Speed, torque, control and efficiency all in a nearly silent, robust and affordable package. Also, what is pretty amazing is that out of the box for almost everyone, this thing requires no programming or tuning. My only tuning, because I am using a 3speed IGH, I figured I need to keep things kind of calm on the acceleration front and frankly, I prefer a smoother ramp up regardless, so I fired up the VESC app on my Android tablet (yes theV2 Ludicrous controller is VESC compliant), extended the ramp-up for throttle and PAS to 2.5-seconds and hit the road. The Eggrider integrates very well and brings a stealthy and easy to control solution to my left thumb. Eric is pretty amazing for him to see the need, design, build and program this controller. Frankly, I am not sure which of those was the hardest to do, but I am grateful he did it.
I have about 50 miles on the bike so far, and in every situation it has performed better than I could have hoped for. Riding with my wife on a leisurely Sunday morning ride, I set the PAS to 2 and it was docile. Wanted to blast into town for some eats, so I pegged it to 9 and zipped along at 40+ mph for a few miles. Took it off road in level 6, and was blown away at the climbing torque. Basically, it is the complete package. The only catch is the ludicrous V2 is mostly available on complete Luna bikes. If you want to purchase it on on it’s own you are going to have to 1.) Know what you are doing and 2.) Convince someone at Luna that you know what you are doing … I am not sure which one is harder. That said, if requirements #1 or #2 are not easy for you … you are far better off simply purchasing a complete bike with a Ludicrous controller from Luna.
OK, this is where many can tune out, as it covers some of the specifics of the build. This is not a “how to” manual on building an ebike, but it will fill-in some of the gaps for a knowledgeable builder about what I did, and how I did it.
The FRAME – When anyone decides to go with a belt drive, you need to put a splitter on the frame, as the belt is a continuous loop unlike a chain which can be split. Best place is usually the seat stay, down near the dropout. I have done this same process before with great results. I cut the frame (a terrifying experience) with a tube cutter for a clean and controlled cut about 1.5” wide. I had a piece of aluminum rod turned to the same outside diameter as the seat stay. They also turned a section the same inside diameter as the seat-stay and machined a “flat” to exactly 1/2 of the outer diameter. Once the rod is turned, I cut it to make two matching pieces that fit the frame opening (1.5”) and drilled and tapped two holes to bolt them together. Then using 3M Scotch-Weld DP420, I epoxied the the inner parts of the frame splice into each of the frame seat stays follow the directions exactly with the DP420. These pictures should make it clear what was done. Clever huh?
The rear hub and belt drive – I went with a 9-speed cassette Sturmey-Archer 3-speed, so I would have maximum flexibility in belt line adjustment. The SA is about as robust as any IGH, aside from a Rolhoff (which I also have, but it simply has too many gears for a mid drive as you end up shifting 3-5 gears at a time). The trickiest part of the hub fitment was making custom chain-tugs to hold the axle under the tremendous force that the BBSHD with a V2 controller can exert. For this I ended up welding some M6 bolts to the keyed locating washers of the SA hub.
Note that I had to then cut the bolts and weld them in board so they would sit within the drop-out, and not outside of the dropout as the washer does. I then cut some aluminum U-stock for end caps that would slip over the rear dropouts. These seem pretty robust and do the trick. Yes my welding skills could use some help…. Just wait until you see the battery box! Up front I used a 55-tooth Gates Di2 pulley. The Di2 version is designed to move the pulley out a couple of millimeters to clear a Shimano Di shifter. Flip it over and you have a pulley that sits inboard a couple of millimeters. Combined with 1.6 millimeters of chainwheel shim, and I was able to get the belt-line perfect!
The wheels-Basic Sun Rhyno-Lite rims with DT spokes. I went with Maxxis DTH 2.3” tires, which only have about 2mm clearance in the rear on either side. We’ll see how good of a wheel builder I am by how well my tires stay off of the chain stays. The Rhyno-Lites are very robust rims with a great braking surface. Yes, I use rim brakes and they are fine. Discs would be cool but aluminum-welding added caliper-mounts is beyond my expertise. And I kind of like the retro vibe of the U and V brakes.
The Battery Box – This was a fair amount of work, but not too complicated. I got some 3.5” wide 18-gauge steel and set about to fit it as I wanted. I screwed a wood 2×4 into the side of my work bench with a 3/6” gap, and used that to help form the bends. Once I had the shape dialed in, I tack-welded the seams while it was on the bike to set the shape.
The mounting tabs for the side panels were made from small steel 20mm L-brackets that I welded in place and added M4 nutserts to to allow for mounting the side panels. The side panels were made from 1/8” ply that I laid up with epoxy and carbon fiber cloth. This is mostly cosmetic, but it is as easy to work with as fiberglass and it looks good. Also you can see I added a tab to the bottom of the battery box to attach the BBSHD to, for maximum stability. Of course it has a Rosenberger charging port for maximum style points. Finally, I added a small removable panel for the unplugging of the main battery, and the attaching of the auxiliary Wolf Pack battery (see below).
The Tank – This is an old Schwinn B6 tank I had laying around (I told you I had a thing for older Schwinns). Repros are available with nice shiny chrome on eBay. The tank is just a great place to consolidate some of the electronics (12V DC/DC converter, the on/off switch for the lighting, USB charging port) and even if I did not need it, I would likely add it for cool points. My tank was a little sad and took some work. I used bondo on a couple of holes and drilled a couple of others, so I ended up with 3 mounting points mirrored on each side. To mount the tank, I simply took some aluminum tube, cut to proper length, inserted some M4 nutserts in the ends, and then screw both sides of the tank into them. They basically clamp onto the frame and are easy to remove as needed.
The carbon fiber inlays are 3M vinyl wrap, and Schwinn was ghosted on by applying some Schwinn lettering decals (eBay) and spraying matt clear-coat over them, and then removing the lettering. Similar technique for the Black Panther lettering on the chainstay.
The Fenders – The stock 2001 Panther had “carbon fiber” fenders but they were basically clunky Wald chrome fenders with very cheap carbon sticker applied. Mine are made from 1/8”-thick steam-bent Ash, with a carbon fiber epoxy lay-up. Very strong and light.
The Rear Rack – The rack is a reproduction of a Schwinn Jaguar, rack but I made it to fit my Panther frame perfectly. The rack also has tabs for the mounting points for my BOB trailer and can accommodate a Luna Wolf pack 12-Ah battery as an auxiliary battery. As a final touch, I got ahold of some reproduction Jaguar rack reflectors (crazy expensive on eBay) and added small red LED lights inside of them, and ran the wiring underneath the rack. They come on when the lights are switched on for maximum visibility and bonus style points.
The Lights-The front and rear lights are “Light and Motion Vis-E”. They are pretty well made, ebike friendly and they allow for remote switching.
The Internal Wiring-Also, all wires are internal. Just added to the clean lines and did not take too much effort. Where items like the rack and seat-post can sometime require removal, I used 2.1mm DC connectors. My biggest moment of fear was cutting and splicing the Bafang wiring harness. Although it looks daunting, it was pretty simple. Flux and heat-shrink tubing are your friends here.
The Trailer – Yeah I know…a trailer? I had it already, and decided to make a matching carbon fiber fender and paint it black to match the rig. As it turns out, with all of that torque, those beefy 2.3” Maxxis and my extended range capability, the trailer makes this the perfect gravel/dirt road machine for exploring and camping. I live near the Blue Ridge mountains and there are literally hundreds of miles of forest roads just waiting to be ridden. Although the Wolf Pack battery fits nicely on the rear rack, I made a 5’ cable so I can place it in the trailer when I’m off-road exploring, as it lowers the CG and gives the bike more stability in the rough stuff.
That’s it. Just the highlights of what was a 14-week build. Simply put, there has never been a better time to be an ebike builder. Grateful to everyone I have learned from, and a special shout out to Eric and Luna for all they have done for the industry.
Written by Adam Livingston, May 2021
The methods used to build a battery pack from cylindrical cells are fairly well known, but pouch cells don’t seem to have a decent article yet, to describe the “best practices” on how to build them. I’ll do my best to show what I found when I recently did a search.
Why Pouch Cells?
When you stack-up cylindrical cells, there is some airspace between each cell, and that’s wasted pack volume that could be used for more active battery material. This isn’t a huge problem for ebikes, but it’s a matter of deep concern for electric motorcycles builders.
A 52V pack that has 20-Ah of range is considered a pretty significant ebike battery pack, but that’s not enough power or range for a scooter, much less a full-sized E-motorcycle. The Zero motorcycle company uses 28S / 104V nominal, and as far as range, their smallest pack is currently 44-Ah, with larger packs available. That’s roughly four times the total energy, and most Zero E-moto owners have chosen the larger battery pack option. E-motorcycle builders need to cram as much active battery material as possible into a frame with limited frame-space.
A 12V Jumpstart Box
I haven’t built an electric motorcycle, but I do occasionally need to jump-start a piece of equipment at work. I had read that the cells that put out the most amps are from hybrid vehicles. This is because EV’s that have a large battery pack (like Tesla) will be able to make more than enough amps, so…they can use cells that are optimized for long range. A hybrid EV has a back-up engine to provide long range, and that means that the expensive battery pack can be small. However, this EV hybrid pack must be able to provide full-vehicle performance from a small pack, sooo…these cells put out high amps.
I saw a youtube video where an electrician tested the amperage draw on his 4-cylinder gasoline car. In the first half-second, it drew a peak of 200A to get the motor spinning, but settled into roughly 100A to continue spinning (the spark and fuel were turned off) . He was putting together a super-capacitor jumper-box to help him start his car when the weather was below freezing, and it would turn over the engine for up to 8 seconds. Chemical batteries like lead-acid and lithium will lose a lot of their amp-producing capability when they get extra cold, but super-capacitors work just fine in cold weather. This is part of the reason Tesla cars have a battery warming/cooling system.
I briefly considered building a super-capacitor bank, because they can provide very high amps in a tiny package, and as stated before, they can operate well in sub-zero weather. However, super-capacitors bleed-down their voltage over time, so I would have to re-charge the super-capacitor pack just before trying to start a vehicle with a dead battery. The “filler” battery was typically a 4S Lithium-Iron pack that is pocket-sized, and even an 18V cordless tool battery can be used. It would take a few minutes to use a cordless tool pack to “fill” the super-capacitor bank (the filler battery pack is kept warm inside the house until needed). Then, you would take the super-capacitor bank out to the frozen car to help start it.
I opted for the simplicity of building my own “jumper” suitcase with only large lithium cells, and no super-capacitors. This way, it would have more “range” when using it as a power back-up during a power outage (I live in tornado country, with frequent wind-storms). I wouldn’t call it small, but it is just light enough and fairly compact, so that I can store in the warm indoors. In an emergency, I can also use it to keep my phone, laptop computer, and my rechargeable USB flashlights charged. If you are considering building a large and expensive battery pack for an E-moto conversion, I recommend building a jumper box first. It will teach you everything you will want to know, before spending the big bucks on the full-sized project.
I found some threaded aluminum bars that would work, to clamp the tabs against each other (seen in the pic above). If I hadn’t found these, I was about to buy some bar stock, then cut, drill, and tap threads into them myself. Since the tabs are touching each other in this configuration, the bars could have been steel to maintain stiffness and provide an even clamping pressure.
When using common lithium-Ion cells that are fully-charged to 4.2V per cell, three of them in series produce 12.6V. Lithium-Iron Phosphate cells (LiFePO4) run at a lower voltage, and they are fully charged at 3.6V each, so a LiFePO4 pack that is used to provide 12V is typically a 4S pack, and 14.4V when fully charged. The LTO chemistry is rare, and at 2.4V per cell, a 12V pack would have six cells in series, for 14.4V.
Chris, in Perth Australia
Chris races electric motorcycles (endless-sphere forum member “jonescg”). When he started, there was nobody around to build a large battery pack using a custom voltage and shape, so…he learned how to build them himself. He has tried a few different methods, and he has agreed to allow me to show them here.
The first step for the method Chris uses is to order a thick PCB, made to the design that he drew on a computer. PCB’s are usually used as a circuit board to attach electronic components to them, but this is a clever use of an existing product-line to use a PCB in a way in which it was never intended to be used. The hex-shapes on the bottom side of the PCB’s were chosen to make a place that threaded nuts could be soldered-to in those locations, rather than using epoxy or some other method, like rivet-nuts or thread-serts.
In the pic above you can clearly see the slots that the tabs slide through, and also the soldered nuts that the clamping screws attach to. The nuts do not require a lot of “anti-turning” hold, so soldering the edges provides plenty of strength. However, the nuts are nickel-plated specifically to make soldering easier.
I’ve seen some battery pack builders use a paper-hole punch to make a hole in tabbed pouch cells, but these cells were ordered with thicker collector-tabs for the high current that Chris needed. The “punch” is made from a rod that has had a hole drilled into it sideways. This results in a shape that works well for punching holes through his alignment jig, shown on the left.
Chris uses thick copper bus-bars to connect the tabs to each other. The pic above shows how the tabs slide through the PCB plate, and are folded over. The actual production plates and buses have more holes than this, but Chris was eager to post a pic of a prototype. Copper buses that touch the tabs not only have low-resistance for carrying high current, they also act as heat sinks to cool the battery by pulling the core-heat out through the tabs.
Just under the head of each stainless-steel screw is a stainless split-lockwasher to provide constant compression on the tabs, during heat and vibration cycles. Then, a stainless washer at the bottom to spread-out the compression. This pack uses three cells in parallel, 3P.
Chris made an assembly box out of non-conductive clear acetal plastic. The pack shown above lays-out the cells in a U-shape, so the positive and negative end-terminals are both on the same end, on the left.
The pack above uses larger cells, but the same method as before. I used a similar “head-plate” configuration to make my jump-starter suitcase battery, but since I was only making one of them, I simply used a thin kitchen cutting-board that was made from HDPE / high density poly ethylene, instead of a special-order PCB board.
At this stage, you would still need to attach a BMS, and a sturdy housing to protect the cells. You can see how to attach a BMS by reading here, and also we wrote a few things about what a BMS does in another article, which can be found here.
Written by Ron/spinningmagnets, May2021
The story begins with one guy, Neal Saiki (SY-kee). He is an engineer who worked for NASA for a while, which shows that he is a pretty smart guy, and also why he was always up to date on new tech developments.
Before starting Zero, he worked at Santa Cruz mountain bicycles, and it seems he has always been developing new products (either full-time or on the side). The early articles about him seem to indicate that he was most interested in the new battery developments that were blossoming just before he started Zero in 2006. In fact, he continues to work to this day with an electric bicycle company he and his wife Lisa formed called “NTS Works“. The interesting part of the NTS ebike is the innovative battery case, which can easily be re-filled with new cells by the owner when the original pack is worn out.
This very much endeared me to Neal, because…when a manufacturer produces a proprietary interface, you can only buy the replacement batteries from their catalog. But…Neal’s battery case design allows the customer to source 18650-format cylindrical-cells from anyone (found in laptops and cordless tools), and easily swap them out in their own garage.
Also, when the cells are worn enough that the owner is only getting about 80% of the range compared to when they were new? The old cells still have 80% of their capacity. There is a growing movement of home-tinkerers who are using near-free “used” lithium cells to make a power back-up battery pack for their home, so…this is not as kooky as it may at first sound. Now…let’s get back to Zero motorcycles…
I know this article is about Zero, but nothing happens in a vacuum, and it’s useful to remember that the Tesla car company was started in 2003, so…by 2006, Tesla had quite a bit of buzz about how the new batteries that were then available were actually making EV’s viable. Tesla had been showing their Roadster prototypes to investors, and it was clear that they were aimed at competing with FUN cars, rather than only “being green” (like the failed GM EV1).
Neal is a brilliant guy, so it wasn’t a big leap for him to realize that motorcycles were a much smaller investment risk, compared to an electric car. If investors were excited in 2006 about electric vehicles that were fun and performed well, then…the time was ripe for someone to start a company that could show the world about everything that an electric motorcycle could be.
He was right, of course. The booming US economy of 2006-07 meant that it was the best possible time for investment and risk-taking on Wall Street. However…apparently nobody could see the devastation of 2008 that was looming just around the corner. Opportunity is a two-edged sword that cuts both ways. By accepting venture capital to fuel growth, Zero now had the money to produce a high-quality production prototype, and scale-up some type of manufacturing facility. However…by accepting that money? he now had to answer to a corporate board of directors, and…it was just at a time when the economy was about to take a huge dump…
The stakes were high, and Zero appeared to be the first company that would produce a viable mass-produced electric motorcycle. Anyone involved was swimming in “buzz”, and as far as anyone knew, Zero was going to be the next Tesla. However, the severe recession of 2008 let the air out of the bubble and even established companies were suddenly struggling. During this difficult time, Zero managed to produce its first production model in 2010, using the Agni-Lynch motor and the very safe LiFePO4-chemistry cylindrical-cells from Molicel.
in February of 2011, the Board of Directors appointed Gene Banman as the new CEO. Neal Saiki stayed on as Chief Technical Officer / CTO. Remember when I mentioned some turmoil? Just a few *months* after Banman became CEO, he stepped down (but stayed on as a member of the BoD). One encouraging note is that at the same time, Mark Blackwell joined the BoD, and he had previously been VP of motorcycles at the global Polaris company. Polaris is best known for its snowmobiles and “Indian” motorcycles.
For a while, the leadership gaps in Zero’s history were managed by their Chief Operations Officer / COO, Karl Wharton.
In July of 2012 (amid various start-up production and design glitches, plus a few recalls), Richard Walker became the CEO. The entirely revamped 2013 lineup was already being designed, so…it looks like he took over just as Zero had found its stride. Walker did a good job, and it appears he held onto the leadership of Zero until July of 2016 when he was replaced by Sam Paschel. Walker is now part of a venture capital investment group for new tech companies.
2013, The Breakout Year
Every year at Zero, the product was improving in a variety of ways. However, there was a distinct change in the 2013 models. The 2010 models had used the Agni axial-flux brushed motor. By using brushes, the controller could be more simple, along with being less expensive. However, in spite of the efforts to make choices that enhanced reliability and kept the price affordable, there were enough problems with design issues and quality…Zero needed to consider big changes.
Potential customers seemed to really like Zero motorcycles, but…customers wanted more power and range, and Zero needed to fix all of the reliability issues, or…recalls would bankrupt them. The system voltage was doubled from 14 cells in series (14S / 52V), to…28S / 104V. And. In 2013, the motor was now a radial-flux inrunner brushless PMAC / IPM design.
The higher voltage and new motor provided a BIG bump in power, and the new battery using Farasis flat foil cells provided more range and a much longer pack life. Although the early QC issues were painful to their reputation (and investor confidence), it forced Zero to improve quality much more than they might have done otherwise.
In 2015, Zero upgraded their products with Showa suspension parts, Bosch anti-lock brakes (ABS), and an optional battery range of 186 miles…this shows that the large global suppliers of motorcycle parts now wanted to be associated with Zero, rather than being afraid of taking that risk with them.
Sam Paschel, CEO since 2017
I am very encouraged about Zero motorcycles today for several reasons. Sam Paschel has been there for a couple of years, and it appears he is likely to be staying on for the near future. That alone retains the stability that they achieved under Richard Walker at a very important crossroads in Zero’s growth. Plus, there are several things I really like about Paschel…
In the business world, many power brokers truly believe that a good manager can guide a company to growth and diversification, no matter what the product is. Case in point…when Steve Jobs stepped down from Apple Computers to start a different company in 1983, he persuaded the head of Pepsi Cola (John Sculley) to take over Apple. High-end computers are not soda, and…it did not go well.
How does this relate to Sam? He is an engineer (Swarthmore), and…he has been riding motorcycles ever since his dad had gotten him a Suzuki RM80 (and didn’t tell his mom), and…that happened when he was 8 years old. He spent several years after college trying to get a used Honda CB550 to run, and also to stop shedding parts all over the highway.
This executive rides.
I also found a couple of behind-the-scenes guys who played a part in the Zero motorcycle evolution…Abe Ashkenazi, and also Luke Workman.
Abe worked at Buell Motorcycles from 1995-2009, so…in in spite of his recent success being in charge of the tech used in Zero motorcycles, he is a motorcycle guy from way back. The fact that he has been at Zero from 2010 to the present day shows that he is one of the people who has seen the entire transformation of Zero from its beginnings into the successful company that we see today. The steady series of promotions on his record tells me that he played a crucial part in the successes of this company during that time.
Next up is Luke Workman. Luke is a “true believer” in the electric vehicle movement, and his personality is an infectiously positive attitude that absolutely radiates everywhere that he goes. Luke was an old hand at wrenching and riding superbikes, and also turbo Honda cars. Somewhere along the line, he became fascinated by electric vehicles, and…Luke isn’t the kind of guy to do anything “halfway”…he dives in with both feet.
Luke is listed as Zero’s Sr. Engineer Battery Specialist from 2010 to 2016. He is currently the senior tech officer at Redivivus (click here).
A timeline of Zero motorcycle tech, starting in 2006
Neal Saiki started a company in 2006 called Electricross, and his first product was an electric dirt bike, which I think was a great decision. It was important (at the time) to quickly find out the performance data that he needed in order to make important decisions about the design and also parts supply, and a dirt bike was the fastest way to do that, since it didn’t need to meet Department Of Transportation / DOT street rules, like having lights and turn signals.
Motocross Action Magazine wrote about the “Electricross Drift” in September of 2006, and the first prototype used four lead-acid batteries in-series for a nominal 48V. When they went into production, they immediately provided an optional pack made from lithium-Ion cells, which has a more stable voltage throughout its full-to-empty range, and also provided more miles of riding. I don’t have any information about the Lithium cells they used from 2006-2008.
The entire dirt bike weighed roughly 140-lbs, and the rolling chassis with no battery was 80-lbs. The base-model retail price was $5500. The stock top-speed on smooth roads was 44-MPH, but that could be changed by simply swapping a sprocket, if you wanted better hill-climbing, and didn’t mind a lower top-speed.
In 2007, the Electricross company was renamed “Zero”, for Zero emissions. The name change was needed because they decided that…the overflowing response from customers, magazines, and potential investors had persuaded them that they needed to also make a street version, and they wanted a new name that was less motocross-specific. The pic below is the only one I could find of the 2006 shop, from 3111 Scotts Valley Drive, in the city of Scotts Valley near Santa Cruz, on the central California coast.
The motor was an existing brushed motor in the axial-flux configuration called the Agni, which was designed by Cedric Lynch in the UK. By using a brushed motor, the controller could be fairly simple and affordable. Here is a video of adjusting the timing of the brushes, which also shows the internal configuration. The two side-plates hold stationary magnets, and the central armature spins, which hopefully helps to shed coil-heat to the air.
Here is a video from Jozztek which shows the disassembly and internal configuration of the Agni-R (The R stands for reinforced, so that it can withstand higher RPMs). From the beginning up until the major changes in the 2013 model year, Zero used axial-flux motors. An axial-flux is like plates that are spinning side-by-side, rather than a radial configuration. A radial-flux motor is more like a cup spinning inside another cup.
Zero did not begin shipping the Zero-S “Street” model is significant numbers until 2010. So the tenuous 2006-2008 period was focused on the off-road and “dual sport” models.
Zero during 2009-2011
It’s sometimes hard to find information and pics of an older model that is no longer in production, but I stumbled across an unusually good write-up on a 2010 Zero DS (Dual Sport). 2010 is also the year that significant numbers of the Zero-S “Street” version began shipping.
The brushed controller for the 2009-2011 Zeros was the Alltrax AXE4855, and the name indicates that this model has a nominal rating of 48V (it will accept voltages up to 60V), and can provide temporary peaks of up to 550 amps (48V X 550A = 26 kW). The controller was later upgraded to the AXE7245, rated for 72V and 450A = 32 kW (25% more total power, but slightly cooler due to fewer amps).
The 2009 battery was made from cylindrical cells from Emoli / Molicell in the 26700 format (26mm in diameter, 70mm in length), and the pack had twelve cells in series (12S).
The world is a better place because of the modern Zero electric motorcycle company, so we are fortunate that these early versions were “good enough” to allow the company to survive. The performance was “as good as could be expected”, but to be honest, there wasn’t a lot of competition during this era. The 44V battery packs that Zero used were reasonable in view of the batteries that were available at that time. However, they do not compare to the exceptional performance of the modern versions. [note, this is 2018, and I recommend that any Zero battery pack from 2011 or earlier should be immediately retired]
If you know of anyone who has one of these early versions, the Agni motor and Alltrax controller are worth keeping if you only task it with 10-HP / 7.5-kW jobs. The battery is another story. Once it is too worn to continue using, Zero does not make any new legacy battery packs replace them. There is some good news though…you might be able to replace a worn 2009-11 Zero battery with salvaged Nissan Leaf cells. Thoroughly research this before attempting it.
Sooner or later, all EVs (like the Nissan Leaf) will have a few examples become involved in a wreck. If the battery pack is in good shape and has only a few miles on it…there has definitely been a market developing for lithium cells of all types. These Nissan Leaf modules contain 58-Ah, and are capable of 240A continuous, and double that as a temporary peak. Replacing a 2009-2011 battery pack with modern Nissan Leaf cells. Each Nissan Leaf module has two cells in series (2S), so a 7-module array is a total of 14S, for roughly a nominal 52V
Big changes in 2012
The model year 2012 saw some big changes, and they were significant improvements. The 2012 cells were flat pouch cells made by Energy Innovations Group (EIG) from South Korea and were configured as an 18S pack, which had a nominal voltage of roughly 65V (this is the only year that this voltage was used). From this point forward, Zero has used flat pouch cells. The chemistry was Lithium “Nickel Cobalt Manganese” / NCM. The EIG corporation calls these their 20-Ah “C020” cells. Packs could be made from paralleled modules in 1P / 2P / or 3P, for capacities of 20, 40, and 60 Amp-hours.
However, there is a warranty recall in-place for the 2012 packs, concerning some of the insulation wearing away from vibration. If you have one of these, please contact your closest Zero dealer to have this fixed.
This was also the year for big changes in the motor. The Agni is very efficient and also fairly affordable to manufacture, but hot rodders have stated that it is near its “long term” reliability limits at around 10-HP, and running it at higher power levels reduced its reliability and life-cycle. Customers wanted more power, and the Agni was only an efficient 10-HP scooter-motor. (it could put out 20-HP, but it simply couldn’t last very long when doing that)
The 2006-2011 Agni axial-flux brushed motor, shown above in a 2010 Zero DS. The brushes are on the left side, and the flex-hose is blowing air onto them to clear dust from brush-wear and then blow air through the motor case for cooling the armature. Brushes are a form of rubbing electrical contact that physically touch as they slide from one contact to another. In order to significantly raise the power, the decision was made that Zero’s next motor was going to be brushless.
The Motenergy ME0913 brushless motor has two stators, and one rotor. The new motor for 2011 was based on the Motenergy ME0913 motor but was made to the exact specifications provided by the Zero Motorcycle engineering team. There were significant design changes and improvements required by Zero to make a motor that met all their requirements.
When Zero had specified that the new motor must be brushless, it required a more sophisticated controller than the brushed Agni, and a robust unit from Sevcon was selected. This turned out to be a good choice, and Zero’s relationship with Sevcon continues to this day.
A 3-phase motor with a single stator typically has three fat cables mounted to it, but…as you can see in the pic below, this model is identified by the fact that the three electrical mounting posts are split into six cables that are routed to both sides of the motor. This version uses two of the 3-phase stators and a single central permanent-magnet rotor.
Motenergy’s previous name had been “Mars” electric motors, and they are a well-regarded engineering firm, which can provide motor solutions that meet unusual specifications. By having two stators and a central rotor, the hottest parts of this model are placed at the outer sides, where air-cooling can be the most effective when mounted on a motorcycle as an application.
The new ME0913 motor from Motenergy is still an axial configuration, but it is a brushless design, which helps with several issues. Temporary burst power can be increased without producing a lot of the conductive arc-dust from the brushes that had occasionally caused issues in the Agni. Also, It’s possible to water-proof a ‘brushed’ motor, but by using a brushless style, you can easily have a robust water-proofing along with also having the dramatically improved air-cooling.
Another difference is that the configuration of the Agni had a central spinning armature, which worked with fixed permanent magnets mounted inside the side-plates. The Motenergy has a single central magnetic-rotor, and two stators, with one on each side. The temporary-burst power of a motor is created by increasing the amps to the stators, which causes a lot of heat. By having the stators located on the outer sides of the motor, it was easier to air-cool them in a more effective manner.
The Agni remains a respected motor for modest power applications (due to low cost, ease of manufacture, and high efficiency), but a major difference between the Agni and the Motenergy is that even though armatures and stators have many similarities, the electromagnetic fields that can be turned on and off will only extend outwards a certain distance. I point that out so that I can say this; to dramatically increase the amps that the motor can use, you must dramatically increase the copper mass. If we simply make the Agni armature fatter (to add more copper) the armatures’ magnetic fields would not be more effective. The Agni only has one armature (in effect, a single spinning “stator”) coupled with two magnet-array plates. The Motenergy uses a single magnetic rotor, with two stators. It has double the copper mass, if not more.
All of these changes result in a motor that is only slightly larger, and yet it can survive a temporary burst of 400A. When using the nominal 65V of their new battery pack (18S), 400A would result in a burst capability of 27-kW / 36-HP for acceleration (although the 2012 battery packs could not actually provide 400A).
Side note: I was curious about the naming system at Motenergy, and John [the owner] told me that the “09” means that this model was designed in 2009, and the “13” meant that it was the thirteenth model-variation they had designed that year.
This was the year that the batteries at Zero took a great leap forward. They use flat pouch cells from Farasis Energy, They raised the voltage to 28S / 102V. Zero designed the new totally enclosed radial flux motor in-house for the 2013 model year, and they chose Motenergy as the supplier for their new design.
From this era forward, each flat pouch cell increased in capacity over time. Starting with 25-Ah, then 27, and today finally achieving 29-Ah’s each. They started with a 4C continuous current rate (100A from a 25-Ah cell) and ended up today with 29-Ah cells that can peak at 10C (290A). This means the 2P packs have 58-Ah’s and are capable of a 580A temporary burst.
A polymer-embedded 14S module made from flat Farasis cells (pulled from a wrecked Zero motorcycle). The polymer fill makes this block unrepairable if it ever has an issue, and doing this was only possible by producing a pack that is more reliable than any other on the market. The potting also provides extreme vibration and shock resistance, plus a level of waterproofing that is exceptional. The module is shown still worked.
Farasis Energy Inc has its design headquarters located in Hayward California (in the southern San Francisco Bay), with a manufacturing facility in Guangzhou, China (near Hong Kong). Their CEO Yu Wang has led the company to a recent financing expansion that exceeded $1 billion USD (with a “B”), and I’m certain the success of their Zero battery cell technology has played a major part.
Farasis recently opened a new office in Stuttgart Germany, in preparation for building a new advanced battery pack factory. The Zero battery pack design team was headed-up by CTO Keith Kepler. Other engineers involved in the Farasis cell success are Jackson Edwards (“Farfle”), Chase Nachtmann, and Zero battery specialist Luke Workman (“Live For Physics”).
2017, The Motor Gets An Upgrade to IPM
The new 2013 radial motors from Motenergy worked great, but there was some room for improvement. The permanent magnets found in these types of motors are a solid chunk of metal, which differs from the laminations, which are a stack of many thin silicon-steel plates. Whenever a solid chunk of metal passes rapidly through a magnetic field, it can suffer from eddy-currents, which results in extra heat. The new radial 2013 Zero motor had surface-mount magnets, and even though it performed very well…the 2017 model was upgraded to a configuration that places the magnets slightly farther away from the surface of the rotor laminations, which allows the magnets in the rotor to run cooler. This is called an “Interior Permanent Magnet” / IPM style.
When you put a permanent magnet next to steel laminations, the steel laminations become magnetized up to a small amount of depth. This means that the magnets can be moved a little farther away from the air-gap between the stator and the rotor, without lowering their magnetic effect on the motor. Doing this reduces the eddy-current heat in the magnets. Permanent magnets can be partially de-magnetized by accident if they get too hot, so magnet-heat is one of the limiting factors in the number of amps and power that a motor like this can survive.
Radial in-runners attach the coils in the stator to the stationary aluminum shell, which does a pretty good job of passively shedding the stator-heat to the ambient air. The engineers at Zero are familiar with all of the known liquid-cooling systems for motors, so this was a conscious choice rather than a compromise. The two biggest reasons I can think of for choosing passive air-cooling is the simplicity, which has a clear element of reliability, and secondly to also to keep the costs affordable. There are 12 poles on the stator (3-phase), and 10 poles on the rotor
Here’s part of a discussion that I found interesting from a technical forum
“I just noticed that the Zero S and SR have the same motor Z-Force 75-7 and that motor is rated 40-kW in the Zero-S but 50-kW in the Zero-SR. Previously, I said “Zero-S power is limited by the motor, not the controller or pack” but that’s not correct. Actually, Zero-S power is limited by the controller, not the motor or pack.
Here’s my reasoning: The same motor model–the Z-Force 75-7–appears in the Zero S and SR. In the Zero SR, the 75-7 motor is rated 50-kW, so we know the motor can handle 50,000W / 102 V = 490 A. But the Zero-S specs state that the Zero-S controller can deliver 420A (no more). Or in terms of power, the Zero-S controller can handle 102V x 420A = 42-kW (no more).
To recap, the Zero S has a 50-kW motor, 42-kW controller, and a 57-kW (102 V x 567 A) battery pack. I.e. Zero-S power is limited by the controller, not the motor or pack. When you choose a Zero SR over a Zero S, you’re getting a more powerful controller and a larger battery pack that can support the Z-Force 75-7 motor at 50-kW”
“As of 2016, a reliable source indicated that Zero was producing 17 electric motorcycles a day”
If you already own a Zero and are interested in a couple of upgrades, you can change one of the pulleys to enhance the low-speed torque. Doing this also lowers the top speed a bit, but many feel it is worth it. Another easy option is to upgrade the rear shock absorber.
The closest Zero dealer to me was Letko Cycles, near Kansas City…which can be found at 12535 Rogers Rd, in the city of Olathe. They are a very experienced dealer of Zero and KTM motorcycles, and I’d like to thank them for their help, and also for the test rides..
Sine Cycles is a custom builder that is proud to be using a completely stock Zero drivetrain. I can see jet skis and snowmobiles starting to do this soon too…
The physical size and performance of the Zero SR have been compared by others to the current crop of high-performance 600cc sportbikes. I’ve never ridden a 600cc, but I have owned a 1981 750cc sportbike, and the 0-60 acceleration torque from the SR felt the same to me. Here’s a quote I found: “67-HP and 106 ft/lbs at the motor. Acceleration is 0-60 in 3.3 seconds. Top speed of 102 mph. The range is 171 miles with the 2.8-kWh Power-Tank option”
If you are already a fan of EV’s I don’t have to emphasize how eerie it is to be riding something this powerful, but the motor and belt-drive only emits a faint whine. The tire contact on the asphalt was as loud as any other noise coming from the bike.
As much as I also like vintage motorcycles, I once owned a motorcycle with spokes that could loosen, and tubes in the tires that were easily flattened…and I never will again. The SR I tested has tubeless tires, cast wheels, and a low maintenance belt-drive. If you’ve ever had to clean, lube, and tension a chain to help it last as long as possible…you know what that means.
No more changing engine-oil, oil-filters, and fuel-filters. No more cleaning out clogged jets and adjusting the carburetor. No more clogged fuel injector nozzles. I recall years ago when I removed a gas tank and dumped the gasoline into a bucket through a T-shirt to filter out the rust flakes. When I was young, I liked fiddling with antiques, but now…I like my antiques on display at a museum.
Written by Ron/spinningmagnets, May 2021
The title says it all. Luna Cycle has their main facility in Southern California, and they now have a second facility in Bend, Oregon. The “Deathbike” from EV pioneer Luke Workman (Live for Physics / LFP) was being stored there, and has been stolen. Luna Cycle has offered a $1500 reward for its return unharmed.
Luna Cycle can be contacted at:
1330 E Franklin Ave.
El Segundo, CA 90245
Published May 5th, 2021
There’s been a struggle simmering in the background concerning the laws that affect ebikes, but here’s one story with some good news. For a variety of reasons, New York has been forced to recognize some reasonable laws, and Propel ebikes has played a part. Let’s take a look…
Chris Nolte is the guy who took on New York. He started Propel Ebikes (click here), as a retail outlet for electric bikes only, since there were already plenty of pedal-only bicycle shops around. I met Chris many years ago at the Interbike convention (click here), and he probably doesn’t remember me, but he made an impression and I remember him. I was casually interviewing any ebike shop owners I could find, in order to get a feel for what customers where actually buying. (The BH eMotion was popular at the time, and also, click here).
Chris has a shop in Long beach, in Southern California, but the store at the middle of this controversy is the shop in Brooklyn, just across the bridge from Manhattan, in New York. Ebikes have been under pressure for a long time in NY. I think some of it was due to aggressive food delivery riders, who used ebikes to get between cars in the crowded traffic.
Also, I have a theory for which I have no evidence. I visited Manhattan several times for a job I did. There is a shockingly high cost to park your car there, and it appeared as though they didn’t want anyone to drive their own car into the city. I was working in New Jersey, right on the border. When the weekend arrived, it was surprisingly cheap and easy to park my rental car in a very secure guarded car-park, and then ride a train into the center of Manhattan.
The island of Manhattan is long and narrow. the widest part was about ten blocks, which is fairly easy to walk. If the average citizen (or tourist) wanted to travel lengthwise, the subway system worked quite well, and was also affordable. It appeared as though the fairly wealthy citizens there used taxis and limousines to travel around Manhattan (think stock brokers and celebrities).
Remember the movie where Tom Hanks played pilot Sullenberger? They took off from the airport right next to Manhattan and both engines immediately went out (due to a flock of birds going into the intakes), so he had to land the plane in the river right next to Manhattan. That’s how close the airport is to Manhattan, but…the taxi fare from the airport to Manhattan was recently listed as $52. My point is…the City of New York does not look kindly on anything that takes money away from the subway or the cabs…like maybe…electric bikes?
A smart investigator once wrote “follow the money…always”. The city of New York derives significant income from the subway and taxis. That sounds like an offhand comment, but…to get a taxi permit (called a medallion) is very high. In 2019, sixteen medallions were offered at auction, three of which sold for $137,000. That is before you earn a single dollar in cab fares. Due to Uber and Lyft “ride sharing” programs, the New York Taxi population is down to around 13,000 (not a typo). Now, I’m no math genius (googles furiously), but that means the city of New York is getting over a billion dollars a year (with a “B”) from just granting “permission” for taxis to operate there.
Covid-19 (COrona VIrus Disease, 2019)
The Covid-19 virus has had far-reaching effects. As of March 2020, New York eased up on ebike restrictions specifically to address the need for more food delivery services. For any worker, it would be horrifically expensive to actually live in Manhattan, so the majority live within commuter-train distance, and ride into work each day. With tourists staying home during the Covid-19 pandemic, and few people going to restaurants (when it was even legal to do so), there was an explosion of food delivery orders from grocery stores as well as cooked food from restaurants.
This made New York realize where they had shot themselves in the foot. There had been a long-simmering battle between New York cops and food delivery drivers who were using stealthy ebikes to ride a delivery bike 12-hours a day. Before ebikes were invented, New York was the center of the “fixie” culture. The name refers to a fixed-gear bike (a single-speed bike with no freewheel).
Manhattan is very flat, so bike delivery riders didn’t need gears, and even if they had a few gears, a malfunctioning derailleur could make the bike inoperable. No gears, because simple is reliable. Plus leaving a bike outside makes it easy to steal, even if it has a lock on it. As a result, delivery bikes were very simple and and very light, so they could be carried inside. Of course, as soon as ebikes were invented, a few delivery guys started to try to get away with using them…
If a regular bicycle was involved in a 20-mph crash, its not good, but…if an ebike is involved in a 20-mph crash?…all of a sudden its a public hazard. Laws were passed, and the police cracked down. It’s bad enough if the police hurt food delivery, but they even made the sale of ebikes illegal. That’s just nuts. What if I want to buy one, but then take it with me to some trail in Connecticut on the weekend? No, New York said, you can’t ride them here, and you can’t even sell them here. Or…did they?
Chris already had a working ebike store, but thought that eventually a shop near Manhattan would be positioned well for some eventual easing of rules to a more reasonable state of affairs. It was risky. However, the actual law concerning ebikes addressed the use of hand-throttles. European format ebikes rarely allow hand-throttles, and their reasoning in the past is that a lack of hand-throttle leaves the rider ready and able to apply the brakes with one or both hands. This means that EU-format ebikes use a Pedal-Assist-Sensor, or PAS. You only get power to the wheel if you are also pedaling.
However, an old friend said that if a cop decides he is going to arrest you…even if you win in court later, you aren’t getting out of the ride to jail, with the free picture and fingerprinting. Soon after opening a Propel shop in Brooklyn, Chris was cited with a $25,000 fine (again, not a typo). This story is great, but Chris does a better job of telling it, so check out this 13-minute video he just published.
Written by Ron/spinningmagnets, April 2021
This last year I’ve seen a boom in motorcycle conversions to electric drives, and after scanning the components used, it looks like the Quan Shun / QS cylindrical motors seemed to be popular, so let’s take a dig into what’s behind this trend.
“Twin Spar” Frames
One of the reasons there are so many electric drive conversions for dirt bikes right now, is that a lot of them are getting to an age where they are showing up on the used market very cheap, with a worn engine that needs a rebuild.
Ten to fifteen years ago, there was a major shift in most frame designs that accidentally benefitted electric conversions. The biggest change is that the big frame designers evolved from a single back-bone top-tube, to a split top called a “twin spar”. The great thing about this change is that it makes the design of a large single battery pack MUCH easier.
Previously, I have seen custom battery packs used that were very oddly-shaped, or were made from two or three sections (click here) that were plugged together. Regardless of what construction configuration was used, it always ended up with the pack being smaller than would otherwise be possible.
This is no small consideration, since dirt bike E-conversions will never have super long range under the best conditions. In the pic below we wrote a feature on Jeremy Webbs Husqvarna conversion (click here), and the battery shape and size really impressed me. At the time, I thought a frame like this might be rare or hard to find, but I recently started digging around, and…all the big manufacturers are now using the twin spar design.
We first heard about the Quan Shun company, it was for their large hubmotors back in August of 2015 (click here). Their popular model is the one that uses a stator that is 205mm in diameter, and 50mm wide. In China, the motors are measured on their side, and the “H” is the magnet “Height”. So it’s called the 205/50H. It is labeled as a “3000W” motor (*continuous), but hot rodders have often taken it to a 7,200W as a temporary peak (72V / 100A)
The pic above is one of their scooter conversion hubs, which do not take spokes for a bicycle/motorcycle. This one mounts the rim directly to the hubmotor to seat a 10-inch ID scooter tire (click here). China has been cracking down on the pollution in their largest crowded cities, so there has been a tidal wave of electric conversions, instead of junking your vehicle and buying a new one. The QS web-catalogue can be found by clicking here.
I occasionally get messages, and a frequent question has been “Which frame makes a good conversion?”, or “I have a brand-X frame, so what motor and controller would work?”. I have lots of hands-on experience with electric bicycles and kits, but…E-motorcycles are hard to find for a test ride.
My typical response has been to recommend a turn-key E-motorcycle. For a youth or pre-teen, the Kuberg seems to be popular (click here). For a mid-sized adult dirt bike, the Sur-Ron has taken the US by storm over this last year (click here). For a full-sized off-roader, Zero has the FX model (click here).
My previous reluctance to recommend any particular motorcycle frame or conversion kit wasn’t simply due to my lack of experience at that, it was also because of the limited choices of components, and the fact I haven’t seen any owners publishing the data since the last time I looked into this. I don’t want to say that something is popular, only to find out it has a lot of customer complaints. Sometimes a new product is selling well simply because it has no competition.
So…I recently started collecting data on motorcycle conversions. Which motor, controller, and battery types were the most common? There has been every possible motor you can imagine squeezed-in to whatever frame a builder could get for cheap. However…I noticed that recently, quite a few were using QS cylindrical motors. So far…the QS builders seem to be happy with them.
I typically recommend using the largest motor that would fit for a given conversion. For many builders, the sweet-spot for a small dirt bike is the QS 3000W (3kW) 138/70H V3 with a 45T/19T gear reduction built into the case (2.3:1 reduction).
The 3kW motor above is the popular 138/70H , and this motor is also available without a geared-reduction, using just the standard splined shaft. But if you want more torque, they also have the 138/90H motor, rated at 4kW continuous. However, the larger 90H does not yet have any optional reduction for it.
According to the builder: “…It’s the first time I rode an electric bike and I am hooked! It so much fun…It hits very hard, maybe I will need to de-tune it a bit…weight is 255 lbs, but it feels way lighter than that when moving. Feels lighter than my KTM….I have a lot of tuning to do, but as of now it feels like a 125 but with more torque down low…“
In the pic above, the 138/90H has a DIY chain reduction on the right side, which drives a jackshaft to the 13T final-drive sprocket on the left.
The 138/90H motors’ external dimensions are 217mm length, and a 162mm diameter (*8.5-inches long by 6.4). The shaft protrudes an additional 14mm (1.4-inches). It’s unlikely that anyone would fit a larger motor onto a dirt bike frame, but…if you want to look that up, the next size up in the QS catalogue is the 8,000W 171/100H (these numbers are the rotor size in millimeters, and in theory the bigger motor would be roughly 33mm larger in diameter, and 10mm longer).
On the QS website, they will sell you a 138/90H motor paired with any one of four controller brands. The APT, Kelly, Sabvoton, or…models from their partner Votol. Most builds I’ve seen use the Votol due to it’s affordable price, and they have two amp levels available that are tuned to work well with the two 138-series motors. The EM-150, and the EM 200.
For this controller series, the numbers in their name are the amps of continuous current, and the US retailer stated that the EM-200 can provide a temporary peak of 250A. The EM-150 is well-suited to the 138/70H, and the EM-200 is a good size for the 138/90H motor.
In the pic above, the EM-200 uses 24 MOSFET’s, which would be 8.4 continuous amps per FET. The dimensions of the baseplate are 285mm X 168mm, the thickness is 70mm (11.2 X 6.6 inches, 2.8 ” thick). The capacitors are maxed at 90V, and a 72V / 20S battery charged to 4.2V per cell is 84V, so there is a 6V safety margin for voltage spikes.
If you go to the Votol.net website, they also have controllers that are smaller like the EM-30 / EM-50 / EM-100, and aso…an even larger model, the EM-300
To be fair, I did find a video of a QS 138 motor being driven by a Kelly QS-KLS-8430H controller, seen here:
When researching for more info on QS motors and their dirt bike conversions, I found several builds that had gotten their swap kit from “Electro and Company”, who goes by the web-avatar “Electro-Braap” on Facebook. They are located in Oakley, which is 51 miles east of San Francisco. You can get your questions answered on their Electro Braap facebook page (click here), or their Electro & Company website (click here).
Austin at Electro Braap verified that he keeps 10T and 11T sprockets in stock for the motors they sell.
Whether you buy a turnkey dirt bike, or build an E-conversion, this last year has seen a huge bump in the number of people who have discovered electric dirt bikes. They are so quiet that they can be ridden anywhere, and their low-end torque numbers are phenomenal (don’t measure the electric motors’ power by the kW/horsepower, it doesn’t scale).
All that matters is the wheel-torque. To see what I mean, check out this video of a Kawasaki KX350 conversion with the QS 138/70H motor.
Riders of the popular Sur-Ron have said that it has the nimble lightness of a 125cc dirt bike, with the wheel-torque of a 250cc. The listed electric motorcycle horsepower numbers of all brands are ridiculously low for technical reasons, so the way to compare gassers to electrics is with the wheel-torque. That is the number one performance area where experienced riders say the E-motorcycles are clearly better.
When the flag drops, there is no clutch or shifting, and the electrics have full torque from the first RPM…Of course the low noise, and not needing a frequent top-end valve rebuild, clutch replacements, and oil changes is nice, too!
How much can these motors do?
I found this youtube linked below, and at one point you can freeze the image and blow it up to see it’s a QS 138/70H “72V 3000W” with no gear reduction (he has small-diameter tires, so that’s OK). This isn’t a dirt bike, but I’m including this because he runs it at 96V and allows 300A peaks!
Brushless motors are very tolerant of a wide range of voltages, and the downside of adding more volts is finding the room to add more cells in series to raise the volts. The big amps using the Votol EM-300 have to be taken VERY carefully to ensure you don’t overheat the motor. In fact I would definitely pull the motor apart and add a heat sensor for a digital readout to warn me when I need to let things cool down a bit.
96V x 300A is 28,000W, which is a little nuts, but…an engineer from Zero told me that if you have enough power, the vehicle will accelerate so fast that you are not drawing max amps for very long.
The DIY battery pack is made from 26 BMW hybrid cells from a salvaged car wreck (26S). Full electric cars have a large pack so they can use cells that are optimized for long range, but a hybrid car has a gasoline engine along with a small battery pack, so the small pack must be able to provide full performance, which means hybrid cells are optimized for high amps without overheating.
The 96V of the pack is the nominal average voltage, and when fully charged to 4.1V per cell, its 107V, which is right at the edge of the voltage that the Votol controllers can use.
And the Facebook page for Electro Braap can be found here
And their company site is Electro & Company, (click here)
HEY!…If you liked the videos of E-conversions, here are 80 of them from the Electro Braap youtube channel (click here)
Written by Ron/spinningmagnets, April 2021