Heavy cargo hauler project

[amberwolf]

Amberwolf: A mid-drive system might end up being needed. I am not too keen on the stokemonkey system, as I'd prefer a freewheel between the motor and the pedals (or I'd go nuts). The Urbancommuter setup spinningmagnets mentioned in one of his cargo posts seems like a good contender, but I'd need to tweak it for a non-longtail system. Hiding it inside the cargo box would be ultra stealthy. Hmmmm.

One way of making a freewheel between motor and pedals is to use an IGH that has room on it's input shaft for two chainrings. One, for the pedals, gets a freewheel. The other, for the motor, does not need a freewheel but you could add one if there's room and need.

If you have no motor freewheel, and the IGH does not freewheel internally, then you also have regen braking available (which can indeed be helpful slowing from higher speeds with a load!). I recommend a perfectly-straight loop chainline for doing regen, without tensioners and such (using motor position to tension the chain).

Doing the above means two chainlines, one for pedals and one for motor, and the motor could be placed wherever is convenient for you and it and chain routing, but right in that triangle between your knees is still the best place, with a "box" around the whole thing for stealth and weather-proofing (and keeping bits of you out of the chains).

Other motors like powerchair motors could even be installed *under* the bike, below the cranks, with the output of the right-angle gearbox feeding the wheel from underneath. I did something like that for CrazyBike2's original road-tested drivetrain. But most of the cheap-and-easily-available-used ones are brushed motors, and most of the brushless ones I've run across are huge direct-drive monsters weighing 20-30lbs! (I have one of these, too, which will probably be tested on the new bike driving an IGH).


There are lots of ways to do a middrive.


I've also used a hubmotor on the front wheel of CrazyBike2, and I'd say that it's been much more reliable than my built-from-junk middrive, but if I were to go back and build the powerchair middrive *correctly*, I would probably see equal reliability and notably higher efficiency (given the losses in brushed vs brushless) in my frequent start-stop usage on the heavy bike, especially when cargo hauling (commonly multiple large dog food bags, but often enough other heavy and awkward things).

Fortunatley I don't have to deal with hills *and* cargo, but if I did, a middrive is definitely the way I'd approach it.

 

[Cargo_Tom]

I like reliable

I am definately feeling the allure of Dogman's 9C2812 drive. If it ends up being slow on a 20" wheeI I could just up the voltage to compensate And if I need to fabricate fork mounts for a disk brake anyway, I might as well look into the possibility of using moped/motorbike parts from the get-go.

As for steer tube failure I reckon some kind of head tube reinforcement jig might help alleviate part of the stress above the crown. In the event of a failure I'd get an opportunity to completely redesign the front end.

As for mid-drive options I definitely need to educate myself some more. Cheers guys for the appetizers.

-Tom

 

[amberwolf]

And if you want a big disc brake on a front 9C, you can do that, too:
http://www.endless-sphere.com/forums/viewtopic.php?f=2&t=28410
I don't have mine completed just yet, but I'm just short some bolts to fasten it on, and welding the caliper mounts to the fork, more or less.

Using a double-crown fork, like motorcycles use (and quite a few bikes) you can make failure of fork/steerer interconnect much less likely. Reinforcement plates between headtube and top/downtube like the Mongoose Basher has will also help that connection:

 

[dogman dan]

Sounds like as long as you build it slow enough, nobody is going to have a clue what your wattage is. In a 20" wheel, and 48v, you'll be nice and slow with a 2812.

2810 winding would likely hit your target speed at 48v. With a 48v 20 amp infineon controller on that you'd see real world wattage on your CA around 1200

I run my 2812 on 72v 20 amp controller in summer, when it's important to keep cool. 110 F is common. 1500w seen real world on the CA.

I consider a 1500w draw to be pretty much the practical continuous maximum for the 9 c motor, so 72v 20 amps should work nice. However, in winter, once I'll be riding in less than 85F, I switch to a 40 amp controller, and see about the same continuos wattage most of the time, but can have 2000-2500w if I'm climbing a steep enough hill. At the point I turn around in that vid, I get stopped by the rocky surface, and 20% grade. 72v 40 amps in that vid.

You likely know this basic thing already, but the simple way to think about it, is volts determines your motors top no load rpm. Wattage determines your torque. more watts more torque, regardless of the motors no load rpm. So you play with various voltage and amperage combinations with the chosen motor. Need more speed, add volts. But you may want to lessen amps if you add volts, to keep the wattage in a range the motor can tolerate longer. At some point, you cross a line and the motor just heats up crazy fast. Below 2000w, 9c motors seldom melt down very fast. Above 3500w I've melted a few.

 

[dogman dan]

I was just browsing Cellmans Emissions Free website, looking at the slower version of the Mac 500w kit. That is the 10 T motor.

In the text, he mentions that there is a 12T version, even slower winding. Sooo, perhaps a Mac could be a good option too, if cellman can supply you a 12T motor.

Suddenly I'm more interested than ever in trying a Mac. Of course, I have three slow 9c's to use for awhile.

But if getting your hands on a slow dd motor is not possible, I'd give this 12t a whirl. I think between the slow winding and the small wheel your gears would not be a problem. You wouldn't be overheating the motor so bad that the gears melt.

At the bottom of the page, some data. Looks like 12T 20" wheel, 48v would be pretty dang slow. http://www.emissions-free.com/catalog/i1.html

 

[kmxtornado]

For budgetary reasons, perhaps you can phase the work. Start off with either the IGH or the motor and let that determine the specs of the other. If you're looking for a motor, I'm guessing you're convinced you need one. So perhaps that's the better way to start. Get the motor, then get the IGH if needed. Personally, I'd go the other way around though b/c for me, the bike's drivetrain comes first and anything additional is extra. But after all, you are in a e-bike forum so maybe go with the motor first? Dunno. Either way though, you'll have an improved bike you can smile about.

 

[spinningmagnets]

double crown...reinforcement plates between headtube and top/downtube

Amberwolf has done a LOT more steel welding than me (I have not tried aluminum brazing yet), and I agree with this, if you use a front hub and regen. Adding a reinforcement plate to the side of a joint is a common strengthening solution, and any fabrication shop would be familiar with this and easily able to add these for you. I think it would be called a "gusset".

I also agree regen is quite valuable as an added brake, because it keeps the wheel brakes cool on a long downhill with a heavy load (more important than the small amount of re-captured battery charge)

 

[John in CR]

A 10% grade at 15mph with 650lbs requires about 2200W, so you can forget about using any single hubmotor with that load, at least if it's used in wheel. Hubmotors make great mid-drives due to their silence, and you'll want a powerful one for that kind of load, so you can also forget about the slowest wind motors. Get a faster one (because they're capable of greater power with our reliable voltage limit of 100V), and the highest voltage you can afford for better cooling due to the higher rpm and lower current requirements, which means more affordable and dependable controllers. Just gear it down to the speed you want, and you'll get billygoat hill climbing too even with a fixed gear ratio that will also enable regen. You'll want regen braking with that kind of load or you'll end up having to go motorcycle brakes too, more weight and expense, but regen is free and weighs nothing.

Since you're limited enough in speed, you'll actually want to gear for peak power on that steepest hill and be able to do the speed limit up hill. That's going to mean setting up for about 25mph top speed to achieve peak power and your 15mph speed up hills. Then use a speed selector switch to limit speed to the legal limit on the flats.

Think of your bike like a car, and riding down the highway at full throttle and speed. What happens when you come to a hill? You slow down of course. Unfortunately for almost all ebikers that's exactly how they are forced to ride due to their low power over-geared rigs. They ride around at full throttle, and speed drops way off as soon as they encounter a hill. If you gear your bike for 25kph max speed the same will happen to you. With a proper set up and sufficient power it doesn't have to be that way. I rarely use full throttle except during acceleration and on steep hills, so when I encounter big headwinds or moderate grades, I just give it more throttle to maintain the speed I want.

Also, forget the geared hubmotors with that kind of load. You'll inevitably hit it with extreme shock loads that will the destroy the plastic teeth on the planetary gears. You're on the right track thinking about torque, but you can't leave power out of the equation, because to maintain a certain speed up hill it requires a given power for a given load, not just torque. You said yourself the bike isn't a good performer at very low speeds anyway, so 25kph up the steepest hill with max load is your appropriate target. Here's a good calculator to help in that regard. If you fill in the speed, then it spits out the power requirements. http://www.kreuzotter.de/english/espeed.htm

Good luck with it. I can't wait to see your bak fiets.

John

 

[Cargo_Tom]

Auraslip: That is such a screwball idea But instead of a sticker how about getting a nice big CNC engraving in the hub cover saying “250W”?. Esp if subsequently painted white on a black hub

Dogman, amberwold, KmxTornado, Spinningmagnets & John_in_CR: Thank you for sharing your real-world experiences. To sum up a hub solution would keep me in the 1-2kW power band, at the lower end if opting for a planet gear version. These would struggle or conk out at inclines above 5% with my target load. On the plus side would be good slow speed handling, simplicity of installation and maintenance (if kept cool).

A mid-drive solution would get me the handling I desire through gearing design (How about a rear hub with cassette and derailleur as a mid drive solution, for ad hoc adjustment?). If designed above 2kW I could get real hill climbing ability and power surplus on demand. The downside would be less stealth, a bigger initial cost and harder repercussions, should I get busted.

One particular concern is slow speed handling. Bicycle infrastructure and numbers are quite different from in the states. I will not be sharing the road with cars, but rather sharing bike tracks with unpowered cyclists. I am also living in a dense urban environment, so there will be lots of stops and starts at traffic lights.

Here is a sampling of what it feels like to be a bicyclist round this neck of the woods:
youtube link (Ignore the hippie propaganda)

With that in mind what would you advise?

I am currently leaning towards a hub solution for the following reasons:

• I can halve the incline on the 300ft hill by taking a small detour. My local vicinity is flat as a pancake.
• Typical total load will be in the 500-550lbs range.
• Front wheel drive should alleviate slow speed wobble as well as winter handling concerns. I'd prefer the added ebike component weight to be in the front anyway, as the bakfiets style is massively unbalanced towards the rear to begin with.
• Easy regen for braking help
• This wont be my last ebike. Starting simple and tweaking from there will give me the satisfaction of experiencing shortcomings myself and working from there.

Cheers
- Tom

 

[dogman dan]

Ok, finally get it about the low speed handling issue. Front hub will eliminate that.

John is really educated about motor performance, and straigtens out me on motor theory constantly. But I really do believe this to be true, if you want to ride slow, you need the slow winding ( on a hubmotor) or low gearing to keep the motor in it's less inefficient zone. John is correct that you will need a beefier motor, to pull that load up 10% at 15 mph. What I had in mind was a motor choice that could survive that climb at 8 mph. or perhaps as slow as 5 mph. John is right, that a really powerful motor with a faster winding would have more avaliable power. Mabye for the next build. Cromotor, and 5-10,000 w 8) As I was understanding it, you wanted about 15 mph top speed on the flat to blend in, and ability to get up the hill with no particular speed in mind. Just get up it without melting down.

I think the slow 9c motor will do you, with 1200-1500w. You'll get up that 10% grade with the bucket empty, but I bet you'll much prefer the long way when the bucket is full. Nice thing about an ebike, the long cut becomes quite ok. Plus the hill is not so tall, so you can get up it in a short enough time to keep the motor cool enough.

For sure though, a hubmotor is not choice one for this application. But a slow hubmotor will be able to do the climb slower without as much overheating. With 1000-1500w, you won't climb that hill fast. If you fail and melt the motor, you should be only out the cost of the motor itself.

But if you can source a more powerful motor in a slow winding, go for it. Not the new clyte HT, but the old 5305, or maybe even 5304. An x5 clyte, you could routinely feed more watts to, and get up the hill nicer that way. Keep watching the for sale used section, a good x5 goes fast now that you can't buy one new.

 

[Cargo_Tom]

Cheers Dogman. I reckon I'm getting my head wrapped around the motor perspective now.

I haven't ordered anything yet, but having itchy trigger fingers I took the cargo bed off this afternoon. Compared to the initial pic I have installed apehanger handle bars and a split ergo seat.

I'd like to pull the seating position back abit, so I will be ordering an angled seat post. I'm pretty sure the apehanger will get in the way of the kid seating spot when turning, so I will likely replace the apehanger with a BMX style handle bar on a taller stem.

The bike was born with a slightly larger rear wheel, so I will also be looking into getting a taller rear tire for some extra ground clearance.

Cheers for all the input guys. The thinking cap is getting nice and toasty

 

[John in CR]

Low speed handling is all about the geometry not the drive wheel, unless you're just talking about how the front wheel would pull you around turns, which would be a nice thing on a long bike like the one pictured most recently above. When I think about low speed handling being poor, I think along the lines of my cargo/kid carrier, which has almost 4" of trail. The result is rock solid tracking at high speeds, but below 10kph or so it's difficult to keep dead straight for any meaningful distance. In your first post you mentioned low speed issue, but the bike is great at speed, so it sounds like the same geometry created issue that no motor placement can fix.

If you can avoid the steepest hill with max loads, then one of those slow wind motors in a 16" wheel will probably work out fine with a high enough voltage to hit your speed target as long as you understand the 25kph on the flats will turn into about 15kph on the hills. On flat ground is where hub motors really shine with silence and durability with almost any load. You could put a few adults in the "box" and a DD hubbie won't care on the flats. Acceleration will of course decrease, but it will cruise just fine.

In fact, yesterday I had my daughter on the back of my bike for a flat and downhill only return home. Sure my bike was set at 6kw peak at the time, but the motor was looking at a load of right at 480lbs, yet we were zipping around passing cars like any normal load ride, only a bit less acceleration.

 

[dogman dan]

Can't imagine that monster every handling "good" If you load enough weight on the front basket of a bike, you get some wobbly feeling at low speed. Had that on a paper route bike once.

I think the motor pulling on the forks may help some. Only one way to find out.
Perhaps there is some play in the steering linkages that doesn't help things any? A motor pulling should dampen shimmy some.

I wonder if some of the shimmy on the front is caused by frame flex? Mabye gusset the head tube, with a gusset 2 feet long creating a box beam?

 

[999zip999]

With motor power and the ease and speed and less effort to peddle taking the long flat won't take long at all. I like in a up hill or down hill and allways take the long haul that knot steep to keep it easy on my 10c konion's ( 6c ) maybe 60v off the charger. 48v a must have or more. Ping has a 60v pack ?

 

[Chalo]

Can't imagine that monster every handling "good" If you load enough weight on the front basket of a bike, you get some wobbly feeling at low speed. Had that on a paper route bike once.

When a load is coupled to the steering assembly, it's a lot harder to manage than when it's mounted on the frame. "Real" cargo bikes carry loads on the frame, even in front. For example:

Long John style bikes like the OP's are not as sweet-handling as normal bikes, but they are a lot better when loaded than a normal bike with a huge load hanging from the handlebars.

Chalo

 

[dogman dan]

Very true. I would think with cargo Toms bike loaded up, anything below 5 mph starts to be like a track stand. Get the gyros going and much better. Gotta be better to pull that weight than push it. Ever wheebarrowed cement with a really floppy wheel on the barrow? No es bueno.

 

[Cargo_Tom]

The handling issue is mainly weight distribution I reckon. Just look at the thing. It becomes fairly solid once it gets loaded with cargo. Unpowered the trick is then getting it fired up, and that experience is wobbly until you get a little speed on it.

A gusset would indeed help with stiffness and could be wrapped around the head tube for extra joint strength. I'll look into it

Well, having poked around some more in the ebike technical subforum, I ran across the "2WD faq" thread, and apart from a treasure trove of motor theory (thanks guys) I ran across the tip that the 9Cont 2810 could be guesstimated in the simulator by taking the 2805, and halving the voltage while doubling the amps.

If this approach works then It suddenly becomes possible to look at different setups for both the 2810 and 2812.

Playing around with those a few things become clear. At heavy loads the 10% route is just not doable. The 5% detour works ok for the 2812 by cutting max speed in half.

The efficiency sweet spot for 25kph on flats fully loaded is 64 volts at WOT. So a 72 volt system might be in order.

I reckon getting a 40 amp controller and using CA limiting would be the way to go for poking around with torque/range tradeoffs in practice.

Looking more and more like I need to pester Methods in the near future
-Tom

EDIT: and just to clarify. I've not been consistent in sticking to either rim size or total wheel diameter when mentioning wheels.

The (Moped) rims are 16x2" and 19x2" for front and back respectively
With tires that then becomes 20" and 23" wheel diameter respectively.
Ive amended the initial post to clear up the confusion. My apologies.

 

[999zip999]

CA is a must have with a controller that is already setup for it 6 pin plug.

 

[SamTexas]

Playing around with those a few things become clear. At heavy loads the 10% route is just not doable...

I'm glad you find that out now even if it's just on the simulator. You might want to overdesign, overplan everything by 10 to 20% to make sure that reality closely match simulation. Have fun.

 

[llile]

Spinningmagnets hacked mid-drive is ok, but why not just buy a stokemonkey mid-drive that is all set up? I have one hacked onto a recumbent, and have never had any trouble with any grade of any description. We have some short sections of 20% grade around here (not kidding) that the stokemonkey hauls me right up in a not-very low gear. Set it up with any gear ratio you want and walk right up everest, if you so desire.

--Lawrence

 

[John in CR]

Cargo_Tom,

You catch on quick. You're on the right track and coming up with pretty solid answers.

I'll start construction on a new cargo/kid hauler soon with the pilot in back, but must mix with traffic and have no restrictive rules. It will have a hubbie in front and a mid-drive rear with a narrow aerodynamic pod for kids to sit inline. Their rocket pod will have a removable canopy to keep them dry and hopefully shield me some too in rain, so it will be mounted higher with their feet extending forward above the front wheel. Plenty of batteries will go in the inline bench seat to maintain more forward weight, though I'll have to make it quite aero to side winds due to the higher "box", which is my one real unknow regarding how that will affect handling. Monotracer handles big side winds with a quite large area to side winds, so hopefully my test rides won't end in disaster. :lol:

John

 

[dogman dan]

I still suck at interpreting the sim, and have years of just trying it and finding out on the road.

Clarify please why you think the 2810 wont cut it on the 10%. Just not capable of enough watts to lift that weight up that grade? Overheats too fast? What specificly is the problem? Bear in mind you get to pedal, which greatly helps a wheel get up a hill, and at 300' of vertical, your hub is very likely to have enough thermal mass to absorb the heat for about 5 min.

Of course, a smart man would just take the 5% route with a load. I just want to learn what you just did.

I know for a fact that a 2810 can tolerate climbing hills at 10 mph. And the 2812 perhaps 8 mph. Regular hubmotors best kept above 15 mph. And that data is for 26" rim. So in 20" if you can maintain 6-7 mph up the hill, it may do it fine for that distance. I wouldn't hesitate to offer a 9c a lot of power for a short time. Even if lots of it is making heat, you can do it for a short time. So maybe 72v 40 amps is needed, but for the majority of the ride, you'll still have decent low speed throttle controll with those slow motors, and only use low wattage except for the hill part of the ride.

Too bad the 5305 is hard to get now. But playing with 9c's will at least be affordable.

 

[joe tomten]

http://clevercycles.com/products/stokemonkey/

Stokemonkey overview
Note: As of October 2011, Stokemonkey is unavailable. Facing a near doubling of component costs and minimum order quantities from our suppliers, we are evaluating how to make Stokemonkey available again. We cannot respond to requests for status updates; we will publish all news. Thanks for your understanding.

 

[auraslip]

I think the simulator downplays a very important fact about a hub motor setup: Efficiency is derived from rpm per volt, and the voltage the motor sees is a function of the PWM from the controller. With full throttle the motor will see full voltage. If you limit the speed via a 3 speed switch or throttle modulation, you can lower the "sweet spot" of efficiency to what ever you want.

Also, I'm confused about the slow wind motors. They have more resistance, and thus draw less current right? So they have less power available?

I don't think the 10% grade is doable.... not with out a bigger hub motor and a high voltage battery pack.

 

[dogman dan]

It's complicated, and I only half understand. So let me just say this. The slower winds, when climbing, seem to draw less power. I see this on the CA. Is it less power to the wheel? It does not feel like it to the rider, hence the myth that the slow winding has more torque. My best guess is that the slow winding is wasting less power, and could have equal or even more power going to the ground than a stalled fast motor.

When on the flat, the start up seems nicer on slow windings. Again, I think it's just leaving the stalling, power wasting part of the curve a few feet sooner, which makes the start spike less big.