Lightweight folder - build thread

[Jeremy Harris]

Two questions:

Where are you getting these Kt numbers for any of these motors? It seems very useful, but I have not seen it anywhere else.

You can derive Kt from Kv, as it's a fixed relationship. If you work out the effective Kv (the rpm per volt factor) for the motor, then Kt is derived from Kt = 60 / (2 x PI x Kv) (in N-m/A). For example, if you take the 328 rpm at 36V Q100 motor, then the effective Kv (including gearing) is 328 / 36 = 9.11 rpm/V and the effective Kt (including gearing) is 60 / (2 x 3.14159 x 9.11) = 1.048 N-m/A. This means that if you put 1 A through the motor windings, then you will get 1.048 N-m (about 0.773 lb-ft) of torque at the wheel.

I'm very tempted to build a Swift. I'm a reasonably good cyclist, but an even better eater. Given my own weight (120kg), the hilly terrain around here and the fact that I may need to haul it up a couple flights of stairs in the train stations I expect to have to trade some speed for torque and some range for weight. I plan to pedal hard on climbs, the assist is more to keep average speeds up for commuting. I'd like to be able to sustain 32kph with about 100w of pedaling and I'd iike to be be able to climb 12% for up to 1/2 mile and 6% indefinately with about 200w of pedaling. I can deal with HK lipo but would prefer to keep it in the range 10-14s so that I can charge without taking the pack apart (Hyperion 1420). I've been all over the simulator and the BPM can do this. But, it is pretty heavy and I'd like avoid having so much weight in the ends of the bike .I think a rear drive might be better, but then there is the whole issue of getting enough gearing. The stock Swift setup seems pretty reasonable, but it is cassette and there are not many acceptable 11 tooth freewheels. I'm a little leary of the Q100 because of the bolt together axle setup. So what motor?

BPM rear?

SWXH rear?

SWXK front?

Comments please.

I reckon you probably need to sacrifice a bit on motor weight and go for the CST motor, fitted with a decent cassette. This will give you the ability to fit a wide gear range, wider than you can get with a freewheel. The additional weight of the motor would probably be offset by the better gearing you could get, as you're a keen pedaller. The only problem would be the one of carrying the bike up stairs, but perhaps you could shave weight off elsewhere to make this easier. Overall, the CST fitted with a cassette would be a little over 2kg heavier than a Q100 fitted with a freewheel, but would be able to deliver a higher torque for longer before overheating, which might be important as you have a lot of hills to climb.

The SWXH is sort of midway between the CST (which is really a BPM internally, I think) and the Q100. I'm pleased with mine on the other folding bike I have, as it's quiet and reliable, but then I only weigh 82kg and I don't often ride it on hills. I'm not sure how the gears in either the SWXH or Q100 would cope with the greater weight and a lot of hill climbing, but would guess that they might start to complain before too long.

The downside of the CST motor is the high price. They are charging a lot extra over the price of the BPM just for fitting the cassette splines.

 

[Alan B]

Is the CST the only motor around with cassette splines?

 

[-dg]

Two questions:

Where are you getting these Kt numbers for any of these motors? It seems very useful, but I have not seen it anywhere else.

You can derive Kt from Kv, as it's a fixed relationship. If you work out the effective Kv (the rpm per volt factor) for the motor, then Kt is derived from Kt = 60 / (2 x PI x Kv) (in N-m/A). For example, if you take the 328 rpm at 36V Q100 motor, then the effective Kv (including gearing) is 328 / 36 = 9.11 rpm/V and the effective Kt (including gearing) is 60 / (2 x 3.14159 x 9.11) = 1.048 N-m/A. This means that if you put 1 A through the motor windings, then you will get 1.048 N-m (about 0.773 lb-ft) of torque at the wheel.

Seems counter intuitive ... [ponder... ponder ... bu bu but what about size ... that can't be right ... hmmm ... well power = V*A ... hmmm. I just don't want to believe this .... power = torque*rpm .... er, ok, I guess ....]. Apparently it has to be true or the math does not work.

So with an ideal motor with no losses and no saturation then for a given Kv then any motor size will produce the same torque given the same current. More torque requires more current and size is just a proxy for current handling capability. In the real world flux saturation comes in here somewhere, and resistance change due to heat, and mechanical failure, but ignoring those, the difference with a larger motor is that it can handle more sustained current. So you could use a SWXU at 40A and it would climb as well as the same RPM BPM until it heated up or failed. I still feel like there is something else missing here.

Any idea how close to saturation these motors run?

Can any confidence be placed in these "201" and "328" number from GBK/BMS battery? Are those no load numbers, or loaded numbers, or do we simply not know?

 

[Jeremy Harris]

Seems counter intuitive ... [ponder... ponder ... bu bu but what about size ... that can't be right ... hmmm ... well power = V*A ... hmmm. I just don't want to believe this .... power = torque*rpm .... er, ok, I guess ....]. Apparently it has to be true or the math does not work.

So with an ideal motor with no losses and no saturation then for a given Kv then any motor size will produce the same torque given the same current. More torque requires more current and size is just a proxy for current handling capability. In the real world flux saturation comes in here somewhere, and resistance change due to heat, and mechanical failure, but ignoring those, the difference with a larger motor is that it can handle more sustained current. So you could use a SWXU at 40A and it would climb as well as the same RPM BPM until it heated up or failed. I still feel like there is something else missing here.

Any idea how close to saturation these motors run?

Can any confidence be placed in these "201" and "328" number from GBK/BMS battery? Are those no load numbers, or loaded numbers, or do we simply not know?

These calcs take no account of the motor power rating, so don't tell you if a given motor will burn out if used at a particularly high current. They do tell you the torque the motor will deliver at any current, but realistically you have to set an upper bound on the current you allow through the motor, using the controller current limit, to prevent overheating and damage. Most of these motors won't self-limit, they will happily draw pretty much whatever current the torque load demands unless the controller limits it.

The smaller motors, like the Q100, are less able to tolerate continuous high current than the larger motors, like the BPM/CST. This is largely down to size and the ability of the motor to shed heat, but a part of it is because the smaller motor is less efficient at high current and generates more heat. I've not seen a hard limit stated for current, but my gut feeling is that 20A is probably maximum for the Q100, maybe 15A if you're going to ride up long hills. The BPM/CST might take 30A, maybe 40A for short periods. The SW motors get warm on 25A, but will take 30A as long as it's not sustained for long. I've been running an SW rear motor on 30A for a couple of years and it's fine, but I don't do a lot of steep hills with it and know that it does get warm pretty quickly whenever I do climb a steep hill at that current.

I don't think that magnetic saturation is a particular issue, as I suspect the motor would overheat from I²R losses before that point was reached. Overheating is the major issue when running at high current.

The accuracy of those numbers is unknown, but they aren't far out, at least for the GBK supplied Q100 I have. I have a 36V 328rpm rear Q100 (GBK100) and it performs on the road as if that 328rpm figure is for a modestly loaded motor. I've not yet checked the no load wheel rpm to see if it gives a figure that's greater than the 9.11 rpm/V (derived from 328rpm/36V) but suspect it will. Be aware that the BMS Battery are completely clueless when it comes to the BPM motor spec. The BPM motors come in a wide range of different speeds, and ordering a particular rpm one from BMS is no guarantee at all that this is what you will get, plus the rpm figure they quote don't tally that well with those from the manufacturer, Suzhou Bafang. I ordered a Code 8 motor from them (the highest rpm version), double checked with them by email, and received confirmation that the motor they were sending was indeed a Code 8, yet the motor that arrived was a Code 10, a somewhat slower wind. There is a table on here somewhere that lists the various BPM motor winds and speeds.

 

[Jeremy Harris]

Is the CST the only motor around with cassette splines?

I believe so, Alan. It's a pity, as it would be really useful to have the ability to fit a wide range cassette on an ebike, especially one with smaller wheels where getting high enough gearing to pedal at speed is a problem.

 

[amberwolf]

No, the Fusin 1000W (peak) geared hub also has cassette splines:
http://www.endless-sphere.com/forums/posting.php?mode=reply&f=3&t=42133
http://www.endless-sphere.com/forums/viewtopic.php?f=3&t=38825

So far it has held up to all the stuff I've put it thru, for a bit over 500 miles. As soon as I have some time I need to build the tool to take the cover off and get pics of the insides, and then stick it on CrazyBIke2 just to see what happens to it.

 

[Jeremy Harris]

Good info AW, thanks. I'd not seen that Fusin also had a spline equipped motor. Hopefully more motor manufacturers will start doing this before too long.

 

[Erikjan]

http://cnebikes.en.alibaba.com/product/584722936-212862822/newest_wonderful_cassette_electric_hub_motor_36v_250w.html

I have this one.

I dont test it yet. Tomorrow i now more.

 

[-dg]

I reckon you probably need to sacrifice a bit on motor weight and go for the CST motor, fitted with a decent cassette. This will give you the ability to fit a wide gear range, wider than you can get with a freewheel. The additional weight of the motor would probably be offset by the better gearing you could get, as you're a keen pedaller. The only problem would be the one of carrying the bike up stairs, but perhaps you could shave weight off elsewhere to make this easier. Overall, the CST fitted with a cassette would be a little over 2kg heavier than a Q100 fitted with a freewheel, but would be able to deliver a higher torque for longer before overheating, which might be important as you have a lot of hills to climb.

The SWXH is sort of midway between the CST (which is really a BPM internally, I think) and the Q100. I'm pleased with mine on the other folding bike I have, as it's quiet and reliable, but then I only weigh 82kg and I don't often ride it on hills. I'm not sure how the gears in either the SWXH or Q100 would cope with the greater weight and a lot of hill climbing, but would guess that they might start to complain before too long.

I'd love to see an update on your Swift.

I'm closer to building one for myself. I built a 700c bike for my wife with a front SWXK on 12s with the 6fet controller bumped to 22 amps. Based on testing that and the way I plan to use the Swift I think that size will be enough. The BPM/CST seems like too much weight in the end of the bike to carry it easily. Plus, I have an 11-34 megarange 7sp freewheel that I could use so I'm pretty tempted by the SWXH. Another choice would be the SWXK in front, but I only see 201rpm winds available and that is too slow in a 20". I looked at the Q100 328 in the simulator and it did not really seem up to it. If you were building a second Swift, would you prefer the SWXH or the Q100 and why? I'm not talking about power handling so much as user experience. Thanks!

 

[motomech]

If I was to put, say, 20A through a Code 10 motor (Kt ~ 1.09 N-m/A) then I'd get (ignoring the small gear and frictional losses) a torque of around 21.8 N-m at the wheel.

If I was to put the same 20A through a code 16 motor (Kt ~ 1.79 N-m/A) then I'd get (again, ignoring the small gear and frictional losses) a torque of around 35.8 N-m at the wheel.
...
If you want to get even more extreme, then take a look at the little Q100 with the 328rpm wind. That has a Kt of around 1.06 N-m/A, so at the same 20A it will give a torque at the wheel of around 21.2 N-m, pretty much exactly the same as the code 10 BPM.

Two questions:

Where are you getting these Kt numbers for any of these motors? It seems very useful, but I have not seen it anywhere else.

I'm very tempted to build a Swift. I'm a reasonably good cyclist, but an even better eater. Given my own weight (120kg), the hilly terrain around here and the fact that I may need to haul it up a couple flights of stairs in the train stations I expect to have to trade some speed for torque and some range for weight. I plan to pedal hard on climbs, the assist is more to keep average speeds up for commuting. I'd like to be able to sustain 32kph with about 100w of pedaling and I'd iike to be be able to climb 12% for up to 1/2 mile and 6% indefinately with about 200w of pedaling. I can deal with HK lipo but would prefer to keep it in the range 10-14s so that I can charge without taking the pack apart (Hyperion 1420). I've been all over the simulator and the BPM can do this. But, it is pretty heavy and I'd like avoid having so much weight in the ends of the bike .I think a rear drive might be better, but then there is the whole issue of getting enough gearing. The stock Swift setup seems pretty reasonable, but it is cassette and there are not many acceptable 11 tooth freewheels. I'm a little leary of the Q100 because of the bolt together axle setup. So what motor?

BPM rear?

SWXH rear?

SWXK front?

Comments please.

I'm your size and I can tell you that to climb those sort of hills without busting a gut or cooking your motor, you would need to go up a size from the mini's to the 3.8 to 4.2Kg. geared midi's.
One thing nobody has mentioned is that the CST allows one to pass over the DNP 11T freewheel, which is a serious chuck of weight. A CST/light weight cassette would probably have the same weight as a mini with a DNP.
The obvious down-side to the CST is that one can buy 3 Q100's for the price of 1 Bafang. Plus it doesn't come pre-built with wheel.
What I would do if I were you, is run a pair of 328 Q100's for 2 WD. The cute is more or less silent when the amps are kept at or below 17A(12S). Both D8veh and I have runs these motors 2WD in mountain bikes, he, using both the 201's and 328's, me just 328's, and we both like this approach a lot. I don't even use torque arms and installing 2 mini's is easier than installing one midi.
two 328 Q100 in 20" wheels on 12S would do 24-25 mph(no pedal) and climb well(although not as well as a single midi).
14S would put the folder in the 27 - 28 mph range.
And just use a 7 to 9-speed DNP on the rear with the Cute.
Then you could auction that Mega-Range off on Ebay and finance the conversion with the proceeds

 

[-dg]

Well, I finally bought a Swift, so now I really do have to choose a motor. I'd like to keep it as bike like, lightweight and stealthy as possible given my size and local terrain. The idea would be that it would help me keep my average speed up, not replace pedaling.

My current concept is a front 201 rpm Bafang SWXU, 15A ku63 type controller and 2 x 7s 5Ah nano-tech tucked into the frame on top of the chainstays between the rear wheel and the seat tube. This gives 52 nominal volts and 260 watt-hours and if my calculations are right a no-load speed of 23 mph. If I can throttle back a little to cruise around 18 mph with 100 watts pedaling it looks like it should have 12ish miles of useful range and the whole system should come in around 4kg. If I need more range I can carry a 2cd pack somehow.

I see some opinions in favor of the Q100 from users of the q100, but I did not see anyone claiming to have directly compared the SWXU with the Q100. I'm not set on it, but given they are about the same weight my reasons for favoring the SWXU are:

- Having seen and used the SWXK I was impressed with the performance, design, and visible construction quality, particularly having the cable exit on the side plate and using larger bearings. The SWXU appears similar.

- In the online photos I've seen of the SWXK and Q100 internals, the Q100 looks crudely finished compared to the SWXK. I'm hoping the SWXU is as nice as the SWXK.

- I'm slightly concerned that the split axle of the Q100 might fail, or that making it reliable will mean the motor has more of its weight devoted to structure and less to magnets and copper.

- The Q100 sun gear looks like 11 teeth, too small to be efficient, and the double stack gears make for narrower gears and less room for the actual motor parts. The SWX gears are substantially larger particularly the sun gear.

- The simulator results for a 328rpm Q100 are marginal for my application as it overheats on even modest slopes. I'm hoping by using the slow wind SWXU to get more climbing torque and to get enough speed by using 14s .

I'd really like to hear from someone who has used the SWXU especially if you can directly compare it to the SWXK or Q100. I'd also like to hear from Q100 users as to the points above.

Comments on the general concept of the build are welcome too.

 

[jerrysimon]

This is mine

This is a very light assist (200W/10amp max, 1.4kg motor) but also very quick. This keeps the bikes carry weight to under 14kgs (battery in an attached bag/ruck sack). Full details here

http://www.bikeforums.net/showthread.php/83711-swift-folders?p=15107222&viewfull=1#post15107222.

Its worth reading that whole thread for ALL things Swift.

I would add this is not going to be any powerful hill climber, but on the flat it gets me up to 18-20mph quick and just gives that added help when its windy/horrible weather. For that kind of speed you also need to upgrade the 52T main chain ring to at least a 56T (as shown) to keep cadence comfortable. Again full details in that thread.

Jerry

 

[motomech]

Well, I finally bought a Swift, so now I really do have to choose a motor. I'd like to keep it as bike like, lightweight and stealthy as possible given my size and local terrain. The idea would be that it would help me keep my average speed up, not replace pedaling.

My current concept is a front 201 rpm Bafang SWXU, 15A ku63 type controller and 2 x 7s 5Ah nano-tech tucked into the frame on top of the chainstays between the rear wheel and the seat tube. This gives 52 nominal volts and 260 watt-hours and if my calculations are right a no-load speed of 23 mph. If I can throttle back a little to cruise around 18 mph with 100 watts pedaling it looks like it should have 12ish miles of useful range and the whole system should come in around 4kg. If I need more range I can carry a 2cd pack somehow.

I see some opinions in favor of the Q100 from users of the q100, but I did not see anyone claiming to have directly compared the SWXU with the Q100. I'm not set on it, but given they are about the same weight my reasons for favoring the SWXU are:

- Having seen and used the SWXK I was impressed with the performance, design, and visible construction quality, particularly having the cable exit on the side plate and using larger bearings. The SWXU appears similar.

- In the online photos I've seen of the SWXK and Q100 internals, the Q100 looks crudely finished compared to the SWXK. I'm hoping the SWXU is as nice as the SWXK.

- I'm slightly concerned that the split axle of the Q100 might fail, or that making it reliable will mean the motor has more of its weight devoted to structure and less to magnets and copper.

- The Q100 sun gear looks like 11 teeth, too small to be efficient, and the double stack gears make for narrower gears and less room for the actual motor parts. The SWX gears are substantially larger particularly the sun gear.

- The simulator results for a 328rpm Q100 are marginal for my application as it overheats on even modest slopes. I'm hoping by using the slow wind SWXU to get more climbing torque and to get enough speed by using 14s .

I'd really like to hear from someone who has used the SWXU especially if you can directly compare it to the SWXK or Q100. I'd also like to hear from Q100 users as to the points above.

Comments on the general concept of the build are welcome too.

Good, just get the bafang. This internal debate has reached the point where you just need to buy something and try it.
We're only talking $70 motors here, not a new car.
Let us know how it works out.

 

[-dg]

Good, just get the bafang. This internal debate has reached the point where you just need to buy something and try it.
We're only talking $70 motors here, not a new car.
Let us know how it works out.

That is what I will do. As far as I can tell no-one here has actual experience with the SWXU, so I'll be the guinea pig for this. Most likely either would be fine. And as you say, they are not that expensive. The only real thing I stand to lose on is the time spent building the wheel, and this is a hobby anyway.

I'll report back once I have it installed.

 

[-dg]

Here's mine.
...
http://www.bikeforums.net/showthread.php/83711-swift-folders?p=15107222&viewfull=1#post15107222.

Its worth reading that whole thread for ALL things Swift.

I would add this is not going to be any powerful hill climber, but on the flat it gets me up to 18-20mph quick and just gives that added help when its windy/horrible weather. For that kind of speed you also need to upgrade the 52T main chain ring to at least a 56T (as shown) to keep cadence comfortable. Again full details in that thread.

Nice looking bike, I like the mono-chromatic effect of all silver. I really really like the ones on the Swift gallery that are polished bare aluminum instead of paint. Alas, my second hand swift is blue. Not too bad though.

I have been reading that thread a few pages a day for weeks now, I'm up to page ninety-something. It was part of my motivation to try the Swift.

I'm planning on a slightly larger motor (Bafang SWXU) front drive with the battery inside the rear triangle on top of the chain stays between the wheel and the seat post. I'm looking at 52V (14sx1p) and 5 or 6 Ah. I'd really like to fit the 8Ah nanotechs in there, but they are too long. I noticed in your build thread that you are using even smaller batteries. How is that working out and how much range are you getting?

14kg is incredibly light. Does that include the battery? I'm trying to keep my under 18kg with racks and bags and fenders and battery and motor and it not be that easy.

 

[jerrysimon]

14kg is the carry weight (incl motor and controller fitted) of the bike which incl guards, big apple tyres fittings etc. The battery is in the rucksack which clips on the Klickfix fitting on the seat post. I have a choice of DIY 1kg 3Ah or 2kg 6Ah batteries.

I live in Cambridge so my 10 mile round commute is pretty flat. The 1kg battery will give me 10-15 miles and 2kg one 25-30 depending on how much assistance I put in. As explained this a light 1.4kg/200W motor, so certainly no hill climbing power hub monster.

Regards

Jerry