Weight Weenie Ebike

[LewTwo]

I made a point of getting a MetroCard and exploring the gains Houston has made towards public transit.

Personally, I am not impressed. Usually one of the big draws for the mass transit systems are the airports. I once looked at getting to the airport from my former work location via bus: 4 plus hours vs. 30 minutes via car. Houston mass transit also shutdowns the service and roll up the sidewalks at 6:00 P.M. My comparisons are to New York, Washington DC, London, Paris, Bangkok and Singapore, the latter being the best by far. To be fair in the other cities I mentioned there are huge numbers of people who use the transit systems which increase the level of service that can be provided. At the time I was not interested in how "bike friendly" these other services were and I can not imagine maneuvering a bicycle among some of those crowded conveyances.

You only need 5 miles of range, so get a battery that does that and no more. A battery in the 9-12 amp-hour range should suffice.

The LEEDS PBJ system has a 24 volt 2.6 Ah battery paired to a Bafang front hub (no idea which one). They claim a weight of 7 pounds and range of about 4 miles (15 -23 mph). I believe the controler is a simple on (full power)/off system. I can well imagine the battery, controller and display of that system integrated into a single box that clamps to the handle bar. That was kind of my concept starting point. I would like a tad more battery ... but not much.

Reference (PBJ E-BIKE KIT): http://www.e-bikerig.com/products/PBJ-e-bike-kit-samsung-li-ion-front-wheel.html

[arkmundi]

The LEEDS PBJ system has a 24 volt 2.6 Ah battery paired to a Bafang front hub (no idea which one). They claim a weight of 7 pounds and range of about 4 miles (15 -23 mph). I believe the controler is a simple on (full power)/off system. I can well imagine the battery, controller and display of that system integrated into a single box that clamps to the handle bar. That was kind of my concept starting point. I would like a tad more battery ... but not much.

Arrghh... you'll quickly find that inadequate I believe. A safety factor of 2x your expected is my rule-of-thumb. So you don't get anywhere near full depth-of-discharge. And factor in losses as the battery ages. But that 2.6 ah battery is too little to begin with. I'd go back, re-figure and plan again. Or not. Best of luck!

[Alex W]

Arrghh... you'll quickly find that inadequate I believe. A safety factor of 2x your expected is my rule-of-thumb. So you don't get anywhere near full depth-of-discharge. And factor in losses as the battery ages. But that 2.6 ah battery is too little to begin with. I'd go back, re-figure and plan again. Or not. Best of luck!

It depends heavily if you are building a pedal assist bike or a lightweight motorcycle that requires no human power.

For the former I think I'd get a range of 10 miles on that battery. 24v is too low for most hub motors though to do anything useful. On my 8fun hub motor I think I'd stop having assist around 14mph, which is trivial to ride at on flats.

However the heavy part of a light e-bike is the hub more than the battery. 36v4ah battery should be small and light enough to be the minimum being considered. It'll still be half the weight of the lightest hubs.

The other problem is all of the wiring and components on the best DIY builds. Integrating the driver and CA on my bike would reduce the wire count, number of clamps and zipties, and clean up the bike visually.

[chas58]

What kind of current draw are you thinking of? even at 15 amps, that is 3C-4C. that is more than most (non lipo) batteries will be happy with.

However the heavy part of a light e-bike is the hub more than the battery. 36v4ah battery should be small and light enough to be the minimum being considered. It'll still be half the weight of the lightest hubs.

[LewTwo]

What kind of current draw are you thinking of? even at 15 amps, that is 3C-4C. that is more than most (non lipo) batteries will be happy with.

Well I was thinking 10s2p using Panasonic 18560B cells. If my figures are correct (20 x 46.5 grams) then that should be just over 2 pounds (plus BMS, wires, connectors, etc). This is one of the many areas I am profoundly ignorant. I do not know if the discharge rate is sufficient. The subject gets 'muddier' the more I read.

Sufficient for a Q100 or Bafang 250 watt front hub at nominal 36 volts.

P.S. I did send an email to Paul regards a 10s2p battery pack.

[chas58]

It will be interesting what he says. I was thinking of asking him the same question. The reply I expect:
Most of these cells will not handle the current draw. 4ah pack is going to be burdened at 8amps, anything above that will really kill the cells.


From his site:
Whilst the 22P cell is 3C rated and can deliver 2C or more continuously, it is recommended not to run them at these levels on a routine basis. The 29E cell is 2C rated and we recommend a Max of closer to 1.5C. If your setup and usage means that the pack will be operated at high continous Current levels, consider the High Power 20R or 25R Pack which can deliver 5C plus discharge rates without strain

for a 2S build, you really need the 5C rate of the 20R or the 25R. That way, a 4ah pack can put out 20 amps comfortably. I'm thinking of getting something like that for a light fast battery pack.

FYI, his bare SDI INR18650-29E 10S, 4P pack weighs 2.1kg at 11ah. That is my battery pack of choice for the weight weenie.

[LewTwo]

It will be interesting what he says. I was thinking of asking him the same question. The reply I expect:
Most of these cells will not handle the current draw. 4ah pack is going to be burdened at 8amps, anything above that will really kill the cells.

From his site:
Whilst the 22P cell is 3C rated and can deliver 2C or more continuously, it is recommended not to run them at these levels on a routine basis. The 29E cell is 2C rated and we recommend a Max of closer to 1.5C. If your setup and usage means that the pack will be operated at high continous Current levels, consider the High Power 20R or 25R Pack which can deliver 5C plus discharge rates without strain

for a 2S build, you really need the 5C rate of the 20R or the 25R. That way, a 4ah pack can put out 20 amps comfortably. I'm thinking of getting something like that for a light fast battery pack.

FYI, his bare SDI INR18650-29E 10S, 4P pack weighs 2.1kg at 11ah. That is my battery pack of choice for the weight weenie.

I am guessing that putting more cells in parallel and less in series resolves the the discharge rate problem --- yes/no ?
So am I wrong in guessing that a 7s3p at 24 volts might be an alternate ?

My Problems:
Most of the motor Hub adverts I see do not include specifications for the current draw other than the nominal 250/350 watt rating.
Watts = Volts x Amps ==> Watts/volts = Amps
250 Watts / 24 volts = 10 Amps (aprox)
250 Watts / 36 Volts = 7 Amps (aprox)

How does LEEDS manage a 24 Volt 250 Watt motor with their 7s1p Samsung cell pack ?
7x3.6 = 25.2 volts nominal 2300 mAH

How does "C" relate to any of the above (point me toward the resource I should be reading)?
----------------------------
Well now at least I think I have figured out what 20R, 25R, 22P, 22# indicates:
These are all some variation of Samsung SDI ICR18650-??? cells.

[Alex W]

My 250W Bafang/8Fun motor will draw down 350W (under 10A) in normal use, but I've never seen it spike much above that. I'm using a 15A controller. I'd be comfortable limiting it to 10A when using the weight weenie battery. It seems like this would likely work with a 22P battery. The longest that I'd have a 10A draw for in regular use is about 2 minutes.

I'd be open to a 10S3P battery if that would last much longer. It's not a huge increase in weight.

Running it at 24V would limit the top speed to be fairly useless. I (and hopefully anyone here) can pedal a half decent bike at 14mph on flat ground.

2.1kg for a battery is not weight weenie enough. That puts the weight penalty of the whole system above 4.5kg/10lbs.

For me the "weight weenie" builds are about adding a boost to a bike that is already enjoyable and sporty to ride on it's own without e-assist. This makes battery life long because the rider is still providing most of the power in regular riding and just using the assist to boost over hills. It also means the bike should be enjoyable to ride without having the assist turned on, which means keeping the weight reasonably low overall.

I've been thinking about using the Greenworks batteries for an experiment with one of these smaller packs, but also emailed em3ev earlier today to ask about a small capacity pack.

[chas58]

d.

2.1kg for a battery is not weight weenie enough. That puts the weight penalty of the whole system above 4.5kg/10lbs.

For me the "weight weenie" builds are about adding a boost to a bike that is already enjoyable and sporty to ride on it's own without e-assist. This makes battery life long because the rider is still providing most of the power in regular riding and just using the assist to boost over hills. It also means the bike should be enjoyable to ride without having the assist turned on, which means keeping the weight reasonably low overall.

That was mostly a reference point for 10s4p 29E cells. 2p would be half the weight.

My 250W Bafang/8Fun motor will draw down 350W (under 10A) in normal use, but I've never seen it spike much above that. I'm using a 15A controller. I'd be comfortable limiting it to 10A when using the weight weenie battery. It seems like this would likely work with a 22P battery. The longest that I'd have a 10A draw for in regular use is about 2 minutes.

That is a good point. Personally, I would be more comfortable going off the current rating of the controller. But yeah, you can do it. You will have voltage sag when you try to pull too much current (and affect battery life) but that doesn't seem to bother everyone. Not to long ago there were plenty of 1C 8.8ah bottle batteries running setups like that. Its not optimal but it can be done.

[chas58]

I am guessing that putting more cells in parallel and less in series resolves the the discharge rate problem --- yes/no ?
So am I wrong in guessing that a 7s3p at 24 volts might be an alternate ?

My Problems:
Most of the motor Hub adverts I see do not include specifications for the current draw other than the nominal 250/350 watt rating.
Watts = Volts x Amps ==> Watts/volts = Amps
250 Watts / 24 volts = 10 Amps (aprox)
250 Watts / 36 Volts = 7 Amps (aprox)

How does LEEDS manage a 24 Volt 250 Watt motor with their 7s1p Samsung cell pack ?
7x3.6 = 25.2 volts nominal 2300 mAH

How does "C" relate to any of the above (point me toward the resource I should be reading)?
----------------------------
Well now at least I think I have figured out what 20R, 25R, 22P, 22# indicates:
These are all some variation of Samsung SDI ICR18650-??? cells.

For a primer, there is a wiki here, but I read everything I could find on electricbike.com. Lots of the articles are written by moderators and long standing members here.
https://www.electricbike.com/lithium-cobalt-manganese/ for instance.

Two things that are critical but not obvious to the beginner:
motor RPM is key
Battery C rate is key.

a cheap battery is 1C, meaning it cant pull more current than its amphour rating.
a 2.0AH battery at 1 C can only pull 2 amps. at 5C it can pull 10 amps.
Yep, you start putting them in parallel, and you are increasing your capacity
That 1C battery at 1p (one set in parallel) can do 2amps
2 parallel packs can do 4 amps
4 parallel packs can do 8 amps, etc.


Motor speed is determined by the motor design (windings) and the voltage. A 24 volt battery will run half as fast as a 48volt battery on the same motor. It is unusual to find a 24v build here because they are slow and don't have much power. Most people here like 48 volts (I've stuck with 36v to keep my weigh down).

[LewTwo]

For a primer, there is a wiki here, but I read everything I could find on electricbike.com. Lots of the articles are written by moderators and long standing members here.
https://www.electricbike.com/lithium-cobalt-manganese/ for instance.

Two things that are critical but not obvious to the beginner:
motor RPM is key
Battery C rate is key.

a cheap battery is 1C, meaning it cant pull more current than its amphour rating.
a 2.0AH battery at 1 C can only pull 2 amps. at 5C it can pull 10 amps.
Yep, you start putting them in parallel, and you are increasing your capacity
That 1C battery at 1p (one set in parallel) can do 2amps
2 parallel packs can do 4 amps
4 parallel packs can do 8 amps, etc.

Motor speed is determined by the motor design (windings) and the voltage. A 24 volt battery will run half as fast as a 48volt battery on the same motor. It is unusual to find a 24v build here because they are slow and don't have much power. Most people here like 48 volts (I've stuck with 36v to keep my weigh down).

I may be a senile old dog (and I have been called a lot worse --- sometimes with good reason) but I can learn new tricks ... just takes a bit longer these days.
EDIT: http://batteryuniversity.com/learn/article/what_is_the_c_rate
"The capacity of a rechargeable battery is commonly rated at 1C, meaning that a 1,000mAh battery should provide a current of 1,000mA for one hour. The same battery discharging at 0.5C would provide 500mA for two hours, and at 2C, the 1,000mAh battery would deliver 2,000mA for 30 minutes. 1C is also known as a one-hour discharge; a 0.5C is a two-hour, and a 2C is a half-hour discharge."

So If I am reading this correct (debateable):

I will answer my question about the LEEDS battery pack first: They are obviously using the 5C discharge rate cells.
5 x 2.3 = 11.5 amps which is greater than 250 watt / 24 volts = 10.4 amps.

In theory one stack of Panasonic NCR18650B cells rated at nominal 3200mAh should deliver 3.2 amps at a 1C discharge rate.
In theory two stacks of Panasonic NCR18650B cells rated at nominal 3200mAh should deliver 2 x 3.2 amps= 6.4 amps at a 1C discharge rate.
In theory three stacks of Panasonic NCR18650B cells rated at nominal 3200mAh should deliver 3 x 3.2 amps= 9.6 amps at a 1C discharge rate.


I am looking at the spec sheet for Panasonic NCR18650B. Is has a "Discharge Characteristics" chart with curves for 0.2C, 0.5C, 1.0C and 2.0C. So all other things being equal and everything being 100% copacetic 2 stacks of should be able to provide 12.8 Amps at 25 degrees Centigrade.
A 250 watt motor at 24 volts should have a nominal draw of 10.4 amps at leaving a 23% margin.
A 250 watt motor at 36 volts should have a nominal draw of 6.9 amps at leaving a 84% margin.
A 350 watt motor at 36 volts should have a nominal draw of 9.7 amps at leaving a 32% margin
Of course we all know it ain't a perfect world.
I am looking at is the middle line and that seems like enough margin for variances, spikes, wear, etc.

Perhaps I should have asked Paul about a 10s2p based on Samsung sdi icr18650-22p cells:
10 x 3.6 volts = 36 volts nominal
2 x 2.3 ah= 4.6 amps nominal
4.6 x 5C = 27 Amps Max draw

[dogman dan]

Though I tend to not recommend RC stuff, you won't be bothered much by the weight or size of 12s 5 ah lipo. I have done a "lighter than most" off road bike that ran on 14s 5 ah. Enough for the short rides I do to run the dog with tons to spare.

Still came in about 50 pounds, but compared to the off road bikes that did longer distances with 3x the power, pretty light.

Got your thread confused with another who had the bike all tricked out with high end components. I thought you were riding the bike now. Make yourself a cool one speed with a tiny rear motor, and a battery so small you can carry it in a larger seat bag. Gear it tall.

I was going to do that with that mixte bike of mine, but found it pedaled so nice once I put some crazy light wheels on it, I changed my mind.

[MadRhino]

For me the "weight weenie" builds are about adding a boost to a bike that is already enjoyable and sporty to ride on it's own without e-assist. This makes battery life long because the rider is still providing most of the power in regular riding and just using the assist to boost over hills. It also means the bike should be enjoyable to ride without having the assist turned on, which means keeping the weight reasonably low overall.

For me the "weight weenie" builds are about:

- Optimizing power to weight ratio (A 70 Lbs bike can be a super weight weenie, as one that is 25 Lbs can be very heavy )

- Building with the lightest components that are meeting performance requirements ( meaning it can be weight saver but not a winner)

- Optimized geometry and weight distribution (A bike is never as light as when it flies and land naturally)

[Alex W]

The pack that em3ev quoted me is using 20 of these cells in a 10s2p configuration for a 4.9 AH battery: SDI INR18650-25R

They are rated for 20A draw (8C) each cell, so 40A draw in a 2p config. It seems like a pretty good setup to me. Battery weight will be around 1kg including BMS and it will easily fit into a small saddlebag. I'd be using it well under the max design current.

I monitored my battery usage during my ride home yesterday. This includes a steep hill (Golden Gardens for those in Seattle), rolling hills, and I had the assist level turned to high (match my input power). Peak draw was 13A at one point during the climb and I was under 10A average on the climb. I was rarely above 5 amps during the rest of the ride, sometimes it would peak near 10A as I stood to leave a stop light, but that was about it. When I'm at my cruising speed of around 20mph the power output is normally around 1 amp.

I rode about 18 miles yesterday and used 3 AH total. This is the normal ride that I built the bike for (bus to the city, get quickly from bus to after work sailboat racing, then ride back home).

[LewTwo]

The pack that em3ev quoted me is using 20 of these cells in a 10s2p configuration for a 4.9 AH battery: SDI INR18650-25R
...back home).

Thank you. That actually sounds pretty close to want I am looking for.

By the way what are you using for a motor?

[Alex W]

8Fun/Bafang SYX-D01F

This is the same motor that is used on the Faraday Porteur. It is not that easy to source in the US, but my great LBS (G&O Cycles) imported a few so that I could get one. It is 250W (nominal, I see more going into it quite often), around 2kg, and has properly located disk mounts unlike many other options.

Other e-bike parts are a Thun BB, 15 amp motor controller from ebikes.ca, CAv3 for logical control, and the e-bike version of B&M's IQ Cyo Premium headlight.

This is my only e-bike. I build bicycle frames as a hobby and have around 7 or 8 other bikes, 3 or 4 of which I ride on a regular basis. I mostly use the e-bike to get around faster on summer evenings to try and fit more activities into the day.

[LewTwo]

It will be interesting what he says.

Hi Lewis,
A couple of options for small 36V packs using high power cells. Including a 20A BMS and a 2A charger (not incl. shipping):
-----------------------
cell = SDI INR18650-20R
Configuration = 10S, 2P
V(nom) = 36
Ah(nom) = 3.8
Whrs = 136.8
Weight = kg = 1
-------------------------
cell = SDI INR18650-25R
Configuration = 10S, 2P
V(nom) = 36
Ah(nom) = 4.9
Whrs = 176.4
Weight = kg = 1
-------------------------
Such packs are fine for 20A discharge or a little more.

Note to self:
INR = lithium iron phosphate cylindrical cell (“LiNiMnCoO2” or “NMC”)
I = lithium ion
N =nickel/manganese/Cobalt oxide cathode
R = round cell type

So pretty much what I was figuring except not the Panasonic cells. I think that the clear winner here is SDI INR18650-25R in a 10S2P configuration. The 4.9 Ah should be adequate for my purposes. If there is occasion to require more Ah then I will strap my Golden Motor 12Ah battery pack to a removable seat post rear rack (always good to have a plan "b").

That leaves me with deciding between the Bafang mid drive (36V x 250 watt) and a Q100 front hub.

In my mind the Bafang option considerably simplifies the build as I have to obtain and install a crank/BB anyway. It also means that I can use any wheel set. I still do not have a weight number for the Bafang system that I have any confidence in (so much for the US based support hyperbola).

The QT front hub on the other hand requires 36 spoke rims and an external controller to be mounted somewhere. In that case I favor the construction of a single handle bar package that incorporates the small battery, motor controller and display (possibly a CYCLE ANALYST unit). Of course that adds a bit more complication to the build.

[tln]

In my mind the Bafang option considerably simplifies the build as I have to obtain and install a crank/BB anyway. It also means that I can use any wheel set. I still do not have a weight number for the Bafang system that I have any confidence in (so much for the US based support hyperbola).

This thread has the weights for BBS01... http://endless-sphere.com/forums/viewtopic.php?f=4&t=64837

[chas58]

I rode about 18 miles yesterday and used 3 AH total. This is the normal ride that I built the bike for (bus to the city, get quickly from bus to after work sailboat racing, then ride back home).

I'm just trying to figure out if I can use the shore power line to charge my bike while I am out sailboat racing...

[chas58]

The QT front hub on the other hand requires 36 spoke rims and an external controller to be mounted somewhere. In that case I favor the construction of a single handle bar package that incorporates the small battery, motor controller and display (possibly a CYCLE ANALYST unit). Of course that adds a bit more complication to the build.

good ideas.

I'm not a big fan of a front hub motor as if you do anything that cracks or brakes the drop outs (which are not designed for a motor), people crash hard when the front wheel gives out. The safe bet is a steel fork. A Q100 at low amps is no torque monster, so you probably won't have any problems.

As for the wheel build, I just buy a bare hub and have a local shop build it to my own wheel as shipping a fully built wheel from China costs a fortune (and isn't worth it considering the end product).

A hub motor could work - might look something like this:

[LewTwo]

I'm just trying to figure out if I can use the shore power line to charge my bike while I am out sailboat racing...

You are going to need a very long watertight extension cable
You probably want to add lead weights to the cable to keep it from floating up and interfering with the other boats as well.
Do not forget to include the drag effect on you racing speed.

[chas58]

The bike stays on shore with the shore power outlet...

[chas58]

My favorite weight weenies, for your inspiration:

Fahrader von Herr Nader:
viewtopic.php?f=6&t=19872

17lb swiss mountain bike (Herr Schiltknecht)
https://www.electricbike.com/lightest-bike/

Kepler's carbon fiber commuter down under (see if this doesn't inspire you!)
https://www.electricbike.com/kepler-super-commuter/

more light weight designs:
https://www.electricbike.com/10-light-electric-bikes/

Nader and Kepler really inspired my builds.

[LewTwo]

I'm not a big fan of a front hub motor as if you do anything that cracks or brakes the drop outs (which are not designed for a motor), people crash hard when the front wheel gives out.

Been there, done that: Summer of 1966, Sun Valley Road, Birmingham, Alabama in front of the post Office. The front wheel dropped off the 3 speed english frame (that is what we called 'em in those days) when I was doing a bit of a wheelie. Fortunately I was able to 'unicycle' it to a softer grass covered landing area. Not something I am ever likely to forget.

In this case the rear drops out are Aluminum as well. Bonded steel torque washers would be advisable in either case.

[mfj197]

As tln says, this thread has all the weights for the BBS01. It isn't really a weight weenie option - whilst the motor and controller unit itself is 3.8kg all the other necessary components aren't featherweights. You will be replacing crank arms and chainrings with some fairly weighty items even though you are removing the front derailleur. I did a reasonably weight-conscious conversion myself with the BBS01 and the bike went from 10.9kg to 18kg - that was using a 10s4p self-build pack from Panasonic NCR18650PF cells which weighs 2kg in total. So if you used a 10s2p pack weighing 1kg you would still be adding 6kg (over 13 pounds) overall to the bike.

It is a great motor though. In many ways I'd prefer a lighter weight crank drive motor as I only use it at half power, but it's nice having full power if needed (which is actually over 500 watts despite it being the 250w motor!). I commute daily just over 40 miles which it'll do with the 11.6Ah pack, although to save deeply discharging the pack I charge at work as well as at home.

Build thread of my Cube road bike here on Endless Sphere or here on the UK pedelecs forum which also has follow-up reports (I'm at nearly 2k miles now).

Michael