New Bafang Crank-Drives

[tungsten2k]

Well it looks like the extreme chain-line stretch of the 1x front Bafang CR while on lowest 32T gear on the rear SRAM cassette took out this old Surly Narrow "SP" chain gave up the ghost lickety-split:



I there any consensus on the best chain to get for our higher-than-human 1x applications with somewhat severe chain line at the extremes ? Or should I just get the KMC X9SL as I would for a normal high-quality 9-speed rear cassette ?

-=dave

 

[jkbrigman]

Hi tungsten2k!

(UPDATE: tln is correct, the BBS01 is 36v and the BBS02 is 48-52v. I did not concern myself with answering that question, since it is easy to understand if you only browse http://www.em3ev.com. Your other questions interest me more.)

BBS02 can be purchased in 36V or 48V versions but each version can't run at the other voltage.

It's a firmware change only, available that way only from Paul at EM3EV. This is just BBS01 vs. BBS02 talk.

1) When I see voltage decrease on my battery down into the 65-66v range, (near discharge) it takes MORE WATTS for me to maintain a given speed than it does when voltage is higher. That is:

Full charge (72v): cruise 23-25mph, 280 watts
Low charge (66v): cruise 23 mph, 450 watts

Are you sure your instrumentation is correct? You should see more amps but similar watts.

I hear you on that. I thought the same thing when I first saw this happen.
I have a straight-up CA v2.23 with ebikes.ca controller. I saw same phenomenon on a 9c and a Crystalyte motor.
So, wait: what do you use to measure, and do you see same watts? Got Data?

BTW: If you want to move this discussion to it's own thread, I'd go for that. We are kinda hijacking this one so I won't say anything more.

3) However: if you have a lower voltage but more batteries in PARALLEL, you can CHARGE FASTER. This is because what limits charging speed is how much current you can put into a battery. If you parallel, you can put much higher current into the total battery than you can into each brick.

Not true. There is a maximum charge current per cell. If the cells are all in series then that is the maximum charging current for the battery. If you rearrange those same cells so there are two cells paralleled and each of these pairs are connected in series, you will have halved the battery voltage and doubled the maximum battery charge current which means power is supplied to the battery at the same rate. Charging current per cell will be unchanged as the current is split across two cells and therefore charging time will be unchanged.

Most cells are allowed to charge at 1C. More parallel paths, more allowable aggregate current for the battery from the charger. Kirchhoff's law, from the perspective of the charger only. That's all.

Over and out.

 

[Samd]

Would the bbs02 controller fit on a BBS01 casting Ken? Probably, I can check if you like. I got your question before it was derailed.
They are potted up with flexy silicone goo - just like the Aprilia controllers. No wonder they overheat.

I have a spare casting from a BBS01 if anyone wants to try dropping a compact 6 fet in there. It's a curvy shape though. I reckon you could vent the unit and drive a BBS01 hard. Bugger efficiency, these things sip power. Trash ten percent and use the lipo you arent using now anyway.

If you did hack a BBS01, you'd find what we Aprilia owners found when supercharging mid-drives: two things break. The smallest pinion inside the gearbox, and the chain sides blow out.

Speaking of the post above looking for a better brand of nine speed chain, now you know why we Aprilia hackers stuck with six speed chain. The sides of the links are much thicker on 6&7 speed. 8 speed and up has thin walls to make it fit.

Sram makes a nice selector and the X4 can be run 7 speed. I buy a lot of them on ebay when I can find them.

If I won the lotto tomorrow and started a corporation focused on ebike products, I could have atribe of people working on my wish list. About two thirds down that list is a 3 or 4 speed derailleur with chunky everything to handle mid drive torques.

Someone mentioned above the chainring adapter versus speed tradeoff. I think my adapters are helping the motor run in a slightly better rpm range and still maxing out at the same top speed restriction on a 38t for me - I need to back that up with data. Not that it matters (given my point above, these things really use such tiny amounts of battery power).

 

[Ken Taylor]

Would the bbs02 controller fit on a BBS01 casting Ken? Probably, I can check if you like. I got your question before it was derailed.
They are potted up with flexy silicone goo - just like the Aprilia controllers. No wonder they overheat.

Thanks Sam, I'd be interested. I want to keep the weight down as much as possible and for me the BBS01 has more power than I need. However, the low maximum cadence forces you to change your pedaling style and is hard to manage. Changing down too early on a hill climb can mean a complete loss of assist which feels like applying the brakes. I had been wondering why there was a Bafang spec sheet for the BBS01 at 48V but there were none advertised and Raygo got me wondering whether the higher voltage controller would work on the BBS01. Maybe that is where the Bafang data came from but there are just none being sold that way. The maximum current would have to be set much lower than for the BBS02 but that's OK. I notice you have been selling a few BBS01's, perhaps there is a market for higher cadence versions.

I have a spare casting from a BBS01 if anyone wants to try dropping a compact 6 fet in there. It's a curvy shape though. I reckon you could vent the unit and drive a BBS01 hard. Bugger efficiency, these things sip power. Trash ten percent and use the lipo you arent using now anyway.

If you did hack a BBS01, you'd find what we Aprilia owners found when supercharging mid-drives: two things break. The smallest pinion inside the gearbox, and the chain sides blow out.

More power would appeal to many but I want only enough to match the best unpowered riders and the 250W BBS01 can easily do that, except in a sprint. Good riders can produce killowatts for 10 seconds or so but I'm not trying to match that.

Speaking of the post above looking for a better brand of nine speed chain, now you know why we Aprilia hackers stuck with six speed chain. The sides of the links are much thicker on 6&7 speed. 8 speed and up has thin walls to make it fit.

Sram makes a nice selector and the X4 can be run 7 speed. I buy a lot of them on ebay when I can find them.

I've been running a 10 speed rear with a narrow Shimano one way chain for about a month. I was having trouble with the chain occasionally coming off the front chain ring, particularly when changing under power. I've added a front derailleur locked into position and, with that, haven't had the chain come off yet. I change under load to maintain momentum which sometimes crunches but mostly seems OK. Of course, running low power means it is easier for me than most others here.

Someone mentioned above the chainring adapter versus speed tradeoff. I think my adapters are helping the motor run in a slightly better rpm range and still maxing out at the same top speed restriction on a 38t for me - I need to back that up with data. Not that it matters (given my point above, these things really use such tiny amounts of battery power).

The smaller, 46 tooth, Bafang front chain ring is about right for me with a 11-34 rear on 700C wheels. I suspect Bafang has optimised the chain ring for 700C and road riding where a lot of people are using 26 inch wheels and slow off road climbing. The Bafang chain ring is too large for steep grades without assist giving a minimum speed of 11 km/h at a practical minimum cadence of 65. With assist, 11km/h is easily maintained on road grades.

 

[Samd]

In theory it wouldn't be hard to jam a bbs02 controller on a BBS01 if they physically fit ("BBS1.5" is now a thing!). Same plugs inside.

I'll check it out with all the spare time I get in between building the nation's coolest customs.

 

[Rusty123]

One potential efficiency advantage of the 48 volt unit: higher voltage means less current for a given power output means less I^2R losses (heat). On the other hand, slower to charge.

 

[Rusty123]

Haven't seen it mentioned here, but EM3ev is selling chain rings for $8.

 

[GeoKrpan]

Below is a gear chart of three chainring sizes and the highest and lowest gear combinations.
As you can see a 42 tooth rear cog is only 6 gear inches lower than a 34, disappointing. SRAM XG-1199 Cassette, 11 speed, 10-12-14-16-18-21-24-28-32-36-42, $300 plus the cost of an 11 speed shifter and a derailleur that can handle a 42 tooth cog. A SRAM Dual Drive rear hub is a cheaper alternative and would provide a lower low gear.

Chainring 42 46 48
Cog
10-- 108.9 119.3 124.5
11--- 99.0 108.5 113.2
32--- 34.0 37.3 38.9
34--- 32.0 35.1 36.6
36--- 30.3 33.1 34.6
42--- 25.9 28.4 29.6

 

[arclarke]

Below is a gear chart of three chainring sizes and the highest and lowest gear combinations.
As you can see a 42 tooth rear cog is only 6 gear inches lower than a 34, disappointing. SRAM XG-1199 Cassette, 11 speed, 10-12-14-16-18-21-24-28-32-36-42, $300 plus the cost of an 11 speed shifter and a derailleur that can handle a 42 tooth cog. A SRAM Dual Drive rear hub is a cheaper alternative and would provide a lower low gear.

Chainring 42 46 48
Cog
10-- 108.9 119.3 124.5
11--- 99.0 108.5 113.2
32--- 34.0 37.3 38.9
34--- 32.0 35.1 36.6
36--- 30.3 33.1 34.6
42--- 25.9 28.4 29.6

but 6 inches is a 20% change at that end of the range - quite a big jump...on my MTB I use a 40t 10 speed cog, but you can get 42t from OneUp components, with their wide range jockey cage plate for Shimano mechs - really works well and lot cheaper than going to 11 speed.

 

[Ken Taylor]

BSRAM XG-1199 Cassette, 11 speed, 10-12-14-16-18-21-24-28-32-36-42, $300 plus the cost of an 11 speed shifter and a derailleur

tungsten2K broke a 9 speed chain and I notice 11 speed chains breaking quite often on unpowered bikes. While it's fine so far I'm worried that my 10 speed is already not strong enough and run lower power than most here. I'd have thought chain failure with an 11 speed is a high risk with a higher torque BBS02.

 

[Samd]

With thanks to Big Ron Spinningmagnets, this thread is absolute gold on chains and breakage parameters:
http://endless-sphere.com/forums/viewtopic.php?f=16&t=46630

If you observe the side link gauge for >7 speed you will see what I mean...

 

[GeoKrpan]

Below is a gear chart of three chainring sizes and the highest and lowest gear combinations.
As you can see a 42 tooth rear cog is only 6 gear inches lower than a 34, disappointing. SRAM XG-1199 Cassette, 11 speed, 10-12-14-16-18-21-24-28-32-36-42, $300 plus the cost of an 11 speed shifter and a derailleur that can handle a 42 tooth cog. A SRAM Dual Drive rear hub is a cheaper alternative and would provide a lower low gear.

Chainring 42 46 48
Cog
10-- 108.9 119.3 124.5
11--- 99.0 108.5 113.2
32--- 34.0 37.3 38.9
34--- 32.0 35.1 36.6
36--- 30.3 33.1 34.6
42--- 25.9 28.4 29.6

but 6 inches is a 20% change at that end of the range - quite a big jump...on my MTB I use a 40t 10 speed cog, but you can get 42t from OneUp components, with their wide range jockey cage plate for Shimano mechs - really works well and lot cheaper than going to 11 speed.

That's good info to know about the OneUp stuff.

I just realized that the Sheldon Brown Gear Calculator will work for the SRAM Dual Drive hub. The lowest gear it will provide with a 42 tooth chainring and a 34 rear cog is 23.5 GI (26x2.125 tire). Not as good as I thought but as good as it gets. For comparison, a 22x34 combination is 16.8 GI (22/32/42 triple crank).

What I'm trying to do here is to provide people with the raw information so that they can make an informed decision. It looks like 8 speed is the way to go because of the strength of the chain verses 9, 10, or 11 speed. AND, 8 speed is CHEAP!

It looks like the largest rear cog on 8 speed cassettes is 34 tooth. I think it's 36 tooth for 9 speed and I am now informed that it's 40 tooth for 10 speed which can be upped to 42 with OneUp stuff.

I will be building my electric bike with a 46 tooth chainring and an 11-34 eight speed cassette. The bike is my old, old steel 29er hardtail with a rigid steel fork. It currently has a 3x9 drive train but I will buy an 8 speed chain and cassette for it. I will be running a "three speed" handlebar (Wald #8095) flipped upside down so that it drops instead of rises. The bar is compatible with bar end shifters and I already have a set of them (8 speed). I want to run a right side thumb throttle. Using a bar end shifter will make room for it AND in order to shift I will have to slide my hand back, off the throttle. I will never be shifting under power. The PAS mode is not very attractive to me, neither is a twist throttle, or a left side throttle (thumb or twist). What I would prefer to PAS is cruise control where power is interrupted only by switches in the brakes and not by the absence of cadence or torque. It would "resume" when cadence or torque is sensed. A kill switch would be a nice feature on this sort of setup.

 

[GeoKrpan]

BSRAM XG-1199 Cassette, 11 speed, 10-12-14-16-18-21-24-28-32-36-42, $300 plus the cost of an 11 speed shifter and a derailleur

tungsten2K broke a 9 speed chain and I notice 11 speed chains breaking quite often on unpowered bikes. While it's fine so far I'm worried that my 10 speed is already not strong enough and run lower power than most here. I'd have thought chain failure with an 11 speed is a high risk with a higher torque BBS02.

8 speed is the way to go. See my response to arclark's comment.

 

[krunchi]

Well it looks like the extreme chain-line stretch of the 1x front Bafang CR while on lowest 32T gear on the rear SRAM cassette took out this old Surly Narrow "SP" chain gave up the ghost lickety-split:



I there any consensus on the best chain to get for our higher-than-human 1x applications with somewhat severe chain line at the extremes ? Or should I just get the KMC X9SL as I would for a normal high-quality 9-speed rear cassette ?

-=dave

That type of chain failure and most failures experienced by Bafang users are a result of changing gears under load. You will see that same failure with non powered bikes and riders not taking into account shifting under power. I have a 10 speed sram chain that has run 2500ks on a Bafang 500w unit. I make sure i change gear when the PAS is not applying too much power, usually means I speed up my pedalling before i shift next gear. Also takes some planning before hills. When a unit with a torque sensor is released that should solve all chain break issues.

 

[Samd]

That type of chain failure and most failures experienced by Bafang users are a result of changing gears under load.

Can't agree with the use of "most". A lot of failures are the extra torque on a curved chainline. I've been doing this for seven years now.

 

[GeoKrpan]

Well it looks like the extreme chain-line stretch of the 1x front Bafang CR while on lowest 32T gear on the rear SRAM cassette took out this old Surly Narrow "SP" chain gave up the ghost lickety-split:



I there any consensus on the best chain to get for our higher-than-human 1x applications with somewhat severe chain line at the extremes ? Or should I just get the KMC X9SL as I would for a normal high-quality 9-speed rear cassette ?

-=dave

That type of chain failure and most failures experienced by Bafang users are a result of changing gears under load. You will see that same failure with non powered bikes and riders not taking into account shifting under power. I have a 10 speed sram chain that has run 2500ks on a Bafang 500w unit. I make sure i change gear when the PAS is not applying too much power, usually means I speed up my pedalling before i shift next gear. Also takes some planning before hills. When a unit with a torque sensor is released that should solve all chain break issues.

I suspected that that might be the case. I also thought it may have something to do with the single chainring. Since there is only one chainring people would be removing the superfluous front derailleur not realizing that it serves a dual purpose, shifting AND chain retention. I have made the same mistake. Without a front derailleur or some other sort of chain retention device the chain WILL fall off the chainring. That could definately result in a broken chain.

PS But I'm going with 8 speed for my electric bike because it is way cheaper. Cheaper to buy and cheaper because it lasts longer than 10 speed.

 

[tungsten2k]

That type of chain failure and most failures experienced by Bafang users are a result of changing gears under load.

Can't agree with the use of "most". A lot of failures are the extra torque on a curved chainline. I've been doing this for seven years now.

I certainly did not have an upshift under load. I downshift under load constantly, but not under high-assisted load. I might have mentioned, but my riding style so far has been me doing the same full-bore ride that i had been doing w/o the Bafang, and letting the Bafang help only for the steep uphills that would normally drop me to a crawl (understandably). My Tuesday evening ride is 15 miles with most of it relatively flat (rolling long slopes with about 200' of total climb) but about 1/2-mile of it climbs 850'. On a 12S1P HK 5Ahr/222Wh lipo battery I put back in 2350Ah for the entire ride, and that included more BBS02 usage than I would normally because I aligned the chainline to 6th gear (to get it perfectly straight) and then used a fire-hydrant and the metal body of my Gerber knife to pound the link back on and limped it home up the steep hill in 6th, which pretty much meant the Bafang doing a lot more work than it would have otherwise (although I did walk it a few of the steepest blocks).

This failure I think was a combination of the severe chain line angle between the SRAM 9spd 11-32T cassette in lowest gear and the now-1x BBS02 stock CR, a chain that needs to be about 3 links longer (the stock CR on this bike was 42), being on the single steepest block of the trip which is probably pushing 20%+ grade, combined with a worn chain, and a questionable-quality chain to begin with. This was a "Surly" chain, and I've never known Surly to be very much into hardcore R&D, and who knows how old this chain was before I got it (2001 KHS bike).

I have a new Ti-nitride-coated KMC X9SL on the way; we'll see how long it lasts (my upgrade from 12S to 14S I'm working on while my chain is down won't help it's longevity, i'm sure )

With thanks to Big Ron Spinningmagnets, this thread is absolute gold on chains and breakage parameters:
http://endless-sphere.com/forums/viewtopic.php?f=16&t=46630

If you observe the side link gauge for >7 speed you will see what I mean...

I'm not sure which part you're referring to, except maybe that they are skinner chains, but yeah, there's a goldmine of info on that landing page to review. At first glance it sounds like I probably need to revamp to no more than 7 gears in order to keep the chain width up ? I'll have to figure out what to do next after this one breaks.

Being relatively new to serious bike mechanics, and completly new to e-bikes, there's a lot to learn !

UPDATE/EDIT: "When a unit with a torque sensor is released that should solve all chain break issues."

Y E S ! The moment it is released, I'll have a lightly-used BBS02 available for some lucky person

-=dave

 

[KC1983]

Please excuse my relative ignorance, but I'm wondering if I can fit a BBS01 on my 2008 Fuji Absolute 2.0 (http://www.bikepedia.com/quickbike/BikeSpecs.aspx?year=2008&brand=Fuji&model=Absolute+2.0). The bike is stock, and has a Truvativ ISO FLow 3.0 26/36/48 tooth 64mm/104mm crank. I understand the BBS001 requires a bottom bracket between 68mm and 73mm, but I've read that virtually any modern mass produced hybrid should work fine with the Bafang. I've never messed with replacing a bike crank, so this answer may be obvious.

If I should make a new post to ask this question, let me know, but it seems like all the Bafang experts are here.

 

[Samd]

Flip it over and measure the BB shell width, easy to do with a ruler.
http://www.nashbar.com/images/nashbar/productinfo/shellwidth.jpg

The good news is that the cranks you have now are very sexy and solid, with a fancy shaft shape on the end. The bad news is they wont be able to bolt onto the end of the BBS shafts. But the good news is you'd probably sell them for a few bucks to fund your BBS!

Dont worry about using calipers to measure. It'll be close to 68mm, 70, 73mm or 82,90 or 100. Bad news is in blue.

 

[Ken Taylor]

BSRAM XG-1199 Cassette, 11 speed, 10-12-14-16-18-21-24-28-32-36-42, $300 plus the cost of an 11 speed shifter and a derailleur

tungsten2K broke a 9 speed chain and I notice 11 speed chains breaking quite often on unpowered bikes. While it's fine so far I'm worried that my 10 speed is already not strong enough and run lower power than most here. I'd have thought chain failure with an 11 speed is a high risk with a higher torque BBS02.

8 speed is the way to go. See my response to arclark's comment.

I ran 7 speed 12-34 for 4000 km. It was OK but I frequently found one gear was a bit low and the next a bit high on a road bike. I'm not sure if one more would have been enough. I'm finding 10 keeps them close enough together that I don't keep oscillating up and down and so far haven't broken the chain even though I change under load. Adding back a locked front derailleur has been important though, as the occasional time the chain came off was always when trying hard and usually meant being left in the dust of an opponent.

 

[1KW]

If you run these like a gng and just using throttle , would it kill it pretty quick?

 

[arclarke]

I've finally got my Bafang fitted to my Yuba Mundo. I'd decided to go for 350W and ordered from Mr Spacey, but his delivery from China turned out to be BBS02s instead, so that's what I have. Blimey it's fast, not entirely sure it needs to be this powerful! I'm only running on power assist 4 or 5 but hit 52km/h (32mph) on gentle downhill, and about same on flat on 9 - this works out at about 110 rpm so I guess is limit with the gears I have. The Yuba is very solid and has hydraulic discs, so feels fine at this speed but I think I have to take it easy to avoid confusing drivers - and to get a sensible range out of the battery.
Will post some photos when I take some - wiring still needs tidying up and don't have good battery case yet...

 

[arclarke]

Some Yuba pics




battery box yet to come - pack just strapped on for now

 

[Samd]

Nice result - they pull really well!

How many stages of power have you set the display on? I've got a 963 on my yuba and I find it hard to get just right sometimes. Good thing there's a 3540 under the rear as well

Australia needs a national Yuba race

 

[KC1983]

Flip it over and measure the BB shell width, easy to do with a ruler.
http://www.nashbar.com/images/nashbar/productinfo/shellwidth.jpg

The good news is that the cranks you have now are very sexy and solid, with a fancy shaft shape on the end. The bad news is they wont be able to bolt onto the end of the BBS shafts. But the good news is you'd probably sell them for a few bucks to fund your BBS!

Dont worry about using calipers to measure. It'll be close to 68mm, 70, 73mm or 82,90 or 100. Bad news is in blue.

Thanks, Samd, that was easy! Yes, it's exactly 68 mm, so I'm good to go. I hadn't thought about selling the existing cranks. I have a bunch of stuff I should sell on ebay, so if I ever get around to that I'll put the cranks on there. For now I'll likely keep them, in case I ever need to put them back on the Fuji.

Based on a lot of comments here, I'm looking at ordering this: http://em3ev.com/store/index.php?route=product/product&path=45&product_id=183 Seems like a great price, and the vendor gets high marks.