Rohloff with BBSHD - shift kill switch or not?

[kiltedcelt]

So, I was having issues with the shifter on the NuVinci N171 that I'm using on my recumbent tandem trike build. The e-assist is Bafang BBSHD so the choices for geared hub were Rohloff or N171. I bought the N171 because it's WAY cheaper, but the cables are SO LONG that the drag (with or without full housing), is causing them to kink inside the shifter housing and I've already ruined two shifter cables. So, I said "EFF IT!" and pulled the trigger on a Rohloff hub. I guess, consider it a longterm investment as it should outlive several projects. Anyway, on the N171 I was installing one of those in-line Bafang motor cutoff switches that kills power when it senses movement of the shift cable. The problem is, the shift cable runs THROUGH that switch and inside is a tiny wheel that actually introduces some drag into a system (long-ass shifter cables) that already has a crap-ton of drag to begin with. The Rohloff shifter is WAY better designed than the very crude 1.0 version of the NuVinci shifter, and I've been doing a LOT of reading about Rohloff in the last few days. It seems that it can handle shifting under load, as the worse that happens is that it defaults momentarily to I think 14th gear when loaded but as soon as the torque lightens up it drops into the proper gear. It's apparently NOT like an Alfine where parts will actually break inside the hub when shifted under load.

The trike has an independent pedaling system where the stoker can pedal while the captain coasts, and the way the BBSHD works, this means that even if the captain is coasting, the stoker can still be forcing the BBSHD to apply power, unless the captain instructs them to stop pedaling during shifts OR a motor kill switch is employed to cut motor power under shifting. Of course this still doesn't change the fact that while a cut-off removes motor power, the stoker could still force the hub to be shifting under load. I am thinking that maybe I just avoid the B.S. of cutoff switches and it just becomes understood that the stoker stops pedaling when the captain calls out "SHIFTING!" or maybe "GEARS!" since in the right circumstances, "SHIFTING!" could maybe be mistaken for something else. The other thing complicating this decision is that Rohloff recommends (and ships the hub), with what is essentially spiral wound brake cable housing as the shifter housing, and they recommend with the newer hubs full housing from shifter to hub. Although, I saw on the Co-Motion YouTube channel them installing a Rohloff and having only housing coming out of the shifter and going into the hub. The longer stretch on the downtube had the shifter cables exposed and cable stops were used. So, I'm thinking if I use even partial housing but use the brake cable housing, then it won't go into a Bafang cutoff switch anyway, not to mention I don't really like how that switch puts noticeable drag on the actual shift cable running through it. The other thing is I read about a one-button cutoff switch that you press when shifting, or using a brake cutoff switch and tapping your brakes momentarily while performing a shift. Both of those situations seem kind of janky. So, I guess - is a cutoff switch really needed - especially if I can also go into the BBSHD program and limit it's make power output and thus it's torque? Also, the low-tech solution of telling the stoker to stop pedaling? Of course that won't always work and it's entirely possible powered shifts will eventually be forced through the Rohloff due to miscommunication. Basically I don't want to blow up a $1600 hub that would have to be sent back to Germany to be repaired.

 

[COAR]

I didn't bother installing the gear shift cutout. I have the brake sensor set up so it will trigger as soon as the lever is touched and before the brakes grab. So I essentially have a kill switch I can use whenever I want. I'll activate it sometimes when shifting or if I want to stop the assist without touching the display.

So no Rohloff specific advice but a simple option you could consider.

 

[kiltedcelt]

I was planning on having the brake cutoff switch since the independent pedaling system means that the stoker could keep pedaling even when I'm trying to brake, thus continuing to engage the motor. It's pretty funky, but with the BBSHD in the front of the trike at the captain's position, and with the independent pedaling system, you can pedal the middle (stoker) set of cranks and continue to engage the front chainring on the BBSHD. I don't know yet whether the motor will continue to output though. I mean it seems like it *might*, but since the captain can hold their cranks stationary the stoker could still be sending power through their crankset to the rear wheel, and since the whole drivetrain is connected they continue to spin the freewheeling BBSHD chainring. I'm thinking though, that only the captain pedaling will engage the motor because it's only the PAS system which senses pedaling from the captain's crank that will engage the motor. This might eliminate some of the concern of motor power being dumped into the Rohloff during pedaling if the stoker doesn't stop pedaling during a shift. It's a weird situation though, and I don't know that anyone else has done this so I have no guidance on how the BBSHD will behave if driven by the stoker but when the captain is not actually pedaling.

 

[COAR]

If it works like I'm imagining, you should only get PAS when the captain is pedals. If the stroker can't turn the cranks up front when they pedal, you won't activate PAS. And the brake switch kills everything - throttle and PAS. Once it's released there's a second or so delay before you get assist again. Another option is something like this (dedicated kill switch): https://electricbike.com/forum/forum/kits/bafang-mid-drives/72950-bbshd-kill-switch-from-gear-interrupter

I'm assuming your trike is something like this. That is quite a lengthy chain!

 

[COAR]

I was curious so I just went into my garage to confirm. If you move the chain by hand, the chainring (on my BBS02) will move, but it does not activate PAS. PAS only comes on when the cranks are physically turned.

 

[TNC]

After living with a BBSHD for awhile now, I'll make this observation on the shift sensor. I built my bike setup for relatively aggressive mountain biking. If you're aggressively attacking the trail in corners, short step-ups, and other trail obstacles where you're often on and off "the gas", you'll being doing more shifting than most of the pavement or smooth trail applications. I notice the shift sensor to be very functional when needing to shift in those on again, off again situations. It's subtle but it's there and does some good.

Now, when I'm playing around on smooth ground, pavement, and such, I don't think it's that necessary. As long as you're using your head to think about not jumping up and standing on the pedals for a hill and then shifting under that load, I think one can do without the shift sensor. One might think you'll want/need to do an aggressive downshift under power to quickly cross a road or deal with traffic...but frankly...that's what the throttle is for.

I think the shift sensor is "almost" a requirement for aggessive off road. Pavement, dirt roads, smooth trails...maybe not so much.

 

[kiltedcelt]

Thanks for that confirmation on the "pedal engagement necessary to activate motor" thing COAR. My tandem has been under construction now for nearly two months and whopping two test rides were both unpowered with the e-assist not hooked up. The stoker can pedal while the captain coasts, but your right in that the stoker pedaling does not cause the captain's crankset to turn which means the motor should not engage/provide assist except when the captain is pedaling. In response to TNC's comment about the shift switch, I have read that many people find it sensitive enough that it can be inadvertently engaged. My biggest concern is that it may cause drag that will affect how the Rohloff shifter functions. I do believe the Rohloff shifter is WAY better designed than the cheap-ass design and construction of the NuVinci N171 shifter though so I may not be an issue. However, Rohloff ships the hub with a "shift cable housing" that is actually spiral-wound brake housing, or something very like it. I'm not sure how necessary this type of housing is, but then again if it's what Rohloff specs I'd rather stick with that for warranty purposes and brake housing WON'T interface with the shift sensor, best I can tell. My other issue is I'm already running cables that are over meters long so I see any potential drag on one cable as a problem, and the shift sensor clearly puts some noticeable drag on the cable that runs through it.

 

[COAR]

After living with a BBSHD for awhile now, I'll make this observation on the shift sensor. I built my bike setup for relatively aggressive mountain biking. If you're aggressively attacking the trail in corners, short step-ups, and other trail obstacles where you're often on and off "the gas", you'll being doing more shifting than most of the pavement or smooth trail applications. I notice the shift sensor to be very functional when needing to shift in those on again, off again situations. It's subtle but it's there and does some good.

don't want to derail this thread too much but what assist level/current % would you say you generally ride rougher trails at?

 

[TNC]

COAR, on the tighter, more technical local MTB trail I frequently ride, it's usually assist level 5 which is at current limit 40% and speed limit 100%...with all my speed limits except 0 set at 100%. But really there are places where I'll tap the plus button to 7 (70%) or 8 (80%) for a more open stretch with a climb. It's one of the things I like about the 500C display over on the left end of the handlebar...ease of selecting more or less assist quickly.

However, this is dependent on the nature of the particular trail. Another location is a 100 mile, singletrack, off road motorcycle venue where it's faster, less tight, but more technical in spots. Overall in that location the assist level remains at a higher level more consistently. So it's like the old hackneyed real estate slogan...location, location, location...LOL!

 

[Grautuefel]

Speaking as one running a Rohloff with a BBSHD, I would stick with the shift sensor engaged. When you have the higher PAS levels or are using the throttle the gear change will bind up if you shift under load. Plus its a small price to pay to not strip the teeth out of your shiny expensive little bag of gears.

 

[geosped]

I agree the shift sensor has been fantastic. I would only add that it works much better when you have the proper cable length around the rear derailleur, "the loop" before it attaches to the derailleur if it's too short shifting may still seem very abrubt and stiff as in my case. But once you have it dialed in shifting has been smooth sailing. IMO this is one of the reasons most people complain about the BBSHD. When I first got my kit installed with out the shift sensor I thought my bike was going to explode. IMO it's not complete with out the sensor and should be included as standard as part of any kit.

 

[k.m]

Speaking as one running a Rohloff with a BBSHD, I would stick with the shift sensor engaged.

As the Rohloff has 2 cables, one for each shifting direction, do you run the shift sensor in the cable for lowering the gears (lower speed) or in the cable for the higher gears (higher speed)?

 

[Grautuefel]

Either, they both move simultaneously as they act as one big push/pull loop