Has anyone tried Bafang's 5v torque sensor?

Hey gang,

It's been a while since I've posted in here. I've been lurking, but haven't contributed much, for which I apologize.

I'm building a new bike. It's going to be an ultra-minimal single-speed ebike. I'm using a q100h rear hub and a Phaserunner to control it. I chose those primarily for their small size. I also bit the bullet and ordered a spot welder to make my own battery pack which will be a 14s2p pack that fits in a small seat bag. It'll be 51.8V, 7Ah, 20A which is all I need for the q100h and how I plan on using this bike.

The real issue I'm having is figuring if I can set this bike up without anything at all on the handlebars. I definitely don't want a dashboard, and I'd prefer not even having a throttle. So that leaves me to a cadence sensor or a torque sensor BB to run this. The cadence sensor is easy, since they all run on 5v and output within the range of a standard hall sensor throttle. However, since the Phaserunner is using a torque throttle, I'd really like to use a torque sensor on the BB. The big issue is that most of the torque sensors I've found all need something like 8V-12V inputs and have multiple outputs (for cadence, torque, etc). They either need a CA3 or some sort of custom Arduino setup to work with a throttle input on a controller (the Phaserunner included).

The BEAMts looks like exactly what I'd want, but I can't find it for sale anywhere. It's been out of stock on ebikes.ca for a while now.

So while I've been searching around, I discovered that Bafang has a few different torque sensing BBs available and the PA211.32.ST looks like it would work great. Unfortunately, I can't find any info on this BB and I haven't found anywhere to order one to try it out. I submitted an enquiry through Bafang's website, but they expect a minimum order of 1,000. I commented that I just want one, but haven't heard back. I was wondering if any of you here have seen one in person or know where to get one.

These are the specs for the BB torque sensor:

Any input you guys may have would be greatly appreciated. Also, if you know of a source to get one, I'll be happy to be the guinea pig and I can update this thread with how the sensor works (or not) with a Phaserunner.

I made mention of this sensor a few years ago in a different thread, but got no answers..
The fantastic thing about this sensor is it does measure the torque on both, the left and right pedals.

I know from a friend who is working in a near ebike company that they got one sample from bafang, but i believe they never have installed or tried it out. Not sure it has torque sensing or only rpm signal.
Will ask him about it

Thanks madin88! Any info would be great.

I think I may have figured out a work-around. I found some super small momentary buttons on mouser which are small enough to fit inside the handlebar tube. I'm planning on drilling two holes into the bar (one for the button next to the grips, and one near the stem for the wires to exit). I'll run a resistor on the 5V from the throttle in to the button, so that when I press the button, it'll provide 3.5V to the throttle input. This will be spliced in-line with a PAS on the pedals.

This way, when I'm at a stop or I'm going up hill—where my pedal cadence will be low—I can press the button for full torque to the motor. Then once I'm moving, I can release the button and use the PAS to set the throttle input and motor assist. I do this now on my current bike using a traditional throttle and PAS, so I think this should work OK. With the Phaserunner's programability, I can adjust how abruptly the power comes on at full throttle so that I'm not doing unintentional wheelies from a stop.

I just wonder if the combined input of the momentary button at 3.5V and whatever lesser V the PAS is providing while pedaling slowly from a stop will be additive and cause the Phaserunner to think there's an over-voltage situation. I'm not quite sure how that part would work. If any of you with more electrical knowledge know the answer (and a potential work-around) I'd love to hear from you about this idea.

Thanks.

The two signals won't be additive, but one source could backfeed into the other and cause damage. If you place a small diode on each one, there is no possibility of back feeding so the controller will just respond to whichever signal is higher. Diodes will cause some voltage drop but I think you can calibrate to compensate for it.

Ahhh. That's the knowledge I needed! It's good to know that the voltage won't be additive. Good to know about using diodes to protect against backfeed. Yeah, I think I can figure out the diode situation now that I know what I need to protect the parts. Thank you fechter.

Here's what I'm thinking in terms of the wiring. The Hall sensor in the diagram is coming from the PAS.

I still need to figure out what diodes and resistor I need to get for this. I don't know what amps are running out of the throttle connector, so I'll need to measure that when I eventually order and receive the Phaserunner. I know the diodes will drop the voltage on the signal line, so I need to figure out how much of a drop it will be and if it'll still be in the range for the Phaserunner.

Side question:
Is a PAS just an on/off sensor, or does it provide variable signal depending on cadence? On my Bafang BBSHD, it seems like the PAS is more just an on/off situation with a soft ramp in the power output. If a PAS is just an on/off signal, I don't think I really need one; I'd probably just use the micro button I got since it's so stealthy. Worst case scenario (and for more control) I could just suck it up and get a small thumb throttle. I'm hoping that's not the case though, I really want totally clean bars on this build.

I've been doing some further searching here since I figure pretty much everything possible has been already discussed and I found some helpful threads. I'm putting them here both so I can get back to them quicker, and in case anyone else comes upon this and would like to know more.

General PAS control discussion:
https://endless-sphere.com/forums/viewtopic.php?f=3&t=69843

How a PAS works:
https://endless-sphere.com/forums/viewtopic.php?f=2&t=72948
This thread's last response was most helpful:

d8veh said:

Not all PAS sensors work the same, though all the common ones use hall effect sensors in them. Some have one and some have two hall sensors. The output is mostly the same - a signal that pulses 5v when the pedals are turned. Some are handed and all are directional, so the direction that you pass the magnet and the orientation of the magnet is important. If you mount the magnet disk the wrong way up, it won't work when you pedal forwards, nor will it work by pedalling backwards with it mounted the right way up. The handed ones won't work when you fit it to the wrong side of the bike. There's 4 combinations of magnet disk orientation and rotation direction. If they're handed, only one works. If they're not handed, two will work.

Only a few controllers relate power to crank speed. Most only respond to pulsing or not, in which case they give a fixed power, which would be adjustable through a control panel of it has one.

Click to expand...

Difference in PAS sensors (number of magnets):
https://endless-sphere.com/forums/viewtopic.php?f=2&t=78962

So it sounds like I most PAS work as a directional on/off system, and any variability in throttle would be handled by the controller if it supports it. Worse for me and what I want to do, you can't just plug a PAS into a throttle input because the PAS is providing pulses in sequence with the magnets when the magnets pass in the right direction. The PAS doesn't provide a variable voltage signal like a throttle does.

Sort of thinking out loud here, but I guess there's always the option of going with a TDCM BB and CAV3, and just installing the CAV3 into my seat bag with my battery pack. I could mount the CAV3 at the very rear, so I can just zip open the bag to access the controls when needed. The only downside I can see to this is that I would either need to set the PAS level on the CA ahead of time, or run a remote dial up to the handlebars. On my current BBSHD, once I've set the PAS, I really never need to change it while riding, so I don't think this would be a deal-breaker for me if I skipped running the remote dial.

If I abandoned the idea of using a Phaserunner, and went with one of the S12SH controllers from BMSbattery I could get the PAS input, but those controllers are huge bricks and there's so much spaghetti coming out of them. I'm fine with cutting off leads at the board to reduce the wire mess, but that box is so big, there's be nowhere stealthy to mont that thing.

I don't want this bike to look like a RadioShack on wheels. :?

I think I have a solution that solves all my problems. I guess I was overthinking this.

I have a few of these micro momentary buttons. What I'm thinking is that I can run two buttons—one on each side of the bars—that would give me 50% power or 100% power. using resistors to adjust the appropriate voltage, or a divider if I wanted to get fancy. It would basically function like a PAS with a 2-position PAS-level switch, but much cleaner since I wouldn't have a big switch hanging off my bars and I wouldn't need a spaghetti controller that has PAS capabilities built-in. These are just NO (normally open) single pole micro buttons, so there isn't any real fear of backfeed. I'd just need to get used to letting go of one button for the other, but that shouldn't be a big deal.

I still want to run these inside the bars, but if that proves too hard, I may mount them in little brackets that I can clamp with the brake lever clamps. I'll post any developments in my build thread: https://endless-sphere.com/forums/viewtopic.php?t=91309

Is there any new information about the Bafang sensor available? I found the related patent, where the mechanical setup is explained, but the function of the 6 wires is not described anywhere

regards
stancecoke

There's a bafang sensor in this thread that uses 5v, but it does still require a controller that understands a torque sensor, or a cycleanalyst. https://endless-sphere.com/forums/viewtopic.php?p=1393179#p1393179

I managed to figure out the pinout now.
The sensor is therefore suitable for use with our open firmware, but difficult to get and quite expensive. I think Sempu is the better choice.

You can find some further information in the german forum.

regards
stancecoke

Thank you so much for sharing that stancecoke, its really helpful. I think I'm starting to agree that the Sempu is the better unit in general. The sole exception being that this only needs 5v. Maybe this isn't an issue any more, but I seem to recall the Cycleanalyst having trouble with power load on some of the torque sensors that needed higher voltages. That info could be totally out of date and incorrect now though.

I just used the Bafang 68mm Imitation Torque Sensor successfully in my Arduino project.

It had 6 pins, but nothing seemed to work logically. Upon cutting the cable that connects to the 6-pin connector, I found only 3 wires. Simply put, I connected the self-explanatory red, black, and orange wires to my Arduino, and I was able to get a perfect digitalRead().

The sensor is indeed very sensitive to even the slightest rotations (not really torque). I plan to combine it with a simple HC-020K sensor to capture pedal rotation, hoping for a perfect setup.

zro-1 said:

Maybe this isn't an issue any more, but I seem to recall the Cycleanalyst having trouble with power load on some of the torque sensors that needed higher voltages. That info could be totally out of date and incorrect now though.

Click to expand...

The issue with the CA is that the higher the voltage, the lower the current available on the 10v supply (to maintain the watt limit on the regulator parts in the CA). AFAIK that hasn't changed, but you can verify with ebikes.ca if this is the case with units made these days.

You can simply use an external 10v (or whatever) DC-DC from your battery to power things if your setup would otherwise be outside the limits.

The V6 phaserunner has a 6 pin PAS connector. I'm assuming it's 10V but Grin's documentation isn't great.

The torque sensors T17 and T15 operates with 5 to 15 volts.