Poltergeists in my BBS02

[eBlur]

My BBS02 started acting very weird...

My setup:
- Luna hotrod programmed BBS02
- Luna mini display
- No shift or brake sensors installed (wires cut and heat-shrink-closed)
- Speed sensor installed
- 28T chainring
Mods to stock:
- DIY hack to fix throttle-with-pedal (https://endless-sphere.com/forums/viewtopic.php?t=93977&start=50)
- Gear-lube swapped for Mobil80

All my riding has been on single-track and I've probably put over 600 mi on the motor at this point.

Last week it started behaving weird - I got what felt like slipping in the drive unit, kind of a pop-pop like plastic gears slipping. I was pretty sure it wasn't chain-slipping on a work cassette as I didn't feel any noise back there, but I did have a filthy chain at the time. I cleaned the drive-train and rode today and it was mostly better, but did slip occasionally.

Then the weirdness started... the motor started running slightly - even in PAS0 with no throttle applied. It felt like PAS-2 level power was being applied and sometimes when as I peddled it would increase substantially, to what felt like 80% throttle. Without the brake-sensors installed - you can imagine this was quite an interesting ride on single track. I had to pull power to stop it. Sometimes after a power-reset it would behave normally again for while - and then start applying phantom power again. At one point I saw a 04H error (Throttle no homing) and so I tried running with the throttle disconnected, but the motor would run again - even on PAS0 and disconnected throttle.

Inspected all wires - they seem normal and undamaged. I've been riding a lot in wet conditions and perhaps been a little sloppy when hosing my bike down near the motor. My leading theory is water ingress shorting something out, so I'm planning to take the controller out and let things dry out for a while to see if it helps.

I can't find any posts with similar phantom power symptoms, curious if anyone has run into similar circumstances.

 

[amberwolf]

When throttle ground is faulty, this sort of thing can happen to any controller that doesn't have a shutdown for that event.

But if ti happens with the throttle disconnected, then it's more likely to be leakage current between the supply voltage and the throttle signal, inside the controller itself (or at the connector of the controller output to the throttle).

 

[eBlur]

Interesting theory on leakage current. Wonder if it’s a side effect of my DIY throttle fix. Can’t wait to dig into the potting again!

 

[DogDipstick]

Maybe the magnetic PAS ring behind the final drive ( big 44t crank sprocket) broke and is giving wrong signal... It is a magnetized multi pole ferrite ring glued into plastic that is fragile, and tell the controller when PAS is active..

 

[eBlur]

@dogdipstick - Do you mean the pedal cadence sensor that senses if you are peddling? If that were giving a false positive it seems like in PAS0 it still should have no effect. The leaky throttle current seems most likely as it feels like applied throttle. Pulled controller apart - can’t can see any obvious dampness or water ingress in controller but possible some water in gear side. The nice clay red mobil80 gear grease has turned muddy.

I’ll give it a few days to dry out and see if it does it again. I’m not quite sure how to investigate leaky current further - I suspect it might be a side effect of the DIY throttle fix where I tapped into throttle wire to ground PAS sensor.

 

[amberwolf]

Before digging into anything, simply connect a voltmeter to the throttle signal wire from the controller.

There should be no voltage on it with a throttle not connected, unless there are pullups or voltage dividers on it inside teh controller that put a voltage there.

If there is a voltage, especially if it's above around 0.8v, and especially if it is not constant, it could mean there's a leakage between that signal wire and a voltage source (uusally the throttle power voltage).

 

[amberwolf]

Maybe the magnetic PAS ring behind the final drive ( big 44t crank sprocket) broke and is giving wrong signal... It is a magnetized multi pole ferrite ring glued into plastic that is fragile, and tell the controller when PAS is active..

Even if the PAS sensor was problematic in some way, at worst it would never tell the controller that pedals were spinning, with a failure of pulses to go high enough in voltage (or long enough to activate). All it is is a pulse train, generated as the pedals spin.

If the ground voltage was too high (broken ground, etc), then the pulses woudn't drop low enough to cause the controller to see a "zero", so it wouldn't then see it cahnge to a "one", so wouldn't detect the pulses and engage the motor for whatever PAS was set to.

If the suply voltage was too low, same thing, in reverse.

If the signal voltage itself was partially shorted to something else, or partly open, same. If open, then no singal, and no PAS response.

If the magnet ring was only intermittently triggereing PAS, then you'd only get the same thing as just pedalling a little, stopping rotation, repeat etc.

There isn't a failure mode of the pulse train that would cause continuous PAS operation, except for something that shorts another pulse train of similar signal type/speed to the signal wire of the PAS.

Even that wouldn't cause operation beyond the power set in the chosen PAS level.


If there was an analog voltage coming out of the sensor, like wiht a torque sensor, *and* the controller responded to those proportionally, *then* it could have faiulure modes like a throttle. But if the controllers still had a power (or other) limit on each level of PAS response, it still couldn't exceed that limit of the chosen level.

If the controller is setup so that the throttle can override the PAS level limit, then a throttle signal failure of the right type can cause the problem being seen.

 

[Voltron]

It's very very much a complete maybe, but water in the level selecting buttons? Continuously holding down the up button engages the walk mode, so maybe it thinks the button is being held down?

 

[Voltron]

Also...

 

[eBlur]

Before digging into anything, simply connect a voltmeter to the throttle signal wire from the controller.

i'll try this tomorrow, however i'm not that optimistic that it will reproduce easily "on the bench" - as the behavior didn't kick in until 30m or so in the ride and seemed to be induced by pedaling. i also noticed sometimes when i pulled power and left it alone for 10 minutes it would behave normally for a little while again, which made me suspect it being related to a certain heat-level.

i'll do my best to get it to kick on, otherwise will hardwire the voltmeter in, strap it to the bars and go for a ride

just so i'm clear - i'm looking at the voltage across pin (7) and (8) on the throttle wire coming out of the controller that the throttle itself plugs into?

 

[eBlur]

It's very very much a complete maybe, but water in the level selecting buttons? Continuously holding down the up button engages the walk mode, so maybe it thinks the button is being held down?

interesting you should mention that... when the bike was taking off on it's own - i played with a few things to see what would happen. while frantically bashing the minus button to get to PAS0, i ended up holding it down and going into walk-mode - i noticed that this seemed to get priority and immediately cut the motor to a slower rate (walk-mode rate) but it didn't last long, a second or two later the phantom throttle kicked back in. it's possible i may have let up on the walk-mode at that point too as i had a lot going on trying to turn corners and avoid the trees on a runaway bike

 

[amberwolf]

Maybe time to hook up those ebrake lever switches?

Doesn't even have to be on the brake levers, it can be a simple on/off switch placed anywhere you like, but on hte bars next to the throttle is a good place. If the throttle has a built in switch not presently in use, it can be used for this.

If you happen to have ebrake levers you didn't use from the kit, you can take the switch out along with the connector, and wire it up so you can press and hold it to keep the motor engaged, and releasing it would cutoff the motor. The switch on the cable can be replaced with any on/off switch, toggle or button, etc., either latching or not, that behaves the way you need it to.

i'll do my best to get it to kick on, otherwise will hardwire the voltmeter in, strap it to the bars and go for a ride

Most likely that's what you'll have to do.

just so i'm clear - i'm looking at the voltage across pin (7) and (8) on the throttle wire coming out of the controller that the throttle itself plugs into?

Yes, from signal to ground.

 

[Voltron]

Next maybe maybe is water in the PAS sensor itself? Do you have the newer kind where it mounts directly to the bottom bracket, with no magnet ring?

P.s. I was doing that thru heavy crosstown traffic the first time my bike jumped to full blast by itself until I physically reached down and yanked out wires while dodging a bus until it stopped. I got a bar mount cutoff after that..

 

[amberwolf]

Next maybe maybe is water in the PAS sensor itself? Do you have the newer kind where it mounts directly to the bottom bracket, with no magnet ring?

Since it's still just a pulse train (rather than analog voltage), and since it still can't cause power to exceed whatever level was chosen at the PAS, this still couldn't directly cause the problem.

It *could* cause it if the contamination caused voltage transfer / current leakage into the throttle signal.

 

[Voltron]

Maybe something in your throttle mod? Transistor leaking signal current?