Sensor (throtle/pas) disable-switch

[underdog]

Hello.

I am experiencing some weird behavior from my controller when trying to make a basic on/off-switch system for pas and throtle on one of my bikes. If I only have a 1-phase on/off switch, which lead should it cut to the sensors (red/black/green) ?

Throttle:
If I cut the black (-Grnd) on the throttle, the bike goes psycho full-speed mode, so thats no good at all...
PAS:
If I cut the green (signal) on the pas, the bike gives me "error 03", and goes in to some kind of protection-mode, partially disabling throttle too...

Is it a viable solution to just cut red (+positive) with the swith on the sensors instead?

 

[alsmith]

Some controllers are for use with PAS only, are you sure it is capable of working both ways?
What controller make and model? Do you have an installation manual / pdf for it?

 

[mlt34]

I generally wire the pushbutton to cut the red (+5v) wire to the throttle, when there is no ignition wire on the ebike. No juice, no gas.

Sounds like your bike has one of those fancy shmancy displays that freak out at you when you try to custom wiring jobs though...

 

[d8veh]

The controller will give an error if it doesn't detect a throttle, so you can't cut any wires. If you want a switch to cut the power on both the throttle and PAS at the same time, put it on the brake connector. If you only want to cut the PAS, put the switch on the signal wire.

 

[Drunkskunk]

As D8veh said, if you want to disable the bike, put the switch on the brake wire.
The controller is looking for ~1 volt when the throttle is off, and ~4 volts when you are at full throttle on the green wire. If the controller gets anything outside of that range, it assumes a fault and shuts down. In theory, if the red or green wires get cut, you would see 0 volts, and if the black wire gets cut, the green would go over 4v. no idea why yours went full throttle, but it can happen if there is too much of a voltage drop on the red/green wires.

If you wanted to disable just one, you'll need the switch to put the voltage on the green wire to equal exactly whatever the 0 throttle voltage is, and override any voltage that comes in over that point.

 

[underdog]

As D8veh said, if you want to disable the bike, put the switch on the brake wire.
The controller is looking for ~1 volt when the throttle is off, and ~4 volts when you are at full throttle on the green wire. If the controller gets anything outside of that range, it assumes a fault and shuts down. In theory, if the red or green wires get cut, you would see 0 volts, and if the black wire gets cut, the green would go over 4v. no idea why yours went full throttle, but it can happen if there is too much of a voltage drop on the red/green wires.

If you wanted to disable just one, you'll need the switch to put the voltage on the green wire to equal exactly whatever the 0 throttle voltage is, and override any voltage that comes in over that point.

Thank you. This was clarifying.. And kind of logical.. I just thought it would be possible to do a switch-thing since the controller I use also work without throttle. But that is probably just because it is disconnected before startup... I guess the same 1 to 4 -volts apply to the PAS..(?)

 

[alsmith]

Thank you. This was clarifying.. And kind of logical.. I just thought it would be possible to do a switch-thing since the controller I use also work without throttle. But that is probably just because it is disconnected before startup... I guess the same 1 to 4 -volts apply to the PAS..(?)[/quote]

The PAS isn't a variable device like a throttle, it counts pulses triggered by a magnet passing a detector- the more pulses the faster you are pedalling.
It could be switched though to stop pulses getting to the controller.

 

[teklektik]

I am experiencing some weird behavior from my controller when trying to make a basic on/off-switch system for pas and throttle on one of my bikes.

If I cut the black (-Grnd) on the throttle, the bike goes psycho full-speed mode, so thats no good at all...
PAS:
If I cut the green (signal) on the pas, the bike gives me "error 03", and goes in to some kind of protection-mode, partially disabling throttle too...

Is it a viable solution to just cut red (+positive) with the swith on the sensors instead?

Taking this without further explanation, it sounds like you are trying to disable both throttle and PAS at once.

The best solution is as mentioned above - put the switch across the ebrake wires. This will work for any controller not only because it makes good design/safety sense, but because it's the law anywhere with pedelec regulations. With this approach, poor contacts/connection will leave your bike operational.

But answering your specific question about the +5v disconnect strategy:

Short form:
A switch interrupting the +5v to both the throttle and PAS will almost certainly work for the throttle and should work for the PAS but is a bit of a crap shoot - particularly because of the behavior you reported. Unlike the preferred solution, with this approach, poor contacts/connection may disable your bike.

Long form:
Different controllers handle disconnected inputs differently and so without knowing the controller type it's simply not possible to give a 100% positive answer to disconnecting anything except the throttle sense lead - any simple disconnection strategy may or may not work depending on the controller.

Here's some thoughts:

• For safety reasons, controller designers pay particular attention to the failure modes for intermittent or broken throttle connections - not so much for PAS or other connections. Controllers have internal pull-down resistors on the throttle sense line for safety so that either a broken +5v or sense wire will stop the bike because the throttle input will be dragged to 0v by the resistor. So - in your case, switching either +5v or the throttle sense connection will stop motor power by controller design. It's expected that the throttle input can be pulled below 1v since there's no safety risk to a stopped motor and this also allows the controller to be operated without a connected throttle if desired (PAS-only).
  
  • Regarding your question: as a matter of principle, it's a not good design to add switches to wires carrying small analog signals. Poor or intermittent connections in the switch will affect the signal and degrade normal operation. This is not as big an issue with the throttle +5v so switching that is sort of okay, but switching the sense lead is not a good plan.
• A broken throttle Gnd connection is arguably the most dangerous failure on an ebike because the +5v connection pulls the throttle sense lead high to WOT. You saw this when you severed the black throttle wire. This is a common behavior in many controllers and makes switching the Gnd wire generally unworkable.
  
  However, some controllers protect against this Gnd failure in either of two ways:
  • The voltage on the throttle sense leads causes an 'input fault' if it exceeds a safe 'guard band' voltage over the normal 4v max throttle voltage.
  • Much more rarely, the controller monitors the current drawn from the +5v lead to ensure the throttle is plugged in. This means that switching the +5v will always stop the motor - the question is whether this introduces an error state that can only be cleared by power-cycling the controller.
• There's no reason for controllers to do anything special with PAS inputs beyond perhaps a current-limiting series resistor - there are no rider safety concerns. Here the input is typically a digital hall sensor that sends pulses with voltages at or close to the rail voltages. You experienced an error when disconnecting the PAS sense lead, which frankly surprises me unless they did something special to the normal hall output voltages so disconnect is detectable (doubtful) or the floating input sent too-rapid transitions because of ambient electrical noise. With no knowledge of the controller or Error Code, it's a crap shoot guessing why this didn't work.
  
  • Regarding your question: if the error was due to the floating input and not some special voltage-based detection, then disconnecting the PAS +5v should pull the PAS sense input close to 0v which should appear as a legitimate no-pedal input condition. This should kill the PAS without error - but this is a reasoned guess, not a fact.

 

[underdog]

I see that I have much to learn, and over-simplified solutions based on uneducated guessing is not getting me anywhere. So sincerly thank you teklektik. This gave me a better insight in the world of controller behavior.

Idea for easy DIY separate on/off-switches for throttle and PAS scrapped for a CA3...

 

[docnjoj]

As an example of a good controller response: We had come to a rest stop and I shut off my switch on the display of my S12S controller. We stayed about 10 minutes and then were starting out,so I switched on the display and the controller but my controller wouldn't work. I examined my throttle trigger and found it stuck in the middle. The controller would not let the trike take off. I fiddled with the throttle and it sprung back to off. Then I pushed the throttle to about 1/4 and the trike started off smooth as silk. I believe this controller has a failsafe. It worked. A bit of lube on the throttle and it was back also smooth as silk. Kudos to the designer.
otherDoc