Brake sensor and Impulse Evo

Hello All. I have an Impulse Evo drive (bought it for a project) and would like to figure out where the brake sensors are supposed to be connected. There are 6 connectors sticking out from the unit. I thought the upper two would be for the two brakes, but it turned out they deliver 6V for the lights. The red plug, I don't know what for (maybe configuration?), there was a jumper inserted already. The big one is for the power and CAN bus, the next for display and the lower one is connected to the speed sensor and the shift wire sensor. But where are the brakes??? Will highly appreciate if someone could help me to resolve this puzzle.

 

I

https://www.ebikemotorcentre.com/

 

they service those motors so should be able to help

There are no brake switch connections on that motor. They're not needed. It has torque sensor only.

There are no brake switch connections on that motor. They're not needed. It has torque sensor only.

Ok, thanks for the answer. That is what I was afraid of. In fact, the torque sensor on this motor is broken and I now tried to run it by simulating the torque sensor's signal. The motor starts at 90 rpm, but does not stop immediately when the simulated signal is off. That is weird - not safe to rely only on software when the motor must be stopped.

Ok, thanks for the answer. That is what I was afraid of. In fact, the torque sensor on this motor is broken and I now tried to run it by simulating the torque sensor's signal. The motor starts at 90 rpm, but does not stop immediately when the simulated signal is off. That is weird - not safe to rely only on software when the motor must be stopped.

Best thing to do is take out all the electronics and run the motor from an external converntional Chinese controller with LCD, then you can have throttle, pedal sensor, brake switches and whatever you want. I've not seen it done before, but it should be possible as all motors work the same. You would have to do some tests to see how fast the motor runs because Bosch ones run too fast at 36v when you use an external controller because the Bosch system steps the 36v down to 24v. I've no idea if the Kalkhoff motor does the same.

 

Have you got a Kalkhoff Impulse bike to put that motor in, otherwise how do you plan to install it?

Best thing to do is take out all the electronics and run the motor from an external converntional Chinese controller with LCD, then you can have throttle, pedal sensor, brake switches and whatever you want. I've not seen it done before, but it should be possible as all motors work the same. You would have to do some tests to see how fast the motor runs because Bosch ones run too fast at 36v when you use an external controller because the Bosch system steps the 36v down to 24v. I've no idea if the Kalkhoff motor does the same.

 

Have you got a Kalkhoff Impulse bike to put that motor in, otherwise how do you plan to install it?

 

I am building a recumbent trike and wanted to install it on that. I've also got an original display and a battery, so my goal is to utilize everything as much as possible. Chinese controller was actually my plan B from the beginning, in case I don't succeed. But for now I will still try to override the broken torque sensor somehow. Just need to understand the algorithm it works on. Seems it expects the pedaling force to be in sync with the RPM readings. Now it's not in sync, so the motor starts at 90 rpm and goes to 120 rpm when detects the movement of the wheel.

Thanks for the information about 24V - will check that.

ebike motor centre might have parts to fix it or know how to bypass it.

ebike motor centre might have parts to fix it or know how to bypass it.

 

Thanks, will contact them

The motor starts at 90 rpm, but does not stop immediately when the simulated signal is off. That is weird - not safe to rely only on software when the motor must be stopped.

a torque sensor has 4 wires: ground, 5V, pulsed PAS and analog torque. Sometimes the analog torque output is also pulsed, but that's rare.

To simulate the torque, you need to use the PAS signal as the override cut off switch to stop the motor when the pedalling stop. For sample code, there is a thread about adding a throttle to Halfords Carrera that uses Suntour HESC kit. If you are going to use an external PAS, just follow saneagle's advice and move the electronics to Chinese kit. It'll save time on adapting and debugging the code for just one case.

Edited July 6, 20232 yr by Woosh

a torque sensor has 4 wires: ground, 5V, pulsed PAS and analog torque. Sometimes the analog torque output is also pulsed, but that's rare.

To simulate the torque, you need to use the PAS signal as the override cut off switch to stop the motor when the pedalling stop. For sample code, there is a thread about adding a throttle to Halfords Carrera that uses Suntour HESC kit. If you are going to use an external PAS, just follow saneagle's advice and move the electronics to Chinese kit. It'll save time on adapting and debugging the code for just one case.

 

In Imuplse Evo, the torque sensor has three wires and there is a separate PAS sensor which is looking at the big gear's teeth. The motor does stop when it's missing the PAS signal, but after a second. Which is better than 8 seconds delay when PAS signal present and torque signal switched off, but still not how it should be.

Thank you for the suggestions. I think I am now getting closer to implementing my plan B.

Best thing to do is take out all the electronics and run the motor from an external converntional Chinese controller with LCD, then you can have throttle, pedal sensor, brake switches and whatever you want. I've not seen it done before, but it should be possible as all motors work the same. You would have to do some tests to see how fast the motor runs because Bosch ones run too fast at 36v when you use an external controller because the Bosch system steps the 36v down to 24v. I've no idea if the Kalkhoff motor does the same.

 

David Henshaw of A to B magazine did this successfully with the Panasonic unit in a Giant Lafree.

.

In Imuplse Evo, the torque sensor has three wires and there is a separate PAS sensor which is looking at the big gear's teeth. The motor does stop when it's missing the PAS signal, but after a second. Which is better than 8 seconds delay when PAS signal present and torque signal switched off, but still not how it should be.

Thank you for the suggestions. I think I am now getting closer to implementing my plan B.

You can rewire the new PAS signal to where the old PAS goes.

In Imuplse Evo, the torque sensor has three wires and there is a separate PAS sensor which is looking at the big gear's teeth. The motor does stop when it's missing the PAS signal, but after a second. Which is better than 8 seconds delay when PAS signal present and torque signal switched off, but still not how it should be.

Thank you for the suggestions. I think I am now getting closer to implementing my plan B.

Can't swear to it but as I recall my Evo II used to assist smoothly even with slow pedalling, which would need a fairly long time constant. If the system's evening out the power delivery when the rider ghost pedals isn't that a good thing?

Can't swear to it but as I recall my Evo II used to assist smoothly even with slow pedalling, which would need a fairly long time constant. If the system's evening out the power delivery when the rider ghost pedals isn't that a good thing?

depending on the programming, some systems let you use PAS mode some require torque signal to be over an 'analog zero' voltage before starting the motor.

I suppose you can replace the dead torque sensor with a throttle and fit a brake sensor to the existing PAS so that it will cut the PAS when you brake. That would be the lowest cost option.

Can't swear to it but as I recall my Evo II used to assist smoothly even with slow pedalling, which would need a fairly long time constant. If the system's evening out the power delivery when the rider ghost pedals isn't that a good thing?

But how fast would it cut off the power when you brake? Do you remember any delays after you stop pedaling completely?

I suppose you can replace the dead torque sensor with a throttle and fit a brake sensor to the existing PAS so that it will cut the PAS when you brake. That would be the lowest cost option.

Unfortunately, that does not work. The processor seems to monitor torque pulses and does not react on DC. The torque pulses must be somewhere between 0.3 - 1 Hz, otherwise the motor does not start.

There is also a red LED blinking inside the motor box, right after power on. I suspect that it somehow detects the sensor's health state and switches to a special mode - that can explain why the motor always runs at constant speed, 90 or 120 rpm... Just trying to investigate it a bit more while waiting for answer from Ebike motor center.

how do you simulate the torque signal at present?

how do you simulate the torque signal at present?

I generate a sine wave 3V and pass through a full wave rectifier.

But how fast would it cut off the power when you brake? Do you remember any delays after you stop pedaling completely?

When I stopped pedalling it would run on for maybe half a second - useful for hill starts etc. The constant motor speeds (and TS) suggest it's a current controller.

There are no brake switch connections on that motor. They're not needed. It has torque sensor only.

When you stall the motor (by braking) the current should stay the same, no sweat, until you stop pedalling.

Edited July 8, 20232 yr by AntonyC

most of these motors use can bus signals like bosch brose ect and yamaha motors use uart same as the bafang mid drive diy motors.

 

you can get in to the can bus with a Arduino setup

 

https://www.youtube.com/watch?v=utdwbQVrKlI:760

I generate a sine wave 3V and pass through a full wave rectifier.

is it microcontroller based? If it is, then it may be worth your time creating the code for it. @SW, that video you linked to is in German. I just wonder if you speak German.

no but subtitles sort of work in English but with can bus you can see the signals but it wont let you wright to the controller as the fkn thing is locked.

 

will be less hassle to rip out the controller and use a phase runner ect or see if peter can get the parts to fix it but peter is on a road trip atm round the eu but think hes back next week.

most of these motors use can bus signals like bosch brose ect and yamaha motors use uart same as the bafang mid drive diy motors.

 

you can get in to the can bus with a Arduino setup

 

https://www.youtube.com/watch?v=utdwbQVrKlI:760

Sure, but to interpret the messages is another thing. They just send a hex code, no human readable text.

is it microcontroller based? If it is, then it may be worth your time creating the code for it.

I have a function generator. Once I understand the logic behind this, can bring it to microcontroller or FPGA.

What kind of wave form (sinusoidal, sawtooth, trapezoidal or Square wave) starts the motor? Does it expect a positive minimal value to avoid limping mode?