Wow this is awesome news my participation this weekend has been poor once kids are in school tomorrow I’ll be free to test this looking forward to it!! Thanks @Woosh @Evian1040 @vfr400 for everything and to the original author.
My next mission is to build an appropriate enclosure for the Arduino I don’t have access to a 3D printer.
My next mission is to build an appropriate enclosure for the Arduino I don’t have access to a 3D printer.
Well done everyone.
Now see if you can hack the speed limit. One guy has already done it, but it could be tidied up with a bit more code and another input and output. He did it by splicing in to the speed sensor wire in the motor cable and using a speed sensor on the crank, which gave about 1/3 of the normal speed signal frequency. The only problem was the LCD showed the meaningless crank speed instead of the bike speed.
If you could cut the white wire and use the motor side as the input and the controller side as the ouput, then use a simple routine to read the pulse frequency and output it at 5/8 of the frequency, the 16 mph limit would become 25 mph, which is probably above the motor's max speed, and the LCD set to km/hr would actually be miles/hr.
Apart from just adding a throttle, this device could be very useful to give an alternative to the very unreliable torque sensor. just having the throttle will at least keep people going, but it would be brilliant if you could connect a normal PAS too to get a mid level power that can be over-ridden by the throttle.
All this is what the Speedict used to do. Read this if you want some ideas:
Now see if you can hack the speed limit. One guy has already done it, but it could be tidied up with a bit more code and another input and output. He did it by splicing in to the speed sensor wire in the motor cable and using a speed sensor on the crank, which gave about 1/3 of the normal speed signal frequency. The only problem was the LCD showed the meaningless crank speed instead of the bike speed.
If you could cut the white wire and use the motor side as the input and the controller side as the ouput, then use a simple routine to read the pulse frequency and output it at 5/8 of the frequency, the 16 mph limit would become 25 mph, which is probably above the motor's max speed, and the LCD set to km/hr would actually be miles/hr.
Apart from just adding a throttle, this device could be very useful to give an alternative to the very unreliable torque sensor. just having the throttle will at least keep people going, but it would be brilliant if you could connect a normal PAS too to get a mid level power that can be over-ridden by the throttle.
All this is what the Speedict used to do. Read this if you want some ideas:
That is already on my plans as when i disassembled the bike entirely to clean it i was trying to see where the speed sensor was and it was either inside the main connector from the controller to the bike (that sits outside the frame, next to the back wheel) or the motor.
This is definitely something to look forward as we get a better understanding how the bike was designed.
I think a big thanks should go to Tony @Woosh for his input in to this with the codes etc as not many who are in business would do this freely.
Just tested version 8.2 of the code. Works perfectly just as @Evian1040 has reported. Brilliant. I just need to tidy up the wiring and get a road test complete.
Completely agree with @Nealh big shout out to @Woosh wouldn’t have achieved this without him.
If I am ever in the Southend On Sea area I’ll buy @Woosh a drink or 2!
I think the speed derestrict is next.
Completely agree with @Nealh big shout out to @Woosh wouldn’t have achieved this without him.
If I am ever in the Southend On Sea area I’ll buy @Woosh a drink or 2!
I think the speed derestrict is next.
Well went to work and back with the bike and most of the journey was on throttle. (10 miles in total)
I can say it works beautifully, the battery is suffering now since after 5 miles it dropped one bar (4 bars in total) so maybe a battery upgrade from 8Ah to something higher is probably in order. Specially if we find the connector to de-restrict the speed.
I can say it works beautifully, the battery is suffering now since after 5 miles it dropped one bar (4 bars in total) so maybe a battery upgrade from 8Ah to something higher is probably in order. Specially if we find the connector to de-restrict the speed.
That’s awesome I’m going to aim to get a finished product on my bike tomorrow had to order another throttle.
A note I'm adding here for the future of de-restricting the bike.
I calculated that the carrera battery at 36V 8ah will last about 15 miles for the average rider on throttle all the time and today i can conclude that its correct.
If we manage to de-restrict the bike we need to upgrade the bike as it will deplete its energy way quicker, what i also noticed is that the battery has other connectors going to the controller with other voltages so this is a bit interesting why the controller is getting other types of power instead of just the main lines.
I calculated that the carrera battery at 36V 8ah will last about 15 miles for the average rider on throttle all the time and today i can conclude that its correct.
If we manage to de-restrict the bike we need to upgrade the bike as it will deplete its energy way quicker, what i also noticed is that the battery has other connectors going to the controller with other voltages so this is a bit interesting why the controller is getting other types of power instead of just the main lines.
you need one with an on/off button for safety in case the throttle gets jammed.
unfortunately, there is some sort of comms between the battery and controller, so upgrading would be complicated. Another project could be to reverse engineer the comms code and send it from the arduino. Somebody did that for the Yamaha system.
That is indeed bad news.
Would the comms be a battery level check or something more?
Probably a bit more. One of the problems with these system including comms is that the details of what it communicated and how is generally not publicly available.
I didn't know what vfr400 posted above that it had been reverse engineered for Yamaha. Part of 'right to repair' should include requirement to publish such data.
The reverse engineering will be from the likes of the pedelec forum.de, the diy ebike scene is very large compared to the UK. As it also cover all the other parts of German speaking Europe esp the alpine and low countries.
What’s the best way to hide the Arduino?? I’m trying to work out where I can squeeze it I want to avoid getting an enclosure but I may have to! Maybe a small waterproof junction box
Here is mine (3d printed box):
EDIT: I attached a zip with the STL in case someone wants it. I modified it to have a small gap to install a rubber ring, with this i tested and concluded that the water and dirt does not get in.
All you would have to do is make a hole at the bottom for your wires (i left it closed on all sides so the user can make the hole at any side and size)
Last edited: