Hot glue when you have it at the time you want would take care of that
But I doubt it
1 Like
That’s a good point, I will try to find one that’s the hardest to turn. I think that while it may turn, the weight of it (the spinning part) is too low to overcome the friction. But we will see thank you for reminding me!
1 Like
Very nice project! For me auto off will be an important feature since I removed all leds from the board (in order not to attract pesky cops). Now the only way to tell if it is switched on is looking at the capacity tester. Someone recently brushed against the on button while board was lying in living room, board was already pretty empty. Rainy days, stayed on for up to a week I guess. Cells at 2.45-2.5v. Fml
1 Like
I think you could save the cells with low current charging, I think this because datasheets for cells measure capacity down to 2.5V.
Thank you! You could add a small 3mm LED somewhere on your board that is not that easy to spot. There is an output for LED on the Button connector on the PCB.
Hello everyone, just a quick update.
I have received all the parts today and here is the first unit that I am going to iron out all the wrinkles in the firmware as well as add the features we discussed above. It is missing the MCU and needs cleaning but firmware first.
Now as I received a lot of requests of customers to bypass the fuse for them, I am thinking that I will do that as a standard, only a few people wanted the fuse and pretty much everyone has over-current controlled either by the BMS or ESC. On the picture that is 100A fuse “wire”
7 Likes
That looks as tiny as the one you sent me - how is the size really? And how much will this cost when you start selling them?
This one is a little bigger, it is 25.5mm (1") * 40mm (1.57") *9.5mm (0.37") The price will be higher than the regular switch since this has a lot more components but at the moment I can not give you an exact number since the BOM is not finalized. But I think it will be around 35-40€ I will have to calculate all the costs to give an exact number, it will be on the product page as soon as I have a stable number and an update here.
3 Likes
Hello everyone,
I have the first batch of switches ready but unfortunately it is not for sale yet. Long story short, I dont feel confident with them enough to release them. What happens is that the firmware was tweaked for my dual setup using VESCs 4.12, FOCBOX, FSESC 6.6. I assumed that that the setup would not change much apart from the battery voltage being on the output of the VESC when in OFF state because the gates of the FETs are pulled to ground, I accounted for that but what I was not expecting is that the motor rotation sensing did not work as predicted on 1WD, 2WD and 4WD setups.
So I am sorry but since I want this to be reliable because some of you want the switch “take control” over your battery almost completely I dont want to release a half working product. The project is still alive and will be worked on by us, what will happen is there will be a new revision of HW and FW soon to improve on the new discovered problems.
If you have any more suggestions I am open to discussing them with you. 
Martin.
9 Likes
Nice job there, keep it up!
One question though : does it also act as anti-spark? Do you plan on adding push ON / hold OFF?
Thank you!
It is just like as if you took a regular anti-spark switch but replaced the regular pushbutton with a microcontroller that senses the rotation of the motor, if it starts rotating it will turn the board on as if you pushed the button. There will be 2 modes, one will be a timer that will turn off after some (adjustable) time and the other mode will only turn the board off if you press and hold the button for 2 seconds. Of course, the modes will be user changeable.
1 Like
Just recently realized that @JTAG has this feature now also in his BMS (which is open source) - maybe check how he was doing it and learn from there?
2 Likes
Thank you for letting me know, I have been following that project almost from the beginning. I will have to take a different approach however. I will see how it turns out, I am sure I will figure something out or as you say learn from other people maybe I am missing something. Most probably thats the case 
Good stuff Martin that’s very responsible.
GitHub has @JTAGs source and it should be findable
But not sure if licensing slows for a commercial reuse like this
1 Like
The problem is that I think JTAGs BMS uses the UART to communicate with VESC regarding the motor rotation sensing which is not possible if the VESC is OFF completely. I may be wrong, I need to catch up on the improvements of his BMS. It may have been as much as a month since I looked at it closely.
And yes there might be an issue but I am not sure since I am not using the same hardware as him. If so, there is more than one way to tackle the problem, I think I have solved it about an hour ago but I will leave that to tomorrow. Solving these things at 1am never ends well for me
2 Likes
Canbus get a little power from the vesc when the wheel is turned afaik. So although not using the canbus protocol, the bms senses the power passed through and turns on the board, again afaik
1 Like
Oh yeah thats true, as the motor spins the voltage powers the VESC on as discussed in other threads before. Thats a pretty clever solution. I will take a look and experiment tomorrow, if it would need too much hardware and hence increase the physical size of the switch I will have to find a “better” way for this application and keep it small.
Yeah “my” (actually @willumpie82 's ) method is indeed extremely complex!
2 Likes
Great news!
I have finished all the necessary changes in hardware and software as well as bench testing and real world testing. The problem turned out to be the precharge circuit, now the switch is tested and working flawlessly with 13S and 11000uF capacitors which is equivalent to 8 FOCBOXes in parralel on 13S!
Here is a quick (and quite shaky 
) video from todays testing. I have been starting and stopping the switch for several hours in both modes and I am happy to say that it works well.
I removed the audio track because it was so windy that you could not hear a word anyway.
At the beginning of the video I point to the rear light that is powered from the 5V from the VESCs to demonstrate it is off, same for the remote. This was recorded in the pushbutton mode which requires to be held for 2 seconds to turn off, than it can be started again with pushing the board.
You can expect better looking instructional video soon! The switch will be available at the beginning of next week! I am so excited!
Thank you all for being so patient and understanding, Martin.
15 Likes
A bit off topic, but how you like the alien power remote? Can you recommend it?
It is pretty good, I have not had the signal fail in a long time. The downside is that it is not slim, I am planning on 3D printing a better shell for it. I just struggle to find the free time right now
1 Like