Well so I spoke briefly with Harlan .... he said any replacement would require reprogramming the controller. So that's out for me. ...
back from the dead - with the Technical Answer (as opposed to the short answer.... which end up being the same but may be of some interest)
Reprogramming of the controller is not that big of a deal.
Very minimal changes need to be made to the throttle map to convert back to 5K pot from Hall.
You can rent a CAN dongle for $50, pull the DCF, change it quickly, and flash it back.
No change to behavior on the bike... other than a slightly different throttle mapping.
Since the hall sensor was dual output... you may have some trouble with the MBB... which is another story. With the two signals they had the opportunity to either run both into the Sevcon as a differential input or run one to the Sevcon and one to the MBB. I would have to pull up the DCF to see which direction they went.
Safety is always first, but it seems that as time goes on the Zero is becoming more and more dependent and integrated with the MBB. This is fine for meeting requirements but it goes in the opposite direction of modularity and any chance of getting a foothold in other markets. It also impedes owners ability to perform modifications... which is close to my heart (no desire to ride anything stock).
My understanding is that the two hall signals are inverted.
The one that goes into the Sevcon is straight forward.
If the other goes to the MBB... for a comparison... they would either have to send *both* analog signals or the MBB would have to read the throttle state off the CAN... which would just make for CAN clutter... so I am guessing that both signals go to the MBB (if they do). In that case they may just be doing difference math (which means a dual pot would work) but more likely they have it bound top and bottom... which means to make the dual pot work (without reprogramming the MBB) you would need to run inline resistors at both the top and the bottom to bound the voltage swing such that it matches the hall output swing.
There would be no way to measure the linearity (Pot is likely linear while the hall has a curve to it).
So the hack is:
1) Remove dual hall throttle
2) Flash new DCF
3) Attach dual pot throttle
4) If no errors, great. If errors then calculate and add inline resistances to top AND bottom of both pot's
5) Test thoroughly... (dont just go out and wrap it with your fingers crossed
)
For Honda's and the like infinite "black box" solutions popped up over the years to allow folks to do as they please with minimal fuss. With something as sketch as the throttle on a 80hp peak electric bike... I would be hesitant to attach any black boxes that were not very well documented. If one existed and it was documented it would look like this:
1) Potted box (must be IP67 or better)
2) Pigtails coming out that emulate the dual hall throttle
3) Pigtails coming out that mate to the dual pot throttle (or cut both to cover multiple years)
4) Box reads the 5K voltage swing and maps it to a Hall swing
5) Done - guaranteed to work perfectly
6) Disclaimer - off road use only... do not use ... etc.
7) Cost? After initial engineering a box like that costs $20 to make so $50 retail... until the Chinese copy it... then $30 retail.
From a modification standpoint... one wants access to the MBB software. Never seen it myself other than looking over the shoulder of Peter or Nate (I purposely never checked it out from the repository as my major is in Software and I did not want to see any conflict of interest crop up. Whats in the box is not rocket science by any means.). I could duplicate it and replace the MBB but it would be 6mo work (after) the initial reverse engineering. Not worth it. Most people just reflash their DCF to eliminate the RPDO's, rewire the contactor, and have complete control over their bikes - tho with some feature compromise. A bit more rewiring may be required for anything running through the MBB... like profile select.
We currently do not include anything like an MBB in our Sevcon powered equipment... but we are getting there... as the limitations of the Sevcon bump up against the safety requirements of such a powerful system. There is a big difference between putting out something OEM and modifying (safely) something that is OEM. One shoulders all the responsibility while the other is "off road" or experimental.
Someone, somewhere could probably hack an older model MBB into your 2017... one which was suitable for a 5K pot... but that would be a pretty bass ackwards way of doing it. Likely fail.
Sounds like Harlan was right.
Its a No-Go.
Bummer... because I know that the best Zero engineers are Modders at heart.
I have a 2013 SR but I dont own it and I am not interested enough to mod it on hobby time.
If anyone wants to see a sketch of how its done... let me know.
I would not do it with an Arduino unless you have lockouts... its one of those rare instances where you keep in the analog domain... as its a simple level transition (scaling) with upper and lower bounds.
-methods
(see pic)