I have a bit of a mystery here -- a throttle that is putting out a max of about 2.5V according to my CycleAnalyst 3, and therefore only registering half throttle. Let me give some history:
* The bike is a two year old GM Magic Pie build that performed flawlessly with a 36V LiFePO4 battery even older than the build.
* The throttle runs through the CA3 using the method that breaks the green wire at the (Thumb) throttle and runs the throttle end to the CA3, the squid cable end to the controller as it would have before the break.
* Recently I replaced the 36V battery with a 52V LiMn triangle pack from Luna. I tested this build over about 40 miles and there was no problem ... very powerful with max speeds you had to be careful with, long battery life (per charge). I read the posts describing the interchangeability of throttles re voltage and encountered no issue
Now the problem:
* I just did a 'clean up build' that mainly involved reharnessing and binding wires ... the only real changes were resoldering the green wire from the squid end of the throttle cable to the CA3 connector
* I'm now seeing that the throttle indicator on the CA3 does not go over half ... that is, when I fully depress the thumb throttle the visual CA3 indicator only goes halfway
* Checking the CA3 throttle settings I find them unchanged, with a 1V low threshold (when it starts noticing action) and a 4.75V cutoff for full power, a little less than the nominal 5V. This has been the setting for ages.
* Switching the CA3 to show throttle voltages I see at full throttle ... about 2.6V in, and ... about 2.6V out. I conclude that the CA3 is operating properly and giving the controller all it has, but it only has a max of 2.6 volts to work with
* Moreover the throttle itself is getting very hot. Which I may not have noticed before, since there was no problem, and may just be the 36V LEDs.
After a chat with Gary I tried some isolation techniques, cutting the CA3 out of the loop by reconnecting the green from the throttle to the green to the squid harness.
The problem with this approach is that I now have no way to measure the throttle voltage --- a voltmeter from the green connection from throttle to any ground I've tried shows nothing. Which probably means I just can't find a decent exposed ground relative to the throttle. I don't want to cut more throttle wires because the connectors seem impossible to locate for replacement.
So ... I tried a simple analog test on the road:
* Before these problems the bike would do well over 19MPH on the flat. If I let it.
* Running through the CA3 I get a maximum of 16-17MPH on a specific stretch of road
* Running direct through the controller and bypassing the CA3 I get maybe 1MPH more on the same stretch, far less than a full voltage would give
I'm attributing the very minor gain in throttle in the bypass CA scenario to cutting out the CA3-s ramping function ... if I were really getting a full throttle I believe the difference would be much more.
What are your thoughts on this?
* Have you heard of an older well-used throttle not putting out a full 5V at full, either for mechanical or electrical mechanisms.
* Do you have a better way of measuring throttle output bypassing the CA3 without savaging the connectors, e.g. find a decent ground relative to the green wire?
* Do you agree that my MPH test shows the CA3 is innocent of wrongdoing? I admit it was crude
* Do you know *anywhere* I can get some 5 pin GM female and male connectors so I can measure voltages decently? I'm in the market for dead throttles if anybody has them
* Are there other throttles that work adequately with the CA3?
Appreciate any ideas, thoughts or places I may have gone wrong. The bike is quite usable as is, but it worries me that it's not quite right, might not have the power to avert some accidents, and the heat thing bothers me. Or it did until I accidentally shorted the LEDs in a test and they fried. No heat now (same behaviour). Can I just cut that circuit board out?
Much appreciated any thoughts.
R.
