Cycle Analyst V3.14 Delayed Speed Measurement Bug

[fordtheriver]

I'm having trouble reading and datalogging vehicle speed during initial vehicle acceleration with the Cycle Analyst. I think there might be a bug in the Cycle Analyst V3.14 firmware. Here is a screenshot to show what is going on.

The areas circled in orange show battery current and vehicle speed as I accelerate the vehicle from rest. The cycle analyst is able to capture battery current immediately, but doesn't capture vehicle speed until about 3 seconds after I've started. This is shown via the speed measurement instantaneously jumping from 0 kph to 12kph at 43 seconds.

The rest of the chart shows that the speed and battery current measurements are smooth and responsive once the vehicle begins moving. The issue I'm talking about can be seen by the sharp jumps in speed when the vehicle is accelerating from 0kph.



Here's are the relevant parameters:

Number of poles: 9
Wheel Circumference: 2262mm
Data Output Rate: 10Hz
Display Averaging Period: 0.08 seconds
Firmware Version: 3.14
V3.0 Cycle Analyst

Is anyone also seeing this issue? Is there a fix for it?

 

[amberwolf]

What are you using for a speed sensor, and what triggers it?

If it's a separate wheel-mounted magnet and frame mounted sensor, with just one pole, it can take a moment to register changes in speed (though not sure it would take 3 seconds).

Also, if the magnet is not optimally placed, it may not always trigger the sensor validly, so the CA may discard some or many readings (filtering them as noise, presumably; not something the user has access to change), and the reading may be delayed and may also be incorrect at times if not always (reading too slow a speed).

Similarly, if there is electrical noise (induced from higher current wires nearby, for instance) on the speedo wire(s), it could make it harder for the CA to correctly read the speed until the current on the other wires drops.


If it's a hall sensor in a hubmotor, with many poles, it should be "virtually instant" readings.

But the electrical noise issue can be even worse there (which can also cause the controller grief reading the motor position and cause rougher / noisier starts).


I do recall some delay in one or more values / responses in a version of 3.x firmware, when using the High Range shunt mode (I had up to 8 seconds of delay), but as that was a few years back I can't recall the details. I am sure I posted it in my SB Cruiser thread, and probably in the big CA v3 "beta" thread as well, though, if you think the info might be useful.