I installed an Ultimate9 evcX controller a few days ago. Here is a picture of my install:
The processor is on a very short pass-thru harness, so there's no easy option to mount it out of sight. Mine is currently adhered to the plastic footwell panel with double-sticky tape - I may look at moving it behind that panel. The wire harnesses are very short - I wish they were a few inches longer to provide more flexibility. I did armor the exposed wires with split loom. Even out in the open like this, the controller isn't obvious when getting in and out of the vehicle unless you squat down in the open doorway.
I routed the USB controller cable up across the dash and back down to my mounting location. Inside the dash, I protected the USB cable with some 3/8" fiberglass loom to makes sure it isn't chafed by the metal dash structure and also used the loom to store the extra cable length (about 18" extra) out of harm's way. The controller is mounted to the lower dash panel with double-sticky tape. As others have mentioned, the smart phone application is nice to use, so I may dismount the controller and tuck the USB cable out of sight. The USB cable is not detachable from the processor.
I started out with the factory setting and didn't notice much difference in the throttle response. Then I tried the adaptive setting for about 10 miles. I didn't really notice much on that setting except that the initial dead spot - where you press the throttle pedal and nothing happens for the first few percent of input - that dead spot seemed to be gone. While in adaptive mode, I turned onto a major road from a stop and - with minor and what I would consider normal throttle input - accelerated from 0 to 45 mph, ending up in 5th gear in about 3 seconds (the speed limit is 35 mph). It was an unexpected result, out of line with previous throttle behavior, WAY out of proportion to my throttle input . . . just way too aggressive, almost to the point of approaching traction limits. NOT COOL. So I pulled over and, as a counterpoint to the craziness I just experienced, tried the low-traction mode. I started from a stop on a 6° grade and could not accelerate above 19 mph . . . even at full throttle, the maximum speed I could achieve was 19 mph. My wife was holding my phone with the app open and she couldn't find the factory setting for a few seconds. Traffic started piling up behind us and she's freaking out . . .
"Honey, it's okay. Take a deep breath and look for something like Factory." She finally found it and we resumed speed with a normal throttle response.
My next experiment was to try the "Ultimate" mapping on setting 5. This mode also gets rid of the initial lag and provides what seems to be exponential throttle curves above and below a linear throttle. Here's my mathematical approximation of what the controller is doing:
In the first graph, I'm approximating the evcX Ultimate mode and setting 5. The throttle output curve (in red) is superimposed on the OEM linear throttle curve (in dashed gray) except the delayed response for the first few percent of pedal input is removed. You can just see the delay in the dashed gray line near the graph origin.
The second graph models setting 7 where the throttle pedal input is modified so the output rises
faster than the input. The effect here is the throttle is more sensitive to input near the beginning of pedal travel, leading to the impression
"my car is more powerful" because for an increment of throttle pedal input, you're getting more power than a linear curve would deliver.
The third graph models setting 3 where the throttle pedal input is modified so the output rises
slower than the input. The effect here is the throttle is less sensitive to input near the begininng of pedal travel, leading to a driver's ability to more accurately meter the throttle since larger pedal inputs are required to effect the same power ouput that a linear throttle curve would deliver.
I settled on the Ultimate mode, setting 6 for on-road operations. I experimented with setting 4 on-road and think it will be what I use off-road. I have noticed the simulated high-stall effect in reverse (especially reversing uphill) that
@Clark Kent commented on previously, at least in the Ultimate mode.
For those interested in some math, the kind of curves I've modelled above are commonly called "exponential" or "expo" for short. An exponential transformation of the input might look something like
Code:
exponent = 1 / (1 + (setting - 5) * k)
output(%) = (input / 100 ^ exponent * 100)
where setting is the Ultimate setting 1 to 9 and k is a constant, ~ 0.3. As k increases, the curve rises faster above the linear curve. This kind of transformation exaggerates the slope of the output line near the origin.
A sine transformation can be used to provide a more symmetrical rise and fall above the linear curve. The approximations above use a sine transformation like this:
Code:
A = (setting - 5) * k
output(%) = input + A + sin(input * π / 100 )
where k is a constant to scale output, ~ 6.3 in the data I showed.