When I tested 5102tii I had no gear gauges on TS. I did a lot of testing over the weekend, I loaded 5102ti and changed Max. Lever Position parameter value from the current value of 1 to a value of 4 this time I noticed an improvement. At a medium/slow speed the faults quit and I took a log changing speed and positions P to 4, R to 4, P to N and as long as I went over 2 gear positions at a faster speed it faulted.The log showed no common elements
when current gear faulted to "6" but what ever you changed in 5102ti seemed to help!
My trans (Subaru 4eat phase 2 version 2 - 2007) uses a switch for lever positions a contact for each position and a wiper. I put resistors between the positions and one to ground and it gave me the voltage drops, 5v on the first resistor and Ford voltage input to the wiper.
I looked at liner pots and I could connect one without to much trouble but their expensive, do you think something like this would work?
I do appreciate your help on this.
- (676.12 KiB) Downloaded 159 times
Linear pots are nice, but they are expensive and have some other potential issues, so if this was mine I would probably try to make what I had work. But I don't understand 100% what shift lever hardware/signals you are dealing with either...
I thought about the linear pots and your right it wouldn't work much different than what i'm using now.
I didn't remember having this issue with 4146 so I loaded it and didn't have the 6 gear fault that i'm having with
5102. Not sure whats different but 5102 doesn't like my shift lever setup I'm using 4146 for now but the load doesn't work
as well as 5102!
I added 2 pics to show the difference, shifting from park to 4th and back to park.
- 5102_log.JPG (53.25 KiB) Viewed 3802 times
- 4146_log.JPG (49.55 KiB) Viewed 3802 times
I will have a look in the code.
In the meantime you might want to set the minimum valid voltage to approx. 0.9 Volts, and the maximum valid voltage to about 4.1 Volts. That should constraint the errors if the voltage momentarily spikes.
Then if this was mine, as an additional measure I would set the number of gears to 8 temporarily, and set the voltages for each of the unused gears (5,6,7,8) to around 4.9 Volts before setting the number of gears back to 4.
The manual doesn't have the information for the minimum and maximum voltage settings yet, but I will add those shortly.
I looked into the code. It was checking for error states in the voltage based shift lever code, but it doesn't look like it was doing this properly ( - I wasn't doing this properly, actually!).
The attached 5.102v test code should be better, but I haven't tried it yet.
If you decide to try it, use a lot of caution!
I would still set the maximum and minimum voltages as above, though.
Here's a version (5.102vi) that should be even better. It has the lever checks rearranged to be more efficient in the lever code, initializes the target gear better when shifting from PNR to D, and also checks the target gear validity more often in the main loop of the code.
This code uses the same 5.101v INI as above.
Again, use with caution.
I loaded 5102vi entered only the lever parameters when I shifted no 6 gear fault but Need burn light came on cleared when cycled, back on when shifted. loaded rest of the parameters and 6 gear fault came back changed gears to 8 set (not used) gear values to 4.9 and got fault back except it cleared it self by stepping down 6-5-4-3-2-1 ( pic for this ) changed back to 4 gears and same problem as before interesting results though on 8.
I hope this feedback helps!
- (59.5 KiB) Downloaded 158 times
- 5102vi.JPG (45.25 KiB) Viewed 3773 times
Yes, this is very helpful. I will need to dig around in the code for a bit, but I will let you know when I have this fixed.
I notice in your MSQ that you have the number of forward gears set to 8 under 'General Settings -> Gear Ratios'. You need to set that back to 4 (since you have set the 5-6-7-8 voltage appropriately) as I believe your transmission has 4 forward gears (you set it to 8 only long enough to change the voltages for the invalid gears).