4126b and earlier issues with voltage gear lever

[Bernard Fife]

ashford,

Here is some changed code. These changes weren't specifically to address your issues, but they might help (they deal with how the lever and solenoids are changed in PNR shifts). These were done in conjunction with other users privately. It's helpful in troubleshooting if we all use the same code with the latest changes as much as possible.

Monitor_4127q.abs.s19

Uses same INI as 4.127p above.

(71.72 KiB) Downloaded 486 times

Also, I'd like to hear from anyone else using the voltage based lever configuration, particularly whether or not others are (or are not) having similar issues to ashford's. And if people are having trouble with the digital gear lever in these newer codes, that would be useful to know too (there have been many changes to try to address ashford's configuration, and it is possible that these changes may have screwed up the 'normal' operation of the code for other users).

Lance.

[Bernard Fife]

Here is code with a new function:

Monitor_4127r.abs.s19

Use INI below

(71.76 KiB) Downloaded 483 times

GPIO_MShift_4127.ini

Use with 4.127r test code

(215.22 KiB) Downloaded 457 times

This code adds minimum and maximum valid voltages to the voltage based lever code. These values can be set under 'Shift Input Patterns -> Configure Inputs'. The defaults are a minimum of zero and a maximum of 5 Volts, making all possible voltages acceptable.

However, if your lever signal drops to zero between shifts, and the minimum valid voltage for *any* gear is higher than zero (1.00 Volts, for example), you can set this value to a little below the lowest valid voltage (ex. 0.70 Volts), and any lower voltages will be completely ignored by the lever position determination code. Similarly, if the voltage jumps to 5.00 between gears, you can set the maximum to something between the highest valid voltage and 5.00 to filter out the invalid states.

One other thing I could add, if it is necessary, is to poll the input lever multiple times, and only 'declare' a lever position once the same gear has been established a certain user-specified number of times in a row (the shift buttons work this way).

This code is largely untested, so use only with great caution.

Lance.

[ashford]

my car is down for a few days while i fix/replace a torque converter.
however i have finished a 4l80 install on a friends car, is there any revision of the code you want me to try while i still have access to the car?

[ashford]

i tried out the "r" version today, it seems to be an improvement, it felt drivable 1-2 2-4 but felt like hyst needed to be broken to get the table gear. it may just be the timing of what i do or the time delay for error detection. i took 2 logs the earlier one was using lots of lever shifts the second of me just normal driving in 4th.

in the first log at 236 i shifted from D to N then back to D then it was stuck in 2nd tried going to 2nd and 1st then on the way to 2nd from 1st it cam out of it at 286. i still have P-R-N solenoid outputs set the same as 2nd gear if this helps out at all, with previous codes the "stuck gear" was always what the sol output for prd was and iam assuming the same is still true being transgear=0 or -99 when it gets stuck.

at 316 i did a 4-1 shift and it seemed to work fine.

a few other points of interest are 310-315 where it cleard by itself or a result of breaking hystersis.

another is 325.

[Bernard Fife]

I'll have a look when I get a chance.

[ashford]

test drove with 4129c no noticable problems with just a test drive other than 2-4 manual shift took a few seconds to let out of 2nd gear.

[Bernard Fife]

ashford,

That sounds good, it seems better than it was, at least.

Lance.