Page 2 of 5
Re: 4126b and earlier issues with voltage gear lever
Posted: Tue Jul 23, 2013 8:40 am
by Bernard Fife
ashford,
The following code could fix this.
The issue was one place in the code where if the trans was not in a forward gear (less than 1) the target gear was set to 1 for when the lever was moved into a forward gear. However, the code should have been checking for a lever position less than 1 AND greater than -99, spot that an error would not cause the target gear to go to 1.
This code is not tested, use with caution.
Lance.
Re: 4126b and earlier issues with voltage gear lever
Posted: Tue Jul 23, 2013 10:07 am
by Bernard Fife
Here's a better version:
Use with caution!
Lance.
Re: 4126b and earlier issues with voltage gear lever
Posted: Wed Jul 24, 2013 5:21 am
by ashford
i have the trans out at the moment waiting on a part, it may take awhile (crosses fingers)
Re: 4126b and earlier issues with voltage gear lever
Posted: Fri Jul 26, 2013 10:22 pm
by skihnia
I seem to have similar problems.
I just did first tests on a new install on a ZF 4HP24. Tested with 4.126 and 4.127 code.
digital input pattern:
P - 0-1-1
R - 0-0-1
N - 1-0-1
D - 1-0-0
3 - 0-0-0
2 - 0-1-0
1 - 1-1-0
When I move shifter one step at a time all is well. If I do quite common move from N to P (or D to P) it seems to take a while to change the input state and the input B goes high before input A goes low so I get a momentary state 1-1-1 and the code stops responding to gear lever position until power cycled. I got round the problem by changing #1 state to 1-1-1. Only now I can't use the #1 position.
I don't have a log available now but I can get one Tuesday.
Re: 4126b and earlier issues with voltage gear lever
Posted: Sat Jul 27, 2013 1:42 pm
by Bernard Fife
skihnia,
I don't think this is the same issue as ashford, as you appear to be using the digital inputs. Have you tried increasing the 'Digital Switch Debounce' value:
http://www.msgpio.com/manuals/mshift/V41tune.html#gs You can set it as high as 65535; higher values reduce momentary glitches.
Lance.
Re: 4126b and earlier issues with voltage gear lever
Posted: Sat Jul 27, 2013 3:06 pm
by Bernard Fife
skihnia,
Also, the controller isn't really designed to be shifted using the shift lever, so you don't really need to use the #1 position.
The manual lever position is only used to set the highest gear to which the controller will shift. If you put the gear lever in 1st, it will always stay in first, which is the same as manually selecting first gear using the shift buttons.
So if you want a specific gear, you should shift to it using the shift buttons - that's the way the controller expects you to select gears manually.
Lance.
Re: 4126b and earlier issues with voltage gear lever
Posted: Sun Aug 04, 2013 4:29 am
by ashford
i tried out the 4127b. it seems to have made it worse. if the lever is touched while moving the trans goes to 1st gear and the only way to get it to shift again is to come to a complete stop shift to neutral then to drive. i still have an oil leak coming from the bellhousing so i have to pull the trans again

.
Re: 4126b and earlier issues with voltage gear lever
Posted: Sun Aug 04, 2013 10:14 am
by Bernard Fife
it seems to have made it worse
Ashford,
Worse than what code version? Do you have a datalog and MSQ for us to look at?
Why are you using the lever while moving? The controller expects you to shift using the shift buttons, not the lever. It was set-up this way for a number of reasons, including the problem you are have of indeterminate inputs.
Or perhaps you mean you can't even put pressure on the lever (rest your hand on it, for example), in which case I'd say it is worn out or broken.
Lance.
Re: 4126b and earlier issues with voltage gear lever
Posted: Sun Aug 04, 2013 3:19 pm
by ashford
Lance wrote:it seems to have made it worse
Ashford,
Worse than what code version? Do you have a datalog and MSQ for us to look at?
from about 4116. 2200 the lever functions corectly
Why are you using the lever while moving? The controller expects you to shift using the shift buttons, not the lever. It was set-up this way for a number of reasons, including the problem you are have of indeterminate inputs.
various reasons, the trans has no engine braking in first and second unless physically in that gear position.
drag racing- dont want it to upshift or freewheel(no engine braking) when the throttle is lifted during tire spin, burnout, neutral when finished is passed.
sometimes i want more rpm to get ac cooling better in slow driving.
many people prefer the feel of manually shifing as opposed to swhich shifting which has no physical feedback.
Or perhaps you mean you can't even put pressure on the lever (rest your hand on it, for example), in which case I'd say it is worn out or broken.
Lance.
the switch works as designed. it is not a potentiometer, it has fixed resistors with pads for the wiper arm to contact with a small window of no contact at all
Re: 4126b and earlier issues with voltage gear lever
Posted: Sun Aug 04, 2013 3:51 pm
by Bernard Fife
ashford,
Here is code that does not set a flag (manual gear = -99) when no gear is found in the variable voltage lever mode.
The 'get lever()' function in this code simply reports the existing lever position until a new valid gear is established. It isn't the best solution, but you can try it and if it is better we can work towards something closer to ideal (maybe a timer so that when a valid gear isn't found, it is ignored for a user specified time before reporting an error state).
Lance.