Page 1 of 2

non-CANbus will not come out of bottom of shift table

Posted: Wed Feb 05, 2014 1:16 pm
by 1991blazer
Been running my system for about a month now in my truck and working great except for one very irritating problem. About 30% of the time after the vehicle is started the controller will not shift out of 1st. To fix the problem I pull over and turn off the ignition and restart and it will work correctly after that. What is happening is that the non-CANbus load using only the bottom of the shift table, it will not go above 20 on the load regardless of speed or load. And yes I am using a TPS for non-CANbus load since this is on a diesel. Is there something in the code that will cause it to lock up like this? It only happens after starting the vehicle, never while driving.

Aaron

Re: non-CANbus will not come out of bottom of shift table

Posted: Wed Feb 05, 2014 2:19 pm
by Bernard Fife
Aaron,

It possible there is something odd in the code. If you post your msq (i.e. your settings file), and ideally a datalog of the problem, I will have a look.

Lance.

Re: non-CANbus will not come out of bottom of shift table

Posted: Wed Feb 05, 2014 2:52 pm
by Bernard Fife
Aaron,

Without seeing your MSQ or datalog, there is one thing I can think of that *might* be happening: if your non-CANbus load parameters don't reflect your TPS output, and it might give you a false very high load value when looking up the next gear. That could keep the trans from shifting. So you might want to make sure you load values are correct. See: http://www.msgpio.com/manuals/mshift/V41tune.html#gl I believe the default values are zero for the intercept (Load at zero volts) and 20 for the slope (Multiplier).

The reason I suggest this is that a very low load value should make the trans shift earlier than normal (that's what is normally in the bottom row of the table), which isn't what you are describing.

A datalog and an MSQ would really help, though.

Lance.

Re: non-CANbus will not come out of bottom of shift table

Posted: Mon Apr 14, 2014 6:33 pm
by 1991blazer
I have datalogged the problem and seems to be getting worse at times. I have to pull over stop, put it into Park and shift into D, will work about 30% of the time. May have to pull over about 3 to 4 times to get it to reset. Doesn't seem to be a boot problem but the system not allowing AutoMode even though it is clearly in auto on the screen. Other than this one problem it runs real good now that I have the shift points and pressure tables just right.

Re: non-CANbus will not come out of bottom of shift table

Posted: Mon Apr 14, 2014 6:43 pm
by 1991blazer
My load values are right on where they should be so the Multiplier shouldn't be a factor. Everything looks stable as far as VSS, Tach, Gear Input and Output, and Gear selected. Just doesn't make sense that pulling over and putting in Park then shifting to Drive will make it go away. Doesn't look good pulling out into traffic with only 1st gear and my truck hits the diesel governor at 25 MPH in 1st.

Re: non-CANbus will not come out of bottom of shift table

Posted: Tue Apr 15, 2014 3:59 am
by tobmag
Everything looks stable as far as VSS
Not really when I look at your log, you have the same noice that I had as in my GPIO built for 4L80E.
This can make the code "crash" and lock up in first gear until you re-boot the ECU.

Below you see what I mean about noice in the VSS signal.
First is when you come to a stop and you get a huge jump in the VSS signal
VSS1.png
VSS1.png (15.35 KiB) Viewed 16635 times
Second is when you start again and you get an even bigger jump;
VSS2.png
VSS2.png (17.24 KiB) Viewed 16635 times


I suggest that you cure that before doing anything else please read this thread on how to fix it.
viewtopic.php?f=4&t=678

Dont know what code you use but there has been some issues with non-CAN Load in the V5 code but its cured now....


BR//Tobmag

Re: non-CANbus will not come out of bottom of shift table

Posted: Tue Apr 15, 2014 6:05 pm
by 1991blazer
Thanks for the info. I did read your particular problems with your VSS and I know my problem stems from that circuit. Attached is a log of the problem of the spiking VSS causing shifting fluctuations. Just happened to have my laptop at the time. The log is just around town driving with slow acceleration and slower speeds. If I am stopped and then release the brake the truck will move forward just a couple of inches and then the ECM goes nuts. It will shift all the way to 4th when I am just doing 2 or 3 mph then back down to 1st and then it shifts great from there. What is causing this is an immediate spike to 100 mph (max setting for speed) once the vehicle just moves just a few inches. No other problems are present and the VSS is rock solid after about 5 mph with the mask set at 80.

Since this is an older GM truck with a DRAC unit that outputs the correct VSS signal to the ABS, factory speedo (I'm not using the Mega as the speedo drive since this system works perfect as is) and cruise control, I may be able to use the output on this device as my VSS input if Mega Shift. But it is converted at that point to I believe a DC signal so may have to do more research on this.

The other problem I notice is that the Tach speed is way off from what it should be. When cruising at 65 mph I should be turning 1,900rpm but Mega says 2,600, seems weird.

Any ideas?

Running 4.135 code on a 4l80E

Aaron

Re: non-CANbus will not come out of bottom of shift table

Posted: Wed Apr 16, 2014 6:13 am
by tobmag
If I am stopped and then release the brake the truck will move forward just a couple of inches and then the ECM goes nuts. It will shift all the way to 4th when I am just doing 2 or 3 mph then back down to 1st and then it shifts great from there. What is causing this is an immediate spike to 100 mph (max setting for speed) once the vehicle just moves just a few inches. No other problems are present and the VSS is rock solid after about 5 mph with the mask set at 80.
The reason for this is that the current build of the VR circuit on the GPIO cant handle low frequenses, its not working accurate below 50hz and this is what has to be fixed.
This is the reason why you have the huge spikes when starting or coming to a stop.

There are 2 solutions for this in the thread, the one I posted viewtopic.php?f=4&t=678&start=30#p3779
And the circuit that Ashford posted viewtopic.php?f=4&t=678&start=40#p3791

I have tested some ready built VR conditioner that uses the MAX9926, excellent circuit but very sensetive to external noice so in my car it was useless. It might work in your application though.
The other problem I notice is that the Tach speed is way off from what it should be. When cruising at 65 mph I should be turning 1,900rpm but Mega says 2,600, seems weird.
Are you using the ISS sensor as well? I think Lance needs to answer this one....


You should also consider to load the latest stable code that is 4.139

BR//Tobmag

Re: non-CANbus will not come out of bottom of shift table

Posted: Wed Apr 16, 2014 1:25 pm
by Bernard Fife
1991blazer,

I haven't had a chance to look at your log yet, but if this is a VSS issue, you might want to check the tips here: http://www.msgpio.com/manuals/mshift/vss.html

Lance.

Re: non-CANbus will not come out of bottom of shift table

Posted: Wed Apr 16, 2014 4:37 pm
by 1991blazer
Been there done that on the tips page. VSS is very stable after about 5mph so it is the very low frequency pulse problem that others seem to have. I just ordered some of the chips referenced in other posts and other stuff from digikey today to get these two input circuits working correctly, other than this one problem I love the system. Will keep you updated with what I find.

Thanks for the help

Aaron