Need some help with Load & tach in a diesel

[mcneil]

I'm having trouble with setting up the load and tach. The application is a VW 1.9L TDI, 4L65e adapted to a NP241OR transfer case in a Jeep. The rear end is up on jackstands and the front driveshaft is out while I figure this out.
I can't get it over 20 mph like this, and I think it has to do with the load vs rpm table.

The load signal is tapped off the electronic pedal, varies from .4 V at zero throttle to 2.1V at max.
I don't have a tach signal connected.

Questions:
1. Will no tach signal prevent proper operation?
2. What's the right way to adjust scaling of the pedal signal?
3. What's a good way to approach the load table for a diesel? Pedal position directly controls fueling, and correlates pretty well to load independent of MAP. I turned down the load smoothing as a start.

[Bernard Fife]

mcneil,

I tried your msq on my bench, and have no issues getting over 20 mph, so the problem isn't the code directly.

Will no tach signal prevent proper operation?

Yes, it could, depending on your settings. However, make sure your are using 'MAP + Speed' for your load table (you are in your posted msq) rather than 'Tach + Load' (which won't work without a tach signal, obviously). Also make sure you have 'Do not enforce RPM checks' enabled (you do). I would also set the 'Error Behavior -> On Error' to 'Ignore Errors' at least for now. You will also want to set up the TPS Load table so that all the columns are the same. If this was mine I would make the table with the load values equal to the TPS values, like this:



I suppose you could put in a 0 rpm bin for the first column (instead of 500 rpm).

There is more on this here:
http://www.msgpio.com/manuals/mshift/V5tune.html#al

What's the right way to adjust scaling of the pedal signal?

This is covered here: http://www.msgpio.com/manuals/mshift/V5tune.html#ol

What's a good way to approach the load table for a diesel?

The best way is to experiment, and find what works for you and your vehicle. TPS shouldn't be a problem at all, and it always takes some fiddling to get the table right (and there are usually some tuning 'blind alleys' that have to be dealt with. But more than anything it takes time and some careful thought. There is more detailed advice here: http://www.msgpio.com/manuals/mshift/tuning.html

Lance.

[mcneil]

I got it - I had a 4WD correction factor in there, but should not have. In my setup the vss is on the output adapter between the 4L65 and the NP241.

I don't have gauges yet so I read the tach off serial/CAN from the VW ECM - I was able to hold 3 speeds (2000, 3000, 3800), and saw that vss was consistently off by about 2.5. Change the 4WD correction factor to 1, and it works great. I probably should have thought that out sooner, but I was set in thinking I would use the 4WD factor to correct speedo scaling when in 4WD low. Which isn't necessary, but nice to have.

Before the 4WD factor fix

2015-12-26_07.41.33.msl

(415.96 KiB) Downloaded 928 times

2015-12-26_07.34.45.msq

(59.16 KiB) Downloaded 915 times

[mcneil]

Here's the after:

2015-12-26_07.45.03.msq

(59.16 KiB) Downloaded 1248 times

2015-12-26_07.45.11.msl

(87.58 KiB) Downloaded 1258 times

This is my arrangement for the VSS & Tach signals. A little convoluted but its mostly working.

HowToDisplayRPMonaJeep2.jpg (87.21 KiB) Viewed 26208 times

[mcneil]

At this point I have what feels like a driveable tune on the transmission and calibrated the pedal scaling. Thank you Lance for all the help on this!

Here's my next round of questions:
4. If I build up VR3 as in the last paragraph on this page, will that work for reading a 0-2.5V tone?
http://www.msgpio.com/manuals/vr.htm

5. Do you see any issues with the above scheme for VSS & Tach? The VW EDC15 is expecting a 0-5V signal where Hz= speed in km/hr. In the Jetta/Golf setup, the Instrument cluster would read the VSS, apply conditioning and scaling, then send to the ECM. I know the 0-5V signal works because I can do it bit banging an Arduino. I would have the Arduino read the speedo signal from the Mshift, but the ECM needs road speed to make all the OBDII features work.

6. Give above scheme, do you recommend building the PWM4 circuit per the instructions or doing anything different?

[Bernard Fife]

4. If I build up VR3 as in the last paragraph on this page, will that work for reading a 0-2.5V tone?

mcneil,

That should work. Building it as a VR circuit should work too. So you could do it either way; if it was mine I would build the digital input as it's a simpler build without the closely spaced transistor leads.

5. Do you see any issues with the above scheme for VSS & Tach?

I don't see any obvious issues at first glance. However, I am not sure what sort of output frequencies you will need, and the scaling will need to be appropriate. There is more here:

http://www.msgpio.com/manuals/mshift/V5tune.html#s

and a speedo output ratio calculator here:

http://www.msgpio.com/manuals/mshift/V5tune.html#ag

6. Give above scheme, do you recommend building the PWM4 circuit per the instructions or doing anything different?

The standard build should be fine for this usage.

Lance.

[mcneil]

Lance, I think there's a conversion error somewhere in the line pressure table. I get good data on 2 point calcs, but the table is off by about 100

2015-12-30_19_LPS.jpg (27.12 KiB) Viewed 26173 times

[Bernard Fife]

mcneil,

Here is 5.104ti test code that multiplies the 2pt line pressure by 100 (it changes a "/10" to "*10" on one line of the 2pt calcs):

Monitor_5104ti.abs.s19

Uses the regular 5.104 INI file

(84.44 KiB) Downloaded 917 times

I have a bad cold right now, and between that and what I am taking for it, my head isn't all that clear (well, actually it rarely is, but it's especially impaired at the moment!). So I don't want to try anything too drastic in the code is this state. If you wanted to try this code and see if it helps, that would narrow the issue down a lot.

Lance.

[Bernard Fife]

mcneil,

I misread your post. I just realized that it is the regular line pressures that are 100x too small, and the 2pt calcs are okay (assuming I am now correctly following what you have said). So don't use the code posted above.

Instead, try this if you like as it seems to work on my bench:

Monitor_5104tii.abs.s19

Uses regular 5.104 INI

(84.53 KiB) Downloaded 954 times

Lance.

[Bernard Fife]

mcneil,

I think the code above might work for you, but I think it will mess up people that are using metric (Bar) pressure units. So this code should work for both (I think...):

Monitor_5105.abs.s19

(84.52 KiB) Downloaded 927 times

It requires a new INI (for the change to 5.105 code signature, and also adjustments in the INI for the line pressure scaling):

GPIO_MShift_5105.ini

(297.42 KiB) Downloaded 943 times

Lance.