input masking problem

[fairmont50]

I am having a problem with the input masking setting. I seems that every setting that I try below 100 will sometimes cause low speed values. This is repeatable on the bench as well. In the car it varied, sometimes when pulling away from a stop it would be correct and sometimes the speed would be 1/3 to 1/2 of the actual value. What is a reasonable setting for this value?

Application: gm lq4 v8 and 4l80E trans with ms2 control.

Thanks,

Chris

[Bernard Fife]

Chris,

I find values between 40 and 80 percent work best in most cases. You may also have to adjust the minimum speed value.

You can also double check the VSS circuit components to verify it is built correctly.

If those don't help, I would look at the wiring, and make sure the connections are good, and the wiring is not near any high current devices (and it doesn't hurt to either use a twisted pair for the VSS, or a shielded pair).

Lance.

[fairmont50]

Lance,
Thanks for the response. I have the minimum speed set to 3mph (the default, I believe). Is this reasonable or should it be changed?

Chris

[Bernard Fife]

Chris,

It is reasonable, and yes, you should experiment with it. Different VR setups can require a fair bit of tinkering to get them right, but are usually reliable once a set of parameters that works is found.

You might also try reducing the maximum errors to reset, at least temporarily. This only causes the VSS masking to be reduced for a time, ensuring that all of the teeth are counted (but the masking must eventually be turned on to prevent false teeth from being counted).

If this was mine, I might also try switching the input capture polarity, just to see if it helped.

All of these settings are under 'General Settings->VSS Setup': http://www.msgpio.com/manuals/mshift/V2tune.html#gv

Lance.

[fairmont50]

Lance,
I checked the circuit and everything was correct according to the 4l80e build instructions. After checking everything I modified the circuit and built it according to the 4l60e build instructions. The 4l60e instructions have a different value resistor in R45 and R37 - not sure why this is. This seems to have taken care of the problem.

Thanks,

Chris

[Bernard Fife]

Chris,

There may be mistakes in the 4L80E instructions - I will have a look and fix it. Thanks for spotting this and posting on it.

Lance.

[Bernard Fife]

Chris,

Yes, you were right of course. The 4L80E instructions had the old values for the VR input circuit resistors R37/R45 (680K Ohms). After considerable experimenting (both with pSpice and in hardware) we determined quite a long time ago that R45 should be 300K Ohms and R37 should be 220K Ohms for best sensitivity to small amplitude signals. I believe the 4L80E instructions were written before the changes were made, and I neglected to update them to the latest specs. This has been done now. Sorry for the confusion and trouble we caused you.

Lance.

[gui67]

Hello,
Could this have caused the speedo output and speed variation on TS gauge with my 4L80? I also use the 4L80 instructions to build my board?
I did not really have time to find a solution to the problem as I am replacing the engine and not driving the car since 4 months.
I will also correct these resistors before trying anything else.
Guillaume

[Bernard Fife]

Guillaume,

Yeah, if this was mine I would try changing the resistors.

Lance.

[ashford]

i just e-mailed diyautotune to ask them to update the gpio trans kit, it is short 2 resistors 220k and 300k.