VSS resetting to 0

[tobmag]

Lance,

What I did was testing the 4126 code but no change....

I did then reload 4129 and decided to make everything from scratch, used the basic .msq rather then mine,
This sorted out the load issue, now works fine following the table and load.

Probably your work around will do as well but since ive done a about a billion changes in my msq I thought I might start over from zero.

Unfortunatly there is no change with the 1300hz issue

BR//Tobmag

[gui67]

Lance,

I did not manage to make the special code work:
when I put the signal on the ISS input, Tuner studio went ofline.
When I put the signal on the VSS Input, the Driveshaft RPM went to 450 and stayed there, with any frequency and even if I took the signal off.

So I put the normal 4.129 again:
still the same, speed drops at 1300Hz.

[gui67]

Gui67

Yes the problem dissapeared to 98% when I set the filterto 0%

Are you still using the additional board you got and not the GPIO on board signal conditioner?

BR//Tobmag

Yes, I am using it at the moment.

I saw that when I put a high value for the filtering % (80% for example) the speed was not linear with the frequency: from 0 to 100km/h, it was OK, but then, it increased quicker:
for example: if at 2000RPM (Converter locked, I was at 70km/h, at 4000RPM, in the same gear, it did not show 140km/h, but more: 155 for example.
I needed to decrease the filtering % to get rid of the problem.

[Bernard Fife]

Tobmag,

Here is new test code:

Monitor_4129c.abs.s19

Test code, use standard 4.129 INI.

(71.98 KiB) Downloaded 638 times

This code fixes the 1300hz glitches, at least on my bench. The problem turned out to be a subtle interaction between the VSS and timer overflow interrupts that only occurred in particular combinations of parameter settings and input values. In the code attached above, timing offsets were added to subsections of the interrupt code to significantly reduce the likelihood of colliding interrupts. Hopefully it works for you. It may just move the problem to another rpm (though I don't see that on my bench..yet). I did do a very slow sweep from 391 to 4340 Hz without problems.

And many thanks for your patience and persistence. Reproducing a problem is always the hardest part of debugging for me; once we can do that a solution is usually straightforward.

Lance.

[tobmag]

Lance.

Fantastic, we do have a winner!!!!

Was running this for about 15mins at 1300hz 50% filter and not a single dip.....
Did also test around all crusial speeds like 30/ 50/70/90/110 kmh and no drops as far as I could see
Im very happy with this and thanks for your pation chasing ghosts in the code

Tomorow its going back in the car for some more live testing

Many thanks Lance, sorry for ruing your weekend.

BR//Tobmag

[Bernard Fife]

Tobmag,

I am really glad it works for you. Any weekend in which a tough problem gets solved is not ruined! I often have trouble relaxing when I know there's something that needs fixing, so now I can go do just that.

Lance.

[ashford]

tried this on a 4l80e- no complaints. gonna try it on my 4r70w later today

[tobmag]

Just got the box back in the car but due to an oil leak I havent been able to run any test.
Did load the 4130 code so Ill get back as soon as the car is back on the ground.

BR//Tobmag

[tobmag]

This thread continous here;

viewtopic.php?f=4&t=678