VSS Filter logic.

[tobmag]

Lance.

Yes that will be a good idea with the pots in a place to set while "on the run". What I still can see is that even though I have a schmitt trigger in there now its still very sensitive to amplitude changes on lower frequensies, hopefully that can be filterd out better later on.

Tomorrow I will have a new live test in the car....actually the only thing that will tell you if it works or not....

Br//Tobmag

[tobmag]

Well tested and no go.........

To much noice and the trigger set way to low.
Does work as adream on the bench though.
Will make it fully adjustible and test again in the weekend.

Br//Tobmag

[gboezio]

It seem to be like the problem I had with my 8 tooth speed sensor and wheel to decoder ratio of 1,410, the signal would skyrocket at low speed. I have posted a diagram of what I have done in the thread, it still works flawlessly from about 6 km/h and up.
All that I have done made sense at that time, but it's all scrambled in my head already, so it was a very good idea to write it down

viewtopic.php?f=4&t=572&start=10

I hope it helps

Giovanni

[Bernard Fife]

I have put some tentative Schmitt trigger info here: http://www.msgpio.com/manuals/mshift/vss.html#schmitt

Lance.

[tobmag]

Ill build this one as well to test in the car..

Br//Tobmag

[tobmag]

Ill build this one as well to test in the car..

Well lastt friday I tried to rebuild the VR1 circuit on the GPIO board according to the new drawing but I must have a cross solder somewhere since I cant get it to work.
Output to the processor sits on +5v all the time guess ist on the backside where I fitted the resistors, extremely tight there to get them in....
If you arent a surgeon its probably better to build this on a normal experimental board.....

I have also ditched the 555 Schmitt trigger and now using a new circuit based on LM2907 http://www.electrokit.com/productFile/download/931 and a optocoupler on the output to the processor.
Will post the diagram as soon as I have it made up electronical. ( Circuit designed and built by my 73 year old father )
This new cisrcuit is rock solid from very low frequenses, around 20hz and up without a single glitch.

I will build the Schmitt trigger as soon as I get some components that are missing and then I will post some logs here using the original VR circuit, the new Schmitt trigger, a circuit based on the MAX9926 (same as gui76 uses) and my new circuit.

Attached a log live from the car where we can see the new circuit in action.....

BR//Tobmag

[tobmag]

And here is the schematics;
Just to clarify on the input, voltage can be with this configuration between 20mV and 28V rms (~80Vpp)

picupp omvandlare.jpg 2.jpg (171.5 KiB) Viewed 11248 times

[Bernard Fife]

Tobmag,

Thanks for posting that!

It seems like 4L80e's in particular have a lot of difficultly with the VSS. I couldn't say at this pint if that has to do with the 4L80e sensor, the wiring, or the MShift settings, code or hardware (for example, it could be the particular pattern and frequency of PWM could be causing noise due to small intermittent voltage biases on the ground leads, or it could be dozens of other similarly likely possibilities). Even the documentation could be wrong, so that everyone is building the circuits to 'spec' but not 'correct' (this may be why your Schmitt trigger build didn't work, for example). All of these need to be looked at until a solution is found.

I understand that we have a test vehicle coming (with an LS2 and a 4L80e), so that ought to help us get this sorted out.

Lance.

[tobmag]

Lance

I can only talk on behalf on what is happening on my installation with my 4L80E, I have followed all instructions when it comes to shielded cables and routing etc.

Exactly where the problem is if now only related to the 4L80E I can’t tell either, would be very good if others could chip in their experiences with other transmissions here.

What I have seen so far is that whatever you build and test on the bench doesn’t necessarily work in the car.
For instance my 555 Schmitt trigger works as a dream on the bench but as soon as it came into live action in the car it was useless....
This is actually where my father came in since he has over 50 years’ experience with electronics and the last 20 before retiring he was building/installing test equipment at Volvo at their engine test cells.
He had many issues with VR sensors over the years and usually a Schmitt trigger would work fine, in the severe cases they used the LM2907/opto coupler, but again whatever was built on the bench didn’t always work in the test cell...

I’m not in any way saying that this is the final solution but so far its working very well, it also comes in a 8dil version so it might be possible to fit it directly on the board.

However I will still build the Schmitt trigger you posted and also test the MAX9926 and also look at the circuit that "gboezio" posted but that will be in a few weeks, right now I need to tune transmission and engine as much as possible before the snow comes.
Snow fall already started up in the north of Sweden.

BR//Tobmag

[Bernard Fife]

Tobmag,

Thanks for all your help and willingness to try things.

The VSS is a bit different from and engine speed VR sensor, because the range is so much larger. An engine might run from 600 rpm to 6000 rpm (10x) or maybe a bit more. Even an F1 engine with a 18000 rm rev limit and a 7000 idle is pretty straightforward. But the VSS must work from 0 Hertz to several thousand Hz, meaning it's range is at least 100 times larger than the tach. It isn't impossible to make a circuit work for every installation, of course, but does take some work.

I'll look at doing a small add-on PCB with your 2907 circuit. It should be a cheap and easy solution installations for some, at least.

Lance.