MAP sensor sampling

[prj]

I am working on a speed density algorithm for the good old Motronic.

I was wondering, if some aftermarket ECU (VEMS?) has the ability to produce a high enough sample rate log, to show the TDC trigger/crank angle as well as the current MAP sensor reading on the same graph?

Something like this:
http://www.msextra.com/doc/ms3/pix/maplog-1cyl.png

I am right now not sure if it's enough if I sample the MAP sensor at TDC and average the last two readings (so realistically having always the MAP from 1 full cycle of the engine at a fixed crank angle), or if I need to poll the sensor more often during the rotation of the engine and then average the readings.

If anyone has any logs like this on idle and maybe at full load, it would be very interesting to see from the standpoint of an I5 engine...

[prj]

Btw, VEMS settings for MAP filter on I5 would be interesting too

[4WheelDrifter]

Hmm.. Interesting.

[prj]

I figured the way to do it right - sample it every X teeth. Will see what X will be and how well it works.

[prj]

Now also wrote the algorithms - MAP * VE(MAP, RPM) * IAT...
Normally ECU stock acceleration/deceleration enrichment is based on load gradient.
Maybe I will need to change acceleration enrichment to use TPS gradient - will need to test.

But sampling every X teeth, then summing it provides a good and stable signal regardless of RPM.

[A80Avant]

are you asking for the timing vs map vs rpm log?

vems is logging roughly 18-19 samples per second from a quick count that i just did on a log

[prj]

Hmm, I think VEMS should be sampling MAP different depending on RPM.

I looked at it a little, and as I understand it, it logs per "trigger".
And "trigger" in VEMS internal division, so if you use DivBy3 for example, it's 3 "triggers" per 135 teeth.

So if there are 10 samples per trigger for example, you would have 30 samples per crankshaft rotation.

So at 800 rpm it would be 400 samples per second and at 8000 rpm 4000 samples per second.
Unless I am missing something...

In Motronic I thought it was a good idea to sync the samples to the cylinder phases.
So right now I am going to test 3 or 6 samples between cylinders, or one sample every 9/18 teeth (135 tooth wheel, 54 teeth between cylinders), 45/90 samples per two crankshaft rotations or 22.5/45 samples per one crankshaft rotation.
At 800 rpm it would be 150 or 300 samples per second and at 8000 it would be 1500 or 3000.

I don't think a very high sample rate will change much. I spoke to Marc about it, and he has experimented with different settings, and there is not really much difference in how much you do per second... so I think 3 samples between each cylinder firing is OK. The goal is to have a stable MAP signal more than anything.

[A80Avant]

Yes that's sounding right

[prj]

I got it working thanks to lots of testing from Marc.

Basically it's enough to do 3 samples per cylinder.
6 samples does not really make anything better.

I also added the speed density calculation to Motronic - so a VE map, IAT correction.
Abusing the stock wall film compensation is enough to do acceleration enrichment and get good response, but I want to add TPS to load calculation...
MAP signal is much slower than MAF as it shows, so TPS is needed too, because wall film compensation only affects injection, not load, so I get knock retard on transients.

But - Marc has a MAFless Motronic running now. Will add the TPS stuff, Marc will verify everything on dyno and after that next test bunny is JP I guess
This also is good news for the USA guys who are near to Marc, as they can get tuned speed density as well.

[nip]

i want maf-less motronic

[prj]

Sure, no problem

[nip]

Sure, no problem

[andy10v]

Thats awesome work!

[prj]

I still prefer MAF on lower powered cars - speed density unfortunately means less precision and also if you change anything on the engine that changes how the engine breathes, you need a remap right way or the fueling/timing will not be quite right.
So it's not only "get rid of the maf" - there are upsides and downsides to everything that must be carefully considered.

[andy10v]

Not for me as the reasons you say, still impressive work though.