MAP behaves oddly during Overrun..
 

I remember a little while ago some people mentioning that they sometimes observed MLV to populate very strange values into certain cells. I myself have seen this occur frequently in the 100kpa area, and I figured out why.

It seems that sometimes, but not always, while I am in overrun (coasting in gear) my MAP value jumps to exactly 100 kpa and stays there for a moment, then reverts back to reality. I have seen this not only in the logs, but also on the gauge display in Megatune. It's never done it while I'm on the throttle, and oddly it does not turn the fuel back on when it happens.

A full log is attached, but here's a screenshot of a typical event:

http://img132.imageshack.us/img132/250/graph2mm1.gif


Has anyone ever seen such a thing before?

You know, I am neither turbocharged, nor at sea level, yet I see this today:

http://img225.imageshack.us/img225/3489/graphsj4.png

I found 4 similar events.
http://i27.tinypic.com/4jhwx.png

http://i32.tinypic.com/2u8c503.png

http://i32.tinypic.com/2iscjtk.png

http://i28.tinypic.com/ohkbjo.png

This morning I turned off Overrun and logged. The exact same thing happened- both in the log and on the gauge display. And again, it seemed not to impact the actual fueling- it's as though this is a display / logging artifact only.

Damned odd. I've started a thread over at MSExtra, which so far has failed to gather much interest. Seems that folks over there don't really care an awful lot about debugging... http://www.msextra.com/viewtopic.php?p=183865#183865

I guess that is why RB26 has jumped up and helped start the DIY EFI project?

I haven't seen this behavior on our own tests, so I'd encourage the others who have to post their similar logs in Joe's thread on MSExtra.com. The more data the developers have, the more they'll have to go on.

I'll upload my data as well to msextra.

Hey guys.
In the meantime until we find a solution to this problem we can use a custom filter in MLV under the advanced options. Just click on "custom filter" and then enter this [MAP] = 100 && [TP] = 0
That will filter out all of the spikes where TP = 0 and MAP = 100. I just tried it out and it works great. Instead of it adding 20 extra VE in my 100 kpa row it made accurate adjustments.

Or if you dont have a TPS you can say [MAP]=100 and [O2]=4.99 too, isn't that right?

Genius! At least someone has been reading and paying attention. :bigtu: That works nicely.

The problem has been properly solved by Keithg: http://www.msextra.com/viewtopic.php...187749#p187749

Short version: it's the kpafactor4250.inc file in the /mtCfg/ directory. The one which comes with MegaTune only has valid values down to 12 kPa, and considers anything below that to be a sensor error. The one which comes in the /hr10x/src/ directory (and is compiled into the bin) has valid data down to 10 kPa. So, every time the MAP voltage drops to ~.05v or less, even though the MS itself is able to interpret this properly, MT considers it an error and throws a 100.

Copy the version of the file that came with the hr distribution into /mtCfg/, and the problem is no more! :bigtu:

shoulda gone hydra, son.

right because hydra would just say "sorry, we cannot fix your version of the software. do not run below 12kpa"