NefMoto

Still very much in discovery mode here... Looking at TVUB scaling information from ME7x S4-wiki, and I am also seeing in other map packs for Audi ME7.1.1 motors that the value in the chip is multiplied by 0.002667 to get the latency time expressed in milliseconds. Apparently the x-axis is scaled by 0.0704 to give us volts.

The headache I am having is that the latency numbers I see in the ME7.1.1 using this method do not align with OEM data that I got from Bosch on the injector. You can see in the attached graph that I made from the data that the overall shape of the TVUB graph matches the OEM data very nicely but the scale is definitely off.

I made the assumption that the OEM data is referencing n-heptane and 3bar operating pressure so my method to project a latency figure at 4bar was to increase the time by a factor of sqrt(4/3). The same as we do for fuel flow changes. I know thats not a perfect method but its certain that latency will be HIGHER with 4bar operating pressure vs 3bar.

This all leads to a few questions:

1 - are the TVUB scaling values correct for all iterations of ME7x ?

2 - assuming the scale is correct, why would the calibration use lower values of TVUB vs the OEM data - is there another scaling factor at play which I don't know about ?

3 - I don't expect that a tenth of a ms difference in latency has any significant effect when we are working at high load points with pulse widths in the region of 10-20ms but if we working at ~2-5ms IPW for idle and small loads / part throttle then this apparent difference of maybe 0.2ms starts to look like imperfect - but of course adaptation will step in to make any tweaks of TE values to end up with the correct pulse width - right ?

4 - I wondered if I was over-thinking all this stuff for such a small deviation in my naive expectation that all these numbers will magically align. Then I got to look at the TVUB table in the C5 RS6 and it is not only the wrong scale but a completely different shape from the OEM data for the stock injector - particularly skewed off at the lower voltages which to be fair we don't much care about. Could the original calibration be so apparently bad on this (because Motronic is so good at adjusting for component ageing and tolerance) or am I barking up the wrong tree ?

5 - Does anyone with IDAPro have the means and the energy to actually verify if the TVUB scaling is correct on ME7.1.1 ?

Any help would be great - I thought I wouldn't be first to ask this question here, but a quick search didn't turn up anything useful on the subject.

Nearly 50 views and nothing. Oh well - lucky I might have found the answer by myself - well thanks to the wonders of the ME7logger utility... I found scalars for TE & TI values of 0.0032 and UB (volts) for 0.0704. When I use a scalar of 0.0032 on the TVUB data found in the BBK ECU then the latency data matches the data sheet perfectly from Bosch. See attachments.

I also got info form Bosch that they say latency is NOT affected by fuel pressure on the EV6 injectors so thats cool as well.

Lesson learned here is to reference ME7logger scales when I find such anomalies again. All credit to whoever wrote that and unpicked all the relevant scaling mechanisms.

Sorry, just catching up here. I haven't seen an issue here with TVUB map definition in the public XDFs, they generally work when you put in aftermarket published values (though I have not personally double checked OEM specs).

Are you saying the map definition is incorrect?

If so, it is likely a function of ECU frequency and many of the XDFs are outright wrong.

See also the KRKTE error as well https://s4wiki.com/wiki/Tuning#Primary

I looked in ME7logger ".ecu" file for the 8D0907551M original file and TVUB scaling of 0.00026667 looks correct there and that aligns with map packs for the B5 S4 motor in the public domain.

The problem arose when I used the same scale on the BBK ECU (8E0910560R) and latency didn't track the OEM data sheet. It was only then I realised the scale vector of 0.0032 called out in the ".ecu" file for that. Applying that made the latency date for the BBK motor fit perfectly.

When I looked on the C5-RS6 I find that TVUB scaling looks like 0.004 so just proves you have to check everything - twice !

As for KRKTE - the scaler for this on BBK is different as well - had to find that in another map pack from a different BBK variant of software.