Icedawg

So I finally learned just barely enough about tuning to install the MS3PNP 2001 Miata and give it a go.
I am posting to ask if it makes any sense at all that the peak VE seems to be below WOT at around 6000 rpm.

Info on the car, stock 2001 Miata, with cat delete, VICS in place, MAF assembly still in place for now, as I learn about how to do this before installing injectors and a turbo.

I used the timing map from trubokitty, just re-scaled to the NA range.
I started with the VE map and AFR target maps that came with the MS for the 2001 Miata.
Max MAP, and barometeric readings are 95.5 or so, even though forecasts say it is really 102 or so.
I have been taking the car for runs holding a constant MAP, while allowing rpm to sweep up, at 20 mm, 18, 12, 8 and WOT of vacuum as typical run attempts.
I log the whole run, take it to Megalog viewer, remove the cold temp data, and then use VE analyze to get a new table. I check it, smooth it a bit, then load the new table in, and do another run. Rinse and repeat.

After the first run peaks and troughs appeared. Every time I smooth them away, they return. These curves do not look a lot like the smoother ones I see posted in here, but most of those cover the boost range too, and do not have so much resolution down low.

What I am asking is if I may be doing the runs poorly, or misusing the VE analyze function in Megalog viewer?
Or, if this bumpy peak in the upper right middle area makes any sense?
I allow VE to analyze the entire run log, after dumping any cold temp data, and let it use the default AFR table, which is not quite the same as AFR Table 1 in the MS3PNP starter, which tends to be just a bit richer.
The edges all look a little poor, but I believe that is because it just does not acquire enough data in those extremes. (For example, it can get to 100 kPa, with a max 95 baro reading, so the data there is never adjusted much, and on over run it is probably not giving reliable readings, plus at idle rpm but 60-95 kPa it probably can't really get there either.) So the edges will be smoothed out, but they are not what I wonder about.

Looking for general advice, not for someone to do a detailed analysis, but I have given the data, so you don't need to ask if you wish to see it.




This results are after 7 tuning runs, with VE analyze changes incorporated, then smoothed a bit, then run again over several days.
The start that came with the MS3PNP is shown below, and was a much smoother curve set. The tuning runs leaned it out a lot, then started to make it richer again.








The current VE7 tune is attached, but for some reason I could not attach the data log

HowPrayGame

Lock any VE cells under 1400 or 30 kPA from being touched by VEAL or else you will get funky values in those areas.

Icedawg

OK, thanks for that, I will look at how to lock the cells when I try VEAL
But right now I am not using VEAL, I am doing it semi-manually through VE analyze on MegaLogviewer. So I can always just edit the ones on the edges.

ryansmoneypit

iTrader: (6)

Peaks and valleys will appear. Veal is just a tool and not a end all.

Things like rapid transit through map sections will cause peaks and valleys as the system lags to respond.

Do about 400 veal runs, increacing to HARD cell change.
fix the odd spots by hand.

Icedawg

Thanks,
As it happens, I just discovered the normal/hard very hard/setting you mentioned, and switched to hard. Then started from the original tune. That seems to have helped, at least the first trial did not create the same odd ridge along the 85 kPa line from 3000-6500 rpm.

Now I have to sort out why the alternator control suddenly started cutting in and out at the end of my last runs today.

DNMakinson

iTrader: (3)

Is ambient air temp consistent when you are tuning?
Are you iterating your MAT corrections between tuning sessions?

In other words, until MAT corrections are spot on, if you tune in the morning, and ambient air ( and therefore MAT at no boost levels ) is low; then tune again in the afternoon when ambient is 20F warmer you can go in circles.

Also, when tuning off line, make sure you load your current tune into Megalogviewer.

Also, I’ve seen worse, much worse.
DNM

Icedawg

I am not sure what you mean by MAT correction. Does the ECU not do corrections for MAT as it delivers the fuel?
If not, then how does one do the MAT correction you are referring to?
In any event, ambient air temp has varied between runs, but the measured MAT is in a 10 degree F range across the various runs, after the CLT goes up to the 200 F range.

DNMakinson

iTrader: (3)

There is a MAT correction cure under the basic tab. Yes MS adjusts, but according to that curve.

shuiend

iTrader: (24)

VEAL and VE Analyzer in MegaLogViewer are the same thing just FYI. VEAL is it just running live while you drive, verse doing the math from a log.

Icedawg

That's good to know. I hoped VEAL might be a bit more sophisticated in its algorithms, since it is doing things live. I suppose it is a bit, given you can lockout cells in VEAL. I could not find a way to do that in VE Analyze, but of course you can later just ignore the changes suggested in those cells.
I am using the manual approach with VE Analyzer in MegaLogViewer as a learning tool for me, to help understand what is happening, and because I won't trust automatic software without some initial experience of what to expect.

Icedawg

I have a question about both of these methods in terms of the AFR targets.
I have a target AFR Table that came with the original tune with the MSPNP3. After look it over I decided it looked conservative and would stick with it for now.
While running VE Analyzer in MegaLogViewer I saw that there is a default AFR table, or I could choose "my" AFR Table. The default table has 14.7 across a larger range of cells.

I am running street pulls with Incorporate AFR on, which must use my AFR Table. But have been using the default table when using Analyze.
Is that going to cause a conflict or confusion for setting the cells?
Does VEAL potentially have this problem too?

Icedawg

Just to update this thread (it can be frustrating to me when resolutions are not added at the end of a series of posts):
I started the tuning over again, used "very hard" on the change settings in VE analyze, used smoothing wherever bumpy lumps began to develop, and analyzed against my actual AFR target table instead of the default table.
The combination has given a much smoother VE table that is more sensible. There are some much smoother peaks and valleys, but they correspond to how the car feels when driven. In particular, that the 2001 intake manifold really tapers off in performance above 6000-6500 rpm.