MAF - Less airflow than stock?

[darkdragon]

I've been trying to tune my MAF (open loop) for a while and I can't figure this situation out:

Car: 98 GTP with CAI, 1.8 Rockers, 3.4" pulley, TOG headers.

Equipment: NGK AFX with Bosch sensor

As I recalibrate my MAF, it appears that I have more airflow than stock up until 8875Hz, but from then on it flows less and less as the frequency rises. See attached tune and log.

I've rechecked this many times and my wideband reads consistently. Every other comparable car flows more than stock with an exponential curve which goes up to 4lb/min over stock at the highest frequency. On mine it reverses and goes negative!

I've cleaned the MAF and get the same response. Any ideas?

[darkdragon]

After looking at more traces, it looks like it's a time or IAC-dependent thing. The longer I am in PE at a constant AFR, the more fuel is dumped in thus lowering the actual AFR. In all my traces, the AFR gradually decreases over time or as IAC increases.

Case in point: in Scan3 I get 11.6 AFR at 8100Hz when at the beginning of the WOT run. In Scan4, I get 11.38AFR at 8100Hz when later in the WOT run.

For Scan7 I disabled AE and COT and started to scan IAC and Injector. Here is seems there is a correlation with higher IACs causing a lower AFR.

Any ideas as to why this is happening?

[Perforator]

Why are you commanding 14.4 afr at idle?

[darkdragon]

Stock SW? I don't see anywhere in the tune where this can be changed.

[Perforator]

I'm sorry, I don't understand, how can you tell how much air your flowing and that's it's lower from the scan you posted? Also I think there is something really wrong. You should be commanding 14.7:1 at idle. Was the car warmed up?

[Perforator]

Ok, I just did a compare of your file to a stock file and see what it is, you have the closed loop enable ect set to 284 so your never going into closed loop. Why are you doing that?

[darkdragon]

Because I am tuning the MAF in open loop. See Russ K's method below:

http://www.hptuners.com/forum/showth...=open+loop+maf

It's not really the amount of air that I am flowing, it's actually the amount of air the MAF is reporting at each frequency point. If you compare a tuned car's MAF curve against the 98 stock curve, typically you will see the difference curve as an exponential line. With mine it starts out that way and then reverses (see attachment).

The cause of that is due to the AFR linearly decreasing over time (or IAC count) during my WOT runs as shown in the plots above. This happens regardless of the MAF curve I use. I am trying to figure out why this is happening.

[Perforator]

I've read it and tried it before. What is the thumbnail from? I checked the bin you posted and your maf table looks fine, nice and linear.

[Russ K]

Did you do this; For the higher freq hz cells of the table that you didn't enter, just enter the error % from the highest airflow hz cell that you did enter.

Also do another WOT scan to your redline now that you have the COT disabled.

What does the IAC have to do with the AFR?

Russ Kemp

[darkdragon]

OK, maybe I posted too much data and not enough explanation. I can't get a run in today since it's pouring rain and I have no traction. I'll try again once the rain stops. In the meantime let's go through the whole story:

1. Prior to tuning MAF, I look at tunes from other people. One typical example is MAF_Curve_typical.png. When compared versus the stock curve, the difference looks like MAF_Curve_typical_vs_stock.png. Notice the upward sloping line there? As I read it, this means that there is more flow at the frequency points than what the stock calibration expects. Right? And this is gets more pronounced as frequency goes higher.

Note: The above is important to understand otherwise the data below is meaningless.

2. I setup my tune to have open loop with AFR 11.7 at all times. Disabled Deceleration enrichment, EGR. MAF curve looks like MAF_Curve_initial.png as a result of previous failed attempts to calibrate MAF. When compared to stock MAF curve, it looks like MAF_Curve_initial_vs_stock.png.

[darkdragon]

3. I go out and do a WOT run. Refer to Scan2.hpl from the first post. Taking the AFR error data from cells 8125 to 10375 (other cells were missing or otherwise unreliable). I paste multiply by % into my tune and fill in the missing cells as per procedure. The resulting MAF curve is MAF_Curve_adjusted.png. When compared to the stock MAF curve, it looks like MAF_Curve_adjusted_vs_stock.png.

So far so good?

4. OK now if we compare my expected result of MAF_Curve_typical_vs_stock.png and MAF_Curve_adjusted_vs_stock.png, do you see anything wrong?

Answer Yes: what is causing this?

Answer No: why does no one else have such a curve?

Wouldn't this imply that I am having less airflow than what the stock calibration is expecting? In both cases, read on...

5. So, next step is that I run again with this adjusted curve. On a WOT run from 0mph to redline just like the original runs, everything is fine.

However, if I go WOT from say 40mph, all of a sudden my AFR is lean and get all kinds of knock. OK. Reduce timing so 9-10degrees so that I don't get knock. Run again and check my AFR. Sure enough I am lean everywhere by 2%. Run again from 0mph and AFRs are OK. WTF?

6. Revert back to the initial tune and take more scans to find out why I am getting differing AFR's. Refer to Scan3 above. In this scan you can see that my AFR decreases from 11.7 at the start of the run, to 10.6 at the end. Hmmm.

7. Ok, out of curiosity, I run again with half throttle (refer to Scan4). In this scan, again the AFR decreases. This time from 11.7 at the start of the run to 11.2 at the end.

Here's the peculiar thing: In Scan3 I hit the 8100Hz point early in the run (see frame 506) and my AFR there is 11.6. In Scan4 I hit that frequency later in the run (see frame 315) and there my AFR is 11.38.

So it looks like my AFR is going down over time!

8. To try to figure out why, I start logging some other parameters. Lo and behold, in Scan7 (again from the above post), I noticed that as the IAC counts go up during the run, the AFR goes down! This should be a coincidence, but so far I can't rule it out.


OK, hopefully this provides a better picture of what's going on.

I have since cleaned out my IAC valve and changed my PCV valve, disabled COT and AE, and restored my VE table to stock. The half runs I did today in the rain don't seem to show improvement (as my IACs go from 60 to 90, my AFR goes from 11.7 to 11.0 and then both stabilize), but I won't know for sure until things dry up.

[pdsq98gt]

Have you tried swapping the MAF with a known good one? It does seem odd that the higher the frequency of the maf that the less fuel it wants. I guess you could have a leaking fuel regulator or an injector that is stuck open.

[Perforator]

Refer to Scan3 above. In this scan you can see that my AFR decreases from 11.7 at the start of the run, to 10.6 at the end. Hmmm.

In scan3, you commanded afr stays 11.7 through your whole wot run. Where are you seeing this 10.6 at? I don't see any wideband output in any of the scans I looked at???

[darkdragon]

The Wideband output is seen by using the PID's in the first post. I've attached a config below which might help.

[Perforator]

Post your vcm scanner.cfg file. That will have your wideband pid's in it.

[darkdragon]

Here you go...but the PIDs.txt already is a cut and paste from the scanner.cfg.

[darkdragon]

OK, I got a couple more traces tonight. On the first one I had to slow down due to traffic ahead. Scan11 looks better though. Even with all the extra stuff disabled, I still get a descending AFR which end up making my MAF curve look wrong.

Anybody have any ideas?

[Perforator]

Ok, now I'm really confused. The user defined paramater Air Fuel Ratio you added is based on PID.7101 which is AC Pressure. Do you have a MVP Pro interface?

[darkdragon]

No, quite a few people use the standard interface with the wideband plugged into the AC pressure or EGR inputs of the PCM. The value shown in HPT matches what is shown on the AFX.

[Perforator]

Personally, if it was me, I would tune the maf in closed loop and tune wot with the wideband. As far as your wideband output, I don't know, I'm not familiar with that wideband or with how you have it hooked up. Could be an issue with either.