RomRaider

I agree with you Mickey. I see your point. The slope of the line will change with larger compressors, but the Mass Air Flow should correlate well in regions of high compressor efficiency.

I updated the spreadsheet with some instructions.

I added the calculation for injector scalar and latency too.

subscribed.


Mickey, my current (huge) obstacle, is that I've changed A LOT at one time, engine-wise..
went from 2.0 to 2.5, stock to stock-modded injectors, stock airbox to 3" CAI, stock turbo to GT32, tmic to FMIC. I 'think' Ive got my injectors really close (scalar @ 800, stock latencies), but now I'm working on the MAF. I cruise this car to/from work, which is about 60mi each way.
I get slight hesitations and such, while cruise/light load driving, but none in WOT instances.
Do you think you spreadsheet will work for my given situation? i must be fairly close with the settings (maf + inj.) as I'm averaging about 22-23mpg in the car, now.

I have a constant issue with stalling after engine-braking to a light, or just letting off the throttle, and clutching it.. But if before I come to a stop, I clutch it and blip the throttle, then leave the clutch in, or kick it in neutral, it'll stay running, albeit rough, but running..
3rd gear, very light throttle, I often get a lean situation, where my wbo2 will oscillate from 14.2-16.9 rapidly, staying above 15AFR most of that time. A slight increase in throttle position makes it go right away, but if I return back to that same position, it leans out again. I'm assuming this is the poorly-scaled MAF? WOT 3rd gear pulls net a mid-high 10s AFR, and 4.5-4.7 MAFv in the 3" intake..

1. The first thing to do is to make sure that you have no mechanical issues. Is your BOV/BPV working correctly? FPR working correctly?
2. You could be pretty far off in scaling and still have decent mpg because the front O2 sensor will make adjustments for you in closed loop (cruising).
3. All the stock-modded injectors that I have checked had slightly higher latency than stock. Even if electrical delay is identical, there will be a slightly increased latency due to time needed to ramp up from 440 to 800. Most people adjust MAF to compensate but I disagree with that methodology.
3a. You can check the accuracy of your latency by driving around (to eliminate heat soak as much as possible) then datalog idle for about 30 seconds (use parameter list for this spreadsheet plus WBO2) then datalog at 6000 rpm constant throttle stationary. If your latency and MAF is correct then you will have 0 corrections at normal idle AND your 6000 rpm run will also be correct.
4. The large diameter intakes that I have seen do not have smooth MAF tables at low MAF voltages. To improve the accuracy of your MAF table, you can increase resolution of your MAF table by getting rid of values that are not needed. Look at your MAF table, you'll see lots of low end voltages that are never used. Post your rom and some logs and I'll change the columns for you if you want me to without actually changing your MAF curve.
5. I have found that the WBO2 is not accurate at low MAF values. The front O2 sensor is more accurate at low MAF. This might be due to distance from engine and/or small leaks. Unless you are bumping into AFR Learning A at 15%, you shouldn't be leaning out at idle.

here's my current rom (400 g/s scaling, via merch)
And a cruise log from a few weeks ago, yet on this above map.
I'll do the idle and 6000 rpm logs this evening.

I'm going to do the driving log today. This morning, I had it at op. temp, and did the stationary idle and 6000rpm logs. Unfortunately, the (i am assuming) CTS reached a level in the 6000 log, that it immediately cut the engine speed to idle, although the pedal remained at ~22%.. I was able to reproduce this, and logs shows temps ~208-210 when it does it.

So here's those, for the time being..

If WB02 is not useful for scaling low MAF values, then how should one perforrm MAF Scaling for low MAF and idle? I've got a 70mm APS CAI.. it seems to idle well enough using MAF values I've copied from other guy's MAF scaling for a similar intake. How can I tell if my changes are ok and do not need more work?

Read the first post. Either use this spreadsheet or keep changing AFR learning ranges to check each range.

Yes, that's why you have to drive around a bit to get rid of the heat soak before trying this. I didn't realize your coolant temperature was going to rise that fast. Mine rises from 190 to 200.

Anyway, it looks like your latency is off.

Is it normal to idle at 1000 rpm even on stock car? I thought manual cars idle at 750 rpm?

You should log IPW direct since it is more accurate when trying to calculate latency.

Change your AFR Learning C/D transition from 50 g/s to 40 g/s and log some closed loop driving as I described in my first post. Add latency to get close and then finish the fine adjustments using the MAF scale. It won't be perfect but I believe you will get a lot closer than where you are at right now.

This is probably due to higher compression, and fresh motor (1900 mi.) 9.5:1


I have to idle it at 1000, since it doesn't want to idle at 750.. This was an instruction from my engine builder.

I made the C/D change, but after I logged this morning. I didnt see this until after I got back..

I logged about an hour, this morning. Given the road conditions, I had to make multiple logs. I tried to log some cruise at higher RPMs (3000-3500) which obviously called for a lower gear, unless I wanted to do over 100mph LOL

Do I just keep adding into the 'input' tab, chaining the logs together, then "filter data"?

Actually, the rpm isn't important except it needs to stay in closed loop. So, just drive fast on the freeway in the top gear at constant throttle.

The first file should be added using the "clear and add data" then all other files are added using "add data".

Can you winzip the files and email them to me? I'm curious to see how it looks. There is a strange jump in your MAF scale.

I'm assuming you have a WBO2? And your scale from 60 g/s -> 400 g/s was scaled when AFR Learning D = 0%?

Yeah, I drove at a pretty constant throttle. Importing them, I did exactly as you just said.
I'm compiling that now, and you should have it in the next 10 minutes or so.

Yes, I have a WBo2. I use/log it via the Fuel Level (V) as it's 0-5V signal. Using a calculated method of "=SUM([cell for Fuel V]*2+10)" ex. =SUM(L2*2+10) and that gives me the WB AFR reading.

The 60 6/s -> 400g/s was scaled, as it is, when AFR Learning D = 0%. I've very, very rarely seen the AFR-D leave 0%.

I just realized i did IPW, not IPW (direct), in the hour's worth of logging this morning...
real nice..

It's okay. It's just not as accurate. And frankly, because you have both MAF and injectors installed, there is no way to accurately calculate the latency anway. We'll just try to get it closer.

Your data is not that unusual.

If you increase your latency and then relog and then adjust your MAF at less than 2.5 volts, I think you'll get better results. Hopefully, that will pull your scatter down as well. Right now, you have a very wide scatter which could be caused by the mix of latency and MAF errors or it could be that the large diameter intake has poor resolution at low MAF.

Try adding 0.22 ms to all your latency values and relog closed loop (drive less than 3600 rpm). DO NOT do any WOT logs.

Yeah, the maf + injector change, at the same time, is definitely biting me in the ass..

I'll add the .22 to the latencies, and give it another go this afternoon. Is it okay to just get ont he highway and log a steady 65mph, or should I log low speeds at a constant, as well, as I did this morning?