As Murfie corrected me an air leak should not cause the MAF curve to be lower. If you are feeling adventurous, I would enter injector data for a larger set and adjust the MAF curve back to the one you calculated based on the area ratio. Based on the percent “off” the MAF is, I would try the 1440 injector data and see what happens. Good luck!
Also what I was thinking. I had thought an air/vacuum leak would cause extremely high fuel trims, therefore raising the curve. My case is the other way around. I remember months ago when I fabbed the turbo kit and took away 22% from stock MAF (the difference between the surface area of the stock housing and the new 3" housing) and the car straight up would not run. It would start, struggle for a second, then die. After a few startup attempts upon dying the car backfired loud as hell and my overall assumption was running way rich, not lean. Took away more and more from the curve and eventually ran much better, resulting in WAY lower MAF values than I anticipated. And now I'm here.
So based on what is being said here and my experience first starting the car after the turbo/return style install, the car is running incredibly rich with the correct data. I'll still look for any possible vacuum leaks but with how OCD I am with my craftsmanship, I'd be very surprised to find one. But I've not anticipated a "running too rich" situation. I'll remove the injectors and the extension sleeves and double check that they truly are what I think they are. After all, I did get them second hand from someone.
If this matters, I do have two fuel pumps running this return style system. One pump is running full time and triggered via relay by FPDM signal, the other pump is running off the FPDM. The car basically can't tell a difference therefore I have no codes and nothing turned off to trick the ECU. For all it knows the system is working as expected since the FPDM is still "controlling" one pump. I can confirm the relay triggering the full time pump is not rapidly turning off and on in relation to the PWM signal from the FPDM while it controls the other pump. This setup doesn't seem like an issue but maybe it is?
Last edited by Artorias; 12-15-2024 at 04:16 PM.
Sounds like a solid plan. Coincidentally, if you multiply your current tune by 1.44 in nearly matches the curve you calculated based on surface area ratios. So……I would enter FIC1440 injector data and change the MAF to the surface area ratio scaled one and see what happens.
Went ahead and did some math on your datalog. I took your MAF period and used it to find a airflow based on the MAF transfer function that should be used (calculated using surface area ratio). Then calculated fuel flow based on this airflow and correlated this to the expected fuel flow based on injector pulsewidth and injector size. I entered different injector sizes until the correlation was 1:1. The FIC1440 injector data resulted in this correlation.
Well, let me try 1440 data then. Haven't had much opportunity to mess with the car the last several days but I'm removing the injectors to verify what they truly are. Looking at past logs when the car had these injectors but had a Paxton 2200sl installed (but was not return style), the data seems correct. 4" MAF housing with the Paxton and values are reasonably higher than stock, injector data is correct. So... I wonder if the return style system is somehow the cause of running extremely rich but I'm racking my brain for how that could be.
I'll remove the injectors and report back whether they are or are not the 1000cc, regardless of the answer I'll also try 1440 data
Edit: Okay so I now remember closer towards the end of my Paxton ownership I installed the return style system so looking at those files, lower 1/3 of the MAF curve is much closer to stock MAF values. 4" housing and 58psi data from the sheet w/ 1:1 reference. This appears to follow what I had to do with turbos and lower the MAF values a lot to compensate for rich condition with what seems to be correct injector data. If this is the cause, maybe it's how I have the system laid out? Regulator is after the rails, not before. So it's not a deadhead setup.
Edit 2: yep, they're definitely 1000cc injectors. Pulled the extension sleeves and confirmed.
Last edited by Artorias; 12-22-2024 at 12:12 PM.
K here's a different approach. And I think this may have been the nicest startup since the turbo install.
I didn't do the 1440 data. After comparing it to the "old" FIC1000 data I decided to give the old data a shot. Here are two excel sheets I posted earlier. I emailed FIC about these towards the beginning of the year and they mentioned one sheet is "old" data and the other sheet is newer. They specified I should use the newer. Now I gave the old data a shot, it doesn't allow any choice for return style other than 58psi. Put in the data, put the MAF curve to stock values then subtracted 22%, fired up the car. It didn't struggle and die like it did with the new data at 58psi, with this older data it ran far better. Here's a log cold start idling to 160, let me know what y'all think and what the next direction should be. I'm thinking minor adjustments to the slopes from here. If I'm not mistaken high slope is used during cold starts.
-OLD- FICLINIC_1000H_FORDDATA_SCT_WEBSITE_RAILTEMP.xlsx
-NEW- FICLINIC_1000_FORDDATA_SCT_WEBSITE_RAILTEMP (2).xlsx
Twin Turbski FIC1000 IMRC - V166 log 1 startup only idle to 160.hpl
Twin Turbo FIC1000 IMRC - V166 stock MAF 22% removed and old 1000 data.hpt
Edit: Added 5% to the high slope (so taking away fuel) and car started a bit better bit still very rich when trims kicked on, tried another round with 15% added to high slope and trims went from about 28% rich to about 16% rich once stft kicked in so that looked better, but the car still idled underspeed. Target idle is 1250 on cold start and it idled at about 1050 and slowly rose to 1250 before going to commanded warm idle of 1000 rpm. Hmmm... increase the MAF values?
Last edited by Artorias; 12-24-2024 at 08:53 AM.
Can you post your stock MAF Transfer function, Stock MAF housing diameter, and current MAF housing INSIDE diameter?Originally Posted by Artorias
Wall thickness is just shy of 3mm so we're at roughly 2.9" inner diameter, nothing trims couldn't figure out but this brings us closer to 75% of stock diameter. Stock MAF housing for a 15-17GT would be 85mm (based on research, I don't have a stock intake to measure), so my stock MAF transfer is just the same as any totally stock 15-17GT. Taking the stock transfer and calculating the difference to the new 2.9" inner diameter, that's about 25% reduction to the MAF values.
With that said, I don't think I can get the MAF transfer any more accurate and any errors I'm seeing in logs would probably be related to injector data I'm thinking.
New transfer, 25% reduction
new transfer 25% reduction.png
Stock transfer
stock transfer.png
Edit: Here's a log of this, 25% off the stock maf so the transfer shown in the photo including a 15% increase to the high slope from 114.3lbs/hr to 131.445lbs/hr. Starts the same as the last logs, still idling under target a little. I did try and hold RPMs and it wasn't easy to control, pedal felt different. Notice how at 2 minutes 27 seconds I barely hold the pedal at 5% TPS and RPMs gently oscillated back and fourth.
Overall I think I'm heading in the right direction, just lost as to where to go from here. With just the MAF adjustment and no increase in high slope as soon as trims kick in she's pegged all the way rich on trims
Twin Turbski FIC1000 IMRC - V172 25% off maf 15% added to high slope.hpl
Last edited by Artorias; 12-29-2024 at 06:38 PM.
Still looks like FIC1440 data would make the rescaled MAF fit. Try this MAF transfer and use the FIC1440 data.
83 72.202
88 59.831
93 50.127
98 42.417
103 36.221
109 30.300
115 25.625
121 21.889
128 18.424
135 15.685
143 13.214
151 11.275
159 9.725
168 8.335
177 7.223
187 6.272
197 5.536
208 4.877
220 4.265
232 3.736
245 3.252
259 2.826
274 2.441
289 2.125
305 1.839
322 1.583
340 1.368
359 1.197
379 1.050
400 0.925
422 0.828
446 0.736
471 0.655
497 0.589
525 0.532
554 0.482
595 0.422
694 0.325
763 0.272
1378 0.000
Alright screw it lemme try it. The OCD individual in me hates that I have 1000cc injectors and 1440 data appears to be correct for some reason, but the low slope between the old 1000 data and the 1440 data isn't far off from each other so that brings confidence. Both dramatically different than the new 1000 data. Worst case I was also going to go back to the first tune I put in here on the first post and calculate the difference between that MAF and the new MAF transfer, then increase the old slopes by that amount with the new scaled MAF. Maybe that's a decent direction too.
I'll try the 1440 data tonight with the rescaled MAF and report back in a few hours.
Side note, any ideas on why the 1440 data seems to be accurate for this? My only reasonable assumption is because I have a 1:1 regulator, not a static pressure or returnless fuel system. This is intriguing to me.
Last edited by Artorias; 12-30-2024 at 06:30 PM.
why is your flow rate low so high? I show it should be low .034490949 and high should be .031736938 with a breakpoint of .000027010 and pw .300 I also show your offset vs voltage is slightly off, if you convert the FIC data from MS to seconds I'm getting slightly different numbers then what you input there.
Unless I'm reading the wrong excel sheet from FIC.
Sorry for the delay on my log per my previous post, I got slapped across the face with a sickness out of nowhere so I'll be trying out the 1440 data shortly.
I'm glad you asked because this is the peculiar part. In my previous posts here you'll see I put two excel sheets for the FIC1000 injectors. Both of those data sheets for my specific injectors I emailed FIC asking which one to use and they specified one of the sheets was just "older" data and to use the newer excel sheet. So in my post above I labeled which one is old and new. The data you mentioned is from the newer sheet, and oddly enough the "older" data seems to definitely start better as in no misfires unlike the newer data. None being logged and I can't even hear any on warmup like I could with previous revisions using the newer data. Smoother idle as well.
Last edited by Artorias; 01-02-2025 at 06:27 PM.
The numbers i put up are also very close to ID1000's, just to confirm on the regulator you set it to 58psi with no vacuum reference on and when the motor starts it's somewhere around ~45psi with the vacuum reference on at idle?
I noticed the newer data is close to the ID1000 data, I tried the ID1000 data with no discernable changes in behavior. Felt extremely similar and logged the same
In terms of fuel trim correction, higher flow rates will raise the MAF curve and lower flow rates will lower the MAF curve. So basically just changing the injector data between these two FIC sheets will not cut it, I can't keep the same MAF values. Fueling will be way wrong. Like Murfie said earlier here, I needed my MAF values higher than what I was originally dealing with. Using the older FIC1000 data that had much higher flow rates than the newer data forced my MAF curve values up higher, that could be why it's starting better.
Yep, regulator set to 58psi. Regulator gauge is off by about 3psi vs what my rail pressure sensor translates to so I set it by the pressure sensor. Vacuum works perfect with engine on, verified through logs pressure drops to around 45psi.
Edit: Trying the 1440 data with a couple different MAF curve sets, can't even get it to idle nicely. The older FIC1000 data with 25% removed from the stock MAF is the best I've gotten it to idle so far. It's just a matter of fixing the slight bog it has at startup with the old 1000 data how it slowly rises to target idle.
Last edited by Artorias; 01-07-2025 at 01:42 PM.
Quick update, I'm probably onto something. I suspect a mechanical issue with my fuel regulator. At 6PSI manifold pressure idling, fuel pressure changed itself by 10-11psi. That doesn't make any sense being referenced to manifold so unless the reported manifold pressure from the car is incredibly wrong, I am suspecting something going on with my regulator or how I have fuel lines routed.
I'm ordering a different setup to try and will report back results.
Reply With Quote