Can we back up a little bit and talk about what problems the vehicle was having with the original tune, with the "wrong" data? It would probably be worth the effort to try and fix that file than go this seemingly ridiculously wrong route.
Why not put the "wrong" injector data tune back in it and data log it?
EFI specialist
Advanced diagnostics, tuning, emissions
HPtuners dealer and tech support
email=[email protected]
In post #51 above the 'C' files are the final SD - Wrong Injector - Car Ran info you asked about. File and log of it. It was after this run that I did the Write Entire and switched it back to the OEM 1BAR MAF OS that AAronC7 hooked me up with. Is that what you were looking for?
The 'D' file set has the banish data.
Last edited by JumboShrimp; 1 Week Ago at 01:01 PM. Reason: added
No. Go way back in the way back machine, before you changed whatever you changed to create the bad behavior. What was the car doing that you were trying to fix? Was it running great and you just decided some injector data wasn't right so it needed changed?
Completely forget about the basis of this thread. Let's talk about how the car ran before any recent changes. Were you happy with it?
EFI specialist
Advanced diagnostics, tuning, emissions
HPtuners dealer and tech support
email=[email protected]
The tuning was coming along nicely. Ron 'noticed' that he had the wrong Injector Data and was pissed at himself for making a mistake. He corrected it and it went straight to hell. The car was running pretty good but had some burble / stumble going into boost. He was trying to fix that. But was wondering if it was fuel hitting the adapter and pooling - that's hard to 'fix'.
In the mean time I had a conversation with Bob at FIC. He's SUPER nice & helpful. He hooked me up with Greg Banish. Both Ron and Greg said I should be running a MAF in a 4" tube - both know the car is FI. So, I put a MAF in a 4" tube.
In talking with Greg Banish the ideal tune uses both MAF and SD - that's the goal here - a Hybrid type of tune. I don't speak the language but That is why it has a MAF. Ron is trying to make that hybrid tune happen. Correct me if I'm wrong but that's it - in a nut shell.
The HPT files in post 51 have both versions of injector data - FYI.
And as far as the plates... I did get a different design that had a milled out area to 'help' with fuel going straight at it. I know they are not ideal but... There a bunch of discussion earl in the thread about them.
Thanks again for your help!!
But here's the thing. You really didn't change the "data". You changed the transient. If that's all he changed and it went to hell in a hand basket, I'd change it TF back.Originally Posted by JumboShrimp
If you were satisfied with the way it ran before, go back to that, whether you think it's right or not. If you want to add in a MAF, do that with your tune that runs well.
Forget about making the numbers what someone tells you they should be. Especially stuff like transient tables. Other than the trans min fuel I normally don't touch the other tables unless there's a problem I think I need to fix that seems to me to be related to actual transient fuel, wall impact, wetting, evaporation, etc.
Go back to your old tune. Turn the MAF back on. Wire it up, start dialing in that table. Give it a try. But forewarning, GEN3 PCMs using a SD COS typically don't work well using the MAF. They just don't. GEN4s do fine setup that way.
This goes back to what I preach constantly. Don't change shit just because you can. I'm not depending on someone's "numbers" as some holy grail.
EFI specialist
Advanced diagnostics, tuning, emissions
HPtuners dealer and tech support
email=[email protected]
I made a stiff drink and did a Write Entire. It went fine - that always spooks. I put the 'C' file in post 51 back in.
The Fkr started right up - 1/2 of a key turn, no complaining, no BS - it just RAN. I am a happy camper. I Rev'd it a few times - it doesn't like cool temps - much better at 180* +
I am logging MAF data so... we'll see. I still think the blended / hybrid tune is that way to go. I know the Gen3 ECM has limits.
It was just nice to hear it run again. here's a warm-up log - till the fans come on - if anyone wants to comment. It is raining CRAZY here or I'd be driving it.
Thanks to all who even cared about my stuff - this is a GREAT place!! Good people here!!
I am going to build you a file that WILL require several iterations of logging and correction to get things right. I will be starting from a mostly stock file with known good injector data and we'll look at errors from there and make corrections. What I can't have is 12 cooks in the kitchen with us while we're doing it. I have done crazier stuff before that drove just fine.
That's fantastic! Thank you!!
Thank you, Greg. Jumbo (AKA Mike) and I appreciate it.
Ron
Y2K C5 A4 Coupe (10.78 @ 127) Cathedral port 6.2
S476 L33 5.3 69 Nova 8.76 @ 158 Drive to the track street race car
06 Cobalt SS bought new
Ok hold on. You are paying a guy to tune your car, and not just any internet warrior, but an OE Calibrator. Yet you are asking the forums?
Do you see anything about money? What's with the attitude?
You know nothing about this build, My relationship with Ron or Mr Banish.
I just got done complimenting this forum / the people on this thread, on how generous some people are with time & talent.
And them YOU show up - WHY are you so Angry. No skin off your nose, right?
Greg Banish and I have a lot in common. He's a good dude and a Rock Star with this stuff. He's trying to help me - Are You? Your comment was toxic. You Are toxic. This forum need less of that - ALL Forums need less of that!
Just FYI...
I did an LS1 / RX7 swap with my son in '16. It was my first dive into the LS Pool. I had to disable VATS - HMMM - my No Start resembled that. Without help from ANYONE I disabled it in my file & the F'n car started. That was FREE - it came out of MY limited experience with 'tuning'.
Troll me not.
End of Rant, please.
You didn't do that FIC data from post 48? How come the beginning of SPA holds for so long on these cals? Physically that doesn't make sense. Flow graph to GM data.
Thanks for the question Sirius.
As far as I know the correct / latest data is written into the file at this moment - I can look at the XL file and try to learn though. I am on a mission to learn as much as I can about this all.
Can you elaborate a bit for me? What is SPA? What doesn't make physical sense?
If it's too Rookie of a question I understand - but I'm the definition of Rookie.
Thanks again
SPA is short pulse adder. It's one of the tables in Fuel>General.
Offset is the delay time between when the injector is told to fire and when it physically reacts. When the injector starts to open there non-linear flow. Think about what happens when you squeeze a garden hose nozzle to full on. There's a brief wide misty spray as the nozzle opens.
In standard injector characterizations, Offset is defined as the immediate physical reaction. In Ford and many other characterizations, offset is the delay in reaction of a theoretical line that linearizes the non-linear region. In GM, offset is defined in terms of the theoretical linear extension of full flow IFR.
Here, true offset is the black line mapping non-linear flow. GM offset is the red dotted line extending through full flow. Short pulse adder is a compensation table for theoretical to physical. It's the green lines bridging the two, measured along the x-axis in milliseconds.
SPA.png
So when you look at the GM SPA in the file it doesn't make sense. The length of the green lines for SPA isn't constant for the first .610ms.
GM SPA.png
Further differences between GM and Ford are illustrated in the graph. High Slope is equivalent the linear flow rate. Low Slope is an average flow for the non-linear region. Breakpoint is the transition from non-linear to linear.
How those terms apply is here. This is an example of a typical characterization. Values are taken at the the standard pressure of 39.15psi. Math can be done to redefine terms at different pressures using the compensation tables.
M-9593-LU47.png
My preferred method of converting Deka data is to use the Continental characterization. Here's Deka 60 Short - FI114962:
Deka 60 Short Continental.png
I then graphically solve for IFR, Offset, and SPA. Deatschwerks happens to supply data compatible with this method. In one tune I was having trouble getting VE to look right. The table was showing typical low rpm/map inflation, accompanied by fueling error and inconsistent STFT's, that indicates incorrect Offset and SPA. Instead of using the supplied GM data I converted based on their supplied point-by-point data:
DW conversion.png
https://forum.hptuners.com/showthrea...l=1#post771172
VE came in a lot more sensibly and fueling could be controlled accurately.
VE before and after.png
Another trick for addressing low rpm/low map issues is to adjust supply pressure. For example, if the injectors are oversized and causing problems then I adjust the fuel pressure and redo the data. The point is to minimize the effect of the non-linear region. I do this on some boosted tunes with uprated injectors since they have an FPR that can be adjusted. Minimization of non-linear contribution to total flow is why it's important to size injectors correctly.
If you've followed everything up to this point the obvious question is, "If we're accurately mapping the non-linear region then why should it matter since fueling is being tracked?" The answer is that at low pulsewidths injectors don't react the same every time. If you look at the Continental data (pdf attached so you can zoom in) the bottom left graph is standard deviation. Below 1.2ms inconsistency from one opening event to the next greatly increases. This is due to both the injector itself and the error in the components of the flowbench.
Which means the other reason is that it seems everyone has a different way of doing injectors. IFR is the easy part and hard to mess up. Even if it's off it can be corrected in the airflow models. Not always the case with Offset and SPA. By minimizing the non-linear contribution, the error from the injector itself and the guy with the flow bench are simultaneously minimized.
The Deka 60 Short data you have from FIConnection was done by Banish. FIC has been using his firm Calibrated Success for characterizations lately. There are instances where there has been "updated data" from FIC (old and new from FIC attached). No explanation as to why. People are able to tune with either set. Thing is they can't both be correct. I've asked questions of everyone involved and have been met with non-answers, so for Deka I rely on the Continental data and graphically solving.
Courtesy to Alvin of PCMofNC for showing me about GM vs Ford injector characterization.
The best advice you're going to get is to contact Alvin and have him tune this for you. Impressed by this post? Seek the Grandmaster.
https://forum.hptuners.com/member.php?5378-Alvin
https://www.pcmofnc.com/
Last edited by SiriusC1024; 6 Days Ago at 09:37 PM.
You are currently about where I was 15 years ago with converting Ford data to GM units (and pressure). That's how I made the files that I delivered along with my original GM tuning DVD's data disc that everyone had so much success with. That data was mathematically derived from the Ford PFSL flow lab data, also publicly released through FRPP. I'd like to say I have a decent handle on how the various OEMs convert the raw mass vs time curve to their ECU data.
Moving forward, I worked through several iterations of my own measurements on my own test bench. I didn't release any of that data until I proved that I could reproduce data that was within ~2% of the OEM data for that same injector across the usable range. Keep in mind that there is also a min PW that comes from this testing where we see that fuel delivery is no longer consistent below some point as the injector doesn't open completely each shot at that small of an activation time. Beyond that, we still see part-to-part variation in the population. OEMs will tolerate as much as +/-6% in production here, we obviously want better "matched" sets in the performance world. But if we test a single set, it's not a surprise to see some variations, especially at short pulses. The good news is that our minimum consistent pulse width here is still well below what we need for idle or cruise.
Regardless, there are literally THOUSANDS of people successfully running my recent data from the new machine from FIConnection, myself, and several others. I started building Mike's current file with that data, and so far it seems to be working just fine, again. Ideally, I would prefer to run the exact set of injectors that will be used on my bench before doing ECU calibration work, but for now, this generic data set is close enough.
Thank you. I spoke to the error at low PW above. I also spoke about how the only response to my questions is the one you just gave - people are able to tune with it. Andy at AACorvette said the same thing when I asked why he still uses the old FIC Deka data and keeps it posted on his website.
email.png
FIC contact could offer no better explanation.
Thousands of people isn't an answer. I also alluded earlier as to why people are able to tune with inaccurate data. I run Volvo injectors in my daily using stock GM Offset and SPA with only IFR changed. 90k miles and counting. Just because you can tune it doesn't mean that it's good for all cases. There's a point where it starts to matter.
I'm aware of what you do and your credentials. I cannot reconcile that GM SPA with the raw data. Flatlining for the first .610 ms is nonsense. You mentioned useable range. Does that include below MinPW? Sometimes that's needed. Not always a safe assumption to make:
gtp updated.png
Unless it can be explained quantitatively I stand with my suspicions. No offense, sir.
What I was getting at before was converting raw data to GM format. It is on that basis that I am suspicious. Showing a Ford cal and using the compensation tables was to help explain what's going on as far as the different ways injector data is generated. Ford's ALOSL,BKPT method of linearizing non-linear behavior obviously works, too, but it's still error isn't it?
Anyway, good on you for working to get the tune right. I'm sure it'll turn out great.
Last edited by SiriusC1024; 6 Days Ago at 02:31 AM.
DAMN! Thanks for the education! I can see reading all that more than a few times. SO much to learn!
This is the kind of thing that makes this place so priceless. No BS, No Drama, just smart, generous people that really try to help.
I got a long way to go!
Thanks ALL!
When converting to GM data from FORD you are doing a huge amount of extrapolating.
Here is a screenshot of our calculator doing this. The blue is the defined FORD data.. The red is extrapolated GM data.
Capture.JPG
Tuner at PCMofnc.com
Email tuning!!!, Mail order, Dyno tuning, Performance Parts, Electric Fan Kits, 4l80e swap harnesses, 6l80 -> 4l80e conversion harnesses, Installs
Yes, most of the error is at low PW. Before getting too excited, I would look at the values involved. If an engine is idling at 0.15g/cyl, then it needs more than 10mg of fuel on pump gas. That means that errors below the 10mg Y-value on the graph become increasingly less important in a practical discussion. Please show me a case where you need a consistent delivery of only 3mg of fuel in a single shot.
As for Andy, I'm not sure why he still has the old, incorrect stuff on his website. I got paid a while ago to build new base supercharged files for him (including proper scaling where necessary) but he still hasn't updated the website. Maybe he's not comfortable with web stuff, or he's just too busy elsewhere.
I get it, and I never ask people to believe me just "because I said so." My point was that we have a massive amount of data points indicating that we have correct data on a practical level. YES, there will be variations in the population, just like the OEMs see in production. The real question is "Are we close enough to use that fuel data with confidence and focus on the airflow/torque data?" Small errors below our practical minimum aren't a big deal. (See above) Beyond that, I'm not really going to publish my exact formulae and procedures for testing and data manipulation. I know my competitors read my posts and would love for me to teach them how to duplicate my work that I spent decades learning. They have yet to catch up, so I'm not in a hurry to give them a free lesson on how to screw me.Thousands of people isn't an answer. I also alluded earlier as to why people are able to tune with inaccurate data. I run Volvo injectors in my daily using stock GM Offset and SPA with only IFR changed. 90k miles and counting. Just because you can tune it doesn't mean that it's good for all cases. There's a point where it starts to matter.
I'm aware of what you do and your credentials. I cannot reconcile that GM SPA with the raw data. Flatlining for the first .610 ms is nonsense. You mentioned useable range. Does that include below MinPW? Sometimes that's needed. Not always a safe assumption to make:
gtp updated.png
If all this still doesn't make someone happy enough, I offer single set testing services for those looking for exact data on their exact set of injectors. This always generates a unique data set with some variation from our reference data for that same part number. But it's the unique data that follows that unique set of samples and brings the errors from <5% down to maybe 1-2% on the injector model for their project.
I appreciate the enthusiasm for finding the truth on these measurements. Please don't be offended if I come off as a bit short on some of this discussion. I'm just an engineer who deals constantly in numbers without emotion.
The errors become increasingly more important because as pulsewidth decreases the non-linear region comprises a greater proportion of total fuel delivery.
I'm not offended at all. I'm glad you seem the same.
Last edited by SiriusC1024; 5 Days Ago at 02:23 PM.
Reply With Quote