Just swapped Injectors, now lean AFR

[CLRD4TO]

I've had a few PM's about this thread so I figured to keep the log going. The "fix" didnt last long. I was able to tune the drivability range really well except deep decel acted like DFCO. In closed loop, the trims went nuts and the car ran worse and worse. Knowing my tune was fine with stock injectors, I set all the VE adjustments back.

Since the flow rates aren't correct in the Calibrated Success spreadsheet for these green tops, I researched more. I found that bumping up the short pulse adder curve up to the 4 msec limit, the lean decel went rich. Ah ha! Im on to something, just no clue what values really need to be in there. I searched here and found some really good related threads about minimum injector pulse. Which would explain my low MAP fueling problems.

I then searched for spec sheets. This one is all over the net http://tunedbyjake.com/docs/m-9593-f302.pdf According to this, the voltage offset is way off from Greg's. I hand smoothed the new offset values in, flashed and scanned. Much different AFR errors, the whole VE table has different results, few low error.

I copied the short pulse adder back to stock and worked on the minimum injector pulse. I can get pulsewidths below 3 msec close to commanded AFR scanning. But its not accurate and higher injPW has error too. Since the spec sheet has 4 tables of multipliers, Im not sure if not only the offset table on the sheet wont work but how do I also figure out the Manifold Vacuum difference on the Offset vs Volts vs Vac?

I read here on older threads to work backwards. My plan is to adjust the whole stock offset table in % to best match VE AFR error. If that has desired results, then I will focus on short/min pulse.

No answers yet but I will post if I can find what works.

[DSteck]

Another good reason for getting the re-sized GM injectors from FIC.

The crap they sell is useless. Redrilled injectors don't just change flow rate and maintain offset and short pulse adder.



The green top data works just fine. I've never had a problem with it on any car I've done with those injectors and Greg's data. I think your wideband is junk. If trims go crazy in closed loop, then obviously your narrowbands and wideband disagree severely on where stoich is.

[CLRD4TO]

If my WB was junk, then where low MAP AFR would be stoich then WOT would be so fat it would fall on its face. It makes great power at WOT so I dont think my WB is bad.

Whether it is or not, can I ask you guys to check one cell from your Greg Banish data. My copy of Greg’s data shows at 0 vac, 14.0 V, offset cell value is 0.090653. Same cell in my stock offset table is 0.3510. Ford’s offset at 39 psi is 0.670, then with 60 psi multiplier is 0.752142

This is a huge split. I would think if Fords offset is higher than stock 28#, Gregs should be higher too.

[eficalibrator]

The point in time that Ford uses for "offset" is very different from the point that GM uses. You cannot simply copy the Ford offsets into a GM calibration and expect them to just work the same.

There's a reason I went through all the trouble to solve for actual fuel mass vs pulsewidth at both pressures and then go back and solve for the right numbers in GM units. The raw data comes from the Ford SciLab. If you have a better or more accurate source that you'd like plug into your car, be my guest.

[CLRD4TO]

Im just concerned I have the wrong data. Im not trying to debate or debunk whats out there already. Im at loss, Ive had the car for a few years and havent been able to use the nitrous that came with car. It shouldnt be this difficult to self-tune.

[CLRD4TO]

Finally figured out what to do, after I figured out what the cause was. I noticed that the lean problem Ive been having is when the injectors are below the short pulse limit. Which with my car, all cruising/idle/decel is below this limit. Above 4 msec, these injectors are fine.

I copied the stock '02 offset, adder, and min pw tables back in. AFR error above 4.0 msec is accurate. I then tuned the short pulse adder at PW's less than 4.0 to best match AFR, and it didnt need much change either. Ive also lowered Min PW to 1 msec but lowest it scans is 1.8 msec for some reason. Now its all tuned within 4% across the board, except for deep decel which is within 10% error now.

I also did move my wideband to the Y-merge. I do have more stable readings now but no drastic changes compared to previous log files.

Im bummed Greg's data didnt work. But Ive found that some green tops work on stock tables and the rest do work on Greg's numbers. Im still glad I purchased his dvd though, great instruction! I hope this thread helps others out there with this same problem.

[DSteck]

I just tuned a GTO yesterday with green tops... it was a breeze to dial in with Greg's data. I'm incredibly curious why it wasn't working for you.