GM offset vs. Ford offset

[Alvin]

I was wondering if anyone else might know a little more on this subject.

I'm working on a pretty neat spread sheet to convert ford data into GM data. It does so by best fitting the pressure and voltage modifiers to breakpoint, offset, hi and low slope into 1st, 2nd, or 3rd order polynomials with best r^2.

The whole sheet is pretty dynamic. Nothing comes from the square root of the pressure difference or a linear interpolation or extrapolation (unless it fits a 1st order equation best)


I'll be frank, where I'm at now looks pretty good.. but it differs from published sheets that do the same thing. I'm at a point where i don't know if I'm doing a better job, they are doing a better job, or maybe we are both off!

I'm going off these assumptions.. am I right or wrong?


The GM definition of offset is where the dynamic injector flow rate (meaning changing with pressure/voltage) intercepts the x axis




The Ford definition of offset is the point at which the low slope line intercepts the X axis which can be found by finding the dynamic breakpoints, the dynamic low slope, finding the low slope equation for a line and setting Y = 0.

I've drawn a simple graph in paint to illustrate my assumptions.

[DSteck]

Use ID data to go back and forth to check your math.

[LSxpwrdZ]

I've got a spreadsheet that does the same things only it doesn't fit the data into a polynomial, it just spits out the raw data.

Now your definition of the offset I'm not so sure is exactly correct. I've alway's been under the assumption that offset (usually vs voltage) is the physical amount of time needed to get fuel from the tip of the injector from the time the coil of the injector is energized. So the offsets "should" transfer right over once the pressure multiplier is taken into consideration.

My spreadsheet basically transfer's that difference in slope from the low slope to the high slope and stops at the breakpoint. Mine is still a work in progress with lot's of notes added into different calculations to further add to the spreadsheet. But I've been using it on all of the car's I've been tuning and the car's seem to be falling right into place with the data that I have generated. I also compared some of my findings with Greg Banish's data and the results are very very close to within decimal points. So I would assume I am on the same track he is.... whether it's right or not I don't know haha. I would really love to have a microcontroller and a flow bench to validate the data I've come up with.

[DSteck]

Offset is not a measured physical attribute of the injector. It is a value taken from the plot just as Alvin has indicated.

ID's website had some good articles explaining how data is generated.

[Alvin]

Well, Here you go guys. This differs from Banish's data. Works good for me.. Question is which works best for you. That is the point.. If someone publish something first it doesn't' make them right or wrong.


I've got the sheet locked but if anyone wants to talk about how I came up with the data I'll be glad to oblige.

Ford Injector Data to GM Gen 3 LSX data

[Alvin]

BTW. This has always been about learning more and not blindly following others... I hope it doesn't step on anyone's toes.. Rather inspires others to find this data themselves.

[DSteck]

I'm going to stick to being an Injector Dynamics whore. Lol.

[Alvin]

I'm going to stick to being an Injector Dynamics whore. Lol.

Does he give out data for ford injectors

[DSteck]

Does he give out data for ford injectors

Not for actual Ford injectors, but supplies data in Ford format. That's why I said to use their data to check your spreadsheet and see how close it gets.

[LSxpwrdZ]

I tried the tool today and it's spitting out some pretty weird numbers for my 60's. The Voltage offset data it is spitting out isn't far off from what I have calculated after I made a calculation adjustment to mine for the correct offset point. The flow rate and the short pulse adder looks a bit off though. I'm going to experiment some more with it to see how some other injectors come out compared to the data I have.

[Alvin]

Like I said, If found the same thing.. That it differed big time from published data on those injectors but to be frank if you put a set of 60's or 80's in a LS1 car with the data that is published you hit a 2.2msec floor that you can't get out with the VE table. The data that I've come up with is pretty close but if your finding offset, SPA data for injectors your doing it by trial and error so I wouldn't expect it to be exact.

There is some error too just because we are extrapolating data that is not there. Ford gives you data from 20 lbs to 60 lbs and we need it from 58 to 70 lbs to cover the pressure differential when in vacuum. We also need some voltage points that are beyond the range of the ford data.


As long as my assumptions (last tab in the sheet) are correct the spread sheet is spot on and seems to be working really good for me.

[eficalibrator]

...if you put a set of 60's or 80's in a LS1 car with the data that is published you hit a 2.2msec floor that you can't get out with the VE table.

That's often a PCM limit based on transient fuel min mass limits, and has nothing to do with the data for the specific injector being used at the moment. Scaling can sometime make this look worse, so you're more likely to have ths issue with injectors that flow more than the older PCM limit of 8.0g/s (63.5lb/hr).

[Alvin]

I don't want to argue but the data I generated myself and the data I'm able to retrieve in this spreadsheet does not have that issue. So if it was a transient fuel limit it would be a problem with anything you put in the offset tables/SPA

[DSteck]

I run into the same problem with the 2.2 limit even when dropping that transient value down to 0.001... Except with IDs. Lol.

[LSxpwrdZ]

I'm running my 60's in a cam and bolt on stock bottom end LS1 (needed larger injectors for the fast102 and had these laying around). With my data I am able to control fueling on decel which hits pulsewidths as low as 1.6ms. Idle usually hovers around 1.8-2.0 with a fairly lean lambda readings.

[edcmat-l1]

I run into the same problem with the 2.2 limit even when dropping that transient value down to 0.001... Except with IDs. Lol.

I can get SD 80/83s lower than that with my pre ID data I hacked up myself years ago.

[DSteck]

I can get SD 80/83s lower than that with my pre ID data I hacked up myself years ago.

Want a cookie? I can do it too using lower short pulse adder values. I'm speaking strictly to the Banish data.

[edcmat-l1]

Want a cookie?

No thanks. Bought a case of Tagalongs from my niece.

[DSteck]

No thanks. Bought a case of Tagalongs from my niece.

Why anyone buys anything besides Thin Mints is beyond me...

[Alvin]

Just a little update.. Turns out the spreadsheet I uploaded had some errors that I fixed in my personal copy but never fixed in the copy I intended for public use. I'll clean it back up and have it under the same link by tomorrow when I get back to work.