Bringing TBI and Multi Port Fuel Injection to a New Level.     EFI Conversions and Tuning! Seattle to Portland! E-mail Tuning Consultant!
Results 1 to 4 of 4

Thread: M-9353-BB302 Injector Voltage Offset work for sharing/ feedback

  1. #1
    Fuel Injected!
    Join Date
    Mar 2015

    M-9353-BB302 Injector Voltage Offset work for sharing/ feedback

    I've wanted to address fueling on my primary LT1 car for a while. The dyno tune from 2009 was good on power and driveability but pretty rich all over. I started checking the tune after learning a lot more about it over the last several years and found some things that needed doing. Primarily, O2s weren't relocated, the VE tables weren't touched. I stumbled across Injector offset voltage and realized I could have some noticeable fuel waste and rich condition. Those that played with it and got it close with Bosch-style Ford 30# reported better all over the place.

    Searching for the answer from someone else's hard work (hey, I admit it) I came up with no clear answers. I have attempted to do this with fairly simple math and Excel's scatter chart functionality to get what I needed, pretty close. Unless I just completely don't get it.

    Ford gives us 8 data points of offset vs voltage. GM uses 32. GM is missing 3 values that Ford gives (11V, 13V, 15V). Ford gives us fuel pressures closest to GM's 43.5 with 39.15psi and 44.95psi. So I divided that into chunks to make up some increments in between to get close to 43.5psi. Then figured out the corresponding multiplier for 43.79psi (close enough, and way better than leaving values meant for the much faster response Rochester 24#ers). That estimated multiplier is 1.05512.

    Next, I laid out 35 values, knowing I'd not use 3 of them (11V, 13V, 15V) but graphing the voltage vs offset curve would help me to find the in-between values I needed to generate and paste into TunerPro.

    Here's the shape with 8 data points, as Ford hands it to us for the M-9353-BB302:

    Now with 35, but with 27 unknowns as zero:

    So I left the chart open while I played with the values in the column. I threw some meaningless fudged numbers in below 6V so that end of the chart is not relevant. Hey, at 6V, it's not even running, right?

    Next, I multiplied these values by 1.05512 to get offsets for those voltages at 43.79psi:

    I've attached the spreadsheet I created, tried to make notes on it as to what I was doing. 2 sheet workbook. Feedback welcome. Column F- red numbers are multipled, blue values are filled to make the graph smooth between REAL DATA points.

    I changed offset voltages, dropped the low pulse adder by an arbitrary 11% and added the O2 relocation values. The car noticeably rips better and harder than before. Throttle is more responsive, revs quicker. Definitely less fat at idle and part throttle. I left PE mode alone as it was already really good. I also dropped the idle from 875 to 800 and it's stable, though I may bump it up a hair. I do plan to datalog it as soon as it stops raining again, to make sure all is good, not leaned out where it shouldn't be. Car is as follows:

    94 M6 Formula 76k mi.
    ported 374 casting heads with 1.94/ 1.56 valves + appropriate springs, 7/16 studs, raised plates
    Comp Ultra Pro Mag 1.6
    short travel 875-16 Comp lifters
    Comp 503 cam
    approx. 10.9+:1 ish SCR
    properly ported stock intake
    TPIS 58mm TB
    Stock bottom end + resize big end + ARP
    Ford Racing 30# M-9353-BB302
    better than stock pump but not overkill 255lph, lol
    Stock regulator
    44psi @KOEO
    Pacesetter long tube headers
    mandrel bent true dual exhaust/ X pipe with glasspacks tucked/ stacked into the old muffler location
    TFS ceramic elbow
    retains unmodified (clean) MAF
    Stock 3-channel ABS 10-bolt with 3.42s
    LCAs/ relo brackets
    Attached Files Attached Files
    Last edited by Gojira94; 04-25-2019 at 11:17 PM.

  2. #2
    Fuel Injected!
    Join Date
    Mar 2015

  3. #3
    Fuel Injected!
    Join Date
    Mar 2015
    I've been agonizing on how to alter the "Injector Offset" parameter table in TunerPro (using EEx 4.1 xdf). I revisited the work above and started doing pen and paper exercises with the Ford datasheet. Staring at a pdf really doesn't help me work the numbers into my brain in a useful way. The top left table of the Ford Datasheet is a set of constants, to be multiplied by the appropriate multiplier in respective tables.

    ALOSL (lbs/sec) multiples against its corresponding multiplier for each given pressure in the FNPW_LSCOMP table
    AHISL (lbs/sec) multiples against its corresponding multiplier for each given pressure in the FNPW_HSCOMP table
    FUEL_BKPT (lbs/) multiples against its corresponding multiplier for each given pressure in the FNPW_BKCOMP table
    MINPW (ms) of .776 is the minimum consistent pulsewidth these injectors deliver at Ford factory pressure of 39.15psi
    FNPW_OFFSET multiplies its whole table by its corresponding multiplier for each given pressure in the FNPW_OFFCOMP table

    Now, since we lack a multiplier for 43.5psi in each multiplier table, we need to find that, or get as close as we can so we can use a reasonable value for each table. Again, I used Excel to graph this and stretch the x axis to get the values as close to correct as I could manage.

    1) For minimum pulsewidth, I fudged by just calculating the proportion of 39.15 to .776 and got 43.5 to .851. I also tried using the Bernoulli equation, which we normally use to alter the flow vs pressure change, to find a relative minimum pulsewidth at 43.5psi. That method returned .776ms at 39.15 vs .8074ms at 43.5. Converted to usec, those are 776 and 807.4usec. Math geeks, please correct me on which of these 2 methods, or neither!! This may be off!! Time or pressure, or a blended value of both, not sure which is better... or even accurate.

    So in TunerPro the STOCK "Injector Offset" table shows:

    (Again, this is the stock table)
    BPW | Adder
    usec | usec
    488 | 381.5
    549 | 351.0
    610 | 305.2
    671 | 274.7
    732 | 244.1
    793 | 213.6
    854 | 198.4
    915 | 167.8
    976 | 152.6
    1037 | 137.3
    1098 | 122.1
    1159 | 106.8
    1220 | 106.8
    1281 | 91.6
    1342 | 76.3
    1403 | 76.3
    1464 | 61
    1525 | 61
    1586 | 61
    1647 | 45.8
    1708 | 45.8
    1769 | 45.8
    1830 | 45.8
    1891 | 30.5
    1952 | 30.5

    If I understand correctly, since the minimum consistent/ repeatable pulsewidth for these injectors at 43.5psi is in the neighborhood of 807-851usec, what we refer to in many cases as a "low" pulsewidth adder falls below 807-851usec. The table shows a parameter data point of 854usec. **This pulsewidth intersects with 14.0V and 14.4V offsets in the Injector Voltage Offset table.** So, I think BPW below 854usec is where the "low" adder is relevant for this injector. Some have had to experiment with different injectors to see what they do vs. what the engine wants with each specific injector/ manufacturer to find where to increase and where to decrease values. Some have zeroed out the table completely and dialed it in by feel/ AFR at low RPM/ part throttle. I guess the point is that at/ below the 793-854usec BPW is where most of the "adding" would be, much less so above, ending in a zero value at 1952usec (max BPW in the table). I'm thinking this can give a much closer place to start with dialing this in.

    2) My 'close enough' value for Low Slope multiplier at 43.5psi is 1.076. Multiply this by the Low Slope Constant and you get .010784748 lbs/sec in the low slope. Multiply this by the length of the pulsewidth in sec/msec/usec at any given RPM vs MAP vs Hz/ g\sec and you get the delivered fuel mass for that pulse in the low slope, if the resulting mass is < .00001336 lbs.

    3) My 'close enough' value for High Slope multiplier at 43.5psi is 1.0584. Multiply this by the High Slope Constant and you get .009041 lbs/sec in the high slope. Multiply this by the length of the pulsewidth in sec/msec/usec at any given RPM vs MAP vs Hz/ g\sec and you get the delivered fuel mass for that pulse in the high slope, if the resulting mass is > .00001336 lbs.

    4) My 'close enough' value for Fuel Mass Breakpoint multiplier at 43.5psi is 1.0091. Multiply this by the Breakpoint Constant and you get a breakpoint fuel mass of .00001336 lbs. as the switch point from low slope to high slope.

    5) These are my slightly tweaked Injector Voltage Offset values with these injectors at 43.5psi:

    Volts usec
    0 | 213
    1.6 | 213
    3.2 | 6011
    4.8 | 6011
    5.2 | 5004
    5.6 | 4211
    6 | 3540
    6.4 | 3112
    6.8 | 2777
    7.2 | 2532
    7.6 | 2319
    8 | 2166
    8.4 | 2014
    8.8 | 1861
    9.2 | 1739
    9.6 | 1617
    10 | 1525
    10.4 | 1434
    10.8 | 1342

    11.2 | 1281
    11.6 | 1190
    12 | 1129
    12.4 | 1068
    12.8 | 1007

    13.2 | 976
    13.6 | 915
    14 | 854
    14.4 | 823
    14.8 | 793

    16.4 | 396
    18 | 122
    19.6 | 0
    Last edited by Gojira94; 1 Week Ago at 10:21 PM. Reason: stock tables corrected

  4. #4
    Fuel Injected!
    Join Date
    Mar 2015
    Let's back up a bit... To understand and apply correct fueling, we need to understand the density of the fuel. Ford datasheets use lb/sec for low and high slopes and the breakpoint (where flow becomes pretty much linear). Injector Dynamics graphs their injectors for pw vs cc/min flow. The also have a good explanation of the Ford datasheets and have a good article on how that looks in a graph as well.

    Water has a specific gravity of 1.00, weighs 8.325lb/gal, and has 453.59237cc per lb. Gasoline varies. Some race fuels have a SG as low as .715 to .728, but pump gas varies from about .710 to .770SG. For this exercise, I'm going to assume a SG of .740 to represent E5-E10 93 octane pump gas. This gas would weigh 6.1605lb/gal. It's helpful to convert flow from weight to volume, from lb/sec to cc/min. To do that, I'll multiply 453.59237cc/lb of water by 1.26 to get how many cc in a lb of this assumed fuel ( = 571.5263862cc/lb).

    Low slope = 0.010023 x 1.076 (43.5psi multiplier) = 0.010784748lb/sec
    x 60 = 0.64708488lb/min x 60 = 38.8250928lb/hr
    38.8250928lb x 571.5263862cc = 22,189.56498186363936cc/hr;
    /60 = 369.82608303106056cc/min; /60 = 6.1637680505176776cc/sec
    /1000 = 0.0061637680505176776cc/ms; /1000 again = 0.0000061637680505176776cc/usec

    High slope = 0.008542 x 1.0584 (43.5psi multiplier) = 0.0090408528lb/sec
    x 60 = 0.542451168lb/min x 60 = 32.54707008lb/hr
    32.54707008lb x 571.5263862cc = 18,601.509344220544896cc/hr;
    /60 = 310.0251557370090816cc/min; /60 = 5.16708592895015136cc/sec
    /1000 = 0.00516708592895015136cc/ms; /1000 again = 0.00000516708592895015136cc/usec

    Using a calculated offset of .8683ms at 14V, we can now get a very close idea of fuel per pulse. Remember, offset is the time it takes for the injector to start to open. The offset time it takes the injector to start to open is accounted for in every pulsewidth commanded by the PCM, adding whatever offset is in the inj. offset table at the batt. voltage detected at the time the pulse is commanded, plus calculating the pulsewidth of actual fuel needed. It's amazing that a 25-26 year old computer can do this 4x per revolution up to 7,000RPM. That's 386-387 times per second at 5800RPM.

    Now, for the breakpoint, the amount of fuel at which flow becomes 'linear' for the injector.
    That's 0.0000132400lb x 1.0091 (approx. multiplier at 43.5psi) = .000013360484lb.
    x 571.5263862 = .0076358691384029208cc.

    Here's where I start to lose how to do this...

    Calculated against the low slope, this mass of fuel occurs at a net pulsewidth, excluding offset, between 1.220ms and 1.239ms 'on' time.
    Calculated against the high slope, this mass of fuel occurs at a net pulsewidth, excluding offset, between 1.464ms and 1.478ms 'on' time.

    If I start calculating delivered fuel using high slope just above 1.22ms - 1.239ms it's of course lower than the delivered fuel AT the breakpoint, by about 73cc/min! What I'd hoped to see is the point at which to stop adding low pulse adder, in terms of ms / usec. In practical terms, I can again fudge and say low pulse adder needs to happen between the lowest data point in the table (.488ms / 488usec) up to 1.24 - 1.47ms (somewhere between 378 - 458cc/min), with minPW somewhere in between at about .818ms/ approx. 305cc/min. Though at least (I think) I can safely say that no adder is needed at all in the cells 1525usec and above in the Injector Offset Adder table in TunerPro. But that could be as low as the 1281usec and above cells.
    Last edited by Gojira94; 2 Days Ago at 12:17 AM.

Similar Threads

  1. 95 lt1 injector offset questions
    By Mr.owl in forum GM EFI Systems
    Replies: 56
    Last Post: 02-11-2019, 02:10 AM
  2. Injector offset vs battery voltage
    By 1BadAction in forum TunerPro Tuning Talk
    Replies: 0
    Last Post: 11-30-2016, 02:41 AM
  3. Injector Offset Adder - Clarification
    By Terminal_Crazy in forum TunerPro Tuning Talk
    Replies: 9
    Last Post: 02-05-2016, 05:04 AM
  4. Injector offset
    By tedjan in forum TunerPro Tuning Talk
    Replies: 2
    Last Post: 02-07-2015, 10:59 PM
  5. Injector Offset/Latency Chart Link
    By lightning-jr in forum Fuel Injection Writeups Articles and How to New and Old
    Replies: 1
    Last Post: 01-25-2012, 03:18 PM


Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts