You are on page 1of 4

AreallFuelInjectorsCreatedEqual?

ByGregBanish

Areallsamesizeinjectorsthesame?Inshort,notachance.
Therearemanydifferentinjectormanufacturersouttherethatallmakeinjectorswithsimilar
flow rates. Even within some injector manufacturers offerings, there may be two different
variantsofa24#/hrinjectoror30#/hrinjector.Forthepurposesofthisarticle,Ivedug
intoahandfulofBoschfuelinjectorsthathavehadtheirdatapublicallypostedontheFord
Racing Performance Parts website. (www.fordracingparts.com) The fine folks at Ford were
niceenoughtoprovideuswithverydetailedanalysesofeachinjectorsbehaviorwithrespect
to opening time, voltage dependency, nonlinear flow characteristics and pressure dependencies. Their Calibration
Summary sheets contain all of this critical information for each unique injector design. The data in these calibration
summaryreportsaretheexactvaluesthatgetenteredintooneoftheirPCMstoallowthecontrollertopreciselyturna
fuelmassrequestintotherightinjectorontimetodeliverthatspecificfuelmass.Itsacriticalstepinmakingsurethe
PCMcanaccuratelydeliverthecommandedair/fuelratio.

CalibratedSuccess,Inc.2009

0.025

0.02

FuelMass(g)

Using the data provided by an OEM test


lab, we can plot the fuel mass delivered
versusinjectiontimeforeachinjector.At
right, we see what the traditional Blue
Top 24#/hr injectors look like alongside
the Red Top 30#/hr injectors when
operated at the normal Ford fuel
pressure rating of 39.15psi gauge
pressure.Noticehowthehigherflowrate
of the Red Top injectors requires less
timetodeliverthesamefuelmasstothe
cylinder.Theslopeofeachlineisdefined
byhighschoolgeometryasriseoverrun,
in our case here thats mass over time.
Thisiswherewegettheinjectorslope
valueofgramspersecondorpoundsper
hour. The 30#/hr injector has a steeper
slopeascanbeclearlyseeninthisgraph.
There is also a small delay before the

0.015
M9593B302"Red
Top30's"

0.01

M9593A302"Blue
Top24's"
0.005

0
0

InjectionTime(ms)

calibratedsuccess.com


injectoropensatallandfuelcanbegintoflow.Amorecarefullookalsoshowshowthattheslopechangesslightlywith
time for both injectors. This is a result of the nonlinear flow of fuel through the holes in the injector. More on this
later
Next we see two different versions
of the Ford Motorsport 30#/hr
fuelinjectors,bothmanufacturedby
Bosch.TheRedTopistheolderfat
bodyversionandtheskinnyisthe
0.02
newer part that has superseded it.
Both injectors flow almost
identically at pulsewidths above
2ms,andthereforetheslopesofthe
0.015
two lines are parallel on the high
M9593B302"Red
end. The big difference here is that
Top30's"
theopeningdelaybetweenthetwo
0.01
differentinjectorbodydesigns,with
M9593BB302
the older design actually opening
"Skinny30's"
slightlyquicker.Below2ms,wealso
see the difference in slopes that
0.005
results in a noticeable difference in
injection time required to deliver
the same fuel mass. If the PCM
thinksithastheRedTopinjectors
0
on the engine, but we really have
0
1
2
3
4
5
6
7
the skinny 30s installed, a
InjectionTime(ms)
calculated fuel mass of 0.010g
wouldworkouttoapredictedinjectionpulseofabout2.9ms.Ifwetravelalongthesolidlinethatrepresentstheskinny
injectorsthatarereallyinstalled,weseethat2.9msofinjectiontimeonlygetsus0.0088goffuel,12%lessthatthePCM
thinksisreallybeingadded.This12%errormeansthattheenginewouldoperateatanair/fuelratioof16.6:1insteadof
our usual target of 14.64:1 unless some other correction was active. The inexperienced tuner simply makes a 12%
adjustment to either the MAF or VE airflow tables to bring the engine back to the target of 14.64:1. This works for
correctingtheair/fuelratioatthisexactcondition(enginetemp,airtemp,airflowrate,andfuelflowrate),butitsbased
onabadassumptionforfueldelivery.Theresultisanairmasscalculationthats12%offallthetime.This12%airmass
errorcangetcarriedovertotheidlecontrolstrategy,makingitthatmuchmoredifficultforthePCMtopreciselycontrol
idlespeedwiththeIACmotororelectronicthrottlecontrolsystem.ThepropersolutionhereistotellthePCMthenew
injectorcharacteristicsthatwillresultinaninjectiontimeof3.1msinordertodeliverthefuelmasstargetof0.010gto
thecylinder.ThismeansgivingthePCMmoreinfothanjusttheoffsetandlinearflowrate.
FuelMass(g)

0.025

This starts to get more interesting as we note the typical idle pulsewidth of 2ms or less for this injector on a warm
engine.Thebiggertheinjector,themorelikelihoodthatthePCMwillcommandapulsewidththissmallatidleorlight
cruisingwherethisnonlinearbehaviorplaysabigroleintheactualfuelmassdeliveredtotheengine.Properlymodeling
thesmallnonlinearregionofflowcanbecriticaltogettingtherightfuelmassdeliveryatidleandcruise,evenwithtwo
sameflowrateinjectors.

CalibratedSuccess,Inc.2009

calibratedsuccess.com


We just looked at two well documented injectors where an OEMs scientific flow lab has properly characterized
thousandsofeachinjectortocomeupwiththedatapresentedinthecalibrationsummarysheets.Itsjustnotpossible
toseethewholestoryifallyouregivenisalinearflowrateandturnontimelikemanyinternetretailerssupply.Ifyou
haveasetofmodifiedinjectorsthathavebeenresizedorredrilled,youreonyourowntoguessatwhatthesespecs
maybe,butitssafetosaythatwecantsimplyassumetheyllbethesameasthestockinjector.
But what about running a fuel
injector at higher rail pressure? Its
well known that the linear flow rate
of an injector is increased as the
0.02
pressureacrossthatinjectorisraised.
Many
amateur
(and
some
professional)tunerswiselycorrectfor
0.015
this change using the Bernoulli
equation(theflowrateincreaseswith
39.15psi
thesquarerootofthepressureratio),
58psi
0.01
but unfortunately stop cold there.
The truth is that the increase in
pressure also affects the both the
0.005
openingdelayandthenonlinearflow
regionofinjection.Inthecaseofthe
skinny30#Ford/Boschinjectors,we
see that increasing the rail pressure
0
addsslightlytotheopeningdelaybut
0
1
2
3
4
5
6
7
raises the slope (flow rate) at both
InjectionTime(ms)
smallandlargepulses.Thepointatwhichtheslopechanges(thelimitofthenonlinearregion)evenshiftsslightlydueto
thepressure.Again,theFordcalibrationsummarysheetclearlyspellsthisout,soitsimportanttoheedtheirdataunless
youvegotsomethingmoreaccurate.(Goodluck!)
0.025

FuelMass(g)

30#SkinnyFRPPInjectors

Doneright,theresnoreasonwhythefuelpressurecantberaisedtoincreaseinjectorflowratewhilestillmaintaining
excellent control over the delivered fuel mass and overall air/fuel ratio. In fact, some of the Roush and Ford Racing
superchargerkitshaveusedthistrickofincreasingfuelpressuretogainadditionalflowthroughthestockinjectorsas
longasthefuelpumpcankeepup.Theyhaveadjustedboththelinearflowrate(highslopeinFordspeak)andnonlinear
characteristics (low slope, breakpoint, and offset) to compensate for this change in working pressure while still
maintainingtightenoughair/fuelratiocontroltopassfederalemissionsstandards.
What about putting Ford injectors on a GM vehicle? Its a mechanical fit thats easy to make, with only a wiring
connector change or fuel rail spacer required in some cases. The trick is that GM and Ford have different ways of
describingthesamephysicsandnonlinearflowcharacteristicofthesameinjector.Theinjectoroffsetcanbedescribed
aseithertheinterceptofthehighslopewithzerofuelmassorasthepointatwhichtheinjectorreallyopens(wherethe
low slope intercepts zero fuel mass). Similarly, GM also uses a short pulse adjustment to make up for the lack of a
secondinjectorslopeatlowmasses.BothstrategiesgetthesamejobdoneandallowthePCMtoproperlymeterfuelat
smallpulsesinthenonlinearflowregionoftheinjectorscapacity.TranslatingfromFordtoGMunitsandtablescanbea
bit tedious, but its necessary if you want to get idle airflow estimates right when calibrating for MAF or VE using a
wideband.
CalibratedSuccess,Inc.2009

calibratedsuccess.com


Sohowbadisitifweignoretheshiftinoffsetandnonlinearmodelingwhenchangingpressure?Ididthemathboth
ways to generate the mass vs. time curves that result from simply replacing the linear injector flow rate in a GM
applicationwithapressurecorrectedflowratewithoutchangingthenonlinearcompensationtables.GMusesaslightly
differentdefinitionofoffsetthanFordanda
0.025
short pulse adjustment/adder table to
replace the low slope and breakpoint that
Ford uses to describe the nonlinear flow
0.02
region. If I keep the stock short pulse
adjustment and offset tables, but swap the
injector flow rate, I end up with a different
0.015
flow vs. time curve for fuel delivery to the
engine. I plotted the results of this half
StockGMSPA&
tuned injector behavior against the Ford
Offset
0.01
supplieddataattheelevatedpressure.Right
CorrectedFord
away,weseethattheoffsetisntthesame
Data
betweenthereferenceGMinjector(another
0.005
Bosch 28#/hr injector in this case) and our
corrected Ford/Bosch injector at 58psi. This
fixedtimeoffsetdifferenceiscarriedacross
0
thewholeinjectiontimerangeandresultsin
0
1
2
3
4
5
6
7
significantdifferencesindeliveredfuelmass
InjectionTime(ms)
at low pulsewidths and still a noticeable
shiftattheupperend.Below2mswealsoseethatthenonlinearflowregionismismatched.Therealsolutionherewill
be to knuckle down and do the homework to convert the known good Ford data into GM units so that the PCM can
accuratelydelivertheproperfuelmass.
FuelMass(g)

30#SkinnyFRPPInjectorsat58psi

RememberthatanytimeyouadjusttheMAForVEtablesbasedonwidebanderror,yourmakingtheassumptionthat
fueldeliverycommandedbythePCMcloselymatcheswhatsreallygoingintotheengine.Whatgoodisallofthework
toperfectaVEtableorMAFtransferfunctionifthefuelinjectordeliveryiswrong?Evenworse,whatifwemakethe
mistakeofassumingthattheinjectorsbehaveonewaywhentheyrereallydoingsomethingelse?Theresultisafalse
confidenceinwhateverairflowcorrectionsweremadeduringthecalibrationprocess.Thisbadairflowestimatemaybe
cascaded into other tables such as idle control or spark advance since theyre ultimately dependent upon airflow
calculationsaswell.
Sowhathavewelearnedfromthisexercise?

Injectorshavenonlinearbehavioratshortpulses;thisrequiressomeadditionaldatatoproperlytunethePCM.
Thedatathatdescribesthenonlinearbehaviorisnotthesameforsimilarflowrateinjectors.
Thedatathatdescribesthenonlinearbehaviorisnotthesameforsimilarbodysizeinjectors.
Errorsintheinjectorcharacterizationwillshowupinthecalculatedairflowtablesandcreateheadacheslater.

Thoselookingtolearnmoreaboutproperinjector,MAFandVEcalibrationcanattendoneofmyAdvancedGMorFord
trainingclasseswithCalibratedSuccess,Inc.ThesecoursestakeacloselookathoweachPCMcomesupwithitsdesired
injection times based on airmass estimates and how we can properly tune these values just like an OEM calibration
engineertogetstocklikeperformanceevenwithasupercharger,largeinjectorsorabigcamshaft.~Greg
CalibratedSuccess,Inc.2009

calibratedsuccess.com