Calculating available injector PW.

[Six_Shooter]

I've been try to get rid of some detonation that is occurring once I'm in boost, usually over 7 PSIG, and I'm trying to determine if I may be out of injector.

Doing some looking around on TGO, the common table is this:

RPM......PW ms
1,000....60 ms
2,000....30 ms
3,000....20 ms
4,000....15 ms
5,000....12 ms
6,000....10 ms

I can't seem to get the same results with the basic formulas that are presented, being (RPM/2)/60 I think there is another part that deals with how many injection events per revolution, but it hasn't been made clear in what I've looked up.

FWIW, according to the reported PW in the datastream, it seems that I'm getting close to or out of injector at about 5500 RPM, and just close above and below that.

[RobertISaar]

well, at 6000RPM, you have 10mS per revolution. if you run the injectors as double-fire, then beyond 10mS at 6000 is doing nothing. single-fire is 20mS.

engine revolution time in mS = (60 / RPM) X 1000.

1 injection event per 2 revs(single fire/SFI) maximum injector ontime = ((60 / RPM) X 1000) X 2
2 injection events per 2 revs(double fire) maximum injector ontime = (60 / RPM) X 1000
3 injection events per 2 revs(???? fire) maximum injector ontime = (60 / RPM) X 1000 X (2/3)
4 injection events per 2 revs(???? fire) maximum injector ontime = (60 / RPM) X 1000 X (1/2)

you should be able to make an injector duty cycle value as an item using BPW and RPM assuming your calibration doesn't switch between firing modes.

[EagleMark]

http://www.thirdgen.org/techboard/di...pping-why.html

HTH!

EDIT: How come your numbers are double what I have been useing?

http://www.thirdgen.org/techboard/po...07-post68.html

[RobertISaar]

not sure what kind of application BMmonteSS calculated those for, but 6000RPM for example:

60 sec / 6000 RPM = .01 seconds per revolution
.01 sec X 1000 = 10 mSec

since we're dealing with 4 cycle engines, the engine revolves twice for a full cycle, so at 6000RPM, there are 20mSec in which to inject fuel before the injector is open at all times.

[lightning-jr]

http://www.thirdgen.org/techboard/di...pping-why.html

HTH!

EDIT: How come your numbers are double what I have been useing?

http://www.thirdgen.org/techboard/po...07-post68.html

I made one for TBI 35-6800, tis in word. If those numbers are 100% that might be your deal 'cause I show 6ms @ 5k 100%
My chart also shows 80 & 85%.

[EagleMark]

not sure what kind of application BMmonteSS calculated those for, but 6000RPM for example:

60 sec / 6000 RPM = .01 seconds per revolution
.01 sec X 1000 = 10 mSec

But it takes 2 revolutions (720 degrees not 360) to make a full cycle. So divide 10 mSec by 2 = 5 mSec

If you have BPW in data stream the conversion is (X * 0.01526 + 0.396) with .396 being the injector bias in the bin. Then make a new value in ADX for Injector Duty Cycle with linked output from X = BPW and Y = Engine RPM
( X * Y / 60000 * 100 )

This works on V8 TBI, not sure if changes are needed for inline 6 and MPFI?

[RobertISaar]

But it takes 2 revolutions (720 degrees not 360) to make a full cycle. So divide 10 mSec by 2 = 5 mSec

why divide by 2? the engine rotates twice per 4 stroke cycle, so 10mSec for rev 1 and another 10mSec for rev 2 for a total of 20mSec in which to inject for the next intake event at 6000RPM.

[EagleMark]

I don't know? But for some reason those numbers posted by Six and calculated by yourself are twice what everyone is using? Now this may have to do with all the information I linked to and have been using is for V8 TBI? So why the difference? Would have to be the TBI injectors need to pulse more, 2 injectors pulsed 4 times each for 8 cylinders per cycle. So batch fire would be pulsed twice. Which would make sense as to 10 mSec over 5 mSec. ?

[RobertISaar]

that would be the difference.

at 6000RPM, a set of injectors that are all fired once per revolution have 10mS per revolution(injection) before they're 100%(20mS total due to 2 revolutions happening for a 4 stroke cycle). this matches pretty much all MPFI/TPI setups.

at 6000RPM, with a firing pattern of 2 pulses per revolution(4 per 4 stroke cycle), you still have 20mS in which to inject fuel for any given cylinder, but since it's cut up into 4 bursts, it's 20mS/4=5mS. having 2(or more) injectors that can fire independant of each other still has the same time constraint, just considerably more flow capability and the ability to inject from more than just a single point in the intake tract for more even fuel distribution. this matches the V8 TBI setups perfectly. rotate the engine twice and 8 injections happen, 4 from each injector.

[1project2many]

Six, I think you're asking about "duty cycle." For batch fire ecms in synchronous mode the number of times the injectors will turn on and off in two crank revolutions is equal to the number of cylinders. If you tell the ecm to keep the injectors "on" for so much time that the "off" events never happen, your injectors are going static. To prevent this you calculate the amount of time between TDC compression stroke on two consecutive cylinders then add in a safety margin. What most people seem to calculate is the amount of time between TDC compression of the same cylinder which is different by a long shot. This formula gives maximum available on time for an injection event in a batch fired six cylinder: 3 injection events per rev maximum injector ontime = (60 / RPM) X 1000 X (2/3). Safety margin is usually 80%-85% so multiply the previous result by .8 for max safe on time.

[RobertISaar]

sure about the "3 injection events per 2 revs"?

i can fire up my bench to be 100% certain, but i always see the reference pulse circuit light up 3 times for every 1 injector pulse(simulating a MPFI 60V6). i get 3 pulses per revolution, therefore, 1 injection per revolution.

be nice if my camera could capture a high enough framerate to see it at anything other than VERY low speeds.

[1project2many]

6 injection events per two revs is correct for older ecm's such as the 165 and 7730. The 7749 also fires every pulse on the apps I've worked with but the recent discussion about fuel mode vs memcal configuration makes me cautious about saying it will always fire that way. If you're seeing 1 injection event then you're looking at sequential or a semi-sequential pattern.

[RobertISaar]

6 injection events per two revs is correct for older ecm's such as the 165 and 7730. The 7749 also fires every pulse on the apps I've worked with but the recent discussion about fuel mode vs memcal configuration makes me cautious about saying it will always fire that way. If you're seeing 1 injection event then you're looking at sequential or a semi-sequential pattern.

i'm using a 9396 on my bench(close enough to the 7727), with or without a MEMCAL installed (and DTC41 disabled when it's gone), i always got 3 ref pulses per injector pulse.

anyways, it fires all 6 injectors on one driver, so definitely not anything close to sequential.

i think i'll try and take a vid at the lowest RPM i can bring my signal generator down to. i really need to look into the soundcard based scope for things like this.

[1project2many]

Are you using the LED's to check pulses? Is that with $A1 and a matching memcal?

[RobertISaar]

LEDs with my nAst1 mask, though i'm sure i could load up a stock A1 BIN and do the same. i've used quite a few different MEMCALs, A1, DF, 2E.... all function the same for what i'm doing.