Fuel Injectors Flowrates/Impedances

[markocosic]

2005 Yamaha R6.


First, some data which might be of use to people:

2005 Yamaha R6 fuel injectors, running Shell Optimax fuel at 20 degree C, on the standard fuel rail and fuel pressure differential set at 3 bar across the injectors. They're 250cc/minute.

(actual value 253cc/minute, though given that they were only run for 12 seconds and the results scaled up that'll be within reasonable margins of error)

Consistency across all four injectors? Firing them into identical 100ml measuring cylinders for 12 seconds, one could not tell the difference in levels, which implies between-injector variance of less than 0.5cc in 50cc, or 1% - so no calibration factors required for each cylinder in your engine management system.

Speaking of the electrical side, each injector will draw ~1A at 13.5V if connected directly to the power source. 250mA is plenty to open the injector against fuel pressure, 100mA is plenty to hold it open.

Measuring the injectors on an automatic (MechE proof) doobery widget, you obtain 20 ohms and 32mH at 100Hz and 72.4 ohms and 13.4 mH at 1000Hz (inductor in series with a resistor model) or 40.5 ohms and 70 mH at 100Hz and 170.6 ohms and 23.7 mH at 1000Hz (inductor in parallel with a resistor model).


Next, some questions:

Injector opening times - how long it takes to open the injector and develop full flow. The engine management system reckons that most are in the range 0.9ms to 1.3ms, typically 1.2ms. Have you measured yours, if so how did you do this, and how critical did it turn out to be once tuning? (presumably it's only an offset that can bite at low load/rpm)

I'm thinking of driving them using a square-wave source with 0.8, 0.9, 1.0, 1.1, 1.2, 1.3, 1.4ms pulsewidths and measuring the fuel flowed in a given time relative to 100% duty flow and trying to develop an opening-time versus fuel flowed plot in teh area of interst - but is this even necessary?

[markocosic]

2005 Yamaha R6.


First, some data which might be of use to people:

2005 Yamaha R6 fuel injectors, running Shell Optimax fuel at 20 degree C, on the standard fuel rail and fuel pressure differential set at 3 bar across the injectors. They're 250cc/minute.

(actual value 253cc/minute, though given that they were only run for 12 seconds and the results scaled up that'll be within reasonable margins of error)

Consistency across all four injectors? Firing them into identical 100ml measuring cylinders for 12 seconds, one could not tell the difference in levels, which implies between-injector variance of less than 0.5cc in 50cc, or 1% - so no calibration factors required for each cylinder in your engine management system.

Speaking of the electrical side, each injector will draw ~1A at 13.5V if connected directly to the power source. 250mA is plenty to open the injector against fuel pressure, 100mA is plenty to hold it open.

Measuring the injectors on an automatic (MechE proof) doobery widget, you obtain 20 ohms and 32mH at 100Hz and 72.4 ohms and 13.4 mH at 1000Hz (inductor in series with a resistor model) or 40.5 ohms and 70 mH at 100Hz and 170.6 ohms and 23.7 mH at 1000Hz (inductor in parallel with a resistor model).


Next, some questions:

Injector opening times - how long it takes to open the injector and develop full flow. The engine management system reckons that most are in the range 0.9ms to 1.3ms, typically 1.2ms. Have you measured yours, if so how did you do this, and how critical did it turn out to be once tuning? (presumably it's only an offset that can bite at low load/rpm)

I'm thinking of driving them using a square-wave source with 0.8, 0.9, 1.0, 1.1, 1.2, 1.3, 1.4ms pulsewidths and measuring the fuel flowed in a given time relative to 100% duty flow and trying to develop an opening-time versus fuel flowed plot in teh area of interst - but is this even necessary?

[ad]

<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">Originally posted by markocosic:
2005 Yamaha R6.


First, some data which might be of use to people:

2005 Yamaha R6 fuel injectors, running Shell Optimax fuel at 20 degree C, on the standard fuel rail and fuel pressure differential set at 3 bar across the injectors. They're 250cc/minute.

(actual value 253cc/minute, though given that they were only run for 12 seconds and the results scaled up that'll be within reasonable margins of error)

Consistency across all four injectors? Firing them into identical 100ml measuring cylinders for 12 seconds, one could not tell the difference in levels, which implies between-injector variance of less than 0.5cc in 50cc, or 1% - so no calibration factors required for each cylinder in your engine management system.

Speaking of the electrical side, each injector will draw ~1A at 13.5V if connected directly to the power source. 250mA is plenty to open the injector against fuel pressure, 100mA is plenty to hold it open.

Measuring the injectors on an automatic (MechE proof) doobery widget, you obtain 20 ohms and 32mH at 100Hz and 72.4 ohms and 13.4 mH at 1000Hz (inductor in series with a resistor model) or 40.5 ohms and 70 mH at 100Hz and 170.6 ohms and 23.7 mH at 1000Hz (inductor in parallel with a resistor model).


Next, some questions:

Injector opening times - how long it takes to open the injector and develop full flow. The engine management system reckons that most are in the range 0.9ms to 1.3ms, typically 1.2ms. Have you measured yours, if so how did you do this, and how critical did it turn out to be once tuning? (presumably it's only an offset that can bite at low load/rpm)

I'm thinking of driving them using a square-wave source with 0.8, 0.9, 1.0, 1.1, 1.2, 1.3, 1.4ms pulsewidths and measuring the fuel flowed in a given time relative to 100% duty flow and trying to develop an opening-time versus fuel flowed plot in teh area of interst - but is this even necessary? </div></BLOCKQUOTE>

Marko,

Great to see you are actually supplying data unlike some of the other members of the forum who wish to have their hands held.

From memory, if this is even of use to you, Carly and the guys from Oakland University measured F4i injectors. There is a paper lying around somewhere.

[markocosic]

Cheers Andrew, I've PMed Carly and we'll see if she's still about! http://fsae.com/groupee_common/emoticons/icon_smile.gif Can't imagine generic high-impedance japanese bike injectors having radically different opening times - probably made by the same manufacturers anyway...

Another scrap of paper here has the DC properties on (12.5 ohms resistance) FWIW.

[murpia]

<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">Originally posted by markocosic:
Next, some questions:

Injector opening times - how long it takes to open the injector and develop full flow. The engine management system reckons that most are in the range 0.9ms to 1.3ms, typically 1.2ms. Have you measured yours, if so how did you do this, and how critical did it turn out to be once tuning? (presumably it's only an offset that can bite at low load/rpm)

I'm thinking of driving them using a square-wave source with 0.8, 0.9, 1.0, 1.1, 1.2, 1.3, 1.4ms pulsewidths and measuring the fuel flowed in a given time relative to 100% duty flow and trying to develop an opening-time versus fuel flowed plot in teh area of interst - but is this even necessary? </div></BLOCKQUOTE>

Check out this thread (http://fsae.com/eve/forums/a/tpc/f/125607348/m/67310241631?r=46710551631) as you might want to factor battery voltage into your experiment as well. And well done for stepping up and doing the work yourself.
Regards, Ian

[markocosic]

<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">doing the work yourself </div></BLOCKQUOTE>

Steady on - I'm still holding out hope that the injector opening time or battery voltage compensation will either be irrelevant or get done by somebody else as even with a taguchi setup it'll be the set variables, squirt, reset variables, repeat ad-infinitum sort of work that drives me up the wall... http://fsae.com/groupee_common/emoticons/icon_wink.gif

A datalog of battery voltage vs time and ancilliries powered would be useful to see if we need bother, but there's 'owt about on here. http://fsae.com/groupee_common/emoticons/icon_frown.gif

This thread mentions one, but it's no longer there it would seem:
http://fsae.com/eve/forums/a/tpc/f/125607348/m/38910068...10910431#11210910431 (http://fsae.com/eve/forums/a/tpc/f/125607348/m/38910068331?r=11210910431#11210910431)
http://www.student.curtin.edu.au/~12419520/yessssShifter.jpg (http://www.student.curtin.edu.au/%7E12419520/yessssShifter.jpg)

[SamB]

Does anyone know the injector firing angles for the standard 2004 injectors?

Cheers guys

[Grant Mahler]

<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">Originally posted by markocosic:
2005 Yamaha R6.

Injector opening times - how long it takes to open the injector and develop full flow. The engine management system reckons that most are in the range 0.9ms to 1.3ms, typically 1.2ms. Have you measured yours, if so how did you do this, and how critical did it turn out to be once tuning? (presumably it's only an offset that can bite at low load/rpm)
</div></BLOCKQUOTE>

I would hope Marco found a way to answer this, but for the sake of posterity, I would like to answer this question based on my (limited) experience.

If you watch the injector current draw, it should look roughly like a fixed-slope line (linear relationship) to a peak, a drop, and then a constant current while open - i.e. peak and hold. If you look specifically at the linear section at the beginning, it should have a kink - the first kink is when the injector is fully open.

I would imagine testing the fuel flow directly to be mildly difficult - the total volume of fuel during a single cycle is small. And if you were to look at multiple openings, they may even out nicely but be crappy individual cycles.