Fueling & Injection - Time for a New Look at Fuel Injectors and Their characterization Data

Let me preface this with a few points. I do not sell fuel injectors, nor do I really care where you bought yours. This report is just the result of a bunch of engineering tests that I recently ran. You can duplicate all of this yourself if you're familiar with SAE J1832 and have a test bench with controls capable of delivering the proper test conditions. The report is just meant to raise the bar for awareness in the performance industry. And finally, no I will not give you my characterization data from any of the injectors tested in this report unless you want to hire me as an engineering consultant. If you're mad because you don't like it, please remember this is just data, math, and science. I just report what I see.

Calibrated Success Fuel Injector Article #2 (http://www.calibratedsuccess.com/Assets/Documentation/WhosGettingDrilled.pdf)

Wow! That was an awesome read. I wish I had half of your knowledge and experience.

Very good read ! I have a question concerning the injector data in the CS dvd. The data in the dvd shows minimum pulse width for SD 80lb injectors as 1.1 ms, however factory data from Siemens VDO (now Continental) shows a minimum pulse width of 1.7 ms. Why the difference ? Thanx for all the info you've posted

That 1.7ms minimum pulsewidth from Conti represents the point at which they think the injector becomes "consistent enough" for reliable use in an emissions compliant, OEM application. Remember that +/-5% guideline I mentioned? It's part of the SAE spec. The injector manufacturer just doesn't want to hear complaints about performance when somebody tries to use the part "outside of the envelope", which includes very short pulses that are very difficult to control consistently. This is also the reason why the car makers don't just slap big injectors on everything. They want the injector to be operating in that linear region as much as possible, especially at idle and cruise where AFR control and emissions are critical.

Since we would be walking the fine line of tickling this pulsewidth at idle on many engines, we move that point down knowing that we may get slightly less consistent behavior but it at least gives us the option to deliver less fuel and not be so rich.

Makes sense, thank you !

What's your opinion of the method to dynamically batch balance the 212lb/hr Bosch injectors used by Injector Dynamics vs a typical "bench matching" of the same 212lb/hr Bosch injector by others?

Generally speaking, dynamic matching should give better results than grouping injectors based upon volume flowed from a 20-30second continuous flow test. We do not run injectors continuously in the car, therefore we should attempt to group them based on how they would actually perform as installed. Grouping them based on total mass delivered at 1.5ms, 2ms, 2.5ms, and 4ms should all give better results in the car than those that were grouped based on continuous flow. It's all in how you do the test since even the ASNU bench can give this kind of data if you take the time to use it properly.

Thanks. I agree and saw the results on a set of ID2000's I tuned in a car that was north of 1100rwhp. The 212lb/hr injectors were more linear than the 120's used earlier...you could even hear the drop in lope since they were closer injector to injector. Having the peaks and valleys in flow below 2ms for all 8 injectors makes a big difference. I was skeptical, but it made me a believer.

Which "brand" of aftermarket injectors typically are the most consistent? THerefore making the tune more efficient/accurate throughout the band? I had no idea drilled were this inconsistent.

Which "brand" of aftermarket injectors typically are the most consistent? THerefore making the tune more efficient/accurate throughout the band? I had no idea drilled were this inconsistent.
Injector Dynamics

Injector Dynamics

Do they deal with high impedance injectors? What injectors would be ideal for HC build on a 346?

Jesus those are expensive. What about a blue collar brand like FAST? Are they fairly consistent?

For your application, just run LS7 or LS9 injectors. Even LS2 would cover you. Anything ID sells would be overkill for a NA 346.

Calibrated Success Fuel Injector Article #2 (http://www.calibratedsuccess.com/Assets/Documentation/WhosGettingDrilled.pdf)

Another great paper, thanks for sharing your expertise.

Can't be bothered to read? Do you like watching TV more? OK...

http://youtu.be/JZmV10pFsCM

This is just a small part of an upcoming video. It will be part of a smaller injector-specific video as well as a longer video version of my basic EFI class. Don't even ask about release dates yet. I'm just getting started.

Nice!

No offense-but was this test "sponsered by" Injector Dynamics?

So the injectors are all made by Bosch and modified by separate companies by machining the ends off using different machines, but in a clean fashion. Now not only is ID performing dynamic testing, but it seems as though FIC is dynamically testing them as well. The data is different due to the different test results on different machines I guess. So now what's the big deal about ID again? Before ID existed there was lots of chatter about injector data being proper, but in the end, I don't know one person that ended up complaining about the way their car performed on the street or track that ended up being an injector issue. Yes there are still people with driveability issues, even with the "proper data" and these great performing injectors. So what now? Good marketing? absolutely!

No offense-but was this test "sponsered by" Injector Dynamics?

As I said in the original post, "I don't sell injectors." I don't have a dog in this fight, I'm just reporting the information that I found through accepted industry standards of testing. (SAE J1832) If it was sponsored, I still haven't seen a check, so I better check with my accountant again. :rolleyes:

Will you feel better if I call out the manufacturer of the BAD units I tested as well? I'm not looking to specifically stir up poop and drama here. I'm just presenting findings based on valid testing. I made the call to give props to the one company that I found to be producing a product that really worked in lab conditions because I think GOOD efforts should be acknowledged and rewarded. Openly dressing down a company selling products that don't meet my standard isn't really going to help the discussion here.

Before ID existed there was lots of chatter about injector data being proper, but in the end, I don't know one person that ended up complaining about the way their car performed on the street or track that ended up being an injector issue. Yes there are still people with driveability issues, even with the "proper data" and these great performing injectors. So what now?
My main thrust in a LOT of my calibration training (Books, Video, Live Classes) has been to try and eliminate as many variables as possible during the testing and calibration process. The more we know about actual fuel delivery, the more precise we can be when it comes time to control air/fuel ratios under all conditions. Any errors in the fuel delivery side (injector characterization, pressure control, fuel composition) get carried over into the calibrations changes that are made to both airflow estimates (VE, MAF) and general AFR control. The closer we get to reality and precision, the better we can expect the car to run under all conditions, just like they do from the factory.

Do you think that GM just guesses at injector data and covers it up with whatever they find necessary later to get "good driveability" and pass federal emissions requirements? Why would anyone NOT want to follow a similar approach to what they're doing at the OEM?

Do you think that GM just guesses at injector data and covers it up with whatever they find necessary later to get "good driveability" and pass federal emissions requirements? Why would anyone NOT want to follow a similar approach to what they're doing at the OEM?

So I guess all the cars tuned without data don't get good driveability and don't pass emissions?

Funny I took a set of 120lb low impedance injectors, plugged them, and got a car to get good fuel economy and pass emissions on a roller with no cats.

Nobody is going to get me to pay $1000 for modified set of $500 dollar injectors and tell me there "better".

There's lots of "sheep" out there-I'm just not one of them.

So I guess all the cars tuned without data don't get good driveability and don't pass emissions?
There is a HUGE difference between passing a local emissions test and being federally compliant as the manufacturer of record. A simple plug in (OBDII) check or even IM240 does not come anywhere near the complexity of the FTP UDDS, HWY, US06, SC03, and Cold CO tests. When I was a calibrator at GM, the goal was to have our tailpipe emissions at 1/2 of the federal standard with 4k mile catalysts for each test. Tier II Bin 5 standards are not always a cakewalk to hit, even with a completely stock new engine. These emissions targets are the reason why it takes an entire team of calibrators about three years to release a new engine and calibration package at the OEM level. Minimizing the variables during the development process is not just a convenience, it's a necessity.

Funny I took a set of 120lb low impedance injectors, plugged them, and got a car to get good fuel economy and pass emissions on a roller with no cats.
I'll be the first to admit that we've all "done what it takes to make it run right" at some point in the performance aftermarket. It never fails to amaze me how dumb I USED TO BE. Learning more about how the control system and physical injectors actually work has been a huge advantage when it comes to making this process easier, faster, and more consistent for me though. Even at my level, I'm never done learning.

Nobody is going to get me to pay $1000 for modified set of $500 dollar injectors and tell me there "better".

There's lots of "sheep" out there-I'm just not one of them.
I think you're still mistaking me for someone who is trying to sell you injectors. I don't care where you bought your injectors. I'm just saying that the process of calibrating the engine is 1000% easier if you actually know what the injector is doing and give the PCM precise control of delivered fuel mass. You are free to continue doing it the hard way. My car runs great.

all the correct data in the world wont help if you cant input the values into the tuning suite cough *HPTuners*...

all the correct data in the world wont help if you cant input the values into the tuning suite cough *HPTuners*...Just to be 100% sure... does this mean I should ditch HPTuners in favor of, say, EFI Live? I'll do it in a heart beat because I've def. had idling issues with larger injectors.

Just to be 100% sure... does this mean I should ditch HPTuners in favor of, say, EFI Live? I'll do it in a heart beat because I've def. had idling issues with larger injectors.

I have no experience with EFI Live, that would be your call


sorry for the hijack Greg

all the correct data in the world wont help if you cant input the values into the tuning suite cough *HPTuners*...

Just to be 100% sure... does this mean I should ditch HPTuners in favor of, say, EFI Live? I'll do it in a heart beat because I've def. had idling issues with larger injectors.

So are you saying you can't input the injector data in HPtuners? :corn:

Just to be 100% sure... does this mean I should ditch HPTuners in favor of, say, EFI Live? I'll do it in a heart beat because I've def. had idling issues with larger injectors.

they have the same tables.... i don't understand this statement.



what I do get is the "delivered fuel mass" that is the trick.

Just to be 100% sure... does this mean I should ditch HPTuners in favor of, say, EFI Live? I'll do it in a heart beat because I've def. had idling issues with larger injectors.

How large is large?

So are you saying you can't input the injector data in HPtuners? :corn:

you can, but the values input are changed. I.e. if you input .0062 into a field you might get .0804 as a final value. Something to do with the hex-dec conversion was the excuse I got. Not isolated to any one table or value field.

you can, but the values input are changed. I.e. if you input .0062 into a field you might get .0804 as a final value. Something to do with the hex-dec conversion was the excuse I got. Not isolated to any one table or value field.

What OS are you working with? I have not noticed this, and I tune a few cars.

What OS are you working with? I have not noticed this, and I tune a few cars.

PC ? XP, also happens on my Windows 7 laptop.

PC ? XP, also happens on my Windows 7 laptop.

No. What PCM OS? What are you tuning that you see that?

12202088

Greg (turbolx), how long does it take to do an injector, meaning full characterization and collection of data?

As long as the injectors are clean and functional, data collection takes about one full day in the lab for me to get what I consider the basics of characterization. This means a complete flow curve at 13.5v for a set of eight injectors at a single pressure and a shorter series of tests at varying voltages in order to find the trendline of offset vs voltage and populate that for the same single pressure.

Adding multiple pressure breakpoints simply adds a complete repeat of the procedure, as does running more than 8 pieces. I typically test 4 injectors at a time to make sure that their flow doesn't outstrip pump capacity or rail distribution, so testing 8pcs means two sets of four. If we are trying to get data for a part number rather than just a single set of flow matched injectors, then we really should test more than 8pcs. There can be build tolerances of +/-6% for production injectors, more for aftermarket. Finding the real "average" behavior requires that we test more samples in order to make sure we get a statistically significant sample.

The going rate for engineering consultation is north of $100/hr (just like a good lawyer, accountant, doctor, etc...) so spending a full day or three on this quickly eclipses the typical cost most associate with "engine tuning." I'm not recommending that everyone has every single injector measured, as this would be prohibitively expensive. But it is certainly nice to be able to have the confidence to say that you really know exactly how much fuel is being delivered on each shot. To some race teams or injector manufacturers, this cost is literally a drop in the bucket compared to the rest of the operation. So for those guys, we try to make sure we have as much GOOD data as possible.

I'll agree with Arun on alot he said.

From what I see and have seen- best data- pay someone to flow every injector you have- huge bill, can be worth it on a big $$ build. I wouldn't trust -it should flow this and here's the data from a set we did awhile ago.

Are my customers going to buy them? NO

The injector data is a big thing now becaue I think the new computers and about 95% of the tuners out there have no clue how to tune them in correctly with these VVE table's. So plugging in data that is close is getting them by.

I'd be willing to say that pretty much no one here cares about emmision output since this board is about performance.

For now I'll stay with the injectors we use. When/if tuning them in becomes a problem I'll look into more expensive injectors.

As long as the injectors are clean and functional, data collection takes about one full day in the lab for me to get what I consider the basics of characterization. This means a complete flow curve at 13.5v for a set of eight injectors at a single pressure and a shorter series of tests at varying voltages in order to find the trendline of offset vs voltage and populate that for the same single pressure.

Adding multiple pressure breakpoints simply adds a complete repeat of the procedure, as does running more than 8 pieces. I typically test 4 injectors at a time to make sure that their flow doesn't outstrip pump capacity or rail distribution, so testing 8pcs means two sets of four. If we are trying to get data for a part number rather than just a single set of flow matched injectors, then we really should test more than 8pcs. There can be build tolerances of +/-6% for production injectors, more for aftermarket. Finding the real "average" behavior requires that we test more samples in order to make sure we get a statistically significant sample.

The going rate for engineering consultation is north of $100/hr (just like a good lawyer, accountant, doctor, etc...) so spending a full day or three on this quickly eclipses the typical cost most associate with "engine tuning." I'm not recommending that everyone has every single injector measured, as this would be prohibitively expensive. But it is certainly nice to be able to have the confidence to say that you really know exactly how much fuel is being delivered on each shot. To some race teams or injector manufacturers, this cost is literally a drop in the bucket compared to the rest of the operation. So for those guys, we try to make sure we have as much GOOD data as possible.

This is amazing to me. It could take days to get the data on 1 set of injectors? Makes me wonder how companies that say they flow and do this characterization to every injector do it, imo I bet they don't, until now I never had an idea how long it would take. How would we know if they did or they did not, lol. That's why I didn't go for spending 1200 for a set of injectors when I got the same injector for 500..

The injector data is a big thing now becaue I think the new computers and about 95% of the tuners out there have no clue how to tune them in correctly with these VVE table's.


Anybody trying to calibrate injectors via the Volumetric Efficiency table (whether virtual or conventional) has no business tuning a lawnmower, let alone a closed-loop automotive engine.

Oh, I have no idea if HPTuners does or doesn't ... I was only asking for clarity. I haven't tried messing with fueling on HPTuners since I got that new cable upgrade, which is probably 5+ years old now. They were 42 lb injectors and simply scaling the table didn't yield the best idling characteristics. So back to stock I went.

Anybody trying to calibrate injectors via the Volumetric Efficiency table (whether virtual or conventional) has no business tuning a lawnmower, let alone a closed-loop automotive engine.

Guess I'll go tune some lawn mowers then.

You don't tune in the whole injector off VVE ass. Most tuners will throw in a injector table and call the injectors bad because they didn't tune in VVE. Get it yet? LOL

Guess I'll go tune some lawn mowers then.


I couldn't tune a lawnmower to save my life. :)

Guess I'll go tune some lawn mowers then.

You don't tune in the whole injector off VVE ass. Most tuners will throw in a injector table and call the injectors bad because they didn't tune in VVE. Get it yet? LOL

Your point on the error that some "tuners" commit is a good one...they just bump the Injector Flow Rate table and then wonder why the things aren't behaving properly.

However, at some point in your education (and we're all learning) you'll discover that injector flow characteristics have nothing to do with an engine's volumetric efficiency as configured in a powertrain control system.

Until then, stay away from my lawnmower. :)

VE tweaking when only the injectors have been changed is a band-aid, and a poor one at that. Fooling the PCM into flowing more or less fuel at a given VE coordinate may get things working reasonably well but it's a bad approach.

The real mojo that amateur tuners miss is in the Injector Offset table as well as an accurate (not algebraic based on the old injector flow rates and percentage increase for the new injectors) Injector Flow Rate table. Short Pulse Adder is also a game-changer when running injectors that have a 2-slope flow rate...one for pulse widths > their defined flow rate breakpoint and a different slope for pulse widths < their breakpoint.

You seem to know enough about engine controls that a read of Greg Banish's (AKA turbolx) other published works on engine control systems would be a big help...you're probably just a short jump away from fully understanding how the PCM manages fuel delivery.

Everything I just regurgitated above is from Banish's other article linked below, along with my own experience tuning a few injector swaps after having read his two books and his DVD: (yes, I'm a fan...they cost $$ but not as much as paying a professional tuner to do it right)

http://www.calibratedsuccess.com/Assets/Documentation/Fuel%20Injector%20Article.pdf

Your point on the error that some "tuners" commit is a good one...they just bump the Injector Flow Rate table and then wonder why the things aren't behaving properly.

However, at some point in your education (and we're all learning) you'll discover that injector flow characteristics have nothing to do with an engine's volumetric efficiency as configured in a powertrain control system.

Until then, stay away from my lawnmower. :)

VE tweaking when only the injectors have been changed is a band-aid, and a poor one at that. Fooling the PCM into flowing more or less fuel at a given VE coordinate may get things working reasonably well but it's a bad approach.

The real mojo that amateur tuners miss is in the Injector Offset table as well as an accurate (not algebraic based on the old injector flow rates and percentage increase for the new injectors) Injector Flow Rate table. Short Pulse Adder is also a game-changer when running injectors that have a 2-slope flow rate...one for pulse widths > their defined flow rate breakpoint and a different slope for pulse widths < their breakpoint.

You seem to know enough about engine controls that a read of Greg Banish's (AKA turbolx) other published works on engine control systems would be a big help...you're probably just a short jump away from fully understanding how the PCM manages fuel delivery.

Everything I just regurgitated above is from Banish's other article linked below, along with my own experience tuning a few injector swaps after having read his two books and his DVD: (yes, I'm a fan...they cost $$ but not as much as paying a professional tuner to do it right)

http://www.calibratedsuccess.com/Assets/Documentation/Fuel%20Injector%20Article.pdf

Dude, do you even know who Don Kinder(Slowhawk)is? He's probably one of the first guys to dial in low impedance injectors the hard way-and of course he attacked flow rate, offset and adders and got 95lb & 120lb injectors to idle like a kitten. Way before the days Greg started to publish his DVDs-all of which I have bought and watched.

Have you ever plugged in injector data on BS3 or Haltech or FAST?

Dude, do you even know who Don Kinder(Slowhawk)is? He's probably one of the first guys to dial in low impedance injectors the hard way-and of course he attacked flow rate, offset and adders and got 95lb & 120lb injectors to idle like a kitten. Way before the days Greg started to publish his DVDs-all of which I have bought and watched.

Have you ever plugged in injector data on BS3 or Haltech or FAST?

I don't know who he is, but advocating use of VE table variables to compensate for fuel injector swaps seems like very weak sauce. I can see where it's possible using WB-02 to watch the exhaust stream and nailing the PCM into open loop during tuning, but still a band-aid fix versus correct injector characterization.

Maybe I misunderstood what he was saying? Because anybody that thinks injectors affect an engine's volumetric efficiency clearly does not understand the calculus...it's all based on airflow and cylinder filling.

Greg B's documented experience developing injector calibration data for OEMs speaks well to his capability and understanding of powertrain controls. I have no reason to doubt the accuracy of his information.

To answer your question, no experience on any of those 3 systems. I learned what I know about fuel delivery and engine setup during high school and college working on an offshore powerboat racing program and a Late Model asphalt racing team, as well as OMC outboard motor school. Carburetors. If you understand carb principles, fuel injector behavior makes perfect sense, no?

Come on eb02z06. LOL

I like getting schooled by new tuners..LOL

Some of these new tuners need to start like we did. LS1 edit and a shitty scanner. No schools, no books, no video's out to teach you anything.

I fully understand injectors,tuning. Scaling,pulsewidths,timing ect ,ect. You do not have to get technical with me.

I like getting schooled by new tuners..LOL

Some of these new tuners need to start like we did. LS1 edit and a shitty scanner. No schools, no books, no video's out to teach you anything.

I fully understand injectors,tuning. Scaling,pulsewidths,timing ect ,ect. You do not have to get technical with me.

Define "new."

LS1Edit, AutoTap, and MS Excel here, since 2003. Before that was crappy DOS-based software and Tuned Port Injection chip burning since 1997. Before that was metering jets, power valves, and accelerator pump cams since 1988. And not a single blown or damaged engine.

Set up a pair of V8 engines to run from Miami to Bimini and back at WOT, lifting throttle only for wave jumps. It's a little more involved than a few seconds at a time on a dragstrip.

It amazes me that people need things handed to them on a silver platter and can't figure this stuff out. Offsets and low pulse width settings are things that tuners have had to back calculate as unknown variables way before injector companies starting giving it out and it's really not that difficult on known fully tuned configurations with the only thing being an injector change. Now I know I'm not selling DVDs or injectors priced at twice the price but this is pretty simple shit for any professional tuner worth their salt who has put in the research and done their homework.

Before everybody jumps my shit I guess I should add that yes I understand it's nice for new people who are learning to tune to have some instruction and it's nice to have data for injectors as it saves time but let's not pretend you can't tune a car and not get damn close on those settings by doing your own research and testing... and when I say close I mean within a couple percent fuel trims of where they should be.

...you'll discover that injector flow characteristics have nothing to do with an engine's volumetric efficiency as configured in a powertrain control system.

This alone is a very good point that perhaps I have not made well enough myself.

We often have TWO complex models (each with varying degrees of freedom) that require adjustment. Even worse, the measurements for each is partially dependent upon the other. All we really want is the correct AFR delivery (all of the time!), but this changes with either airflow estimate or fuel flow characterization.

Ideally, we would only change one of these at a time. It's always easier to take a car with stock injectors and tune for a new camshaft or light-pressure supercharger as long as the OE injectors can keep up. This lets us concentrate on the VE and MAF tables without much concern for fuel delivery. Alternatively, we can change injectors on a stock airpath engine (or one that has previously been correctly mapped for VE and MAF) and assume that the errors must be in the fuel delivery side.

The rub comes when a customer has us tune something where both airflow and injectors have been physically changed. It's possible to come up with multiple sets of solutions that combine some change to injector data and some other change to airflow modeling that happen to work out today. Only one of these sets of solutions however will continue to work as the conditions change for temp, baro, fuel quality, etc. This is why starting with some good reference point goes so far toward minimizing the guesswork for the calibrator. If we know that the injector data is right, we can concentrate on airflow modeling.

Conversely, if you're lucky enough to have a flow bench sufficient to measure the entire range of the MAF (and the entire inlet plumbing path!) then you could generate "THE" MAF curve and never touch it again like GM does.

Have you ever plugged in injector data on BS3 or Haltech or FAST?

Fundamentally, these aren't that different. They're just simpler in that they make no corrections for short pulse phenomena. Any short pulse errors get "baked in" to the base VE table. Luckily that resulting skewing of the VE and airflow estimate don't mess up closed loop idle airflow corrections or knock routines, or torque models like they can in newer OEM ECUs.

Some of these new tuners need to start like we did. LS1 edit and a shitty scanner. No schools, no books, no video's out to teach you anything.

Ack! Those were the days. I don't miss that. :barf: I still remember some of my earliest discussions with OEM calibrators and how violently the light bulb lit up over my head when I realized how these systems were really working versus how the mIRC channel "experts" were claiming they worked. I guess we all have to learn the hard way a bit at the beginning.

Another note on the aftermarket controllers is they have much better injector circuits than an OEM GM ECM/PCM. IMO that alone changes the short pulse and voltage offsets because they normally supply much more current to the coil to get the injector open.

That is another variable that test equipment can't even take into consideration unless you physically use the PCM or ECM as the injector controller as the test means.

Another note on the aftermarket controllers is they have much better injector circuits than an OEM GM ECM/PCM. IMO that alone changes the short pulse and voltage offsets because they normally supply much more current to the coil to get the injector open.
I think you may be confusing the difference between saturated vs peak and hold driver circuits rather than OE vs aftermarket. There have been plenty of OE systems that use peak and hold drivers (almost every direct injection system does this now).

The current supplied to the coil in a saturated driver is largely just a function of V=IR, meaning that for a constant supply voltage and fixed impedance injector coil (which only changes a small amount with temperature) the current is also constant during the length of the "on time". You can see this if you put a scope on the wires. There is a little bit of change in flyback voltage between drivers since they may have varying zener clip diodes in the circuit to limit field collapse in the coil, but mostly affects a very short closing time anyway. Opening time is almost exclusively a function of supply voltage and injector design. Short pulse behavior is part opening time and part dynamic flow through the orifice plate (which is a HUGE part of injector tip design).

Short pulse behavior is part opening time and part dynamic flow through the orifice plate (which is a HUGE part of injector tip design).

Nice info.

I think you may be confusing the difference between saturated vs peak and hold driver circuits rather than OE vs aftermarket. There have been plenty of OE systems that use peak and hold drivers (almost every direct injection system does this now).

The current supplied to the coil in a saturated driver is largely just a function of V=IR, meaning that for a constant supply voltage and fixed impedance injector coil (which only changes a small amount with temperature) the current is also constant during the length of the "on time". You can see this if you put a scope on the wires. There is a little bit of change in flyback voltage between drivers since they may have varying zener clip diodes in the circuit to limit field collapse in the coil, but mostly affects a very short closing time anyway. Opening time is almost exclusively a function of supply voltage and injector design. Short pulse behavior is part opening time and part dynamic flow through the orifice plate (which is a HUGE part of injector tip design).
Interesting! So the grounding circuit within the PCM that has to carry that current wouldn't cause any variation in how the injector opens and how quickly it opens?

It is just very odd that high impedence saturated injectors seem to have much better control over lower pulsewidths than a factory PCM does. I had alway's assumed it was from the beefier drivers within aftermarket ECU's.