Questions Answered

We get a fair amount of email, and time permitting we make every effort to answer questions. This page will show some of the more helpful email question and answer sessions. In all cases, we will remove customers names, and in some cases we may re-word our original response because we can always think of a better description once we look it over again.

We hope you find this helpful.

1. Is a High Impedance Injector With No Resistor Better Than a Low Impedance Injector With Resistors?

This willl be kind of long.

The latency, or dead time of the injector describes an amount of time that the injector is effectively not flowing fuel when it is energized with a pulsewidth. Note that the mechanism that creates this lag or dead time is not so simply explained or measured but for the sake of understanding the end result, this explanation works fine.

It is important to understand that dead time itself is not an indication of quality, or even the ability of the injector to meter fuel accurately at low pulsewidths. It is simply an indication of dynamic flow vs pulsewidth.

Note - It is even possible for an injector to have a negative dead time value! That should keep you scratching you head for a while...

For reference, our EV-14 high impedance injectors have a dead time of about 1000 microseconds (1 millisecond) at 14 volts, and 3 bar pressure. The Rochester and SMP are considerably lower. From memory, they are probably in the .75msec range, although I don't know what it is with a resistor added.

There is a disadvantage to high dead times that is hidden in the background. That is the overall reduction in dynamic flow which will force you to run a higher duty cycle for a given dynamic flow rate as compared to another injector with the same flow rate but a lower dead time. This disadvantage can be offset completely if the recovery time of the injector is low

As you may have guessed, there is more to this, and overall dynamic flow which relates to horsepower potential cannot be completely described by flow rate and dead time. We also have to consider the maximum duty cycle that we can run while staying within the injector's linear operating range. This maximum duty cycle is determined by the recovery time of the injector which can range from 1.75msec (Bosch 846's) to as little as .5msec.

Consider this example. 2 injectors that both flow 1000cc/min. One has a dead time of 1msec and a recovery time of 1msec. The other injector has a dead time of .75msec, and a recovery time of 1.25msec.

When we calculate maximum dynamic flow to determine horsepower potential, we have to give injector #1 1msec of recovery time and it is effectively not flowing for another 1msec, so for any given rpm our total time available to deliver fuel will be cycle time minus 2 msec. (Cycle time = 1 divided by injector firings per second)

For injector #2 we need to give it 1.25msec of recovery time and it is effectively not flowing for .75msec, so to calculate total dynamic flow we again need to subtract 2 msec from our total cycle time.

In the end, both injectors have the same dynamic flow potential.

Now on to your actual question, the resistors. The problem with resistors is not so much that they increase dead time, the problem is that they increase flow deviation between injectors. If you look at a plot of flow deviation across a set of injectors you will find that injectors that match nicely at 14 volts will probably be a mess at 8 volts. Especially at low pulsewidths.

Note - You will never see this with a static flow test.

As the voltage decreases, the flow variation between injectors increases because they were designed to run at approx 14 volts. Since the resistor creates a voltage divider in the circuit, the injector effectively sees less voltage, and so now you have an injector designed to run at 14V that may only be running at 8V, or 10V, and you no longer have a nicely matched set. Some injectors are more sensitive to this than others, but in the end you are adding a component to the circuit that was not intended when the injector was designed.

It looks like you are trying to choose between 3 injectors. My 1000cc, the SMP 1000cc, and the Rochester 1000cc (Roughly 1000cc for all of them)

First off, I would remove the Rochester from the list. It is the most non linear injector that I have tested.

The SMP injector has decent linearity, good low pulsewidth extension, but no atomization. No as in none. It’s just three streams of fuel.

My 1000cc injector has better linearity, slightly better low pulsewidth extension, and vastly superior atomization.

The atomization quality alone is worth choosing it, but there is more. The excellent idle quality and drivability that people have been talking about on the forums is the result of proper flow matching across the pulsewidth range.

We test in batches of 100 to 200 injectors at a time. From that batch, we typically see a total of 4.5% spread in flow. From there we separate them into three groups with a flow spread of about 1.5% within each group. Then we match them within each flow group based on dead time. This insures minimum flow deviation not just at high pulsewidths, but at low pulsewidths as well.

This is critically important because a difference in dead time of even 100 microseconds (One tenth of a millisecond) can add up to a large flow deviation at low pulsewidths. For example. Injector # 1 has a dead time of 1msec, injector # 2 has a dead time of 1.1msec (100 microsecond difference) At a pulsewidth of 2 milliseconds, injector # 1 will be effectively flowing 1msec worth of fuel while injector # 2 will be effectively flowing .9 msec of fuel, for a difference of 11.1 percent!!!

And all of this is happening with 2 injectors that have the same flow rate!

PY

2. Why Are Your Injectors Rated at 1000cc/min When They Only Flow 880cc/min on My Flowbench?

The question was, "why are your injectors rated at 1000cc per minute when they test at 880cc on my bench?" The short answer is that your flowbench uses mineral sprits, or some other "calibrated" test fluid while my flowbench uses gasoline which has approximately half the kinematic viscosity of mineral spirits.

It was proposed that we were being misleading because we were not testing according to "industry standard."

I'm not sure if the question was posed to stir up trouble or simply to get an answer, but it doesn't really matter because it is a valid question.

And here comes a valid answer.

The technology that we call "Injector Dynamics" was developed not from a desire to be part of "the industry" but from a need for accurate injector characterization because it was completely lacking in the industry.

To be more specific, we felt that the fuel injector was the least understood part of a fuel injection system and so we set out to characterize them in much the same way that you would run an engine on the dyno while measuring torque, horsepower, temperatures, pressures, lambda, etc.

We were not looking for a simple test designed to show relative flow differences between injectors, we were looking for a test that would exactly relate dynamic injector characteristics to a running engine.

During the development phase, we worked closely with Motec USA's engineering department, and they offered direction by specifying how the data should be presented. Left to my own devices, the end result would have been streams of engineering data which the end user would then have to sort through.

At this point I should thank Simon Wagner, Head of Engineering at Motec USA for slapping me in the face with a dose of practicality so that my work would be well received and genuinely helpful to his dealers.

Motec requested that the data be interpreted in a manner that would allow their dealers to quickly look at the data, and determine the following; Dead Time Compensation Values, Maximum Horsepower Potential and Useable Operating Range.

For any of these values to be relevant, the test fluid must be the same, or at least very close to the fluid that will actually be powering the engine. The most obvious example is horsepower potential.

The horsepower potential of an injector is determined by considering its dynamic flow rate, and the Brake Specific Fuel Consumption of the engine. Brake Specific Fuel Consumption is a measure of how efficiently an engine turns fuel into horsepower, and is stated as pounds per hour per horsepower. Put more simply, pounds of fuel required per horsepower.

For example, if your engine makes 1000 horsepower, and uses 500 lbs of fuel per hour, it has a brake specific of .5 (500lbs per hour fuel flow divided by 1000 horsepower equals .5)

Using the above example, let's calculate the power potential of an injector using both gasoline and mineral spirits as a test fluid. If we convert the flow numbers from cc/min to lbs per hour we get 95.24 lbs per hour for gasoline, and 83.8 pounds per hour for mineral spirits.

If we divide our brake specific of .5 into our gasoline flow rate of 95.24 we get 190.48 horsepower per injector. Performing the same calculations with mineral spirits we get 167.6 horsepower per injector for a difference of 13.6 percent!

I don't know about you, but in my world an error of 13.6 percent is equivalent to a solid kick in the nads.

To sum all of this up, if the purpose of flow testing is to show relative static flow differences between injectors, mineral spirits, olive oil, or even KY Jelly will get the job done. On the other hand, if the purpose is to relate the data to actual engine performance, the intended fuel must be used.

It should be noted that accurately calculating the horsepower potential of a fuel injector is a bit more complicated than in the above example but in the end, a 13.6% flow error will always equal a 13.6% error in calculated power potential.

In the near future we will post a downloadable horsepower calculator for all of the Injector Dynamics fuel injectors. The horsepower will be calculated based on dynamic flow, dead time, recovery time and BSFC. For those of you testing on an engine dyno where you accurately measure BSFC, you will find that the results are spot on.

PY

3.

Strange title for a question to be answered?

Not really. They say a picture is worth a thousand words.

I've spent the last several days trying to wrap my brain around this next topic, and I keep coming back to this. The proverbial ostrich with its head buried in the sand.

As the myth goes, the lowly ostrich has little brain matter to work with, and assumes that if it can't see anything, either can its predators. Therefore it is invisible, and safe from harm.

This myth has pervaded for years because it is such an apt description for those who cannot see beyond their own view of the world.

To be fair, I have been guilty of this many times myself. In the moment that I realize it, I feel both stupid and relieved at the same time. Hopefully this article will have the same effect on those who inspired it.

From the time that Tony hit the market with the Injector Dynamics injectors we have been claiming to offer a superior product. Since the ID injectors are more expensive than all the rest, I didn't expect anyone to be threatened by it.

I was wrong.

As it turns out, part of supporting Tony has been defending attacks from a myriad of keyboard warriors.

Am I complaining? Hell no. I am a capitalist after all, and this sort of thing is to be expected.

But just because it is expected doesn't mean that I have to lie down and take it. Lucky for me, the attacks are based on misinformation, lack of understanding, flawed logic, or all three. This makes defending my position quite simple.

Thanks to the internet, I can make my point logically and succinctly, and let you be the judge.

The short story is this. The claim was made on an internet forum that it is impossible for us to match injectors at low pulsewidths because someone out there tested several injectors on their $10,000 flowbench, and found only 5% to 10% repeatability at low pulsewidths testing the same injector over and over again.

Apparently the best results showed 5% repeatability, and the worst showed 10% repeatability.

Here are a few quotes from the forum:

"Also while they are supposedly matched better, the repeatability on injector testing is withing about 5-10 percent at lower pulsewidths so if you can't get 1 injector to consistantly flow the same at lower pulsewidths withing 5% then how the hell are you gonna match 4 injectors to within 5% at low pulse widths??"

"We ran 4 ID1000 injectors 5 times, and ALL of injectors each had a variance at low pulsewidths of over 5%, and as high as 10%, over the 5 tests. I'd love to think that is was the flow bench, but i have a really hard time second guessing a $10k machine thats all its design to do is create pulsewidth to fire a set of injectors. i really doubt that our honda ecus produce a better signal than the flow bench, so where does that leave all of this."

"If the tests are unreliable, how can the low pulse width matching of the injectors, which comes from the testing be considered any better."

And this is where we get to the picture above. Is the assumption that my testing results are the same as yours?

If I relied on equipment that offered no better than 10% repeatability, I would properly be termed an idiot.

Would any of you be interested in tuning your engine on a dyno that had 10% repeatability?

Or how about measuring your bearing clearance with a micrometer/bore gauge that had 10% repeatability?

Better yet, how about going to a drag strip where the posted times were stated to be accurate within +/- 5%?

Any takers...I didn't think so.

So let's ignore the obvious for a moment and consider the problem in a logical manner. After all, logic and the ability to reason is what separates us from ostrich's right?

Stating the problem, we have repeatability that fits within a 10% window. There are only 3 variables. The test equipment, the injector, and the test operator. Assuming that the operator performed the test properly, we can only come to two conclusions. Either the injector is inconsistent, or the test results are inconsistent.

Let's consider the first option. If the injectors are in fact inconsistent on the order of 10% as our test equipment suggests, we should see a 10% variation in air fuel ratio at low pulsewidths on a running engine. That means that if we tune for a lambda value of 1.0 at idle, we should expect to see the lambda value vary from .95 to 1.05.

Is this the case? Absolutely not. Any quality injector can deliver a rock solid air fuel ratio at idle as long as the cam is not large enough to cause random misfires due to exhaust gas dilution of the intake charge.

If you disagree with this, you have much bigger problems with your tune than choice of injectors, and buying ID injectors will not help you.

If you agree that your engine holds much better than 10% variation in lambda value at idle, then you must also agree that the injectors themselves are not inconsistent within the 10% window specified by the test.

So what does that leave us with? Errors in the measuring process/equipment.

If we are to remove, or at least lessen the magnitude of the errors, we must first identify them.

Having been through this while designing our test equipment, I can point out at least a few of the more obvious.

1. Temperature control of the test fluid.

If we want repeatable results, shouldn't our test fluid have repeatable properties? If our measurements are based on mass, or volume, we need consistent viscosity, and consistent density.

Let's start with viscosity. If we perform one test with methanol, and another with Mobil 1 gear oil, we can expect to get very different results. We will measure far less flow with Mobil 1 gear oil because it has a higher viscosity.

If our test fluid changes in temperature the viscosity will change, and so our measured flow will change. So how do we keep viscosity consistent? For starters, we test it and change it regularly, but most importantly we control its temperature within very tight limits because as the temperature of a fluid changes, so does its viscosity. +/- six tenths of a degree gives pretty good results.

So what about density, typically stated as specific gravity. Again, it does not matter if we are measuring mass or volume, the density of the test fluid must be held constant. This is also accomplished by controlling its temperature within a very tight range. With the exception of quartz, some composites, and esoteric materials like cubic zirconium tungstate (Say that three times fast!) almost all materials expand when they get hot. As a result of this expansion, the density is decreased which simply means that it weighs the same, but it takes up more space.

If our measurements are based on mass, the injector which is a volume flow device meters a smaller mass of fuel.

So what if our measurements are based on volume? In theory, the results would be the same ASSUMING THAT THE CORRESPONDING VISCOSITY CHANGE DOESN'T ALTER THE RESULTS which we know is impossible. (Unless you think that an injector will flow equal volumes of methanol, and Mobil 1 gear oil)

If our "theory" is expanded to consider dynamic flow characteristics inside of the injector body, we are left with even more to consider. As the valve in the injector cycles open and closed, it moves a small mass of test fluid inside the injector. Minor as it may be, even this has a small effect on the dynamic characteristics of the injector. And we haven't even considered the damping effects of the fluid on the moving valve which is almost entirely viscosity dependant.

2. Consistent Test Voltage.

We all understand that the dynamic response of a fuel injector changes with battery voltage which is why we provide dead time values at various voltages. It should go without saying that the drive voltage to the injector should be held within very tight limits. +/- .05 volts is a good starting point.

Is there any droop in the supply voltage as the injector is cycled? More important than the obvious voltage measurement that you might perform with a volt meter, does the oscilloscope show any dynamic variance? Does this voltage change as the drive circuit heats up? Does the voltage vary depending on the number of injectors being tested?

3. Basic Measurement Accuracy.

If we want to measure anything, we need to have an accurate yardstick. Does it have graduation every 1/6th of an inch, every tenth of an inch, every inch?

Our 1000cc injectors flow approximately 115cc/min at 2 milliseconds cycled at a frequency of 100Hz. At 50Hz this volume is halved. That means that at 100Hz a 2msec test 30 seconds long will deliver 57.5 cc's. If our cylinder has graduations every cc, we need to consider that 1cc out of 57.5 equals 1.7%. Let's say that the operator who is visually determining where the meniscus lines up with the graduation can repeatably see a variation of one half of a cc. That equates to +/- .85%, for a total spread of 1.7%.

At 50Hz, the amount of fluid in the cylinder will be halved, and the same error will be doubled.

And has anyone considered that the cylinder itself changes in length as its temperature changes? How much? Dow Corning lists the coefficient of thermal expansion of Pyrex on their website. A simple calculation will give the answer if the temperature delta is known. That is assuming that the cylinder is Pyrex. If it is plain old glass, the coefficient of thermal expansion will be even higher, and the potential for error will be even greater.

Did you know that for this reason, graduated cylinders have their accuracy rated within a specific temperature range?

And what about the amount of fluid that remains in the cylinder from the previous test? Was the cylinder completely cleaned and dried, or does the machine just open a dump valve and let "most" of the fluid out. How much is left clinging to the walls and how consistent is it? Could there be one or two cc's left? Or could there be one cc left over on one test, and two cc's left over on another? Surely the amount of fluid left in the cylinder is affected by viscosity.

How long does it take to empty all the honey from a jar? How long does it take to empty all the water from the same jar?

Get my point?

To be honest, I have no idea to what degree these measurement variables may have affected the tests described on the forum. When designing our equipment, I considered the above mentioned variables along with several others and decided to not even go down that road. There are just too many variables, and in reality there are probably many that never even occurred to me. We do not use graduated cylinders to measure flow.

The point is that if variability in the test is to be reduced, VARIABLES INHERENT IN THE TEST PROCEDURE MUST BE REDUCED!

4. Fuel Pressure.

This is another obvious one, but getting the required consistency when measuring DYNAMIC flow is much easier said than done. If you're measuring static flow, only pulsations caused by the fuel pump need to be considered.

Well...maybe not. The pressure recovery time after the injector first opens and the resulting oscillations at the systems undamped natural resonant frequency will effect the first portion of the test, but probably not to a large degree.

Dynamically, it's whole different game. Any undamped system will have its own natural resonant frequency which may or may not correspond with the test frequency. At best, it will only affect your ability to produce a linear flow vs. pulsewidth curve. At worst (And most likely) the undamped natural frequency of the system will not correspond with the test frequency, and the result will be a beat frequency which will skew the results to the point that you can't make sense of anything.

A beat frequency is the result of two disturbances (pulsations) occurring at different frequencies. At some points in time, the high pressure portion of both disturbances will be in phase resulting in a high pressure spike. At other points in time, the disturbances will be out of phase, and result in an overall low pressure in the system. This is described as constructive and destructive interference and the result is an absolute mess.

Sound like I'm making it up? Just Google "beat frequency" or "constructive and destructive interference" Better yet, dig out your first year Physics textbook for a more thorough description.

As bad as this sounds, I am only considering 2 frequencies. In the real world, the mass of the valve seated on its spring inside the pressure regulator will have its own undamped natural frequency, the fuel pump will be adding its pulsations at who knows what frequency, the lines in the system will all have their own organ pipe resonance, and guess what? As the temperature of the fluid changes, so does the speed of sound through the fluid and all of the organ pipe resonances change as a result.

Think you can purchase an off the shelf pulsation dampener from Summit or Jegs and solve the problem? Why not purchase a fast response pressure transducer and log it at a minimum of 2 times the test frequency while you're at it.

After seeing the data, tell me if you see some potential for error there!

5. Fuel Conditioning.

Not so obvious at first, but when the injector is tested an awful lot of air is mixed in with the test fluid. Do I have to tell you what will happen if you have no provision for de-aeration of the test fluid?

So is this the end?

Not even close.

Educating the competition is not on my list of things to do today, and so I have pointed out just a few of the more obvious problems inherent in achieving accurate results. The rest is up to you.

The point is that measuring anything that is dynamic in nature is challenging. There are entire industries devoted to it.

I have worked vary hard to achieve repeatable results with my equipment. In fact the majority of the 6 month design period amounted to getting repeatable results, regardless of test pressure, or injector type.

So does this sound like a bunch of stuff I just made up, throwing in some impressive sounding terminology to confuse you? Don't take my word for it, purchase the SAE standard for electronic fuel injector testing. About 3 pages in you will realize that I gave you the short story, and there is far more to the picture.

After repeatedly being accused of ignoring "industry standard" I think it would be good for all of my detractors to read and understand the “real" industry standard. If you find ANYWHERE in the SAE standard that 10% repeatability is acceptable, you will have something to base your argument on.

Otherwise, look to your equipment and procedures and understand that there is a wealth of information and technology available to you if you are willing to go get it.

As it stands now, repeatability of the ID1000's tested at 2 milliseconds on my equipment is +/- .65%. To me, that's troublesome. +/- .2% would make me feel a whole lot better. To others, that level of repeatability seems unattainable, and if it doesn't exist in their world, it can't possible exist in mine right?

As it turns out, I made a fairly major breakthrough just last week. This strategy will be implemented in our new test bench which should be completed by early summer. I am fairly confident that the repeatability can improved two fold.

Note on 8-14-09: As it turns out my guess was pretty close because our repeatability is now +/- three tenths of a percent.

So why am I taking it further if I am already getting good results? Because in this industry if you are not moving forward, you are moving backwards. My goal is to be the absolute best at what I do, which means moving forward constantly.

And that is why I am not relying on equipment with a 10% variation in measured flow.

So to the gentleman that politely accused me of being full of shit, I politely ask that you stop acting like a scared ostrich. Or at the very least do not attack me because of it.

(For those of you who didn't accuse me of being full of shit, thanks for visiting the page and reading my rant...)

PY

An additional note to the above rant. The reason for batch testing so many injectors at once (At least 100 at a time) is that once you modify the injector, the usual Bosch consistency goes to hell. Grabbing 8 random unmodified Bosch EV-14's off the shelf is usually a safe bet. Grabbing 8 random modified injectors is a whole different story. A good match across the pulsewidth range can only be attained through batch testing/grouping. Or throwing away an awful lot of injectors...

4. Do You Have a Low Impedance Version of Your ID2000 Injector?

The short answer is no.

Here comes the long answer.

With a few exceptions from Siemens, the only low impedance injectors currently available are based on technology that is at least 20 years old.

All of the Bosch EV-1, Rochester/Delphi, and Lucas style injectors have a very heavy valve, typical of the older designs.

Because it takes a lot of grunt to get that valve moving, a low impedance, high energy coil is required.

The newer designs from Bosch, Siemens, Denso, Keihin, etc have a valve mass that is several times less than the older designs, and so they are able to achieve the same, or in most cases better low pulsewidth response than the old low impedance injectors.

As an added benefit, the non linearity that results from the switchover from peak to hold mode is eliminated.

The picture below shows the valve assembly of various Bosch injectors as they have progressed over the years.

For reasons I can’t explain, the aftermarket is still 20 years behind while the injector manufacturers are moving their technology forward.

Currently, low impedance peak and hold is used on some direct injection injectors and a handful of specialized racing injectors that few of us can afford, and even fewer of us need.

(TAG Mclaren, Bosch HDEV and Magneti Marelli come to mind)

These injectors are meant to run at very high pressures approaching 1,500 psi, or several times that for the DI injectors.

In the end it comes down to technology. The injector manufacturers are simply doing a better job than they did 20 years ago.

To see just how well the ID2000 handles low pulsewidth operation, follow this link to see the ID2000's delivering a smooth 850rpm idle on a bone stock 1.8 liter Integra

5. Will The ID2000 at 70 psi Work at Idle on My EVO

I am switching to Ethanol and my engine builder says "4 injectors of that size will bore wash hugely at low rpm. It will ruin the engine quickly."

I would like to have 2 maps, regular pump gas and Ethanol.

Will I have a problem at low RPM with the STD gas at 70 psi base pressure?

Bore wash is certainly a valid concern. The idea being that an injector that large will dump copious amounts of fuel into the motor because of poor low pulsewidth control.

Dual fuel arrangement. Hmm...I hadn't thought of that.

Enough fuel for full power on E85, but good enough control to idle the motor at a reasonable air fuel ratio on gas???

And just to make it more interesting, he had to add "at 70 psi" which is what he came up with using our horsepower calculator.

The short answer is yes it will work, and because it will work, bore wash will not be an issue.

The short answer is pretty lame, and sounds like the BS you get from a typical injector peddler so I'll show you how we arrived at the answer.

A few year back we tested his RC injectors to gather the necessary flow and dead time values for his Motec ECU.

We started by requesting a datalog from his ECU so that we could determine his current idle pulsewidth

Looking at the data, you can see that the average idle rpm is 1,265 and the average pulsewidth is 1.841 milliseconds.

You can also see from the lambda, rpm, and manifold pressure traces that the motor has a fairly large cam and corresponding lope at idle.

This is helpful, because this increases the idle fuel flow requirements. A bone stock motor would require considerably less fuel at idle.

Now that we know his idle pulsewidth, we can go back to his injector data to determine idle fuel flow.

Looking at the yellow trace which is the flow curve at 12 volts (The voltage at idle as shown by the data log.) we can see that the injector is flowing 131 cc/min at 1.84 milliseconds.

From here we go to the flow vs. pulsewidth curves for the ID2000 at 70 psi and find the corresponding pulsewidth.

Looking again at the yellow trace (12 Volts) we find that a fuel flow rate of 131 cc/min occurs at a pulsewidth of 1.482 milliseconds, well above the "knee" of the curve.

This bit of breathing room between the required pulsewidth, and the knee of the curve will allow us to account for atmospheric conditions, or inlet air temperatures that require less fuel flow.

So how much breathing room do we have? Again, we just look at the data and find that the fuel flow at the knee of the curve which occurs at 1.40 milliseconds is 114cc/min, or 13% less fuel than the motor is currently using.

And all that can be accomplished with a lambda value of 1.0 Not too shabby.

So in case any of you are wondering how we can make such specific claims as to what our injectors are capable of, now you know.

There's no magic, just the ability to properly characterize an injector, and come to some very obvious conclusions based on that data.

The next "Question to be Answered" has to do with how the injectors react at high duty cycles. Once again I will answer the question by simply looking at the dynamic flow characteristics of the injector.

For now I'll leave you with this chart showing the maximum duty cycle of the ID injectors. Maybe you will be curious enough to come back and find out how these values are determined, and what happens when you "cross the line"

Thanks for reading.

PY

More to come...