General Info: How Injectors Work

[SSpdDmon]

Due to some recent conversations about injector sizes, calibrations, etc., I did a quick Google search and found some pretty interesting (free) info that, IMO, is worth sharing for those who want to learn more about what's happening under the hood. They seem to do a fairly decent job explaining the basics.

FYI - I'm including their whole zip file only because the text/PDF document I speak of references their Excel file quite a bit and I didn't know if anyone wanted to play around with it. As for the company, they do offer injector services for a fee. But, I can't speak to their work.

[eficalibrator]

Paul Yaw is very knowledgable on injector behavior and is one of the very few people in the industry who actually recognize how to properly test the nonlinear flow region of the injector. You'll see a similar discussion in an article I wrote (hey this one is free too!) a while ago before reading Paul's work. The important point is that calibrators (or "tuners", if you will...) should recognize the facts that 1) injectors are not linear, and 2) not all injectors are non-linear by the same amounts.

<LINK TO MY ARTICLE HERE> (http://www.calibratedsuccess.com/Assets/Documentation/Fuel%20Injector%20Article.pdf)

[Aloicious]

Due to some recent conversations about injector sizes, calibrations, etc., I did a quick Google search and found some pretty interesting (free) info that, IMO, is worth sharing for those who want to learn more about what's happening under the hood. They seem to do a fairly decent job explaining the basics.

FYI - I'm including their whole zip file only because the text/PDF document I speak of references their Excel file quite a bit and I didn't know if anyone wanted to play around with it. As for the company, they do offer injector services for a fee. But, I can't speak to their work.

Paul is a good guy, I've had a chance to talk to him quite extensively when I did my MPFI swap, I got my 24# injectors from him and had him do his dynamic testing on them. he provides alot of good info for them, its general scientific data for them, not specific to EFILive, but from his data I was able to extrapolate the oft unknown tables like variable voltage offsets {B3701}, small pulse adjust {B4005}, Small pulse threshold {B4006}, etc. basically most of the injector data tables that alot of us either use stock tables, or guess on since the data isn't widely available from the manufacturer. and its great because not only does he test these variables but other good info as well above and beyond the standard flow testing that injector testing usually provides. if I wasn't so poor I'd have him test my new 39# injectors too, but alas my wallet is on the light side from my whole 24x/whipple/T56 project. I can speak of his work from experience though and he does a very quality job.

I have however recently acquired a quality high resolution oscilloscope, and when I have the time I've been looking into building a rig to test various injector dynamics, mainly out of curiosity. injectors fascinate me, they're simple, yet elegantly complex.

for anyone interested I am selling the 24# injectors with all the info I have for them. some info on them can be foundhere (http://www.pacificp.com/forum/viewtopic.php?t=8977)
however if anyone is interested in them please PM me so we don't derail this thread.

[SSpdDmon]

Paul Yaw is very knowledgable on injector behavior and is one of the very few people in the industry who actually recognize how to properly test the nonlinear flow region of the injector. You'll see a similar discussion in an article I wrote (hey this one is free too!) a while ago before reading Paul's work. The important point is that calibrators (or "tuners", if you will...) should recognize the facts that 1) injectors are not linear, and 2) not all injectors are non-linear by the same amounts.

<LINK TO MY ARTICLE HERE> (http://www.calibratedsuccess.com/Assets/Documentation/Fuel%20Injector%20Article.pdf)
I remember reading this a while back, but didn't remember who wrote it or where to find it. Thanks for posting...some good theory/observations there too. :)

[joecar]

...

I have however recently acquired a quality high resolution oscilloscope, and when I have the time I've been looking into building a rig to test various injector dynamics, mainly out of curiosity. injectors fascinate me, they're simple, yet elegantly complex.

...Cool...:cheers:...scopes teach you a lot, what scope did you get...?

I got a Picoscope from Autonerdz, see my waveform (see image, see the open dip and the close hump)... I haven't had much time to explore this further, but I intend to...

I'm trying to figure out how to measure the amount sprayed...


http://usera.ImageCave.com/joecar/Injector%20Response.png

[SSpdDmon]

I guess the one thing I'd like to see more of after reading this stuff is, how MANVAC affects the injector's flow in the time between full open and full close. I see mention of dead time and compensating for that based on commanded pulse widths and battery voltage. But looking that the tables we have available, GM thought it was relevant enough to add in this one more variant - vacuum - and it doesn't seem to be a part of the discussions above (unless I missed it).

[5.7ute]

I guess the one thing I'd like to see more of after reading this stuff is, how MANVAC affects the injector's flow in the time between full open and full close. I see mention of dead time and compensating for that based on commanded pulse widths and battery voltage. But looking that the tables we have available, GM thought it was relevant enough to add in this one more variant - vacuum - and it doesn't seem to be a part of the discussions above (unless I missed it).

Which tables are you talking about Jeff?

[SSpdDmon]

Which tables are you talking about Jeff?
The battery voltage offset table is 3D and allows for MANVAC. That's what I'm getting at...and when we switch to bigger, slower injectors...I'm wondering if the effects of MANVAC are more pronounced.

[5.7ute]

No worries, just making sure we are on the same page.:)

[Aloicious]

Cool...:cheers:...scopes teach you a lot, what scope did you get...?

I got a Picoscope from Autonerdz, see my waveform (see image, see the open dip and the close hump)... I haven't had much time to explore this further, but I intend to...

I'm trying to figure out how to measure the amount sprayed...


yeah its a Picoscope. I scored it used but in great condition for a steal of a price. but its also not their newest model but should work just fine for pretty much everything I would use it for though, including injector dynamics. thats a great image, I'm looking at the same thing, but I wanna build a flow bench type rig that I can adjust the voltage and pressure on to get a full range of injector characteristics over varying circumstances. SSpdDmon brings up a good point about Manifold vacuum too. I bet I could rig up something to account for that as well. interesting stuff, too bad it'll be a while before I can work on it due to work, school, and research that I have going on....

[joecar]

I have the Pico 3423 and it's fine for automotive use (I'm told the 4423 has more settings and is more complex to use)...

Here some secondary waveforms I took on my wife's other car (1976 Datsun)...

red = capacitive probe on coil wire
blu = inductive sync probe on #1

1st image shows secondary event train.
2nd image shows single event (#3 cylinder, FO is 1-3-4-2).

I'm still learning, but by counting the number of cycles in undershoot/overshoot and the slope/length of the arc line you can make some determination on the soundness of the coil/wires/plugs.

[Aloicious]

I have the Pico 3423 and it's fine for automotive use (I'm told the 4423 has more settings and is harder to use)...

Here some secondary waveforms I took on my wife's other car (1976 Datsun)...

red = capacitive probe on coil wire
blu = inductive sync probe on #1

1st image shows secondary event train.
2nd image shows single event (#3 cylinder, FO is 1-3-4-2).

I'm still learning, but by counting the number of cycles in undershoot/overshoot and the slope/length of the arc line you can make some determination on the soundness of the coil/wires/plugs.

nice. mine is the 2 channel version, but its not USB, its the Parallel port one, which is okay since my laptop has a parallel port. I can't remember the part # off the top of my head. I really havent had much time to use it except some minor messing around with it. I'll have to pick your brain on some stuff once I get my current project finished, and I can really sit down and crank out some data from it.

[eficalibrator]

I guess the one thing I'd like to see more of after reading this stuff is, how MANVAC affects the injector's flow in the time between full open and full close. I see mention of dead time and compensating for that based on commanded pulse widths and battery voltage. But looking that the tables we have available, GM thought it was relevant enough to add in this one more variant - vacuum - and it doesn't seem to be a part of the discussions above (unless I missed it).
Does it have SOME influence? Sure.

Is it significant? I'm not so sure at the moment.

If you look at the change in offset across pressures, it's minimal compared to the change with respect to input voltage. We're talking about small fractions (hundredths) of milliseconds here, where the difference between two injectors is more like tenths of a millisecond. The primary discussion usually centers around injector design versus latency instead of manifold effects versus latency since there's an order of magnitude more importance there.

It really comes down to a question of "Does splitting these hairs give me any benefit to AFR delivery control?" For offset vs voltage for one injector design compared to another, the answer is a resounding "YES", but for manifold pressure variances the Jury is still out and likely leaning toward "not so much."

[SSpdDmon]

Does it have SOME influence? Sure.

Is it significant? I'm not so sure at the moment.

If you look at the change in offset across pressures, it's minimal compared to the change with respect to input voltage. We're talking about small fractions (hundredths) of milliseconds here, where the difference between two injectors is more like tenths of a millisecond. The primary discussion usually centers around injector design versus latency instead of manifold effects versus latency since there's an order of magnitude more importance there.

It really comes down to a question of "Does splitting these hairs give me any benefit to AFR delivery control?" For offset vs voltage for one injector design compared to another, the answer is a resounding "YES", but for manifold pressure variances the Jury is still out and likely leaning toward "not so much."
So in essence, there's not necessarily a need for a 3-D table that allows for MANVAC variance? I'm wondering now if GM included this functionality in the tune due to their fuel system running a flat 4-bar in place of a vacuum ref'd regulator. Looking at the stock numbers, I'm inclined to agree since there is some variance in the atmospheric vs. idle end of the table...but the difference is pretty small.

[Aloicious]

So in essence, there's not necessarily a need for a 3-D table that allows for MANVAC variance? I'm wondering now if GM included this functionality in the tune due to their fuel system running a flat 4-bar in place of a vacuum ref'd regulator. Looking at the stock numbers, I'm inclined to agree since there is some variance in the atmospheric vs. idle end of the table...but the difference is pretty small.

IIRC when I was looking at alot of these injector tables back in the day I came across some that were actually altered for manvac from the factory. I'll see if I can find them again

[eficalibrator]

So in essence, there's not necessarily a need for a 3-D table that allows for MANVAC variance? I'm wondering now if GM included this functionality in the tune due to their fuel system running a flat 4-bar in place of a vacuum ref'd regulator. Looking at the stock numbers, I'm inclined to agree since there is some variance in the atmospheric vs. idle end of the table...but the difference is pretty small.

IIRC, GM is the only major OEM doing this right now. (At least based on my failing memory and limited access to everyone's software) Somehow Ford, Chrysler, BMW, and others are able to pass the same emissions tests without compensating offset for manifold vacuum. This confirms that the effect is minimal at best, as seen by the raw numbers used in most GM applications. There may be the oddball case where some calibrator swagged this table to bring something else in line, but it's rare.

You'd be surprised how (poorly) some of the OE calibrators work. It makes me scratch my head a LOT during the day... :angel_innocent: A few rare, bad examples still exist at times. Properly done, the tables follow the physics.

[Aloicious]

IIRC, GM is the only major OEM doing this right now. (At least based on my failing memory and limited access to everyone's software) Somehow Ford, Chrysler, BMW, and others are able to pass the same emissions tests without compensating offset for manifold vacuum. This confirms that the effect is minimal at best, as seen by the raw numbers used in most GM applications. There may be the oddball case where some calibrator swagged this table to bring something else in line, but it's rare.

You'd be surprised how (poorly) some of the OE calibrators work. It makes me scratch my head a LOT during the day... :angel_innocent: A few rare, bad examples still exist at times. Properly done, the tables follow the physics.

the earlier GM tunes didn't use the vacuum offset in the table, its a newer thing, the earlier LSx tunes only compensated for the voltage offsets, where the newer, (I believe it was some LS2 tunes) had a slight variation in the manivac axis. I agree it would be minimal, if any, need for this. but I would assume it is there for a good reason, and definantly something that would be interesting to test. I did play around a little with it when I was altering that table a year or so ago. changing the Manivac settings seemed to make no difference whatsoever. even extreme changes didn't do much. I believe this may be a factor for the returnless fuel systems, and seeing as how mine is a return system perhaps that is why it didn't do anything. who knows. I wish I could find those tunes. I'll keep looking.

[gmh308]

Cool...:cheers:...scopes teach you a lot, what scope did you get...?

I got a Picoscope from Autonerdz, see my waveform (see image, see the open dip and the close hump)... I haven't had much time to explore this further, but I intend to...

I'm trying to figure out how to measure the amount sprayed...



Which model PicoScope Joe?

[joecar]

Which model PicoScope Joe?
Pico 3423 quad channel.

[gmh308]

Pico 3423 quad channel.

Very nice! :)