the question is really when does it open it?

The PCM calculates or has calibrated a end of injection point, a point (crank angle) where it must inject all the fuel for that cycle if it is to be drawn into the cylinder. It opens the injector "InjPW" time before this target close point. This is a slight generalization but good enough.

I can't remember if the LS1 has a minimum off time that is enforced by the PCM (ie. a period of time the injector MUST be off). Regardless the PCM either holds the injectors wide open or near enough to it.

 

Sorry, don't have any.

I also have 60lb/hr injectors in my car right now and they are only at 70% duty, got quite a way to go before I max them out.

 

So the question remains:
At >100%DC does the PCM spike the injectors closed at the end of each cycle
(which, at 100%DC, is also the start of each cycle)...?
(Or does the PCM just keep the injectors open...?)

 

ha.. it sounds as though you are answering your own question friend. lol


Chris Macellaro

 

The car doesn't know what the duty cycle is so it just keeps pulsing. You can't start another pulse without ending the previous one. I'd have to scope it again to see how long that distance is, but I played with an injector on the bench with my scope tonite. It was responsive in closing down to .5 mS, but that is just at atmospheric with no pressure on it.

For a true test you would need to run the injector at pressure and monitor the spray pattern with a strobe. Different injectors will react in different ways as well.

 

Hilarious, you should be a politician.

Let me restate the questions for you.

Does the injector ever see a static signal from the driver during operation?
What type of proof do you have?
and what was the reported D.C. when this happened?

Thank you

 

This is an interesting thread. Even if the driver firing strategy was simultaneous double fire (SDF) or alternating double fire (ADF) aka "batch" per above, the Period (P) would be 1 crank revolution instead of 2 crank revolutions. But any pulse width longer than P would still overlap the next cycle resulting in a DC exceeding 100%.

It shouldn't matter if its the 1st, 2nd, 3rd ..nth period or any combination commanding the injector open. Open is open and max flow is max flow even if prior pulse widths overlap.

But it seems most everyone agrees with the above, so either:
A) White2001s10 data aquisiton is flawed
B) GM doesnt use standard Sequential Fuel Injection (SFI) per SAE J1832

It helps to have a copy of J1832 in front of me, no I dont memorize this stuff

gameover, how confident are you that GM uses 2 rev = 100% DC and not perhaps 1.5 rev = 100% DC with .5 rev pulse OFF? Also, when is HP Tuner going to include access to the "end injection point" relative to crank position?

Zombie, use a scope to compare injector supply current/voltage and accelerometer glued to the injector. No need for visual fuel in order to measure opening and closing time. Is this injector driver you are developing high or low impedance? Underhood? Curious as I am looking for low and may have found an off the shelf solution.

white2000, thanks for raising a good point with this thread. I've heard of others increasing fuel above 100% so I know you're not alone. Having never experienced this first hand its hard to comprehend.

 

SSpeedracer,

GM PCM returns IBPW (Injector Base Pulse Width);
the injector duty cycle is then calculated by the scan tool using IBPW, RPM, and 2 crank revs.

GM PCM fires each injector once every 2 crank revs.

Joe

 

So for a scan showing 150% DC, PW is six strokes. The only way fuel would increase would be if the OT/CT was removed as the injector goes static. So, maybe get 1.5ms additional flow and that could correlate to ~107%DC maximum at 6000 RPM.

The plot thickens. I want to see a screen scope image also. But who has the time to be a scientist?

 

the only thing in quesiton is if the injector closes or stays fully open, if i get time i will look thru the code but it's likely to be part of the I/O hardware. The pcm generally just loads values into mapped memory locations then the hardware does the rest.

GM doesn't use duty cycle in any of it's code, nor Ford etc. It's just a convenient way to express pulse width in relation to RPM so you can tell when your injectors are maxing out or you need to jack the fuel pressure up a little

Chris...

 

I don't care enough about proving my point to go through all that trouble, but for those that say the ECU just holds the injector open, how does it know to do that if it's just commanding pulses and doesnt know the duty cycle? It is possible that the off time is so short that the injector doesn't have time to close all the way though. I need a car with maxxed out injectors to figure that out though.

As for what i'm working on it's a injector controller that is not vehicle specific. It has the ability to run high or low impedence injectors and give fuel tuning ability on a 500-10000 rpm scale every 500 rpms. It's also boost referenced every 20 kpa from 20 kpa to 240 kpa (20psi). I don't know of anything else on the market that does what i'm doing.

I'm hoping to have the drivers working on a car this weekend. The controller portion is almost done, writing code for LCD menus on a microprocessor is time consuming.

 

Right, that's 100% of one cycle plus 50% overlap into the next cycle;
the PCM may be commanding 150%DC, but injector was already delivering max fuel at 100%DC,
so there's no way to get more fuel by commanding >100%DC (it's not physically possible).

 

Back to the original post, why would anyone want to switch from sequential to batch? With seq we are activating each injector in sequence with each intake stroke for each cyl which of course becomes a moot point at high duty cycles. Batch activates every injector every 180 deg of crank ratation Vs the 360 deg for each injector in a sequential mode. Running batch every 360 causes idle/drivability issues because some cyl's intake stroke have to wait too long for the next squirt. Thus the sequential mode activated each injector half as often with a pulse width twice as long keeping the DC the same (in theory). Activating the injector half as often greatly improves the dynamic range and accuracy of the injectors, so again, why would we want to go back to the dark ages and utilize a batch operation? What am I missing?

I agree with anyone who stated that you cannot increase fuel flow past 100% DC. If I could increase fuel after my scanner indicated 100% I would question my scanner. Next I would scope the pulse width and do the DC math. But then again, how would I know, I would and have never run any engine very long near 100% DC. Anything past 80% usually causes very erratic injector flow rates due to the internal valve's (pintle, disc or ball) ability to fully open and close. According to any injector manufacurer I've ever talked to, beside this being outside the engineered operating range of an injector, running an injector near or at 100% will cause premature failure. Of course this is mostly theory but some has explained some melted pistons we've seen on the dyno and track.

 

The original question was trying to find out if the pcm switched from sequential to batch mode under high DCs. I agree, batch fire injection is a way less precise and economical EFI mode. white2001s10 is seeing cars that run their injectors beyond 100% DC (sometimes way beyond) and yet they are able to run track times that show no loss in power. He's also backed this up with wideband AFR readings. I find it interesting that this is showing up on dry nitrous equipped cars that are running far beyond 100% DC. Could this be a connection?

 

FWIW, the newer Vipers will shut fuel like a rev lmt if the injectors approach a certain duty cycle. It took us awhile to figure that one out.

 

The answer to your first question is no, the injectors don't switch from sequential to batch fire on any LS1 i've seen. The following graphs illustrate this, firstly look at the injector frequency, if it went to batch fire the frequency would double, it is clearly not.

The injector duty cycle shown on the other graph is an actually measured value at the injector (so is the frequency), not a calculated value that you see in scanner software. I can guarantee when the "measured" duty gets into the high 90% range the AFR starts to head north (lean), as can be seen in the example with the IDC just over 90% the AFR is just starting to head lean.

Ignore the first part of the AFR trace, as the vehicle was in early stages of being tuned.

 

Why not increase DC above 91%? Injectors arent static yet, they have some more flow left in them. Obviously not an ideal situation, but what is the true limit of your system?

How is DC measured directly if its a function of P and PW?

 

I wasn't tuning the car, it is just an example I had lying around.
The IDC is measured by a purpose built module designed to measure IDC at the injector, if you put a scope on, it would measure the same number.
The true limit that I have seen is 98-99%, I'm not saying that this is what anybody should do, it is just what can be measured.

 

That's very text-book, but doesn't give all the answers.
The frequency is clearly pure sequential which is the easier way. Your graph seems to be a tad off because 6000 RPM would equal 50hz.


AFAIC there isn't nearly enough evidence there to prove that the increased DC had anything at all to do with the rising AFR. It's too late now to adjust the test, but a specific test would be in order for that.
You should be able to push the test well over 100% DC to see what happens.

 

I will dig up a graph that actually shows this happening, I primarily used the first graphs to show that the frequency did not double at any certain point in response to the original question.

The following graphs show some real world data of a stock LS1, showing all of the information relating to Injector IDC, Freq and Pulse width, as measured at the injector. I even made a maths channel to show the maximum available Injector Pulse Width that would be available based off Time. The example has the IDC at a max of 73%, and you can work that out from the other channels I've provided, this car still had plenty of fuel, the AFR was 12:1 at the top end.

If you use the graph data provided, you can calculate that eventually if the same Injector Pulse Width is maintained, it will have to hit 100% IDC, so there is no way you can go beyond 100% in a truly measured fashion.