Steve, I think bluecat answers your question on the reference periods. Still trying to work through this myself.
https://forum.hptuners.com/showthrea...OI-fuel-timing
Originally Posted by pontisteve
Steve, I think bluecat answers your question on the reference periods. Still trying to work through this myself.
https://forum.hptuners.com/showthrea...OI-fuel-timing
I must have missed this
It's been a long time, but now I remember that Bluecat's post is where I got my initial understanding of how this all works. He did some pretty solid testing, but that -784+normal formula of his just sounds janky. It sounded janky to him too, at the time. And so we've all kind of wondered all this time, is that really it? Why the weird numbers?
I'm not sure I even understand all that high tech stuff he did on the bench there, but I know he's a smart fella, so he probably did it right. I'm not that sharp, so I'm just gonna use my Picoscope and start moving the timing around and recording what happens. If I could record cam and crank sensors and injector #1, along with cylinder pressure, and keep making changes and flashing, I should be able to map out exactly what the heck is going on. I think I'll start by zeroing them all out and see what happens.Then add 1.0 to normal a few times and watch it. Then add 0.1 and see, etc.
I've found plenty of documentation where GM refers to a single crank tooth signal a reference pulse. They could be referring to the cam sensor as well, but there's only one pulse so it's hard to make the math add up that way either. We WILL find out! But I do believe Bluecat probably has it right that a reference pulse probably means 90 degrees. 8x90 = 720.
I mean, just looking at it. Reading it. It says clear 0 to 8 which must represent one full cycle 720*. If we say 0 is compression TDC, a 6 would be bottom dead center intake stroke, and 7.999 is just about to TDC compression stroke again.
With that logic I look at the factory table and see that 5.55 is standard which must be about halfway through the intake stroke, and 6.1 is just after bottom dead center intake stroke. I guess by setting the number to 6.1 or close to that we allow short pulses (like 2ms or 3ms around idle) to be sprayed after exhaust valve close.
My question is, even with 5.55 isn't that plenty late enough? Don't most exhaust valves close much earlier than that... guess not but it doesn't seem that they would be open much around after about 4.5 to 5.0 let alone 5.55. Something doesn't quite add up here... but Im too lazy to do advertised duration numbers atm so
I think you're about right. 0 is TDC compression, 360 is TDC exhaust (4.0), and 0/720 is TDC compression again (8.0), assuming each reference period is 90 degrees. That 90 is such a big number that I seriously doubt Bluecat could have been off about that.
The bigger question is about the formula itself, where these numbers start from, and how boundary affects it. Nothing will beat testing in a car with a scope, while moving around each one, one at a time.
I've seen older Fords put EOIT around 280 to 360 at light loads, and about 464 at high loads. If you consider an 80% injector duty cycle to be what the factory shoots for on a stock motor, that's 80% of 720, or 576 degrees. If the duty cycle at WOT is 85%, that's 612 degrees of injector duration.
We can't start the injection until after the intake valve closes, which would be somewhere just after BDC on the intake stroke, as the piston starts coming up on the compression stroke. I'm guessing somewhere around the 600 degree mark at the earliest. If we start at 600 (120 before TDC compression), and we need to finish 576 to 612 degrees later, that would put us at an EOIT of 456 to 492. In "reference pulses", that would be 5.07 to 5.47, assuming these references start at 0 TDC. Stock is 5.55, or 499.5.
That coincides pretty nicely with what I figure, and maybe just starts injection a few degrees later than what I figured. I'm guessing that Normal works exactly like that, Makeup determines when the injector fires, probably on the other stroke (360 away from the Normal?), and I'm still not sure about Boundary. But I'm guessing it's a max limit of some sort, that keeps the injector firing no earlier or later than x.
The factory idea is to fire the injector and have EOIT end just before the intake valve opens. Since we have longer exhaust duration and lots of overlap, we need to rethink that, and add just enough injector delay to cause a 2 to 4 ms idle pulsewidth to begin just after the exhaust valve closes. I figure that's somewhere around 30 degrees of injection duration. If 360 is TDC, and an aftermarket cam closes the exh valve at say 385, then we need to finish injection by 415 or so.
With older Fords, we can pick EOIT based on RPM and engine load, so we can really move EOIT based on conditions. With these LS engines, it seems they don't want EOIT to vary, so we just get one number, Normal. If that's the case, the injection would take place WAY after TDC anyway, much closer to BDC. That just doesn't seem right to me, so further testing should be done. Meanwhile, a good guess to just blanket add Inj Delay might be about 30 to 50 degrees above stock (Increase Normal by .33 to .56).
I might have missed this but what effect is there if eoit is to high. Say its at 6.5 when should be 5.8. Would the afr read richer as you work to 5.8 then start to lean back out after that as the fuel is now being injected straight into the cylinder and has less time to evaporate?
This is how I interpreted it when I read Blue's original data. The computer is "watching" the current cycle (720) then calculating when to fire on the next. So, the mysterious "extra 720" adds 8 watched reference periods to the formula. 784 and 64 are telling us the same thing, but 784 includes the 720 where the injector firing plan comes from. I don't see how the "plan" for firing the injector could come from within one, current, 720 degree cycle.
Similar to ignition timing: You can not fire the spark plug "before" the sensor gets there, so you really can't advance ignition timing. All you can do is read (via sensing) what is happening on the current cycle, and delay firing a portion of the next complete cycle, so as to give the appearance of being "early" on that next cycle. It's the nature of sensing... I read what's going on right now, and predict where to do it next time around and how soon that's going to occur based on how the last cycle happened.
Last edited by Picklito; 10-04-2018 at 12:31 AM.
That's a good explanation. I buy it.
Thanks for reading it! And right back at you... your in-depth explanations are really helpful. Someone on here, or the referenced threads, mentions that the maximum spark advance is 60? BTDC. It "makes sense" that this is probably a few degrees after the beginning of a "logging cycle" of 720 degrees. Maybe... say... 64?? All speculation for me at this point, as I've done no lab work on this. A really easy solution, which we can't seem to get, would be the reference definitions of boundary, normal, and makeup. Are there really no GM engineers on here who could help us find this?
That is typically how it goes. The crank and cam sensors are made, with their teeth placement in certain spots for a reason. Since fuel injection and spark are both based on these cam and crank signals, they're going to design the reluctor teeth to be spaced out right for both. Almost every engine has a maximum spark advance of around 60 degrees. They just never need more than that, and often somewhere in the 50's is pretty much max part throttle timing. So having a reluctor tooth located around 60 degrees makes sense.
With the LS engine, they wanted fine resolution for accurate misfire detection, so they opted for 60 crank teeth (minus a couple for syncronization), and a cam sensor that's just there to indicate which stroke you're on, intake or exhaust. 360 degrees / 60 teeth = 1 tooth every 6 degrees of crank rotation. When I get around to scoping one of these again, I'll be able to plot each tooth's location against piston position and against cam sensor signal location.
There's likely a little bit of inaccuracy in exact location. The 64 degrees might really be 66 degrees, or 1 tooth before TDC. Who knows. I'm not sure how dead accurate it has to be. Then again, spark is based off of it, so one would think they would need to be pretty spot on.
You are close. Reference degrees are 65 per period
How did you come up with that?
It is that.
So how can you have eight 65 degree periods in a 720 degree cycle? We're gonna need a little more than a proclamation here.
Each value moves roughly 11 degrees which if you divide 720 by 11* gives you 65 sections.
Then keep running around and trying. Not everything is centered around a circle. Gm didnt. You shouldnt either.
Forgive me, but can you base this on something other than theory? GM is not likely to use round odd numbers.
After reading this entire thread over the last couple of days I'm going to weigh in with what I think I understand. Will begin with some assumptions and observations, and then expand from there. Hopefully I can make my thoughts make sense?.
While yellz06 has stated that reference periods are 65 degrees of crank rotation, due to Bluecat having documented how he determined empirically that the reference period is 90 degrees I'm assuming that 90 degrees is correct, as well his equation regarding that EOIT (in degrees) equals the sum of boundary and normal reference period values times 90 minus 784.
Plugging in stock values for boundary and normal (warm motor) of 6.5 and 5.55 gives 300.5 degrees as others have noted, including Bluecat, who noted that was 300.5 degrees ATDC of the compression stroke.
As others have noted, an 80% injector duty cycle corresponds with an injector off duty cycle of 20%. 20% of a 720 degree piston firing cycle corresponds with 144 degrees of crank rotation.
Adding the 144 "off" degrees to the 300 EOIT to calculate when the stock parameters would have the injectors come back on in the 80% duty cycle scenario indicates the injectors coming on at 84 degrees ATDC of the exhaust stroke, which has them coming on well before the intake valve has closed for the previous firing event.
From reading through this entire thread, I've seen several references to needing to start the injector after the intake valve has closed. At first blush It seems as if this wasn't a concern to GM. In fact, even if using the 0.050 point stock LS1 cams "close" at 123 degrees, the number of degrees available for the injectors to wait to turn on is 183 degrees. For the stock parameters, an injector duty cycle larger than 75% pushes the start of the injector time prior to the 0.050 closing position of the intake valve's prior cycle. Maybe GM sized the injectors so they wouldn't go above a 75% duty cycle in a stock configuration?
If however one considers the boundary point to be the limit to where to start the injectors, the normal value then would seem to correlate with the maximum time "on" time for the injectors. Dividing the stock 5.55 value by 8 gives a max duty cycle of 69% before the "boundary" is violated. Did GM size the stock injectors to not go beyond a 69% duty cycle?
S2H, aka soundengineer, has posted dyno test results that indicate that for reasonably sized injectors, the EOIT has little impact on WOT performance - EOIT for reasonably sized injectors does affect low throttle and idle performance however. Obvious conclusion therefore is unless one is running oversized injectors the focus should be on idle performance which from S2H's results indicates EOIT should be between the intake valve's 0.004 and 0.050 positions.
Given that the low throttle / idle performance seems to be optimized when the injector is firing when the intake valve is cracked, and therefore there is exhaust gas flowing out of the cylinder through the intake valve, I will postulate that the hot exhaust gases combined with the injectors' spray optimizes the fuel vaporization and minimizes the amount of unburned fuel that will eventually find its way out of the exhaust pipe.
Last edited by Hoosier Tuner; 02-13-2019 at 03:54 PM.
gm sprays onto closed intake valves because they are hot and help flash any liquid fuel into vapor before entering the cylinder. the whole chasing the valve deal comes into play when you consider the amount of scavenge that a large cam has compared to the stock cam. the idea with a big cam is to spray it at the last possible minute to still allow a proper mixture because the scavenge is going to pull fuel air mix straight into the exhaust before that valve closes.
the small cam spray on closed valve increases emissions and cylinder efficiency when scavenge effects are not so dramatic.
2000 Ford Mustang - Top Sportsman
Scavenging is not efficient at low rpm with a big cam. Reversion pushes exhaust gases out of the cylinder through the intake valve in the early stages of its opening, which pushes any vaporized fuel backwards in the intake port. Since the scavenging at low rpms is inefficient, providing what some have described as natural EGR, a fair amount of these gases will not be exhausted on this cycle. They will get sucked back into the cylinder, along with the fuel that was sprayed before and during the reversion. There is no need to spray at the last moment to prevent unburnt fuel from going out the exhaust. This is why S2H, soundengineer, found that moving the EOIT beyond the neighborhood of the specified valve opening points stopped improving idle. As far as why that point was better than having EOIT at 60 degrees before the TDC exhaust closing intake opening point, I would suspect that vaporized fuel is condensing on the relatively cold walls of the head's intake port. At the relatively low airflow velocities during idle I further suspect that this condensed fuel is not vaporized and a good portion of it doesn't get burnt and ends up going out the exhaust.
Reply With Quote