Injector Opening Time (Academic Queston)
08-10-2011, 02:58 PM
#1
Senior Member
Thread Starter
iTrader: (10)
Join Date: Apr 2007
Location: Ohio
Posts: 782
Total Cats: 4
Injector Opening Time (Academic Queston)
I say "academic" because I'm just curious and not really trying to solve any particular problem. First, I know the post I'm about to reference is discussing MSII and I only have MSI to play with, so it may be moot for all I know, though I don't think it should make a difference in this case.
I was reading this old thread:
https://www.miataturbo.net/showthrea...r+dead+opening
I was reading this old thread:
https://www.miataturbo.net/showthrea...r+dead+opening
In particular, this part of Abe's post:
......
The idle issue:
It's very likely injector settings.
Let's say you turn on the injector at time 0. It doesnt' start to move for 300 us, and doesn't spray significant fuel until 1200 us. I'm picking numbers out of a hat to illustrate.
If you have open time set to 1500 us (1.5ms) then at idle when it tries to give you 500 us of fuel, it actually opens the valve for 500 + 1500 = 2000 us. The computer thinks you got 500 us, but really the valve opened in 1200, so you got the balance in fuel, or 800 us.
This is all great, and it'll idle and you'll never know it's wrong. Then you ask for 5% more fuel (slight throttle). The car goes "Great. 500 * 1.05 = 525 us. Open time of 1500 us, and that's 2025 us."
But what really happened? Before you were getting 700 us of fuel, now you're getting 2025-1200 = 825. Your "5%" increase was really only 3%.
Up top, with 9 ms of open time, you don't care about 200 us, you just tune it out. But right off idle it's very important to understand the opening time.
The fix: In the first settings window, there's a squirts-per-cycle, and I think your choices are 2 or 4. Since the open time is constant, switching from 2 squirts to 4 will double the error in opening time. So if you're idling at 14.6:1, and you go to 4 squirts and you richen up to 14.0:1, then you're opening time is too long. The math:
2000 ms (1500 + 500) becomes 2x (1500 + 250), so the difference between 1500 and 1200 shows up. Basically, go back and forth from 2 to 4 squirts and tweak opening times until there's no change, then go retune your low end.
I thought this was interesting and decided to try "the fix" on my car. As soon as I switched to 4 squirts, it dies - flat on it's face. In order to even expect that it would run on 4 squirts, wouldn't the Req. Fuel need to be changed?
......
The idle issue:
It's very likely injector settings.
Let's say you turn on the injector at time 0. It doesnt' start to move for 300 us, and doesn't spray significant fuel until 1200 us. I'm picking numbers out of a hat to illustrate.
If you have open time set to 1500 us (1.5ms) then at idle when it tries to give you 500 us of fuel, it actually opens the valve for 500 + 1500 = 2000 us. The computer thinks you got 500 us, but really the valve opened in 1200, so you got the balance in fuel, or 800 us.
This is all great, and it'll idle and you'll never know it's wrong. Then you ask for 5% more fuel (slight throttle). The car goes "Great. 500 * 1.05 = 525 us. Open time of 1500 us, and that's 2025 us."
But what really happened? Before you were getting 700 us of fuel, now you're getting 2025-1200 = 825. Your "5%" increase was really only 3%.
Up top, with 9 ms of open time, you don't care about 200 us, you just tune it out. But right off idle it's very important to understand the opening time.
The fix: In the first settings window, there's a squirts-per-cycle, and I think your choices are 2 or 4. Since the open time is constant, switching from 2 squirts to 4 will double the error in opening time. So if you're idling at 14.6:1, and you go to 4 squirts and you richen up to 14.0:1, then you're opening time is too long. The math:
2000 ms (1500 + 500) becomes 2x (1500 + 250), so the difference between 1500 and 1200 shows up. Basically, go back and forth from 2 to 4 squirts and tweak opening times until there's no change, then go retune your low end.
Reply
0
0
08-10-2011, 03:29 PM
#2
Elite Member
iTrader: (1)
Join Date: Jun 2006
Location: Warrington/Birmingham
Posts: 2,642
Total Cats: 42
No, the opening time.
That's how you work out the correct opening time for your injectors, switch between 2 and 4 squirts, when the AFR stays the same you have the correct opening time.
That's how you work out the correct opening time for your injectors, switch between 2 and 4 squirts, when the AFR stays the same you have the correct opening time.
Reply
0
0
08-10-2011, 03:41 PM
#3
Senior Member
Thread Starter
iTrader: (10)
Join Date: Apr 2007
Location: Ohio
Posts: 782
Total Cats: 4
Since 4 squirts causes my car to die, I can assume my settings are not good despite the fact that it runs fine on 2 squirts. So maybe AE is allowing it to work well enough off-idle that I don't notice, I'm guessing.
Reply
0
0
08-10-2011, 03:52 PM
#4
Senior Member
Thread Starter
iTrader: (10)
Join Date: Apr 2007
Location: Ohio
Posts: 782
Total Cats: 4
FWIW, I had my opening time set to 1.2 per Brain's recommendation and it seemed fine. Tried 1ms and it ran like crap. 1.25 works too.
If it's stalling on 4 squirts, does it make sense that the opening time would need to be longer? Shorter certainly doesn't seem to be good.
As I said, I'm not having any noticeable issues, but I guess that doesn't mean everything is as it should be I suppose. I'm just playing around with stuff to learn what everything does.
If it's stalling on 4 squirts, does it make sense that the opening time would need to be longer? Shorter certainly doesn't seem to be good.
As I said, I'm not having any noticeable issues, but I guess that doesn't mean everything is as it should be I suppose. I'm just playing around with stuff to learn what everything does.
Reply
0
0
08-10-2011, 05:20 PM
#5
Senior Member
iTrader: (4)
Join Date: Jan 2008
Location: Falls Church, VA
Posts: 1,361
Total Cats: 17
1. If it ain't broke, don't "fix" it.
2. It's perfectly acceptable to use the same opening time everybody else uses with RX-7 550s, which is 1.2 ms.
3. This thread shows a more mathematical way to go about testing it on your particular car, as opposed to taking somebody else's number on faith, or switching between 2 and 4 squirts while dicking with opening time.
https://www.miataturbo.net/ecus-tuning-54/how-find-your-injector-dead-time-56061/
On page 3 somebody has a plot for RX-7 550s. You'll notice that it's not entirely linear. Opening time makes the most difference (as a % of pulse width) at low pulse widths, like idle and off-idle, so if anything you might want to play with increasing the opening time based on that plot. I believe this jives with you stalling on 4 squirts. Of course, changing the opening time will require retuning your whole VE table.
3a. Do not change reqfuel in an attempt to compensate for changes in injector opening time. It isn't academically correct.
4. RX-7 550s are somewhat shitty and we use them primarily because they are plug-and-play and cheap. There are injectors with lower dead times and better spray patterns that will idle better, but they cost 3-5x as much.
5. Your car will idle and run better if you upgrade to MSII, or better yet go with MS3/X and run sequential. Then you can play around with all kinds of stuff to quench your thirst for learning. There's EAE, injector timing, lots of stuff to optimize with real gains in fuel mileage and driveability.
2. It's perfectly acceptable to use the same opening time everybody else uses with RX-7 550s, which is 1.2 ms.
3. This thread shows a more mathematical way to go about testing it on your particular car, as opposed to taking somebody else's number on faith, or switching between 2 and 4 squirts while dicking with opening time.
https://www.miataturbo.net/ecus-tuning-54/how-find-your-injector-dead-time-56061/
On page 3 somebody has a plot for RX-7 550s. You'll notice that it's not entirely linear. Opening time makes the most difference (as a % of pulse width) at low pulse widths, like idle and off-idle, so if anything you might want to play with increasing the opening time based on that plot. I believe this jives with you stalling on 4 squirts. Of course, changing the opening time will require retuning your whole VE table.
3a. Do not change reqfuel in an attempt to compensate for changes in injector opening time. It isn't academically correct.
4. RX-7 550s are somewhat shitty and we use them primarily because they are plug-and-play and cheap. There are injectors with lower dead times and better spray patterns that will idle better, but they cost 3-5x as much.
5. Your car will idle and run better if you upgrade to MSII, or better yet go with MS3/X and run sequential. Then you can play around with all kinds of stuff to quench your thirst for learning. There's EAE, injector timing, lots of stuff to optimize with real gains in fuel mileage and driveability.
Reply
0
0
08-10-2011, 05:34 PM
#6
Senior Member
Thread Starter
iTrader: (10)
Join Date: Apr 2007
Location: Ohio
Posts: 782
Total Cats: 4
1. If it ain't broke, don't "fix" it.
2. It's perfectly acceptable to use the same opening time everybody else uses with RX-7 550s, which is 1.2 ms.
3. This thread shows a more mathematical way to go about testing it on your particular car, as opposed to taking somebody else's number on faith, or switching between 2 and 4 squirts while dicking with opening time.
https://www.miataturbo.net/showthread.php?t=56061
On page 3 somebody has a plot for RX-7 550s. You'll notice that it's not entirely linear. Opening time makes the most difference (as a % of pulse width) at low pulse widths, like idle and off-idle, so if anything you might want to play with increasing the opening time based on that plot. I believe this jives with you stalling on 4 squirts. Of course, changing the opening time will require retuning your whole VE table.
3a. Do not change reqfuel in an attempt to compensate for changes in injector opening time. It isn't academically correct.
4. RX-7 550s are somewhat shitty and we use them primarily because they are plug-and-play and cheap. There are injectors with lower dead times and better spray patterns that will idle better, but they cost 3-5x as much.
5. Your car will idle and run better if you upgrade to MSII, or better yet go with MS3/X and run sequential. Then you can play around with all kinds of stuff to quench your thirst for learning. There's EAE, injector timing, lots of stuff to optimize with real gains in fuel mileage and driveability.
2. It's perfectly acceptable to use the same opening time everybody else uses with RX-7 550s, which is 1.2 ms.
3. This thread shows a more mathematical way to go about testing it on your particular car, as opposed to taking somebody else's number on faith, or switching between 2 and 4 squirts while dicking with opening time.
https://www.miataturbo.net/showthread.php?t=56061
On page 3 somebody has a plot for RX-7 550s. You'll notice that it's not entirely linear. Opening time makes the most difference (as a % of pulse width) at low pulse widths, like idle and off-idle, so if anything you might want to play with increasing the opening time based on that plot. I believe this jives with you stalling on 4 squirts. Of course, changing the opening time will require retuning your whole VE table.
3a. Do not change reqfuel in an attempt to compensate for changes in injector opening time. It isn't academically correct.
4. RX-7 550s are somewhat shitty and we use them primarily because they are plug-and-play and cheap. There are injectors with lower dead times and better spray patterns that will idle better, but they cost 3-5x as much.
5. Your car will idle and run better if you upgrade to MSII, or better yet go with MS3/X and run sequential. Then you can play around with all kinds of stuff to quench your thirst for learning. There's EAE, injector timing, lots of stuff to optimize with real gains in fuel mileage and driveability.
I really wasn't trying to fix it, and I'm inclined to leave it the hell alone since my VE table is getting pretty decent. It was just one of those things that I read and realized that I never thought about it. I just used the 1.2ms because "everybody else does".
I know the RX7 550s are kinda shitty. I had them all spiffed up and they will do until I can stomach spending more money on this thing. I have a couple parts I actually NEED and then I plan to do the MS3/x thing and go back to sequential. Then Xidas.......
...actually dyno in the spring, THEN shocks!
Reply
0
0
08-11-2011, 08:15 AM
#7
Elite Member
iTrader: (1)
Join Date: Jun 2006
Location: Warrington/Birmingham
Posts: 2,642
Total Cats: 42
So when setting the opening time you'll also HAVE to adjust the VE table to correct.
2. It's perfectly acceptable to use the same opening time everybody else uses with RX-7 550s, which is 1.2 ms.
3. This thread shows a more mathematical way to go about testing it on your particular car, as opposed to taking somebody else's number on faith, or switching between 2 and 4 squirts while dicking with opening time.
https://www.miataturbo.net/showthread.php?t=56061
On page 3 somebody has a plot for RX-7 550s. You'll notice that it's not entirely linear. Opening time makes the most difference (as a % of pulse width) at low pulse widths, like idle and off-idle, so if anything you might want to play with increasing the opening time based on that plot. I believe this jives with you stalling on 4 squirts. Of course, changing the opening time will require retuning your whole VE table.
3a. Do not change reqfuel in an attempt to compensate for changes in injector opening time. It isn't academically correct.
https://www.miataturbo.net/showthread.php?t=56061
On page 3 somebody has a plot for RX-7 550s. You'll notice that it's not entirely linear. Opening time makes the most difference (as a % of pulse width) at low pulse widths, like idle and off-idle, so if anything you might want to play with increasing the opening time based on that plot. I believe this jives with you stalling on 4 squirts. Of course, changing the opening time will require retuning your whole VE table.
3a. Do not change reqfuel in an attempt to compensate for changes in injector opening time. It isn't academically correct.
Without the correct opening time it's hard to get a consistent AFR during small PW engine loads i.e. idle, particularly when the environment/temp etc changes.
Reply
0
0
08-11-2011, 08:56 AM
#8
Senior Member
Thread Starter
iTrader: (10)
Join Date: Apr 2007
Location: Ohio
Posts: 782
Total Cats: 4
Well it will. When reduce the opening time you're telling the ECU that the injector doesn't need to be open for as long. If the opening time is actually 1ms, you've just removed 0.2ms of fuel that you've tuned for. (and Vis-versa obviously)
So when setting the opening time you'll also HAVE to adjust the VE table to correct.
So when setting the opening time you'll also HAVE to adjust the VE table to correct.
So really, you would need to adjust the opening time to where it runs on 2 or 4 squirts, then back off the VE values to get the AFR right, then adjust the opening time back down? Then adjust the VE values again? It seems like a vicious cycle. How do you really hone in on it this way?
I haven't looked at the mathematical method yet ( I was thinking of leaving mine as it is....but now I'm wondering if I shouldn't do it "right".) but it seems to me that if there's a way to work this out to where it's really close using theory, that would be better.
I'm sure I'm over-thinking all of it...as that is my nature.
EDIT:
OK, just looked at that thread on dead time. Lots of good info.
Last edited by soloracer; 08-11-2011 at 09:07 AM.
Reply
0
0
08-11-2011, 10:12 AM
#9
Elite Member
iTrader: (1)
Join Date: Jun 2006
Location: Warrington/Birmingham
Posts: 2,642
Total Cats: 42
It's nothing to do with the AFR values per-say, just get the car to idle at 2 and 4 squirts - whatever VE table values that may be - then adjust the deadtime so that when you switch between 2 and 4 squirts the AFR doesn't change.
Then tune the car for your desired idle AFR.
The idea is to double the amount of times the engine is squirting per cycle, but not effect the AFR (because with 4 squirts each individual squirt of fuel is 1/2 that of when running 2 squirts but the actual amount of fuel injected is the same as the dead time, i.e. is correct so no 'extra' fuel is added)
Reply
0
0
08-17-2011, 07:39 AM
#10
Senior Member
Thread Starter
iTrader: (10)
Join Date: Apr 2007
Location: Ohio
Posts: 782
Total Cats: 4
Spent about 15 minutes playing with this last weekend. I would never have believed how much better the car runs with an injector opening time of 1.25ms vs. 1.2. Went on a 3 hour drive last night and all the burbling it was doing at cruise is gone.
Thanks for all the input.
Thanks for all the input.
Reply
0
0
Thread
Thread Starter
Forum
Replies
Last Post
