Opening time: the overlooked constant

[sebch]

Hi all,

After some tuning and getting some good results, I thought it might be interesting to read again the tuning page of the Megamanual to see if I still was in the move and not too much in the routine, forgetting sth important.

And I actually think I was overlooking sth: he INJ opening time story caught my eyes, as I haven't adressed it before before . Suddently the question came: now what, my patiently tuned VE table could be totally out?

When I first read through the instructions, I had assumed the opening time was not so much important, initially thinking "hey, from what it says, the "typical opening time" is 1.0 ms, so mine must be the same". Now I've read several posts and the section about PWM where this is detailed, and I realize I should actuall look into that. The engine runs fine, but I still have some issues as far as acceleration from idle to cruise is concerned, so it could very well be that the opening time is wrong.

Now, from what I read, there's nowhere I could find simple and precise advice on how to tune the open time constant. There's sth about oscilloscopes, a tchnology I don't have acces to, and I I'm looking for a seat of the pants solution if it does exist. Can anyone point me to some resource (post I haven't found with the search engine) or describe the procedure?

Thanks and sorry for the long post, I get a bit enthusiastic, at times! Séb

[jonfx4com]

Seb,

I have a good solution for this. Say you have the engine running 2 squirts and it it idled ok, move it to 4 squirts and if it goes rich the injector opening time is set too short, if it goes lean the opening time is set too high. once you can go from 1 squirt to 2 or 4 you know the opeing time is set accurately. The logic is that if you have it wrong the error is doubled as you double the number of squirts. Worked for me!

You should be able to move the number of squirts around as much as you like if the opening time is set right. Mine worked out at 0.97ms.

It gets much less critical as you move to higher VE numbers as 0.2 ms out at an idle of 1.5 ms matters more that 0.2 ms out at an opeing time of 10ms.

I have tried a few injectors and it has always been between .92 and 1.1 for me.

Jon

[Keithg]

It'd be nice if this were in the MegaManual as a note when this is discussed.

KeithG
Posted by email.

[2screwed]

Message #23389 on the old Yahoo list also has some good info on how to do this.

From tbitz:

Thought I'd pass on this info.

Here is how I measured the injector open time using the megasquirt
ECU. Firts the engine was brought up to temperature so no
enrichements were active. Accel enrichements were turned off and EGO
correction was bumped up to 100%.

My ECU was setup to run 2 squirts. while stopped I set the engine rpm
to 2500rpm and let the EGO corection do it's stuff. I monitored the
Pulse width when the crossing occured. This happened at 2.9ms pulse
width.

Then I repeated with the squirts set to 3 squirts per cycle. At
2500rpm I again read the pulse width when the O2 crossed. This time I
got 2.2ms.

I repeated again with 6 squirts per cycle (note I run a 6 cylinder
engine, some combinations were not allowed). This time I got 1.5ms.

Since in every case the rpm and engine load are exactly the same I
figured the fuel requirement should be exactly the same for all three
cases?

With this info and a little math, I could then figure out the actual
injector opening time.

The equation I used is:

Open time = [numOfSquirt(a)*pulseWidth(a) - numOfSquirt(b)*pulseWidth
(b)]/ [numOfSquirt(a) - numOfSquirt(b)]

here is the data from my example using 3 and 2 squirts:

open time = [3*2.2 - 2*2.9]/[3-2] = 0.8ms

here is the data from my example using 3 and 6 squirts:

open time = [6*1.5 - 3*2.2]/[6-3] = 0.8ms


The fact that both cases gave me the same injector open time
validates the theory. Note if you run this and your injector open
time is more/less than what you have now, you will have to re-adjust
your VE table to account for the new injector open time.

One might say, who cares if your injector time is not right on. The
VE table factors that in during tuning. VE will correct for the error
in injector open time, however when the engine is run during
different conditions (temperature, pressure) the correcting factors
will be skewed, because gammeE only scales VE and won't scale the
time when MS is dumping fuel during the errored open time. For all
the correction factors to work properly the injector open time must
be right on.

Please comment....

Tony

[sebch]

OK folks, I tried both methods and here are some results:

jonfx4com's technique didn't meet much success on my setupa as there was no way I could simply make the engine idle when altering the number of squirts. Looks like my opening time is way off what it should be. I'll go back to that later, as a test to validate my value.

This morning I tried tbitz method and when calculating from the datalogs, I get an opening time of 0.6 ms, which I think is real low, esp as my injectors are high impedance ones (at least they are advertised as such on several injectors lists I got on this board...). I expected the opening time of my units to be 1.2 or so, but I'll give it a shot and use jonfx4com's method to validate, as i said. As I have a straight four, I can't use as many squirting combinations as with a V6 or 8, though, and the error factor tends to be a little higher. But I'll try again with alternating and simultaneous squirts and see how it goes.

Have any of you ever used opening times as low as 0.6 ms??

Have a good day, Séb

[snillet]

OK folks, I tried both methods and here are some results:

jonfx4com's technique didn't meet much success on my setupa as there was no way I could simply make the engine idle when altering the number of squirts. Looks like my opening time is way off what it should be. I'll go back to that later, as a test to validate my value.

This morning I tried tbitz method and when calculating from the datalogs, I get an opening time of 0.6 ms, which I think is real low, esp as my injectors are high impedance ones (at least they are advertised as such on several injectors lists I got on this board...). I expected the opening time of my units to be 1.2 or so, but I'll give it a shot and use jonfx4com's method to validate, as i said. As I have a straight four, I can't use as many squirting combinations as with a V6 or 8, though, and the error factor tends to be a little higher. But I'll try again with alternating and simultaneous squirts and see how it goes.

Have any of you ever used opening times as low as 0.6 ms??

Have a good day, Séb

I´m about to try this out today, since i most likely have shorter openingtime on my TBI injector.

Btw, what´s wrong with Volvo 544 and 121 ?

[sebch]

Btw, what´s wrong with Volvo 544 and 121 ?

Nothing, pal, nothing

Actually, my volvo originally was (and for the cops officially still is...) a 121. You're right to point it out, I'll change my signature at once!!

The question is: should I include the 140's?? The're getting wanted!

And yes - please keep us posted on how you're doing with your opening time!!

Cheers, Séb

[snillet]

Yes ,i´ve come to a conclusion here, it´s defenitely on the right track for me.

First a little "thinking" that i´ve doned.

The "average" 1ms openingtime is, i guess, for the most common type of injections namely high pressure ones, operating at around 2.5 to 4bar. Those injectors are somewhat more "tough" mechanically to open for the coil in them so the 1ms opening time is probably correct.

The injector in my TBI works at 0.85bar which is FAR less. So my theory is that it actually ISopening quicker then the high pressure types.

Namely i discovered that when i shorted down the opening time to 0.5ms i got rid of my weird values in the idle part of the fuel map and the other values was more or less ok already, the 0.5ms change does not have a very big impact on the values at cruise speed.

So this solved my problem, i´ll have to continue to do some math on this with the formulas mentioned above here and see if i can get closer to the truth then what i am now.

But, changing the opening time seem to be the correct way to go for me.

About the Volvos, yes the 140:s are beginning to "fade" in numbers for real even here in sweden. One model that people has realized that it´s meant for saving is the 164 though, seems like nowadays there are more of the 164:s then the 140:s here A friend of mine has a 1962 121 that he inherited from his grandfather, the only thing that can kill them is corrosion

[sebch]

Hi,
I'vedone some math myself and I havec ome to the conclusion that my opening time must be around 0.6 ms. I typed this in and the idle is A LOT more stable. Also I get very good idle and off-idle running with all the squirt patterns the 4 cyl will run on (2 and 4) without O2 getting lean or rich.

Now 0.6 seems a little low for high impedance injectors operating at 3+ bars. What do you think?$

Long life to the 140's, I like their front end much more than the 164!!

Cheers, Séb

[racingmini_mtl]

Séb,

You have to realize that the opening time is actually a parameter to get the effective injector pulse width and it includes both the injector opening time AND the injector closing time, which is something that is not mentioned frequently. This parameter compensates for the time the injector is not injecting fuel when it is opening and for the time the injector is still injecting fuel after the commanded pulse width.

This means that the parameter that you enter as the injector opening time is actually the opening time minus the closing time. So if for whatever reason, an injector continues to inject fuel for a significant amount of time after the commanded pulse width, the value that you will have to enter will seem abnormally low. I'm not saying that this is what you are seeing but it is one possibility.

Regards,
Jean