Injector timing
Moderators: jsmcortina, muythaibxr
Injector timing
I was just looking into the injector timing setting in MS3. An automotive engineer that is heavily involved in EFI for OEM's and posts on another forum, said that it usually takes about 8 ms for the fuel that is injected to reach the valve and to vaporize. When recalculating this to the number of degrees when the injector stops injecting this is what I arrived at :
Re: Injector timing
From 1000 to 7000 rpm you have a timing shift of roughly 260 degrees. Seems like a lot. How did you calculate it?
Rover Vitesse 3.5
MS3 full-sequential
36-1 triggerwheel
Direct coil control
Innovate LC1 wideband lambda sensor
MS3 full-sequential
36-1 triggerwheel
Direct coil control
Innovate LC1 wideband lambda sensor
Re: Injector timing
1. RPM/60 = rpm per secsd1nl wrote:From 1000 to 7000 rpm you have a timing shift of roughly 260 degrees. Seems like a lot. How did you calculate it?
2. rpm/s x 360° = degree/sec
3. result under 2 /1000 to get degree/ms
4. result under 3 X 8 (since the time is apparently 8 ms delay for the charge to reach the valve)
5. Result under 4 + the initial 360° before ingition event.
Re: Injector timing
So did you get a chance to try that out to see if the car runs any better?
Re: Injector timing
I went the other way with my injector MAP.
I remove degrees as RPM Increases.
Anyone have any correct info on this???
I remove degrees as RPM Increases.
Anyone have any correct info on this???
1968 Dodge Charger, 446ci 513HP/542ftlbs,Full sequential fuel/Spark using MS3X. Jeep CPS, Ford Crank sensor, 36-1 wheel, Coil near plug with D-585 ls2 coil packs. Edelbrock pro-flo intake, 600cc injectors, GM MAP, TPS, IACV, CTS, MAT. 92mm Throttle Body.
Re: Injector timing
That sounds more logical.
Re: Injector timing
I got lost at step 3. After step 2 at 900 RPM equals 5400 degree/ms. What do you do with the 5400?Yves wrote:1. RPM/60 = rpm per secsd1nl wrote:From 1000 to 7000 rpm you have a timing shift of roughly 260 degrees. Seems like a lot. How did you calculate it?
2. rpm/s x 360° = degree/sec
3. result under 2 /1000 to get degree/ms
4. result under 3 X 8 (since the time is apparently 8 ms delay for the charge to reach the valve)
5. Result under 4 + the initial 360° before ingition event.
Re: Injector timing
5400 degrees per second is 5.4 degrees per ms
-
- Super MS/Extra'er
- Posts: 17507
- Joined: Thu Apr 16, 2009 8:08 pm
Re: Injector timing
Here's a video Dieselgeek put together of an engine running on a dyno with correctly dialed in injector timing - correctly dialed in for this motor, at least.redmist wrote:I went the other way with my injector MAP.
I remove degrees as RPM Increases.
Anyone have any correct info on this???
https://www.youtube.com/watch?v=6kQGp-fHuFc
I believe he was using oversized injectors to maximize the effect of fine tuning the timing. Notice how timing advances with RPM.
Matt Cramer -1966 Dodge Dart slant six running on MS3X
Re: Injector timing
Advancing in this sense is that it takes place earlier (futher away from the point of ignition - this is another debate : does the ignition advance have to be substracted or not).
The way I understood it : it's fuel vaporization that is import with regards to the end point of injection. The hot intake valve plays a large roll in this since the fuel vaporizes on it. There is also a delay before all of the fuel reaches the valve (sort of an X-Tau factor) which is dependant on injector intercept angle, spray pattern, wall wetting etc. Like I said, the engineer told me that it takes roughly 8 ms overall for this fuel to reach the valve in a stock injector position. This is how I arrived at the tabel above.
To note is that the position of my injectors due to space restrictions is plain awfull (perpendicular to the runner) so a lot of wall wetting takes place and it takes longer for the fuel to reach the valve.
A point I'm not so sure about is the effect of airspeed at higher rpms on the delay. I see that Scott started around 3000 rpm with advancing the timing and then stopped at higher rpm.
This is highly debatable since each engine is different of course and I do not have a dyno to tell me which way to go.
The way I understood it : it's fuel vaporization that is import with regards to the end point of injection. The hot intake valve plays a large roll in this since the fuel vaporizes on it. There is also a delay before all of the fuel reaches the valve (sort of an X-Tau factor) which is dependant on injector intercept angle, spray pattern, wall wetting etc. Like I said, the engineer told me that it takes roughly 8 ms overall for this fuel to reach the valve in a stock injector position. This is how I arrived at the tabel above.
To note is that the position of my injectors due to space restrictions is plain awfull (perpendicular to the runner) so a lot of wall wetting takes place and it takes longer for the fuel to reach the valve.
A point I'm not so sure about is the effect of airspeed at higher rpms on the delay. I see that Scott started around 3000 rpm with advancing the timing and then stopped at higher rpm.
This is highly debatable since each engine is different of course and I do not have a dyno to tell me which way to go.
Re: Injector timing
Looks like he is injecting fuel when the intake valve is open. Most people inject fuel just before the intake valve opens. The pro's and cons are still not clear to me.Matt Cramer wrote:Here's a video Dieselgeek put together of an engine running on a dyno with correctly dialed in injector timing - correctly dialed in for this motor, at least.redmist wrote:I went the other way with my injector MAP.
I remove degrees as RPM Increases.
Anyone have any correct info on this???
https://www.youtube.com/watch?v=6kQGp-fHuFc
I believe he was using oversized injectors to maximize the effect of fine tuning the timing. Notice how timing advances with RPM.
Rover Vitesse 3.5
MS3 full-sequential
36-1 triggerwheel
Direct coil control
Innovate LC1 wideband lambda sensor
MS3 full-sequential
36-1 triggerwheel
Direct coil control
Innovate LC1 wideband lambda sensor
Re: Injector timing
Hmmm...
Yeah I am confused here. Mine goes the other way.
I guess it makes sense that if it's end of squirt, as RPM increases, and PW increases, you are going to want to tune end of squirt so that it's not still injecting as the intake valve is closing. However you don't want it doing it to early or you are dumping fuel out of exhaust.
There has got to be a formula (I don't do "Math") that can be used with your cam specs to figure out exactly where you want it???
Yeah I am confused here. Mine goes the other way.
I guess it makes sense that if it's end of squirt, as RPM increases, and PW increases, you are going to want to tune end of squirt so that it's not still injecting as the intake valve is closing. However you don't want it doing it to early or you are dumping fuel out of exhaust.
There has got to be a formula (I don't do "Math") that can be used with your cam specs to figure out exactly where you want it???
1968 Dodge Charger, 446ci 513HP/542ftlbs,Full sequential fuel/Spark using MS3X. Jeep CPS, Ford Crank sensor, 36-1 wheel, Coil near plug with D-585 ls2 coil packs. Edelbrock pro-flo intake, 600cc injectors, GM MAP, TPS, IACV, CTS, MAT. 92mm Throttle Body.
Re: Injector timing
As the rpm's go up, you need to have the end of pulse a littlebit earlier. But your table is the other way around. You have to increase the degrees a rpm's go up.redmist wrote:Hmmm...
Yeah I am confused here. Mine goes the other way.
I guess it makes sense that if it's end of squirt, as RPM increases, and PW increases, you are going to want to tune end of squirt so that it's not still injecting as the intake valve is closing. However you don't want it doing it to early or you are dumping fuel out of exhaust.
There has got to be a formula (I don't do "Math") that can be used with your cam specs to figure out exactly where you want it???
Rover Vitesse 3.5
MS3 full-sequential
36-1 triggerwheel
Direct coil control
Innovate LC1 wideband lambda sensor
MS3 full-sequential
36-1 triggerwheel
Direct coil control
Innovate LC1 wideband lambda sensor
Re: Injector timing
It depends on what you have speced it at : end of squirt, middle of squirt, beginning of squirt, but since the squirt has a duration (as in time related) the time to complete the squirt remains the same for a given load, but with increasing rpms more degrees are turned by the engine, so you have to increase the timing as rpm goes up...in theory that is.
Re: Injector timing
Yves wrote:I was just looking into the injector timing setting in MS3. An automotive engineer that is heavily involved in EFI for OEM's and posts on another forum, said that it usually takes about 8 ms for the fuel that is injected to reach the valve and to vaporize. When recalculating this to the number of degrees when the injector stops injecting this is what I arrived at :
I'm wondering if I should adjust my table. I have my on default (360)? What is your combination?
Re: Injector timing
a short stroke long rodded sbc with ITB's and an agressive cam.
The 8ms that I gave above is a supposed to be a delay that the OEM's seem to use, so this entails the stock injector position spraying at the back of the valve. If yours is different from that it may need more delay imo. For instance if your injectors are placed more upstream, not pointing at the valve etc.
Some even say that it might be necessary to limit the duty cycle so that you only spray during the time the valve is not open. If that is true, it means that most of us need a lot larger injectors with the associated problems and so on.
The 8ms that I gave above is a supposed to be a delay that the OEM's seem to use, so this entails the stock injector position spraying at the back of the valve. If yours is different from that it may need more delay imo. For instance if your injectors are placed more upstream, not pointing at the valve etc.
Some even say that it might be necessary to limit the duty cycle so that you only spray during the time the valve is not open. If that is true, it means that most of us need a lot larger injectors with the associated problems and so on.
-
- Super MS/Extra'er
- Posts: 1341
- Joined: Mon Sep 05, 2011 1:20 pm
Re: Injector timing
I would think this should be result under 3 + 8, the time lag is a constant offset, not a multiplier. Your way makes the timing degree spread much broader than seems rational.4. result under 3 X 8 (since the time is apparently 8 ms delay for the charge to reach the valve)
Graduate of EFI University.
I build, repair, install and tune Megasquirt systems in North Dakota and beyond!
I build, repair, install and tune Megasquirt systems in North Dakota and beyond!
Re: Injector timing
The result under 3 is not in ms, it's in degree per millisecond which in itself is dependant on rpm. So if the delay is a constant 8 ms, you would need to know how many degrees the engine turns in 8 ms so you can make up your table which is in degrees before spark. Therefor I multiplied the result under 3 which is in degree/ms X 8 ms to use as an adder.elutionsdesign wrote:I would think this should be result under 3 + 8, the time lag is a constant offset, not a multiplier. Your way makes the timing degree spread much broader than seems rational.4. result under 3 X 8 (since the time is apparently 8 ms delay for the charge to reach the valve)
An example : engine turning at 3000 rpm :
1) 3000 rpm / 60 = 50 rotations per second
2) 50 rotations per second X 360° per rotation = 18000 degrees per second
3) 18000 degrees per second / 1000 = 18 degrees per millisecond of rotational speed
4) For 8ms delay, this means that during those 8 ms the engine turns 8 X 18 = 144 degrees.
5) the point at which in all cases the injection needs to end is 360° before spark, so the injection needs to begin 360 + 144 = 504° before spark when taking into account the delay for the charge to reach the valve.
The 8 ms is debatable since it depends on the actual location and position of the injector. Mine is close to the port but is perpendicular to airflow, so a lot of fuel ends up on the walls and has to return into the airstream + will run to the valve over the floor. Could be that my engine needs a lot more delay than 8 ms.
Now I can imagine that some would want to spray into an open valve, in which case the delay will be shorter as the airflow will make the charge reach the valve sooner.
-
- Master MS/Extra'er
- Posts: 682
- Joined: Tue Jul 18, 2006 3:12 pm
Re: Injector timing
Here are some points that came from a thread here on MS forums (I can't find it right now) :
- You want all the fuel to be injected before the opening of the intake valve.
- You want the end of squirt to end about 10° earlier for each 1000 RPM (the fuel still take about the time to evaporate on a hot intake valve, so you need to compensate for that).
- At mid to high RPM, injection timing doesn't have a lot of impact, so do not spend a lot of time to fine tune your injection timing table.
- The first step is to play with the injection timing in a single cell at mid load and crusing RPM. Find the spot where the AFR is at its richest. This is your ideal injection timing value. Extrapolate in the X axis as per point #2.
- After injection timing tuning, you will need to retune your VE table because it will be a little too rich (that's an indication that you did a good job with your injection timing tuning : combutsion is more efficient because it require less fuel to stay at the same load/RPM).
Sam
- You want all the fuel to be injected before the opening of the intake valve.
- You want the end of squirt to end about 10° earlier for each 1000 RPM (the fuel still take about the time to evaporate on a hot intake valve, so you need to compensate for that).
- At mid to high RPM, injection timing doesn't have a lot of impact, so do not spend a lot of time to fine tune your injection timing table.
- The first step is to play with the injection timing in a single cell at mid load and crusing RPM. Find the spot where the AFR is at its richest. This is your ideal injection timing value. Extrapolate in the X axis as per point #2.
- After injection timing tuning, you will need to retune your VE table because it will be a little too rich (that's an indication that you did a good job with your injection timing tuning : combutsion is more efficient because it require less fuel to stay at the same load/RPM).
Sam
Re: Injector timing
Thanks for posting this, I've been trying to figure out how to dial in my injection timing and I like your method. We can go on and on about our flawed theories of internal combustion or what our buddy said when he was four beers deep, but as all of our installations vary so widely, I think it's best to start simple and tune like you said - the engine will tell you where injection timing should be.pit_celica wrote:- After injection timing tuning, you will need to retune your VE table because it will be a little too rich (that's an indication that you did a good job with your injection timing tuning : combutsion is more efficient because it require less fuel to stay at the same load/RPM).