This....
250/200s sooner than later.....
- Thread starter Lowdown89
- Start date
This....
at 3000 rpms if you were running somewhat aggressive timing of say 15 degrees BTDC you have that plus about 24 degrees after TDC for a total of about 40 degrees to make the fuel actually make power (excluding nitrous' magic properties of allowing power to be made even longer after TDC). Take an average injector delay time of about .7ms from the time it is commanded to the time you actually get fuel out of the nozzle and there is the most pulsewidth you can actually do some good with at 3000 rpms.
Matt,
That was exactly the post I was looking for.
It is this reason why I and others have said when you look at injector performance for making more power you need rate injectors by what they can do at 2.0 ms and less. We can make power @ 3K its when you get up into 3.5-4K+ we are limited. Also why all the research into finding ways to get the IDM to allow more PW for a standard setup was worthless. It gives you WAY more than you can ever use.
That was exactly the post I was looking for.
It is this reason why I and others have said when you look at injector performance for making more power you need rate injectors by what they can do at 2.0 ms and less. We can make power @ 3K its when you get up into 3.5-4K+ we are limited. Also why all the research into finding ways to get the IDM to allow more PW for a standard setup was worthless. It gives you WAY more than you can ever use.
Doesnt the amount of rotella your.gerotor can ultimately move.come into play at this.point? Or how much can be pushed through the injector rather?
If these inj/nozzle numbers were paper-realistic we'd be making a grand on 300% nozzles by now. Although what Bill did with this is pretty damn good and clean.
@lowdown-keep us posted.how.the trans works out. Im needing.to.get one too.
Last edited:
That's a different issue. Yes it plays out as you get higher but that was not what the question was that was being answered. It was about IDM limits compared to the mechanical ones. But the fact is what the injectors flow on the bench is very close to what they are flowing on the running engine. Now can we get a injector to flow faster and more if we can get the oil in and out past that itty bitty valve absolutely. It is the ultimate limit of the injector but we are still in the confines of the limits of the crank window time it is that the injectors will flow more in any given time.
Chvyrkr
Active member
- Joined
- May 18, 2011
- Messages
- 2,204
- Reaction score
- 0
Don't forget refill. That takes time as well, and even firing inside of IDM limits there comes a point where you're not injecting a full shot, as there wasn't enough time to refill a full shot.
Tim @ P.I.S.
Member
Ok Matt, what exactly would 40 degrees equal in ms/pw @ say 5k rpm?
And yes the commanded pw has a mechanical delay, but I never remove that figure from calculations because it is part of the equation. When I state 2.25 ms/pw that is including delay.
Tim @ P.I.S.
Member
It does give you plenty of time. It will allow 2.1 ms/pw @ 5k rpm. And at 2.1ms a 238/200 CAN still flow 215cc of fuel.
But no one is making power at 3k rpm. Not rising or holding on a graph at least.
Tim @ P.I.S.
Member
You have no idea how bad this problem is. Run a set of any brand injector on a bench at 1000rpm. Keep dropping pw until you just have enough time to empty your injector. Not use that same pw at 2k rpm, 3k rpm, ect. It is really sad what you will see.
Tim @ P.I.S.
Member
We will see very soon.
I an not understanding why you keep going back to what the IDM can allow for PW. What that IDM can allow is solely from the point of not allowing a overlapping signal. It has nothing at all to do with what the actual real world mechanical window of the engine is that we have to inject into in time The only reason I mentioned the IDM was to point out that is not the issue and never has been. Its very late or early I guess but I will see if I can find were I noted down the various mechanical windows converter to PW @ various rpms. I will post them up if Matt has not done so by then.
Last edited:
Lowdown89
New member
Feel free to send a set this way, and I will make sure they get a completely unbiased review
Bill, Matt, Dave, Charles, 907Dave etc anyone else that may know definitively the answer,
I have been going over my old chart and redoing a chart to more accurately compute the available PW. Matt thank you for all the phone time tonight. It is so helpful to bounce ideas off someone that has such a strong grasp on the subject matter at hand.
In doing this something came up that maybe just me as not doing any tuning myself has misunderstood. It has to do with the injector delay. Now as I understand it we have a minimum injector delay. This is basically the time from when the injector is told to fire until it actually start to fire and inject fuel. This has been found to be 0.7ms. Is this correct? Is this delay from the time the IDM sends the signal to energize the solenoid till the time where the injector actually starts to spray fuel out of the nozzles? Or is it only when the injector first starts the process to firing? I know things are happening fast but I recall Dave taking the time to post the series of events the injector goes thru before any fuel actually comes out of the nozzle and I just want to make sure this is part of that and if not how much time in ms are we talking about.
Here is why I am asking. I have started to work on a chart to break down just how much of a PW window we have at various RPMs going from 2000-5000rpms. I felt this could also be helpful for other to see. Also have it crossed with data from the 80% IDM values to show how this compares to what the limits are of the IDM signals.
Two things are giving me a little bit of hang up.
1: Injector delay with a static time of 0.7ms
2: SOI Advance BTDC and how corresponding increase with increases in RPMs
Starting with #1 is it correct that this is a fixed static delay and 0.7ms is accurate enough assumption of this delay?
Dave,
Does your modification to IDM effect this delay and if so how much have you found it does? To me if its 0.5ms or more its worth noting and adding into our factors in estimating our PW window. Any info you are willing to share would be greatly appreciated on this.
If anyone else has done testing on this with modded IDMs I am all ears on effects on delay or any effect is has on the injector speed vs com PW.
I assume this delay is part of the IDM signal and thus gets added in to commanded PW as it's part of the PW signal of the IDM
On to #2 issue: I was working off a figure of 15 degrees BTDC @ 3K rpm Based of that I figured a decline in advance as RPMs drop and increase as they increased. Worked out a rate of basically 5° for every 1000rpms in either direction. So basically with 15°@ 3K I worked up from 2K@10°, 2.5K @12.5°, 3K @15°, [email protected]°, etc up to 5K@25° I want to confirm that this sort of rate is correct and is there a limit to this. I found some very interesting associations when following a rate like this and factoring in time.
If anyone can help me with these and some insight on them that would be helpful. Once I have this I will feel more comfortable with the numbers I have come up with and will post up the chart for review and hopefully so good discussion.
I have been going over my old chart and redoing a chart to more accurately compute the available PW. Matt thank you for all the phone time tonight. It is so helpful to bounce ideas off someone that has such a strong grasp on the subject matter at hand.
In doing this something came up that maybe just me as not doing any tuning myself has misunderstood. It has to do with the injector delay. Now as I understand it we have a minimum injector delay. This is basically the time from when the injector is told to fire until it actually start to fire and inject fuel. This has been found to be 0.7ms. Is this correct? Is this delay from the time the IDM sends the signal to energize the solenoid till the time where the injector actually starts to spray fuel out of the nozzles? Or is it only when the injector first starts the process to firing? I know things are happening fast but I recall Dave taking the time to post the series of events the injector goes thru before any fuel actually comes out of the nozzle and I just want to make sure this is part of that and if not how much time in ms are we talking about.
Here is why I am asking. I have started to work on a chart to break down just how much of a PW window we have at various RPMs going from 2000-5000rpms. I felt this could also be helpful for other to see. Also have it crossed with data from the 80% IDM values to show how this compares to what the limits are of the IDM signals.
Two things are giving me a little bit of hang up.
1: Injector delay with a static time of 0.7ms
2: SOI Advance BTDC and how corresponding increase with increases in RPMs
Starting with #1 is it correct that this is a fixed static delay and 0.7ms is accurate enough assumption of this delay?
Dave,
Does your modification to IDM effect this delay and if so how much have you found it does? To me if its 0.5ms or more its worth noting and adding into our factors in estimating our PW window. Any info you are willing to share would be greatly appreciated on this.
If anyone else has done testing on this with modded IDMs I am all ears on effects on delay or any effect is has on the injector speed vs com PW.
I assume this delay is part of the IDM signal and thus gets added in to commanded PW as it's part of the PW signal of the IDM
On to #2 issue: I was working off a figure of 15 degrees BTDC @ 3K rpm Based of that I figured a decline in advance as RPMs drop and increase as they increased. Worked out a rate of basically 5° for every 1000rpms in either direction. So basically with 15°@ 3K I worked up from 2K@10°, 2.5K @12.5°, 3K @15°, [email protected]°, etc up to 5K@25° I want to confirm that this sort of rate is correct and is there a limit to this. I found some very interesting associations when following a rate like this and factoring in time.
If anyone can help me with these and some insight on them that would be helpful. Once I have this I will feel more comfortable with the numbers I have come up with and will post up the chart for review and hopefully so good discussion.
Last edited:
I feel the same way/ I could have worded that better. More than a ultimate curve I think there is a limit for the curve up top. You may be able to have a curve like I have posted but it can not keep going, obviously. It can only go till it hits a certain limit. Much like there is a ATDC limit of 24° after which you are no longer getting all the spray in the bowl.
What I did find very interesting is when I used the numbers I posted for advance without knowing it before hand they actually worked out to end up being the exact same amount of time in MS for each RPM step. That is by multiplying the ms per degree for 2K 2,5K 3K etc going the off set of 2.5 degrees of advance for each 500 rpm step from 2K to 5K they all end up being the same amount of time in ms./ It worked out to be 0.8334ms/ So how is that for a coincidence?
Here is something I was working on. We have a fixed .7 for injector delay. The fact that this is a static unchanging variable in commanded PW means that with each increase in RPM which causes a decrease in the available PW window time the delay becomes a larger percentage of the commanded PW. What this means when you also add in the BTDC SOI advance to the 25°@5K and go no further is the IDM limits actually hit much sooner than most think. Now this limit will never be an issue in useable RPM range even for a CR setup. The point is these limits that are time based take away from that amount. You can understand it better when you see a chart of it side by side. I am double checking my numbers and math. Once done I will post it up. I am sure others such as Bill Dave Matt have done this all long ago. But I think it could be helpful for others trying to understand and learn the limits of the system.
I also while not being a tuner understand there is a injector or injection delay table. This is what is also used to deal with EOT delays as well as the base mechanical delay of the injector. In considering this and talking to other experienced tuners I found some try to use this delay for other control without understand its effect so you end up with an adding of additional delay time. That time works out to being effecting the start of injection as RPMs increase by more degrees per that time amount.
What I did find very interesting is when I used the numbers I posted for advance without knowing it before hand they actually worked out to end up being the exact same amount of time in MS for each RPM step. That is by multiplying the ms per degree for 2K 2,5K 3K etc going the off set of 2.5 degrees of advance for each 500 rpm step from 2K to 5K they all end up being the same amount of time in ms./ It worked out to be 0.8334ms/ So how is that for a coincidence?
Here is something I was working on. We have a fixed .7 for injector delay. The fact that this is a static unchanging variable in commanded PW means that with each increase in RPM which causes a decrease in the available PW window time the delay becomes a larger percentage of the commanded PW. What this means when you also add in the BTDC SOI advance to the 25°@5K and go no further is the IDM limits actually hit much sooner than most think. Now this limit will never be an issue in useable RPM range even for a CR setup. The point is these limits that are time based take away from that amount. You can understand it better when you see a chart of it side by side. I am double checking my numbers and math. Once done I will post it up. I am sure others such as Bill Dave Matt have done this all long ago. But I think it could be helpful for others trying to understand and learn the limits of the system.
I also while not being a tuner understand there is a injector or injection delay table. This is what is also used to deal with EOT delays as well as the base mechanical delay of the injector. In considering this and talking to other experienced tuners I found some try to use this delay for other control without understand its effect so you end up with an adding of additional delay time. That time works out to being effecting the start of injection as RPMs increase by more degrees per that time amount.
At wot, the pcm adds about 8 degrees per 1000 rpms to compensate for the static delay so you can take that out of the calculations and just add that . 7 ms of of pw to the total commanded pw and don't worry about the timing being added electrically because it only brings the true injection timing back to 0 degrees BTDC so the timing tables actually function as real timing tables.
Ok I figured that from what we were speaking about with the table zerod it still was showing advance at RPMs climbed as it had to adjust for the delay.
The thing I found was that because the delay is still part of the IDM signal as the delay (the .7ms) is at the injector itself. So as your actual PW window decrease as RPMS go up that delay becomes a larger and larger percentage of the total available window.
After the issues I had getting the math right I was a bit back to front on if I had everything right in my head as well as on paper.
Give me a few mins and I will try to get the chart to look OK in a post and will get it posted. Its such a PITA to get columns to line up when you are doing a post. I do not understand why posting with spaces is not WYSIWYG. I put 6 spaces between words, then dammit, I want 6 spaces when its unloaded. Not back down to 1 space!! GRRRRR!!!
Let me see how the copy and paste goes.
The thing I found was that because the delay is still part of the IDM signal as the delay (the .7ms) is at the injector itself. So as your actual PW window decrease as RPMS go up that delay becomes a larger and larger percentage of the total available window.
After the issues I had getting the math right I was a bit back to front on if I had everything right in my head as well as on paper.
Give me a few mins and I will try to get the chart to look OK in a post and will get it posted. Its such a PITA to get columns to line up when you are doing a post. I do not understand why posting with spaces is not WYSIWYG. I put 6 spaces between words, then dammit, I want 6 spaces when its unloaded. Not back down to 1 space!! GRRRRR!!!
Let me see how the copy and paste goes.
OK here is the break I think some wanted to see for what the actual injection windows look like. This is the window that is based on what the
crank/piston/cylinder allow. You can then, if needed, compare this to what the IDM limits are. You will see why the IDM limits do not come into play in any reasonable RPM limits of even top end diesel setups compared to the actual engine mechanical limits as are created by RPM rate. IDM limit is only that it can not have a overlapping signal.
To help I will grab a quote from something Bill posted a few years ago. This will allow the two windows to be put side by side to see where they fall as the rpms get higher and higher.
I am attempting to break this down in the way I found it easiest to explain.
First thing was to take a specific RPM and calculate what the actual milliseconds per degree of crank rotation is for that rate is. I am taking this all the way down so it should allow the more people to follow it. It is not meant as a insult to the intelligence of those who understand all of this as some or all of this may be very redundant for some.
We start with RPM (revolutions per minute) convert that to RPS(revolutions per second) then further convert that to RPMS (revolutions per millisecond). Then compute that figure to get ms(milliseconds) per degree of movement based on a 360 degree rotation. So basic formula of:
1/{[(RPM/60)/1000]*360}= ms per degree
In speaking with Matt he gave me what he considered "Aggressive" levels of SOI (Start of Injector) advance for BTDC (Before Top Dead Center) and a rate to work with some advance rate for inside these rpms limits. I hope I got this correct Matt. There is always the chance I miss understood what you were telling me. But that can all be fixed if need be.
Then there is the standard window that starts @ TDC (Top Dead Center) till the point where the injection spray can no longer be kept inside the piston bowl. This comes to 24 degrees from TDC.
Lastly, we have "Injector delay". This minimum is a fixed time figure. As it deals with a actual time delay between when the IDM sends the signal to fire the injector to the moment of the injector firing (spraying fuel). It's not computed by crank degrees but actual time i.e. milliseconds. I still put down what the degrees would be as well as the percent the delay is of the entire window.
With the break down it will also show the difference between commanded PW (pulse Width) window (delay included) and actual window (without it).
These figures are all being used to create what can be considered a "maximum" injection window. I am trying to put this forth as with in reason what the largest possible window can be used for injection. Then having it all broken down so it can be easily seen where it all comes from.
The chart is broken down by RPM starting @ 2000rpm up to 5000rpm in steps of 500rpm. Next to it will be the MS per single degree of movement. Then BTDC degrees and MS for SOI advance TDC degrees and MS. Injector Delay MS and its % of the whole. Finally total degrees PW, then total MS which included the 0.7ms delay for the full commanded PW window that is available. I could add another line to include what the delay adds up to in crank degrees added into the other total crank degrees to give a final on that adds up to all the ms of window if someone would find that helpful. I saw no reason as that is not how the delay is factored. Its already there so all someone has to do is just add the two together if they want it.
RPM--------------MS/Degree---------°BTDC = MS-------------24°ATDC------Total Window °/MS----Delay 0.7ms=°/%----Total Comd Window:MS
2000 rpm------0.08333 ms/°------10.0°=0.8333ms------1.99992ms------34.0°/2.83322ms------08.4° / 19.81%------------3.53322ms
2500 rpm------0.06667 ms/°------12.5°=0.8334ms------1.60008ms------36.5°/2.43346ms------10.5° / 22.34%------------3.13346ms
3000 rpm------0.05556 ms/°------15.0°=0.8334ms------1.33344ms------39.0°/2.16684ms------12.6° / 24.42%------------2.86684ms
3500 rpm------0.04762 ms/°------17.5°=0.8334ms------1.14288ms------41.5°/1.97623ms------14.7° / 26.16%------------2.67623ms
4000 rpm------0.04167 ms/°------20.0°=0.8334ms------1.00008ms------44.0°/1.83348ms------16.8° / 27.63%------------2.53348ms
4500 rpm------0.03704 ms/°------22.5°=0.8334ms------0.88896ms------46.5°/1.72236ms------18.9° / 28.90%------------2.42236ms
5000 rpm------0.03333 ms/°------25.0°=0.8333ms------0.79992ms------49.0°/1.63317ms------21.0° / 30.00%------------2.33317ms
---------------------------------------------------------------
Taking the above IDM PW Signal Window Chart and adding a row for Mechincal PW window
RPM------PW ms@80%DC
3000------4.000----------2.86684ms
3250------3.692
3500------3.429----------2.67623ms
3750------3.200
4000------3.000----------2.53348ms
4250------2.824
4500------2.667----------2.42236ms
4750------2.526
5000------2.400----------2.33317ms
5250------2.286
5500------2.182
5750------2.087
6000------2.000
Let me know if I got something wrong or there are any changes that need to be made.
crank/piston/cylinder allow. You can then, if needed, compare this to what the IDM limits are. You will see why the IDM limits do not come into play in any reasonable RPM limits of even top end diesel setups compared to the actual engine mechanical limits as are created by RPM rate. IDM limit is only that it can not have a overlapping signal.
To help I will grab a quote from something Bill posted a few years ago. This will allow the two windows to be put side by side to see where they fall as the rpms get higher and higher.
I am attempting to break this down in the way I found it easiest to explain.
First thing was to take a specific RPM and calculate what the actual milliseconds per degree of crank rotation is for that rate is. I am taking this all the way down so it should allow the more people to follow it. It is not meant as a insult to the intelligence of those who understand all of this as some or all of this may be very redundant for some.
We start with RPM (revolutions per minute) convert that to RPS(revolutions per second) then further convert that to RPMS (revolutions per millisecond). Then compute that figure to get ms(milliseconds) per degree of movement based on a 360 degree rotation. So basic formula of:
1/{[(RPM/60)/1000]*360}= ms per degree
In speaking with Matt he gave me what he considered "Aggressive" levels of SOI (Start of Injector) advance for BTDC (Before Top Dead Center) and a rate to work with some advance rate for inside these rpms limits. I hope I got this correct Matt. There is always the chance I miss understood what you were telling me. But that can all be fixed if need be.
Then there is the standard window that starts @ TDC (Top Dead Center) till the point where the injection spray can no longer be kept inside the piston bowl. This comes to 24 degrees from TDC.
Lastly, we have "Injector delay". This minimum is a fixed time figure. As it deals with a actual time delay between when the IDM sends the signal to fire the injector to the moment of the injector firing (spraying fuel). It's not computed by crank degrees but actual time i.e. milliseconds. I still put down what the degrees would be as well as the percent the delay is of the entire window.
With the break down it will also show the difference between commanded PW (pulse Width) window (delay included) and actual window (without it).
These figures are all being used to create what can be considered a "maximum" injection window. I am trying to put this forth as with in reason what the largest possible window can be used for injection. Then having it all broken down so it can be easily seen where it all comes from.
The chart is broken down by RPM starting @ 2000rpm up to 5000rpm in steps of 500rpm. Next to it will be the MS per single degree of movement. Then BTDC degrees and MS for SOI advance TDC degrees and MS. Injector Delay MS and its % of the whole. Finally total degrees PW, then total MS which included the 0.7ms delay for the full commanded PW window that is available. I could add another line to include what the delay adds up to in crank degrees added into the other total crank degrees to give a final on that adds up to all the ms of window if someone would find that helpful. I saw no reason as that is not how the delay is factored. Its already there so all someone has to do is just add the two together if they want it.
RPM--------------MS/Degree---------°BTDC = MS-------------24°ATDC------Total Window °/MS----Delay 0.7ms=°/%----Total Comd Window:MS
2000 rpm------0.08333 ms/°------10.0°=0.8333ms------1.99992ms------34.0°/2.83322ms------08.4° / 19.81%------------3.53322ms
2500 rpm------0.06667 ms/°------12.5°=0.8334ms------1.60008ms------36.5°/2.43346ms------10.5° / 22.34%------------3.13346ms
3000 rpm------0.05556 ms/°------15.0°=0.8334ms------1.33344ms------39.0°/2.16684ms------12.6° / 24.42%------------2.86684ms
3500 rpm------0.04762 ms/°------17.5°=0.8334ms------1.14288ms------41.5°/1.97623ms------14.7° / 26.16%------------2.67623ms
4000 rpm------0.04167 ms/°------20.0°=0.8334ms------1.00008ms------44.0°/1.83348ms------16.8° / 27.63%------------2.53348ms
4500 rpm------0.03704 ms/°------22.5°=0.8334ms------0.88896ms------46.5°/1.72236ms------18.9° / 28.90%------------2.42236ms
5000 rpm------0.03333 ms/°------25.0°=0.8333ms------0.79992ms------49.0°/1.63317ms------21.0° / 30.00%------------2.33317ms
---------------------------------------------------------------
Taking the above IDM PW Signal Window Chart and adding a row for Mechincal PW window
RPM------PW ms@80%DC
3000------4.000----------2.86684ms
3250------3.692
3500------3.429----------2.67623ms
3750------3.200
4000------3.000----------2.53348ms
4250------2.824
4500------2.667----------2.42236ms
4750------2.526
5000------2.400----------2.33317ms
5250------2.286
5500------2.182
5750------2.087
6000------2.000
Let me know if I got something wrong or there are any changes that need to be made.
Power Hungry
Member
You guys might like this....
