Idle Misfire and Dead Time
#1
Junior Member
Thread Starter
iTrader: (1)
Join Date: Dec 2013
Location: Ontario, Canada
Posts: 154
Total Cats: 0
![Default](https://www.miataturbo.net/images/icons/icon1.gif)
Hey,
I am running rx7 460cc injectors with a MS1 ecu. I get the odd misfire at idle, caused by my AFR being a hair leaner then 13.3 at times. When my idle is tuned with a 1.0ms dead time the misfire is louder, but when i set the dead time to 1.2ms and re tune idle cells for same AFR the same misfire happens but it is more faint and quieter.
Does anybody know why it would be more faint and quieter with the higher dead time?
I am using 91 octane, 14 degrees timing at idle, idle around 900rpm.
I am running rx7 460cc injectors with a MS1 ecu. I get the odd misfire at idle, caused by my AFR being a hair leaner then 13.3 at times. When my idle is tuned with a 1.0ms dead time the misfire is louder, but when i set the dead time to 1.2ms and re tune idle cells for same AFR the same misfire happens but it is more faint and quieter.
Does anybody know why it would be more faint and quieter with the higher dead time?
I am using 91 octane, 14 degrees timing at idle, idle around 900rpm.
#2
Retired Mech Design Engr
![](/mt2011/images_td/ranks/supporter1.gif)
iTrader: (3)
Join Date: Jan 2013
Location: Seneca, SC
Posts: 5,011
Total Cats: 859
![Default](https://www.miataturbo.net/images/icons/icon1.gif)
You can play with stuff and get things to work at a single point, but that is not the right approach.
Set the dead time to what has been shown by knowledgeable people to be correct for those injectors. Then tune from there.
Set the dead time to what has been shown by knowledgeable people to be correct for those injectors. Then tune from there.
#3
Junior Member
Thread Starter
iTrader: (1)
Join Date: Dec 2013
Location: Ontario, Canada
Posts: 154
Total Cats: 0
![Default](https://www.miataturbo.net/images/icons/icon1.gif)
I have searched this site and everyone uses from .75-1.2 for dead time for the rx7 460cc, i tuned with 1.0 because that is what the base map had. I just want to see if anyone has any idea as to why the misfire would be quieter and more faint with a higher dead time.
#5
Tweaking Enginerd
![](/mt2011/images_td/ranks/supporter1.gif)
iTrader: (2)
Join Date: Mar 2013
Location: Boulder, CO
Posts: 1,809
Total Cats: 366
![Default](https://www.miataturbo.net/images/icons/icon1.gif)
To elaborate on what DNM is saying ...
the relative impact of dead-time on total pw is most significant during low load conditions (decel/idle). Errors/non linearity in dead time can cause significant errors in the tuned fuel equation at idle. Due to this, compensation factors (like EGO PID coefficients) that work in the higher load cells, don't work in the low load cells. With the wrong dead-time, the VE used to calculate the required fuel is not representative of the actual engine VE, so everything gets effed up.
Unfortunately, if I am not mistaken, the RX7 injectors are an old tech that makes the dead time very difficult to compensate for at idle conditions. Newer injectors are much more predictable and easier to work with. IMO, new injectors are totally worth it.
the relative impact of dead-time on total pw is most significant during low load conditions (decel/idle). Errors/non linearity in dead time can cause significant errors in the tuned fuel equation at idle. Due to this, compensation factors (like EGO PID coefficients) that work in the higher load cells, don't work in the low load cells. With the wrong dead-time, the VE used to calculate the required fuel is not representative of the actual engine VE, so everything gets effed up.
Unfortunately, if I am not mistaken, the RX7 injectors are an old tech that makes the dead time very difficult to compensate for at idle conditions. Newer injectors are much more predictable and easier to work with. IMO, new injectors are totally worth it.
#6
Junior Member
Thread Starter
iTrader: (1)
Join Date: Dec 2013
Location: Ontario, Canada
Posts: 154
Total Cats: 0
![Default](https://www.miataturbo.net/images/icons/icon1.gif)
Fully agree with what you guys have said! Thanks for the input. I am looking for better injectors.
So with the 460's im running do you think i am better off with a higher dead time value? What i have figured out is that for the MS1 the dead time value you put in is for 13.2V and there is also a ms/v correction factor which i currently use 0.100 for. So that means if i set my dead time to 1.0 and since my car mostly runs around 14V the actual run dead time would be around .9 right?
Am i better off having a lower dead time with higher VE table numbers at idle, or a higher dead time value with lower VE table numbers at idle. Which would be best for a smooth idle? At 1.2ms dead time my ve numbers for idle are 42, with 1.0ms they are 48 and even lower would probably be in the mid 50's. I know the rest of the VE would need to be tweeked as well but just wondering what would be a better starting point.
So with the 460's im running do you think i am better off with a higher dead time value? What i have figured out is that for the MS1 the dead time value you put in is for 13.2V and there is also a ms/v correction factor which i currently use 0.100 for. So that means if i set my dead time to 1.0 and since my car mostly runs around 14V the actual run dead time would be around .9 right?
Am i better off having a lower dead time with higher VE table numbers at idle, or a higher dead time value with lower VE table numbers at idle. Which would be best for a smooth idle? At 1.2ms dead time my ve numbers for idle are 42, with 1.0ms they are 48 and even lower would probably be in the mid 50's. I know the rest of the VE would need to be tweeked as well but just wondering what would be a better starting point.
#8
Cpt. Slow
![](/mt2011/images_td/ranks/supporter1.gif)
iTrader: (25)
Join Date: Oct 2005
Location: Oregon City, OR
Posts: 14,360
Total Cats: 1,184
![Default](https://www.miataturbo.net/images/icons/icon1.gif)
It doesn't matter, and if you're truly tuning for the same AFR, you shouldn't hear any difference. Not sure what a "loud" misfire is.
Think of it this way:
1+2=3
2+1=3.
That's all your changing.
Except the formula is: PW = REQ_FUEL * VE * MAP * E + accel + Injector_open_time
Look at your pulse width at idle, not dead time. I'm guessing it's the same at both dead times.
Think of it this way:
1+2=3
2+1=3.
That's all your changing.
Except the formula is: PW = REQ_FUEL * VE * MAP * E + accel + Injector_open_time
Look at your pulse width at idle, not dead time. I'm guessing it's the same at both dead times.
#9
Retired Mech Design Engr
![](/mt2011/images_td/ranks/supporter1.gif)
iTrader: (3)
Join Date: Jan 2013
Location: Seneca, SC
Posts: 5,011
Total Cats: 859
![Default](https://www.miataturbo.net/images/icons/icon1.gif)
It doesn't matter, and if you're truly tuning for the same AFR, you shouldn't hear any difference. Not sure what a "loud" misfire is.
Think of it this way:
1+2=3
2+1=3.
That's all your changing.
Except the formula is: PW = REQ_FUEL * VE * MAP * E + accel + Injector_open_time
Look at your pulse width at idle, not dead time. I'm guessing it's the same at both dead times.
Think of it this way:
1+2=3
2+1=3.
That's all your changing.
Except the formula is: PW = REQ_FUEL * VE * MAP * E + accel + Injector_open_time
Look at your pulse width at idle, not dead time. I'm guessing it's the same at both dead times.
I was making the assumption that the difference could be an actual single miss vs a double miss was the source of the more and less pronounced situation.
Hence why I tried to give an possible explanation.
#10
Junior Member
Thread Starter
iTrader: (1)
Join Date: Dec 2013
Location: Ontario, Canada
Posts: 154
Total Cats: 0
![Default](https://www.miataturbo.net/images/icons/icon1.gif)
Bringing this tread back.
Yes PW stays the same at both dead times. My idle PW is around 1.6-1.65ms, which to me seems quite small. Is there anything i can do to increase the PW at idle? My current req'd fuel is 5.8 and my idle cells in VE are 51 which puts my AFR in the low 13's. I am using 1.0ms dead time.
Does incorrect dead time affect hot lean restarts as well? My AFR goes from low 13's normal to around 16-17 for a hot restart. I have done the math and air density correction would only affect the fueling by 3% for a hot restart. 3% less fuel wont cause the AFR to go from low 13 to around 17. Something else is also messing it up.
Yes PW stays the same at both dead times. My idle PW is around 1.6-1.65ms, which to me seems quite small. Is there anything i can do to increase the PW at idle? My current req'd fuel is 5.8 and my idle cells in VE are 51 which puts my AFR in the low 13's. I am using 1.0ms dead time.
Does incorrect dead time affect hot lean restarts as well? My AFR goes from low 13's normal to around 16-17 for a hot restart. I have done the math and air density correction would only affect the fueling by 3% for a hot restart. 3% less fuel wont cause the AFR to go from low 13 to around 17. Something else is also messing it up.
#11
![Default](https://www.miataturbo.net/images/icons/icon1.gif)
Now keep in mind that here is where the difference comes from.
If the MS sends a 1.6ms pulse, it's not the same amount of fuel if the dead time is 0.5ms than if it's 1ms. Which if your dead time is 1ms then I'm confused as of what the **** the MS thinks it's going out.
Does the MS count only the OPEN time and doesn't count the closing time? Or it only counts the FULLY OPEN time?
So from those 1.6ms, if it counts the WHOLE (deadtime + full open + dead time) we are at -0.4ms of injection, because the deadtimes only are 2ms.
If it only counts the OPEN time including opening dead time then you'd be actually injecting 0.6ms at full 460cc/min.
If it only counts how long it's open at FULL. then you'd have this unknown amount of fuel during 1ms, then 460cc/min for 1.6ms and then other 1ms at unknown rate while the injector closes.
Which one is it? I'm probably wrong with all three lol
If the MS sends a 1.6ms pulse, it's not the same amount of fuel if the dead time is 0.5ms than if it's 1ms. Which if your dead time is 1ms then I'm confused as of what the **** the MS thinks it's going out.
Does the MS count only the OPEN time and doesn't count the closing time? Or it only counts the FULLY OPEN time?
So from those 1.6ms, if it counts the WHOLE (deadtime + full open + dead time) we are at -0.4ms of injection, because the deadtimes only are 2ms.
If it only counts the OPEN time including opening dead time then you'd be actually injecting 0.6ms at full 460cc/min.
If it only counts how long it's open at FULL. then you'd have this unknown amount of fuel during 1ms, then 460cc/min for 1.6ms and then other 1ms at unknown rate while the injector closes.
Which one is it? I'm probably wrong with all three lol
#12
![Default](https://www.miataturbo.net/images/icons/icon1.gif)
The dead time is somewhat confusingly listed as opening time in many MS1 settings, but all MS variants really use opening time minus closing time. The MS will calculate how much pulse width the engine "should" use for an injector that opens and closes instantly, then adds the dead time and battery voltage correction.
The data log will show the actual pulse width, after all modifiers, including dead time. 1.65 ms is 1.65 ms of the injector being switched on. How much of that time is spent moving the pintle and how much is spent spraying fuel will depend on the injector.
And the final rule: A good pulse width is what the engine runs well at, not what somebody, including myself, on a forum tells you your pulse width should be.
The data log will show the actual pulse width, after all modifiers, including dead time. 1.65 ms is 1.65 ms of the injector being switched on. How much of that time is spent moving the pintle and how much is spent spraying fuel will depend on the injector.
And the final rule: A good pulse width is what the engine runs well at, not what somebody, including myself, on a forum tells you your pulse width should be.
#13
Retired Mech Design Engr
![](/mt2011/images_td/ranks/supporter1.gif)
iTrader: (3)
Join Date: Jan 2013
Location: Seneca, SC
Posts: 5,011
Total Cats: 859
![Default](https://www.miataturbo.net/images/icons/icon1.gif)
Bringing this tread back.
Yes PW stays the same at both dead times. My idle PW is around 1.6-1.65ms, which to me seems quite small. Is there anything i can do to increase the PW at idle? My current req'd fuel is 5.8 and my idle cells in VE are 51 which puts my AFR in the low 13's. I am using 1.0ms dead time.
Does incorrect dead time affect hot lean restarts as well? My AFR goes from low 13's normal to around 16-17 for a hot restart. I have done the math and air density correction would only affect the fueling by 3% for a hot restart. 3% less fuel wont cause the AFR to go from low 13 to around 17. Something else is also messing it up.
Yes PW stays the same at both dead times. My idle PW is around 1.6-1.65ms, which to me seems quite small. Is there anything i can do to increase the PW at idle? My current req'd fuel is 5.8 and my idle cells in VE are 51 which puts my AFR in the low 13's. I am using 1.0ms dead time.
Does incorrect dead time affect hot lean restarts as well? My AFR goes from low 13's normal to around 16-17 for a hot restart. I have done the math and air density correction would only affect the fueling by 3% for a hot restart. 3% less fuel wont cause the AFR to go from low 13 to around 17. Something else is also messing it up.
Now keep in mind that here is where the difference comes from.
If the MS sends a 1.6ms pulse, it's not the same amount of fuel if the dead time is 0.5ms than if it's 1ms. Which if your dead time is 1ms then I'm confused as of what the **** the MS thinks it's going out.
Does the MS count only the OPEN time and doesn't count the closing time? Or it only counts the FULLY OPEN time?
So from those 1.6ms, if it counts the WHOLE (deadtime + full open + dead time) we are at -0.4ms of injection, because the deadtimes only are 2ms.
If it only counts the OPEN time including opening dead time then you'd be actually injecting 0.6ms at full 460cc/min.
If it only counts how long it's open at FULL. then you'd have this unknown amount of fuel during 1ms, then 460cc/min for 1.6ms and then other 1ms at unknown rate while the injector closes.
Which one is it? I'm probably wrong with all three lol
If the MS sends a 1.6ms pulse, it's not the same amount of fuel if the dead time is 0.5ms than if it's 1ms. Which if your dead time is 1ms then I'm confused as of what the **** the MS thinks it's going out.
Does the MS count only the OPEN time and doesn't count the closing time? Or it only counts the FULLY OPEN time?
So from those 1.6ms, if it counts the WHOLE (deadtime + full open + dead time) we are at -0.4ms of injection, because the deadtimes only are 2ms.
If it only counts the OPEN time including opening dead time then you'd be actually injecting 0.6ms at full 460cc/min.
If it only counts how long it's open at FULL. then you'd have this unknown amount of fuel during 1ms, then 460cc/min for 1.6ms and then other 1ms at unknown rate while the injector closes.
Which one is it? I'm probably wrong with all three lol
Yes, different dead times affects how EGO and ASE, etc controls hot restarts, because those things act on the calculated pulse BEFORE the dead time adder.
So 1.6 mS actual pulse made up of 0.6mS of calculation and 1mS of dead time. If, during hot restart the injector needs 1.8mS of pulse to keep AFR correct, then the EGO must move the 0.6 to 0.8, or 33% EGO.
OTOH, if that normal 1.6 is 1.0 calculation and 0.6 dead time, then when hot restart needs 1.8mS the EGO has to only made the 1.0 into 1.2, or 20% EGO adder.
Not saying you should change dead time based on this. Just answering your question.
I you want longer pulse width at idle:
Call for more fuel with VE if you are not "Incorporate AFR"
Call for more fuel by lowing AFR at idle if are "Incorporate AFR"
Lower the fuel pressure.
Put in smaller injectors.
EDIT: See Matt's "Final Rule" above.
Last edited by DNMakinson; 05-17-2017 at 11:18 AM.
#14
![Default](https://www.miataturbo.net/images/icons/icon1.gif)
The thing is that there is also time for the injector to close, but I guess that's going to be constant no matter what, after every pulse we'll always have that X amount of time of the injector closing down.
Does the MS calculate how much fuel goes in during the deadtime? Approximately it could integrate the area under a linear curve that goes from 0cc/min to 460cc/min in 1ms no? not sure if that's added to the equation. Then it would know almost exactly how much fuel is going in. That with a very well calibrated MAF should provide awesome fueling, no?
So if it needs X cc then it would know that it already has 2ms of deadtime and that amount of fuel, and calculates the PW needed based on that.
Does the MS calculate how much fuel goes in during the deadtime? Approximately it could integrate the area under a linear curve that goes from 0cc/min to 460cc/min in 1ms no? not sure if that's added to the equation. Then it would know almost exactly how much fuel is going in. That with a very well calibrated MAF should provide awesome fueling, no?
So if it needs X cc then it would know that it already has 2ms of deadtime and that amount of fuel, and calculates the PW needed based on that.
#15
Retired Mech Design Engr
![](/mt2011/images_td/ranks/supporter1.gif)
iTrader: (3)
Join Date: Jan 2013
Location: Seneca, SC
Posts: 5,011
Total Cats: 859
![Default](https://www.miataturbo.net/images/icons/icon1.gif)
The thing is that there is also time for the injector to close, but I guess that's going to be constant no matter what, after every pulse we'll always have that X amount of time of the injector closing down.
Does the MS calculate how much fuel goes in during the deadtime? Approximately it could integrate the area under a linear curve that goes from 0cc/min to 460cc/min in 1ms no? not sure if that's added to the equation. Then it would know almost exactly how much fuel is going in. That with a very well calibrated MAF should provide awesome fueling, no?
So if it needs X cc then it would know that it already has 2ms of deadtime and that amount of fuel, and calculates the PW needed based on that.
Does the MS calculate how much fuel goes in during the deadtime? Approximately it could integrate the area under a linear curve that goes from 0cc/min to 460cc/min in 1ms no? not sure if that's added to the equation. Then it would know almost exactly how much fuel is going in. That with a very well calibrated MAF should provide awesome fueling, no?
So if it needs X cc then it would know that it already has 2ms of deadtime and that amount of fuel, and calculates the PW needed based on that.
Go look at the basic MS fueling equation. The number one plugs in for b is the number that makes the line fit the action of the injector. There is no need to break it down any further.
#16
Junior Member
Thread Starter
iTrader: (1)
Join Date: Dec 2013
Location: Ontario, Canada
Posts: 154
Total Cats: 0
![Default](https://www.miataturbo.net/images/icons/icon1.gif)
Just to clarify, i asked about having a higher pulse width because in the MS1 tuning manual it is recommended to try and have a 2.0ms min idle PW. Did not come across it randomly online.
So if i have 1.6ish idle pw that means with a 1.0ms opening time(as its called in ms1) approx 0.6ms of fuel is injected right? Also what if my injector opening time was higher then what it actually should be, say i use 1.0ms but it actually should be 0.8 lets say does that mean fuel will be injected during 1.0ms opening time? And then that would result in less VE needed?
Should i try a lower or higher dead time then? My afr will go from 13.0 to 16-17 during a hot restart.
So if i have 1.6ish idle pw that means with a 1.0ms opening time(as its called in ms1) approx 0.6ms of fuel is injected right? Also what if my injector opening time was higher then what it actually should be, say i use 1.0ms but it actually should be 0.8 lets say does that mean fuel will be injected during 1.0ms opening time? And then that would result in less VE needed?
Should i try a lower or higher dead time then? My afr will go from 13.0 to 16-17 during a hot restart.
#17
Retired Mech Design Engr
![](/mt2011/images_td/ranks/supporter1.gif)
iTrader: (3)
Join Date: Jan 2013
Location: Seneca, SC
Posts: 5,011
Total Cats: 859
![Default](https://www.miataturbo.net/images/icons/icon1.gif)
I'm stepping back. You must post your tune and a log for people to be able to further advise you. We need to see what MAT Corr looks like, what Voltage Correction looks like, what battery voltage is running during idle. Then we could spoon feed you.
Meanwhile, ponder this: Assume you have no Voltage Correction (100% across all voltages), if you were to set your dead time to 1.65 then your idle VE would have to be set to zero to have the MS put out a 1.65mS pulse, which is what you say your car needs. What kind of control would the MS now have to adjust for anything?
Meanwhile, ponder this: Assume you have no Voltage Correction (100% across all voltages), if you were to set your dead time to 1.65 then your idle VE would have to be set to zero to have the MS put out a 1.65mS pulse, which is what you say your car needs. What kind of control would the MS now have to adjust for anything?
#19
Junior Member
Thread Starter
iTrader: (1)
Join Date: Dec 2013
Location: Ontario, Canada
Posts: 154
Total Cats: 0
![Default](https://www.miataturbo.net/images/icons/icon1.gif)
Tune and log posted. You can see idling at beginning of log. One thing that was different with my tune at time of log was my mat correction at 20 and 30 temp correction at time of log were 101 and 106 but are a bit higher in that tune file, i was trying something but will go back to old values.
So from what i have read in MS1 tuning manual the opening time is suppose to be for 13.2V, so im thinking with my .10 ms/v correction and my 14Vish at idle that the injectors are running around 0.9ms opening time then?
Dont mean to ask so many questions but im just getting confused with some info posted in here compared to whats in the manual. Let me know what you think. Thanks in advance!
So from what i have read in MS1 tuning manual the opening time is suppose to be for 13.2V, so im thinking with my .10 ms/v correction and my 14Vish at idle that the injectors are running around 0.9ms opening time then?
Dont mean to ask so many questions but im just getting confused with some info posted in here compared to whats in the manual. Let me know what you think. Thanks in advance!
#20
Retired Mech Design Engr
![](/mt2011/images_td/ranks/supporter1.gif)
iTrader: (3)
Join Date: Jan 2013
Location: Seneca, SC
Posts: 5,011
Total Cats: 859
![Default](https://www.miataturbo.net/images/icons/icon1.gif)
Cruising. This was my first time loading an MS1 tune. I think I got it correct, but I'm not positive. I could not find MAT / Density correction anywhere.
From the log, it indicates that you are not pulling fuel due to MAT as Gair does not go low. As you say, that is not your normal, however. Also, it does not appear that this log shows a hot re-start.
I suggest that you increase ASE from 71C - 110C to combat hot restart lean right at start-up. Choose Cycles, rather than seconds, as you need less time if you rev the engine, than if you let it idle. Then implement EGO to about 15-20% at lower kPa. There is the ability to shift it to lower above a certain kPa. I would suggest 10 cycles, rather than 70, but again, I am more familiar with PID control in the MS3. Then enrichen your idle VE a bit (10%). EGO will pull fuel from that when in the normal state, and then add fuel in a hot restart situation. Presently your log shows no EGO ever.
I am surprised your curves do not extend beyond 71C CLT temps. That seems a low temperature to consider fully warmed up.
If I had to make a suggestion, I would keep Open Time at 1.0 or possibly .95, not 1.2.
EDIT: I'm gone for 5 days, so will not have access to TS or MLV.
Tip: Put your location in your profile. It will show up under you name, over to the left, when you post.
From the log, it indicates that you are not pulling fuel due to MAT as Gair does not go low. As you say, that is not your normal, however. Also, it does not appear that this log shows a hot re-start.
I suggest that you increase ASE from 71C - 110C to combat hot restart lean right at start-up. Choose Cycles, rather than seconds, as you need less time if you rev the engine, than if you let it idle. Then implement EGO to about 15-20% at lower kPa. There is the ability to shift it to lower above a certain kPa. I would suggest 10 cycles, rather than 70, but again, I am more familiar with PID control in the MS3. Then enrichen your idle VE a bit (10%). EGO will pull fuel from that when in the normal state, and then add fuel in a hot restart situation. Presently your log shows no EGO ever.
I am surprised your curves do not extend beyond 71C CLT temps. That seems a low temperature to consider fully warmed up.
If I had to make a suggestion, I would keep Open Time at 1.0 or possibly .95, not 1.2.
EDIT: I'm gone for 5 days, so will not have access to TS or MLV.
Tip: Put your location in your profile. It will show up under you name, over to the left, when you post.