Afr cant reach commanded

[GHuggins]

I haven't used it. I've learned over the years that tools like that are nice to have to start off because even that one is a lot like one we developed years ago, but in the "real world" they don't work in every scenario. The way I outlined above shows me everything I pretty much need to know with how to set the tables up. It takes back pressure, port velocity, piston velocity, compression issues, volumetric inefficiencies, adiabatic inefficiencies and so on all into the equation that paper formulas don't You can't argue with trim and wideband data. You'll even find that when inj is close to optimal the two will start to agree with one another better too simply because the fuel will burn a little cleaner...

Just like the one I was talking about above, it wound up taking normal advanced settings at idle and super advanced settings for the wot stomp off idle. Basically I shaped it so it was advanced - right at 490 for boundary at idle (it was originally made 520 across the board) and then I left coolant stock and rpm 0ed - and then I had it advance with rpm a lot like a stock ls3 does. I'm still cleaning it up and getting the main small details right such as idle torque, min air and timing, but he's already said it smells a lot better, is cranking easier and idling smoother just with the minor changes. Going from 520 to 490 netted 3% richer at idle whereas going WAY advanced got 5% leaner and idled rough, but with the wot stomp it went from being 10% lean at 520 to 10% rich down at 3something - (big change on this one). This is why I don't use tools for injection timing anymore. The engine will show all the variables whereas a "tool" only gives you soi and eoi in relation to valve events...

[LS ROB]

I haven't used it. I've learned over the years that tools like that are nice to have to start off because even that one is a lot like one we developed years ago, but in the "real world" they don't work in every scenario. The way I outlined above shows me everything I pretty much need to know with how to set the tables up. It takes back pressure, port velocity, piston velocity, compression issues, volumetric inefficiencies, adiabatic inefficiencies and so on all into the equation that paper formulas don't You can't argue with trim and wideband data. You'll even find that when inj is close to optimal the two will start to agree with one another better too simply because the fuel will burn a little cleaner...

Just like the one I was talking about above, it wound up taking normal advanced settings at idle and super advanced settings for the wot stomp off idle. Basically I shaped it so it was advanced - right at 490 for boundary at idle (it was originally made 520 across the board) and then I left coolant stock and rpm 0ed - and then I had it advance with rpm a lot like a stock ls3 does. I'm still cleaning it up and getting the main small details right such as idle torque, min air and timing, but he's already said it smells a lot better, is cranking easier and idling smoother just with the minor changes. Going from 520 to 490 netted 3% richer at idle whereas going WAY advanced got 5% leaner and idled rough, but with the wot stomp it went from being 10% lean at 520 to 10% rich down at 3something - (big change on this one). This is why I don't use tools for injection timing anymore. The engine will show all the variables whereas a "tool" only gives you soi and eoi in relation to valve events...

I see. Real world vs theory. Thank you! You mentioned Tq at idle; What should that be roughly? Anyway, the car feels pretty strong but not sure how close those numbers are in that "UP" log I posted. At this point I feel like I can experiment with timing some. Any idea on what max timing may be on this set up?

[GHuggins]

Even though it opens up it's own can of worms on debates, the torque model will indeed influence idle fueling and control. If torque is high, such as after adding a cam and removing a lot from the VE table or especially doing a MAF only tune, then the ecm will try to add fuel to obtain/maintain that torque. This is where it can benefit getting it down around 20 to 40 at idle delivered. You also need to make sure delivered matches something close to what's in your throttle follower table. A lot of the time you'll see logs where it's up in the 60's to 90's after a cam is installed due to the map and airmass shift.

As for timing, just play with it. Set up 2 different tunes with your current or using the latest one I made up as your base then have 30 or 40 added and the other subtracted. Make sure the engine is hot then flash the first. Do the 20 minute idle/stomp then flash the second. Do the same log for the current file so you have something to compare to. Zoom in on the stomp - see which one shows the richest afr and see which one averaged out or is showing the richest at the 20 minute mark of idle time. Doing the test in the same manor gives you a control, so even the rich after flash gets ignored in the test.

[LS ROB]

Awesome. That gives me an idea on a normal range after a cam swap. Thanks for all your help brother!

[Cringer]

...
What I've found is when they're overly retarded you wind up with too much "raw" fuel in the chamber and this raw fuel won't burn the best. This is when it's better to test it with advancing and retarding alike...

I too have found that spraying as early as possible (without short circuiting) gives the best results. Perhaps it is lazy aftermarket cam at idle vs tight factory cam at idle...but what I found is that I am getting better fuel trims by setting Makeup Mode to NONE. This is forcing all the fuel to be sprayed all at once in a single pulse, rather than trying to break it up in multiple sprays over time. This is giving the mixture more time to swirl and tumble in the cylinder with the air. It is a simple test too, just turn it off and collect logs. I do also zero out the Makeup ECT table though.

[GHuggins]

I too have found that spraying as early as possible (without short circuiting) gives the best results. Perhaps it is lazy aftermarket cam at idle vs tight factory cam at idle...but what I found is that I am getting better fuel trims by setting Makeup Mode to NONE. This is forcing all the fuel to be sprayed all at once in a single pulse, rather than trying to break it up in multiple sprays over time. This is giving the mixture more time to swirl and tumble in the cylinder with the air. It is a simple test too, just turn it off and collect logs. I do also zero out the Makeup ECT table though.

I don't remember how those operate. I thought it was discovered some years ago that those settings were only for cranking, but I don't know for sure. Verlon would be the only one I would know to be able to read the code for those and say for sure.

BUT if you can see things changing in the logs then I would assume it's coming into play. I'll have to try it out on the next one I'm dialing in the injection on. Just started on another 431 ls7. Might see what it does on it once I get it close.

Are you seeing the need to advance it more doing this or retard it more?

[Cringer]

I don't remember how those operate. I thought it was discovered some years ago that those settings were only for cranking, but I don't know for sure. Verlon would be the only one I would know to be able to read the code for those and say for sure.

BUT if you can see things changing in the logs then I would assume it's coming into play. I'll have to try it out on the next one I'm dialing in the injection on. Just started on another 431 ls7. Might see what it does on it once I get it close.

Are you seeing the need to advance it more doing this or retard it more?

That would be great if Verlon could clarify, because I have a lot of questions on this setting myself. But I know we have been putting a lot on his plate lately, so I don't want to stress him out!

To be honest, I sorted out all the EOIT stuff on the stock LS3 injectors which have a "laser beam" spray pattern right at the intake valve. I finally have some bigger decent aftermarket injectors that I sent to Greg Banish for flow testing and he said these new ones spray a much wider cone shaped pattern. All of which is to say I probably need to revisit my EOIT with the new injectors. However, from what I have seen in my limited experience using street/logging on my combo, spraying early (advanced) is best for fuel trims and fixing misfires. Now if I had access to a dyno in my garage...I would never leave.

EDIT: if you could try toggling that setting to NONE, SINGLE, and MULTIPLE on someone else's car to verify that would be gold!

[GHuggins]

That would be great if Verlon could clarify, because I have a lot of questions on this setting myself. But I know we have been putting a lot on his plate lately, so I don't want to stress him out!

To be honest, I sorted out all the EOIT stuff on the stock LS3 injectors which have a "laser beam" spray pattern right at the intake valve. I finally have some bigger decent aftermarket injectors that I sent to Greg Banish for flow testing and he said these new ones spray a much wider cone shaped pattern. All of which is to say I probably need to revisit my EOIT with the new injectors. However, from what I have seen in my limited experience using street/logging on my combo, spraying early (advanced) is best for fuel trims and fixing misfires. Now if I had access to a dyno in my garage...I would never leave.

EDIT: if you could try toggling that setting to NONE, SINGLE, and MULTIPLE on someone else's car to verify that would be gold!

OK, so straight off it may not have been a fair test. I didn't add 30 more advance to the timing or 0 out the makeup table (sorta forgot those steps), but there was a 2% change for the worse changing to none at idle. In other words it shifted leaner. Again may not have been a fair test. I did however get a 25% gain in total changing this guys inj timing and he says it's driving smoother. Mode is set to SINGLE. I would "think" single and none should be the same??? BUT it showed a 2% difference everywhere.

[Cringer]

OK, so straight off it may not have been a fair test. I didn't add 30 more advance to the timing or 0 out the makeup table (sorta forgot those steps), but there was a 2% change for the worse changing to none at idle. In other words it shifted leaner. Again may not have been a fair test. I did however get a 25% gain in total changing this guys inj timing and he says it's driving smoother. Mode is set to SINGLE. I would "think" single and none should be the same??? BUT it showed a 2% difference everywhere.

If you still can, test NONE and add the 30 to the Normal ECT table (and also zero out the M/U ECT table for good measure).
Single and keeping the 30 in tact is fine. Single would be two pulses: the pilot pulse + a single additional pulse.

The way I interpret this section is we are giving the ECM 30* of crank rotation to chop up the injector pulse.
The [ECM] 13341 - Makeup Pulse Minimum table is used to determine how long to keep the injector open and the [ECM] 45524 - Injector Off-Time table is used to determine how long to keep it shut. So now we have the definition of the duty cycle for on and off.

The EOIT tool I made has a Makeup Mode worksheet. I plugged in the stock LS3 injector specs as shown in the screenshot.
Injector duty cycle is 0.7969ms + 0.250ms = 1.0469ms for a single pulse.
So so when we plot these milliseconds against the crank degrees vs RPMs we get a chart like this. The Y-axis is showing how many whole number of pulses we can get in our injection window of 30* at this RPM. We can see the stock LS3 setup can't go past 3,200 rpms (at which point it defaults to one large pulse). And at idle we are getting a max of 7-8 pulses (plus the initial pilot pulse). I contend that when people get aftermarket injectors they create problems when they change they change the min pulse but don't change the off time to keep the same duty cycle. This tends to push the make up mode into much higher RPM's than is good. Of course my testing seems to like make up mode off at all RPMs.

I tried to capture the multiple wave form on an oscilloscope but failed. I would love it if someone else with better skills and equipment can can attempt this. If you are interested I go into more details here.

make up mode worksheet.jpg

Some of my questions/assumptions are:
-I assume make up pulse comes AFTER the pilot pulse.
-How does the ECM decide how long to make the pilot pulse? Does it back calculate it after determining how many make up pulses it can squeeze in during make up mode?
-Why is it labeled Make Up Pulse MINIMUM? Can it keep the injector open longer than the minimum? What are the circumstances it does so?
-I also believe make up mode can be problematic since we are opening and closing the injectors are a lot more which means the non-linear region is MUCH more of a factor and thus requires the injector data to be perfect.

[07GTS]

another thing i have noticed is my 2008 OS has the ''injector off time'' populated and my mates 2011 OS has the same makeup settings but has the ''injector off time'' zeroed out from factory

[GHuggins]

^^^ Probably the best description I've seen on what those tables do. Now I'm going to have to focus on that too.

[Cringer]

another thing i have noticed is my 2008 OS has the ''injector off time'' populated and my mates 2011 OS has the same makeup settings but has the ''injector off time'' zeroed out from factory

I wonder if setting the offtime to 0 would the logical equivalent of disabling M/U mode since it won't allow the injector to pulse. Does he have a Commodore ? I wonder if an example can be found in the repo...?

[Josue Galban]

What is the best way to get the AFR to follow commanded? At WOT, I'm commanding 12.13 but only getting 12.5ish. MAF is already showing rich as well as the VE. The only way i've been able to obtain target AFR is to leave MAF a little lean up top so that it carries over into PE. You guys got any tips on how to effectively dial in WOT? Tune and log attached.

Thanks.

Attachment 141032
Attachment 141033

you can try adjusting your inj and bring it closer to your commanded afr, also disabling LTFT can help you find that commanded afr, it could help raise the MAF a little more, really the VE shouldn't give you problems

[07GTS]

I wonder if setting the offtime to 0 would the logical equivalent of disabling M/U mode since it won't allow the injector to pulse. Does he have a Commodore ? I wonder if an example can be found in the repo...?

mine is the LS2 and my mates is the L76 commodore maby the dod affects it if its used at a cruise ? thats bout the only real difference between the engines aside from the fpcm where mine is just constant pump

[Cringer]

mine is the LS2 and my mates is the L76 commodore maby the dod affects it if its used at a cruise ? thats bout the only real difference between the engines aside from the fpcm where mine is just constant pump

So I found an example in the repo and you are right offtime is 0...but what a surprise! I am curious if your mate has the same 964 value in the Normal RPM table?! Probably doesn't matter too much since the RPMs are lower than idle, but still there is probably some interpolation occurring. This is the only 2011 Holden in the repo...
WTF.jpg

[07GTS]

So I found an example in the repo and you are right offtime is 0...but what a surprise! I am curious if your mate has the same 964 value in the Normal RPM table?! Probably doesn't matter too much since the RPMs are lower than idle, but still there is probably some interpolation occurring. This is the only 2011 Holden in the repo...
WTF.jpg

yes it has 964, 964, 964, 994 which can only be for cranking which is odd

[L1FTD]

I wonder if setting the offtime to 0 would the logical equivalent of disabling M/U mode since it won't allow the injector to pulse. Does he have a Commodore ? I wonder if an example can be found in the repo...?

Throwing something else in here I used for info.. https://forum.hptuners.com/showthrea...se-width-table

Inj. Off-Time.png

Inj. Off-Time 2.png

Inj. Off-Time 3.png

The off-time seems like a characterization value. I've tried zeroing before, made everywhere richer, but I saw no need.

The way I was interpreting the M/U pulse in HPT, example:

- Factory LSA, M/U Pulse Minimum = 1.28125ms. Say this minimum pulse value delivers 1 g/s of fuel
- Install injectors flowing twice as much as stock
- Now this minimum pulse of 1.28125ms, is delivering 2 g/s of fuel

Using the M/U Pulse Minimum HPT description, anything under 1.28125ms will be ignored. So you've gone from ignoring anything under 1 g/s of fuel (stock injectors), to ignoring anything under 2 g/s of fuel (2x flow injectors). Now the M/U pulse is less effective and has less range than when stock.

So if you wanted to keep the stock characteristics of the M/U pulse, you'd need to lower the Minimum Pulse value by the value of injector size increase.

Using rough example above:

- Injector size has doubled
- Cut the M/U Pulse Minimum in half = 0.640625ms

Now any pulse less than 0.640625ms will be ignored, so it's back to a 1 g/s of fuel minimum, same range as stock.

I think it's always going to be an approximation when using different injectors. Would need to know the flow rate of old vs. new injectors at your fuel pressure, and then there's all the changed characterization of the new injectors (offsets etc.), and fuel systems with varying pressures (stock LSA/LS9, 1:1 regulated).

Anyway, when M/U Pulse Minimum is set lower, AFR goes richer. So maybe just set to a value that makes sense, dial fuel in off that value, and move on.

[Cringer]

Throwing something else in here I used for info.. https://forum.hptuners.com/showthrea...se-width-table

Inj. Off-Time.png

Inj. Off-Time 2.png

Inj. Off-Time 3.png

The off-time seems like a characterization value. I've tried zeroing before, made everywhere richer, but I saw no need.

The way I was interpreting the M/U pulse in HPT, example:

- Factory LSA, M/U Pulse Minimum = 1.28125ms. Say this minimum pulse value delivers 1 g/s of fuel
- Install injectors flowing twice as much as stock
- Now this minimum pulse of 1.28125ms, is delivering 2 g/s of fuel

Using the M/U Pulse Minimum HPT description, anything under 1.28125ms will be ignored. So you've gone from ignoring anything under 1 g/s of fuel (stock injectors), to ignoring anything under 2 g/s of fuel (2x flow injectors). Now the M/U pulse is less effective and has less range than when stock.

So if you wanted to keep the stock characteristics of the M/U pulse, you'd need to lower the Minimum Pulse value by the value of injector size increase.

Using rough example above:

- Injector size has doubled
- Cut the M/U Pulse Minimum in half = 0.640625ms

Now any pulse less than 0.640625ms will be ignored, so it's back to a 1 g/s of fuel minimum, same range as stock.

I think it's always going to be an approximation when using different injectors. Would need to know the flow rate of old vs. new injectors at your fuel pressure, and then there's all the changed characterization of the new injectors (offsets etc.), and fuel systems with varying pressures (stock LSA/LS9, 1:1 regulated).

Anyway, when M/U Pulse Minimum is set lower, AFR goes richer. So maybe just set to a value that makes sense, dial fuel in off that value, and move on.

Hmmmm great info! Thanks for sharing this!

[Cringer]

I ran across a white paper from Greg Banish at Calibrated Success. He states the importance of controlling maximum duty cycle to be <100% to allow cooling time for the FETs (although I personally suspect they are actually MOSFETs or possibly IGBTs, but I guess that doesn't matter...) in the ECM circuitry from overheating. This also applies to the injector off time in the tune which serves as an additional buffer to prevent 100% duty cycle...it is basically a forced time out to allow some cooling to occur. I sent Greg a message and he said:

the off time protects both the FETs and improves our control of fuel mass. The solenoid itself is liquid cooled, so they rarely fail from duty cycle

So my guess is that for the factory calibrations that have 0ms off time defined...perhaps they have larger heat sinks or better cooling that allow them to shed heat better....and/or combined with larger injectors that will never come near 100% DC on a factory engine.

So the key take away is, the off time is an important factor to consider when you upgrade the fuel demands on a vehicle with stock injectors that are approaching 100% DC...especially if your factory tune is 0ms off time! You may be damaging the ECM, not the motor.

Taking this to the next level would be to understand how this applies to Make Up Mode. While transistors will generate heat in the fully ON position, they also generate substantial heat in the quiescent state (that is transitioning from fully OFF to fully ON and vice versa). A properly designed MOSFET/IBGT driver circuit will minimize the time spent in the quiescent state by slamming the FET open/closed as fast as possible. Its good, but never perfect. So when you think about a scenario when Make Up Mode is enabled, the injector may be spraying 7-9 pulses for a single injection window/intake valve opening at lower RPMs...look at the potential for heat to be generated:
1) that is 14-18 quiescent states (amperage > 0, generating heat)
2) 7-9 fully ON (max amperage, generating heat)
3) 14-18 times we are asking the injector to operate in the non-linear region, which means the injector needs to stay ON for a much longer total amount of time, in other words more heat from being on longer

[dassaur]

My L77 (2014 commodore) has injector off time to 0 but RPM is zero until 1152 when it goes from 4.6 up to 92.9 @ 4096.