|
VE analyze autotune and MAT correction table
I know I know, it's been beaten to death, but I want to understand what is going on.
I've received a little conflicting information, but I think it might be due to how I was asking the question. I asked on MSextra if VE analyze takes into account the MAT correction table while autotuning. The response was yes. Reverent I believe, said no. I think Reverent is right, but help me understand what the algorithm IS and IS NOT taking into account while autotuning. How is the Autotune algorithm handling the ideal gas correction (Gair) and the MAT correction table? Would someone please spoon feed this to me, and then place it in a useful save post. Not understanding this can completely waste a dyno session. |
1 Attachment(s)
I'm almost positive autotune looks at Gair% when making corrections, but I could be wrong.
Why not create an Anti-Ideal Gas Law table in your MAT corrections table? Cause right now, when you start seeing intake temps higher than 80*F the MS is going to want to pull fuel like a bitch. This oculd screw up your tune and any autotuning. You'll end up with this: Attachment 189064 This pretty much causes Gair to stay at 100% regardless of the intake temps (assuming the MAT correction value is 100%). I personally think the code is too aggresive, while it may be the laws of physics - in terms of fueling it doesn't seem to be the case. |
note that you can custom filter for a range of MAT values if you want.
|
Are you zeroing this table while autotuning, or tuning with this table as it is above?
Are you using 100% correction? I have mine set to 50%. I'm not sure if it matters. My plot at 50% is: 20 -2 60 0 76 0 86 1 103 3 145 6 This was based on countering the Gair value from a zeroed MAT correction table. note that you can custom filter for a range of MAT values if you want. |
Mine's close to what you quoted, since mine's at 50% as well. I really need to put it on my stim and set it there, making sure it's accurate in canceling it out.
I'm leaving that table as is, when autotuning or not. I'll adjust it at idle for extreme intake temps if need be. But for now, I do not like the ideal gaw law code and how agressive it pulls fuel with higher MAT temps. I see high temps in boost, so it was effecting me on the dyno and hot restarts were lean like what when they went past 80*F. |
How is this being handled?.......
I'm tuning at 75 degrees ambient, so my IAT is 85 degrees. I want to autotune to full boost and see 125 degree temps at the upper kpa fuel cells. My cruise cells are now being tuned at 85 degrees, max boost at 125 degrees, and some gradient between cruise and max boost. If I autotune with the MAT correction table nulling the Gair (100%), then I would have to add a correction on top of it to actually take into account the changing temps. Nevermind, because as temp decreases it isn't adding as much fuel, so it is taking into account the air correction. By zeroing the table and tuning at 125 degrees it is tuning with significant fuel pulled, but it is meeting the afr criteria. If I then change the MAT correction to add fuel at 125 degrees to counteract Gair, then won't it be rich at 125 degrees? |
I did a lot of testing of underhood temps. I bet your AIT is really seeing 85*F.
In fact, my AIT just off the IC was cooler than a second AIT sensor when sitting behind the firewall, when sitting under my air filter in my cold air box, and when sitting underneath my headlight. and who's to say you really need to subtract fuel when the temps get that hot? the best way to judge this would be a hairdryer on your AIT sensor/intake at idle. If the hot temps, with no corrections, make you go rich, then you need some correction...if not, then you dont. I'll take my chances with it going rich in boost if the AITs get hot enough to require it before I'd want it to pull fuel in boost. As of now, between 45-80*F I don't need any corrections. My AFR is on target at idle between those intake temps with zero Gair correction. |
Well then, I will continue to tune with my MAT table in tact. A good way to determine the values for MAT correction table is to zero the table and start datalogging on a cool morning. You will see the Gair that is being applied relative to the AIT. Make sure you get into boost so that you can get the high AIT temps.
This is probably known to most, but since I wanted this to be an educational thread I thought it might be helpful. |
a stim is easier :)
or take the table i posted (which was done on a stim), put MAT correction value back to 100% and whamo! I just havent confirmed it myself, someone else did this and posted with an MS2. But my values are more or less half those. |
Think you could post you non linear correction table brain? (That you use with 50 percent correction value). I havnt ran mine up to extreme hot temps like 140 yet so I haven't set those values yet...but it'd be nice to get it ballpark close. Like everyone else I go crazy lean when it gets even a little bit hotter.
Edit: nevermind just read your above post, ill just use those numbers with 100 correction value |
im going to try brains at 100. i had to increase mine around 100* to satisfy idle afr, but gair wasnt pulling as much as i am adding. |
I feel like I am having this same issue, but I cannot find MAT correction on TS, Ive searched through TS with no luck. :hustler:
|
yeah if you start to add too much, you'll start going rich like whoa in boost. I had that issue the last dyno day we had in Nov. I arbitrially added in correction to help revent heatsoaking at idle and going lean. didn't realize what effect it had in boost.
I probably should have guessed something was up when I was averaging less than 20mpg. having the stupid voltage offset in TS from my LC-1 didn't help. 1 full point richer from that alone. |
Originally Posted by Braineack
(Post 725244)
yeah if you start to add too much, you'll start going rich like whoa in boost. I had that issue the last dyno day we had in Nov. I arbitrially added in correction to help revent heatsoaking at idle and going lean. didn't realize what effect it had in boost.
I probably should have guessed something was up when I was averaging less than 20mpg. having the stupid voltage offset in TS from my LC-1 didn't help. 1 full point richer from that alone. MAT correction table is located under the "extended" tab. |
yes eventually I figured out that was the issue. my gauge and logworks read the same. TS showed 1 full point leaner. So when I tuned based off that I was 1 full point richer.
When calibrating my AFR table in TS, I use .2v instead of 0v for my 7.3X:1 value. everything matches up well now. Averaging 25mpg city. |
How about Ego correction? Do you have your closed loop O2 on when running AutoTune on TS?
|
definitely not.
I'm revisiting my VE table using autotune.... I turn off EGO, set accel enrichments cutoff to like 999 revs/sec so it will never kick in, applied the above MAT correction values and tuning from there. I changed one of my gauges on the lefthand side in autotune to Gair and mine stayed right at 100 the entire time in my last drive.... might have been flipping to 101 every now and then, but that could have been the tail end of warmup enrichments... Basically I'm under the impression you want to kill all enrichments for VE table tuning... then once it's 90 percent there, add in accel enrichments, switch on EGO control for cruise if you'd like, etc. Just be mindful go easy on the throttle inputs when tuning the VE table... otherwise leanspikes will get compensated for in the VE table from autotune and you don't really want that, you want the VE table to be tuned for steady state conditions, then accel enrichments to take care of the lean spikes. |
i leave everything on! im hardcore like that.
|
interesting, i thought you were "supposed" to turn the extra crap off initially for VE table, then start adding that stuff in. Shows what I know..
|
VEAL will work fine with EGO enabled. I suspect it keeps track of the current VE value plus the EGO correction value and adds them together to get the new VE target.
I would leave your corrections in place and make a custom filter that addresses them. you'll have to figure out the right variables to use if you don't already know them by going to the dashboard and viewing "output channel" for the specific gauge (which you may have to create to see). and then add in your filter. remember that you are filtering for where you don't want VEAL to run. so if you put "rpm < 400" then it will not tune below 400 rpm. here's what I use: (tpsDOT > 50) || (rpm < 1900 && fuelload < 67) || (advance < 1 ) make sure to save the filter somewhere because if you hit "cancel" it goes away. Basically I filter when TPS enrichments kick in, around idle, and when various limiters are enabled (the advance < 1 part) && means "AND" || means "OR" () allow you to nest operations so you could say "( airCorrection > 97 && aircorrection < 103 ) airCorrection is the output channel for Gair. |
what is the "advance < 1" do ?
|
piece of shit.
I did those MAT correction values... Gair is always 100 now...but still lean on heatsoaked/hot start... I dont get it. Car was fine on 80-85 IAT, then when I started it back up after leaving pubelix, it was a little over 100 IAT. Gair was still at 100...but it was running lean. Hot air, and it wasn't pulling any fuel...and it was still lean, im so confused. |
Originally Posted by y8s
(Post 725301)
when various limiters are enabled (the advance < 1 part)
Originally Posted by djp0623
(Post 725397)
what is the "advance < 1" do ?
when I enable my launch control or flat shift, it sets timing to less than 1 degree of advance. you could make it more like 10 or 5 or something but it works. the point is to avoid VEAL thinking that the 19:1 AFR spike during flatshift means "add a shit ton of fuel". |
If you ever watch VEAL, it will flash filters during AE and DE and other conditionals like VE idle.
|
I tried the 100% MAT correction values and the table above. It pulled a crap load of fuel from the VE table. Obviously because it is now adding a lot more from MAT correction table.
I'm still having the same issue below 75 degrees. Although it pulls fuel to make Gair 100% it is still making the car run too lean. Options are to pull a little of the correction, or see if there is another issue. Brain: You said you are using the Mazda CLT sensor preset??? This moring it was 70 degrees and my CLT in tunerstudio read 91. So something is up with this value. It also seems like the car warms up too quickly with these values. I'm wondering if my car isn't as warm as it appears when I start autotuning at 150* CLT. There is a very large difference between how my car runs at 130 degree and 155 degree when I start autotuning. I may need to increase the temp at which I start autotune to 180 or 190. Then see if this has any affect on the overall stability of my VE table relative to MAT. |
yeah that's pretty far off. But yes, I use the Mazda Rx7 (s4 & s5) drop down when configuring the CLT.
the tune shouldnt really change that much from just CLT temps, do you have any other enrichments running? when does your warmup enrichements stop? and they stop at 100% right? obviously, the tune will be richer during warmup. |
WUE stops at 150* at 100% ( I thought in the past you said you used the "mazda" preset. I'll change mine this afternoon.
I just went over a few morning logs. Baro adds a percent, but that is accurate with my altitude. I just can't figure out why a tune would change from with temp form hot to cold it the MAT table is dialed in unless I am tuning at a point when the car is not at equilibrium (warm) The truth is though, I'm much less concerned about reducing how much the MAT pulls fuel while its cool out, then I am about reducing how much it adds when hot. I'm going to tune above a CLT of 185 and see how that works. If that doesn't fix it, then I'll modify the MAT correction < 70* |
djp0623 - where is your AIT located? we have having the same heat soak issues and I am wondering if I should move the AIT to the intercooler endtank
|
Originally Posted by longuyen88
(Post 725981)
djp0623 - where is your AIT located? we have having the same heat soak issues and I am wondering if I should move the AIT to the intercooler endtank
|
im in a pretty intersting discussion on heatsoak with the MS developers. ill let you know what the outcome is. Right now I can disable AIT input during ASE. that allows me time for my wbo2 to warm up, then I can use EGO at idle to add/subtract fuel if needed. but i want more, but we haven't figured out the best approach.
|
Originally Posted by Braineack
(Post 726000)
im in a pretty intersting discussion on heatsoak with the MS developers. ill let you know what the outcome is. Right now I can disable AIT input during ASE. that allows me time for my wbo2 to warm up, then I can use EGO at idle to add/subtract fuel if needed. but i want more, but we haven't figured out the best approach.
|
were you reading my last posts?
Originally Posted by 11:57am
...let's assume I'm genuinely heat soaking and I start tinkering with the MAT corrections table to compensate. So at idle, traffic and hot restarts I can maintain stoich when the ideal law demands you need less fuel; this seems to be what everyone is doing.
When it is truely a hot day outside and I may then need less fuel, true to ideal, since I altered the MAT corrections table I'm going to be adding more fuel and have to rely on EGO to pull it back where the ideal law code might have put it. It will add even more so when I'm in boost and running open loop. Would it then not be a novel idea for a corrections curve that is impletmented only at idle where heat soak sets in, so that you can tweek the corrections a little to maintain stoich, then be able to decay these corrections out with increased RPMs so that they do not effect fueling outside once the sensor is reporting real temps again? Why is there a MAT corrections table implemented to begin with if the ideal law should not be ignored? |
Originally Posted by Braineack
(Post 726013)
were you reading my last posts?
In any case, like wher it is headed. I think it is retarded that we have to adjust he MAT table tomake Gir 100% all the time. We should be looking for a modifie gas law equaion. I thoght the goal of the MAT table was to take into account the users unique AIT sensor characteristics, not create a table that our car depnds on to run correctly. Is the idal gas code not taking into accont the efficiency of the engine? I know we can reduce how extreme it reacts, but maybe it is off because its assuming 100% of the fue is being burned and converted to power. IDK, I'mrambling now so I'll leave it alone. |
you need the ideal code.
The Ideal Gas Law You might remember from high school physics classes that an ideal gas (which air is reasonably close to) obeys the relationship: PV = nRT Where: P = pressure, V = volume, n = number of moles (which is related to the mass of the gas, i.e. 1 mol = 6.023x1023 molecules of the gas, and n = mass (in grams)/molar mass(MM)), R = the ideal gas constant, and T = the absolute temperature. What does this have to do with fuel injection? In order to know how much fuel to inject, we need to know how much air is going into the engine so the chemically correct mixture (called “stoichiometricâ€) can be achieved. So for a fuel injected engine, we use sensors to determine the pressure in the intake manifold and the air temperature. However, the temperature in this equation is “absolute temperature†measured in Kelvins which is equal to degrees Celsius + 273º. The volumetric efficiency (VE) is a percentage that tells us the pressure inside the cylinder versus the pressure in the manifold. We know the volume (V) from the displacement of the engine. Thus we can calculate the mass of air (M) in the cylinder (proportional to n) from n = PV/RT => M = n x MM = PV/RT x MM = (VE * MAP * CYL_DISP) / (R * (IAT-32) * 5/9 + 273)) x MMair Since: P = VE * MAP (i.e. the pressure in the cylinder in kPa), V = CYL_DISP = the displacement of one cylinder (in liters), R = 8.3143510 J/mol K, and T = (IAT-32)* 5/9 + 273 to convert IAT from ºFahrenheit to Kelvin. Note that we can combine the constants R and MMair into one, and we will ignore them from this point on since they can be hard-coded into the assembly language code and neglected after that. Since we now know the amount of air in a cylinder from the MAP and IAT values and the 'tuned' value for VE, we need to know the amount of fuel to inject. We specify this with a parameter called REQ_FUEL. Now that it's getting warmer out I can see if denying the ideal law works. I have a feeling I will find that I may need to add more at idle but subtract at speed. MS1 handled this better where you could decay out the MAT corrections. |
the last part is the type of variable that im thinking about. i as well would prefer not to rely on ego correction to fix the afr, but it does work |
I had to change back to 50% correction value today. It just wasn't working for me a at 100%. The spread of 16% fuel from low temp to high temp seemed to cause problems. datalogging showed that all of the values in my MAT correction table were zeroing at Gair, but at the same time this morning wasn't adding in enough fuel while in boost.
I'm wondering if it is my new AEM wideband calibration. It might be safer to switch back to the standard calibration in tuner studio. It least then I was sure that my car was .3 rich. I can at least tune that out. With a 50% MAT value my fuel spread is only about 3%, compared to 16% with a 100% MAT correction value. This 3% spread is manageable since I don't usually get into boost until the car is warm. |
It seems a though everyone has the same issue of sorts. But I thought the MAT corrections tapers off at like 3K rpm. Is it effective in boost?
|
in MSI, like you have in your MSPNP, you can taper it out at whatever RPM you'd like. in MSII and MSIII, they disabled that feature.
|
Ok, good. I was really confused for a few. Is it good I taper it out or not?
|
Sorry to bring this back- is Brain's anti ideal gas law table on page one still the one to use now, and at 100% correction? :hustler:
|
Originally Posted by Greg G
(Post 763180)
Sorry to bring this back- is Brain's anti ideal gas law table on page one still the one to use now, and at 100% correction? :hustler:
|
I made my own, but it's very similar to Brains. I would start with Brains at 50% correction and adjust as necessary. i didn't like how 100% correction added/subtracted so much fuel. It seemed like a recipe for failure if a sensor went bad and stopped correcting in boost correctly.
|
I'd have to put my MS back and double check. or just look at my map. I have no issues in 100°F weather - with AIT sitting at 130°F at startup.
|
| All times are GMT -4. The time now is 08:04 PM. |
© 2024 MH Sub I, LLC dba Internet Brands