Baro and MAT ( IAT ) correction.
Moderators: jsmcortina, muythaibxr
Re: Baro and MAT correction.
I have read this thread with interest as I am just about to move my IAT sensor from the ally plenum (on the a 3.9 rover V8). I am going to reposition it downstream in the air-filter. I have been having problems with the tune drifting which I believe is as result of heatsoak to the IAT sensor. Hot restarts and traffic jams! It is evident from the logs that IAT is soaking as the temp drifts up as the car idles and then drifts down on the move.
One thing that hasn't been mentioned regarding the airflow over the sensor, and I think may be significant in my case, is that air flow through a pipe or air-box must be laminar. This means that even though an open element sensor might be used the air it is measuring the temperature of will be the slower moving air at the perimeter of plenum that could be at a different temperature to the faster moving air towards the centre. Presumably through the large volume of the plenum chamber the overall air flow will be slower too.
It also seems that a lot of people report an improvement from placing the sensor further downstream since this will give a lower IAT reading anyway (regardless of it's accuracy) thereby reducing the amount of MAT correction applied.
One question though the % values in the MAT correction table. The higher the % value the less effect MAT has in the equation?
One thing that hasn't been mentioned regarding the airflow over the sensor, and I think may be significant in my case, is that air flow through a pipe or air-box must be laminar. This means that even though an open element sensor might be used the air it is measuring the temperature of will be the slower moving air at the perimeter of plenum that could be at a different temperature to the faster moving air towards the centre. Presumably through the large volume of the plenum chamber the overall air flow will be slower too.
It also seems that a lot of people report an improvement from placing the sensor further downstream since this will give a lower IAT reading anyway (regardless of it's accuracy) thereby reducing the amount of MAT correction applied.
One question though the % values in the MAT correction table. The higher the % value the less effect MAT has in the equation?
Rover P6 Estate. 3.9V8 with Toyota Supra 5 speed box. AND FOR THE LAST 2 YEARS!!! working MS2Extra running EDIS 8 ignition. Spends a lot of time in the garage being punished for misbehaving BUT!!! Just had a fresh burst of enthusiasm and I'm determined to nail those last few gremlins!!
-
- Super MS/Extra'er
- Posts: 1004
- Joined: Thu Feb 16, 2006 6:18 am
- Location: George, South Africa
Re: Baro and MAT correction.
As I understand it, a value in the MAT table of say 10% will increase the calculated PW with 10%, iow, the higher the value the MORE effect it has.BBLongman wrote:One question though the % values in the MAT correction table. The higher the % value the less effect MAT has in the equation?
PS: Your signature seems outdated - mentions carb...
Kind regards
Philip
'74 Jensen Interceptor 440ci (EFI'ed with MS2 and wasted spark + GM 4L60e GPIO controlled - both on Extra FW)
Philip
'74 Jensen Interceptor 440ci (EFI'ed with MS2 and wasted spark + GM 4L60e GPIO controlled - both on Extra FW)
Re: Baro and MAT correction.
So that means that 0% is the same as 100%?
What does "iow" stand for EDIT.(in other words )
Will update signature! It's been a long time since I was on here.
What does "iow" stand for EDIT.(in other words )
Will update signature! It's been a long time since I was on here.
Rover P6 Estate. 3.9V8 with Toyota Supra 5 speed box. AND FOR THE LAST 2 YEARS!!! working MS2Extra running EDIS 8 ignition. Spends a lot of time in the garage being punished for misbehaving BUT!!! Just had a fresh burst of enthusiasm and I'm determined to nail those last few gremlins!!
-
- Super MS/Extra'er
- Posts: 1004
- Joined: Thu Feb 16, 2006 6:18 am
- Location: George, South Africa
Re: Baro and MAT correction.
Yes, 100% meaning that which is calculated by the internal MAT algorithm. So 0% on the MAT graph means the internal algorithm value applies at that temp (which seems too much at higher values of MAT). 10% means the internal value is calculated but the result is increased by 10% at that value of MAT.BBLongman wrote:So that means that 0% is the same as 100%?
Kind regards
Philip
'74 Jensen Interceptor 440ci (EFI'ed with MS2 and wasted spark + GM 4L60e GPIO controlled - both on Extra FW)
Philip
'74 Jensen Interceptor 440ci (EFI'ed with MS2 and wasted spark + GM 4L60e GPIO controlled - both on Extra FW)
Re: Baro and MAT correction.
ressurecting this from the dead as i've been beating against this for a while.
while tuning my supra, i came up with the "counter mat corrections" which effectively places mat correction to 100% at all times, observe:
before i get yelled at, yes, i am ignoring laws of physics, don't care.
digging through a few dozen AEM cals and looking at how that is supposed to work, it seemed that the corrections were tuned afterwards, to help with the tuning process.
just today, i wondered if i could coerce the airden scaling into providing me with a flat map that i could use non-linear MAT correction to do what i've been trying to do with the counter mat correction, and yes, it's possible.
in the ini:
change to
sets the minimum airden_scaling to 0, but since the code says (ms2):
we have to set it to non-zero, so 1.
reload TS, set the MAT correction value to 1.
with MS2, this flat out disables airCorrection, it's 100% from 29F to 350F, according to my stim here
with MS3, airCorrection goes from 100% @ -9F IAT, to 99.8% @ 320F IAT.
while tuning my supra, i came up with the "counter mat corrections" which effectively places mat correction to 100% at all times, observe:
before i get yelled at, yes, i am ignoring laws of physics, don't care.
digging through a few dozen AEM cals and looking at how that is supposed to work, it seemed that the corrections were tuned afterwards, to help with the tuning process.
just today, i wondered if i could coerce the airden scaling into providing me with a flat map that i could use non-linear MAT correction to do what i've been trying to do with the counter mat correction, and yes, it's possible.
in the ini:
Code: Select all
airden_scaling = scalar, U08, 826, "%", 1, 0.00000, 50, 150.00, 0
Code: Select all
airden_scaling = scalar, U08, 826, "%", 1, 0.00000, 0, 150.00, 0
Code: Select all
if (flash5.airden_scaling) { // only if non-zero
ccor = (unsigned char)cor - 100;
cor = ((signed int)(ccor * flash5.airden_scaling)) / 100;
cor = cor + 100;
}
reload TS, set the MAT correction value to 1.
with MS2, this flat out disables airCorrection, it's 100% from 29F to 350F, according to my stim here
with MS3, airCorrection goes from 100% @ -9F IAT, to 99.8% @ 320F IAT.
2020 BMW X3M - bm3 - stage1
1994 Supra - ms3pnp pro - j&s
1994 Supra - ms3pnp pro - j&s
Re: Baro and MAT correction.
The above chart helped me cure the hot start issues on my car and also cleaned up my VE table. I put a MAT correction gauge next to the air temp gauge to show how drastically PW can change with the non linear MAT correction table zeroed out. I hooked up my MS2 to the stim board so all other parameters are the same.gurov wrote:ressurecting this from the dead as i've been beating against this for a while.
while tuning my supra, i came up with the "counter mat corrections" which effectively places mat correction to 100% at all times, observe:
before i get yelled at, yes, i am ignoring laws of physics, don't care.
I ran into the same problem with the MS1 setup on my other car and eventually 'resolved' it by moving the MAT sensor out into the bumper. With MS1, hot starts were really lean even with what I thought was a good VE table. In hind sight, the MAT correction was skewing my tuning efforts in stop and go traffic and after hot starts. Here is the anti MAT correction chart I ended up with.
This effectively cancels out the air temp correction to within 1%.
-
- Site Admin
- Posts: 39621
- Joined: Mon May 03, 2004 1:34 am
- Location: Birmingham, UK
- Contact:
Re: Baro and MAT correction.
Could you explain the physics of how that works?
James
James
I can repair or upgrade Megasquirts in UK. http://www.jamesmurrayengineering.co.uk
My Success story: http://www.msextra.com/forums/viewtopic ... 04&t=34277
MSEXTRA documentation at: http://www.msextra.com/doc/index.html
New users, please read the "Forum Help Page".
My Success story: http://www.msextra.com/forums/viewtopic ... 04&t=34277
MSEXTRA documentation at: http://www.msextra.com/doc/index.html
New users, please read the "Forum Help Page".
Re: Baro and MAT correction.
but if it matches where the corrections go ? we then have a dilema, things do not follow physics but make the engine run better. ain't this what tuning is all about, if everything followed laws of physics, we wouldn't need any of the correction tables.jsmcortina wrote:Could you explain the physics of how that works?
James
i've been changing airden scaling to 0-150 range (instead of 50-150) and setting it to 1 effectively cancelling out the SD corrections, then using small amounts of correction where i need them instead of using the builtin SD corrections.
not looking to start a flamewar here, but if it makes my engine happier, i don't really care about the physics of it.
2020 BMW X3M - bm3 - stage1
1994 Supra - ms3pnp pro - j&s
1994 Supra - ms3pnp pro - j&s
-
- Site Admin
- Posts: 39621
- Joined: Mon May 03, 2004 1:34 am
- Location: Birmingham, UK
- Contact:
Re: Baro and MAT correction.
OK, let me phrase that another way. WHY does it work?
I agree to a degree that empirical tuning "works" - sure it does. But the more we understand the "why", the more chance we have of writing code that more correctly represents what the engine needs. I also want the defaults to be "a good start" for most installs. That's one large reason why we have a VE table instead of a raw pulsewidth table. Enter a few parameters and the result is likely to start your engine.
James
I agree to a degree that empirical tuning "works" - sure it does. But the more we understand the "why", the more chance we have of writing code that more correctly represents what the engine needs. I also want the defaults to be "a good start" for most installs. That's one large reason why we have a VE table instead of a raw pulsewidth table. Enter a few parameters and the result is likely to start your engine.
James
I can repair or upgrade Megasquirts in UK. http://www.jamesmurrayengineering.co.uk
My Success story: http://www.msextra.com/forums/viewtopic ... 04&t=34277
MSEXTRA documentation at: http://www.msextra.com/doc/index.html
New users, please read the "Forum Help Page".
My Success story: http://www.msextra.com/forums/viewtopic ... 04&t=34277
MSEXTRA documentation at: http://www.msextra.com/doc/index.html
New users, please read the "Forum Help Page".
Re: Baro and MAT correction.
okay, i see where you're going with that. i wish i knew why this does this. the VE table works, and it works well, so does the AFR target with incorporate AFR, that takes care of majority of the changes i make, so afr target change doesn't take all that much driving with VEAL to get right (although, i have to keep restarting TS to clear out the cell weights, as they stop changing after a while, without really getting to target, but i get to take that up with phil)jsmcortina wrote:OK, let me phrase that another way. WHY does it work?
I agree to a degree that empirical tuning "works" - sure it does. But the more we understand the "why", the more chance we have of writing code that more correctly represents what the engine needs. I also want the defaults to be "a good start" for most installs. That's one large reason why we have a VE table instead of a raw pulsewidth table. Enter a few parameters and the result is likely to start your engine.
James
answer is, i do not know, but i do know that with enough datalogs, it should be possible to extract the correlation of the temperatures to the corrections needed. i'm even to the point of extending water temp corrections past the operating temperature, as that 7-10 degree water temp swing causes some fueling differences from what i'm seeing. but again, no datalogs on this as i keep changing settings far too often, trying to get it to better drivability.
2020 BMW X3M - bm3 - stage1
1994 Supra - ms3pnp pro - j&s
1994 Supra - ms3pnp pro - j&s
Re: Baro and MAT correction.
I think its because there is no perfect spot in the engine bay where sampling the temperature directly relates to the temperature of the air entering the cylinder at all times. Its sort of an impossible problem in my opinion. I struggled with this to no end on my MS1 car and eventually settled on positioning the air temp sensor outside the engine bay. That works pretty well. On MS2 I'm giving this a try because I don't have an effective cold air intake yet. So far it seems to work very well at all points above 70f but its about 3% lean at 40f. More testing is in order.jsmcortina wrote:OK, let me phrase that another way. WHY does it work?
I agree to a degree that empirical tuning "works" - sure it does. But the more we understand the "why", the more chance we have of writing code that more correctly represents what the engine needs. I also want the defaults to be "a good start" for most installs. That's one large reason why we have a VE table instead of a raw pulsewidth table. Enter a few parameters and the result is likely to start your engine.
James
cheers
Justin
-
- Site Admin
- Posts: 39621
- Joined: Mon May 03, 2004 1:34 am
- Location: Birmingham, UK
- Contact:
Re: Baro and MAT correction.
I too have my air temp sensor under the front bumper.
In the MS3 code I added the MAT/CLT blending option with the theory that the air charge is getting more heating from the intake, valves, cylinder etc. at lower rpms. But I've not really tried it myself.
So, my thrust is not "you must use the default corrections" but more a case of why aren't they working and what can we do better to make it easier for all of us.
The MAT sensor position varies wildly in OEM installs and I'm sure I recall Ken mentioning some installs that use two - one in the intake and one out under the bumper.
James
In the MS3 code I added the MAT/CLT blending option with the theory that the air charge is getting more heating from the intake, valves, cylinder etc. at lower rpms. But I've not really tried it myself.
So, my thrust is not "you must use the default corrections" but more a case of why aren't they working and what can we do better to make it easier for all of us.
The MAT sensor position varies wildly in OEM installs and I'm sure I recall Ken mentioning some installs that use two - one in the intake and one out under the bumper.
James
I can repair or upgrade Megasquirts in UK. http://www.jamesmurrayengineering.co.uk
My Success story: http://www.msextra.com/forums/viewtopic ... 04&t=34277
MSEXTRA documentation at: http://www.msextra.com/doc/index.html
New users, please read the "Forum Help Page".
My Success story: http://www.msextra.com/forums/viewtopic ... 04&t=34277
MSEXTRA documentation at: http://www.msextra.com/doc/index.html
New users, please read the "Forum Help Page".
Re: Baro and MAT correction.
Was out messing with the car out there in the snow, and extended the WUE curve past 180 up to 200f water temp, with 200f at 95%. seems to be keeping slightly closer to 100 ego, which is my goal.
2020 BMW X3M - bm3 - stage1
1994 Supra - ms3pnp pro - j&s
1994 Supra - ms3pnp pro - j&s
Re: Baro and MAT correction.
I'm leaning toward doing this as well with the supplied mat correction. I'm just concerned that I'm sucking in air some times that is much hotter from under the hood. I guess the worst that could happen is it'll run slightly rich for a short period. The MAT correction seems to work well with large air flow and work poorly during low air flow. Idle vs load for example. Also reading through a few other threads, I think the fuel in the rail is heat soaking as well. So many variables, not enough information.jsmcortina wrote:I too have my air temp sensor under the front bumper.
...
James
Re: Baro and MAT correction.
After more testing, it looks like canceling out the MAT correction wasn't the panacea that I expected. Cold start in the 30s was lean and hot start around 120f (manifold temp) was rich and staid rich while idling. I changed the table MAT correction back to all zeros and moved the temp sensor to just behind the front license plate. I'll give this a try for a whole and report back. I stand corrected.
Justin
Justin
Re: Baro and MAT correction.
Rather than disable MAT correction, I used After Start Enrichment to help with hot lean starts. If the coolant thermostat opens at 180 F, then increase ASE percentage and taper for CLT = 190, 200 and 210 . It's not a perfect solution by any means, but it always starts now, even on really hot days after a 5-20 minute soak time. FYI, the intake and exhaust are both on the same side for my engine, and I have an unwrapped turbo under it also, so heat soak of the fuel rail and MAT sensor are common occurrences.
Daniel
'73 240Z
Castle Rock, CO USA
'73 240Z
Castle Rock, CO USA
Re: Baro and MAT correction.
Has anybody determined the best location for boosted applications? The main reason I went away from the stock ECU was to better control IAT fueling since those settings were not very good.
My MAT sensor is right before the throttle body measuring the hot compressed air, normally 25 degrees above ambient when at full boost ~10psi. I would actually prefer a slightly rich AFR as result of higher MAT temps to avoid detonation.
My MAT sensor is right before the throttle body measuring the hot compressed air, normally 25 degrees above ambient when at full boost ~10psi. I would actually prefer a slightly rich AFR as result of higher MAT temps to avoid detonation.
MS3X on 2010 Camaro LS3 + Whipple @ 10psi + HPT + AEM 30-0333/0334 WB
MS3 Knock Module + JBPerf Dual VR v2.1 (Removed), didnt play well with LS3 DBW Motor, works for VSS.
OLD: 1997 Chevy Z71 Vortec 350, + BOSCH 0261231036 WB Knock Sensor
OLD: LT1 TA 700hp + MS3X
MS3 Knock Module + JBPerf Dual VR v2.1 (Removed), didnt play well with LS3 DBW Motor, works for VSS.
OLD: 1997 Chevy Z71 Vortec 350, + BOSCH 0261231036 WB Knock Sensor
OLD: LT1 TA 700hp + MS3X
Re: Baro and MAT correction.
jsmcortina wrote:How about a code feature to scale down the correction?
MS1 used a hardcoded air density lookup table, so that was hard, but MS2 does this:
Could be changed to include something like. (This variable doesn't exist but would be EASY to add.)Code: Select all
int aircor_eq(int mat) { // returns air density correction from equation in % (100 is no correction) // mat in deg F x10 or deg Cx10 /* Following equations used: For Farenheit, air density (pounds/cubic feet) = .0391568 x map (kPa x 10) ------------------------- ((mat(degFx10)/10) + 459.7) If we take 70 degF as standard temperature (no density correction) and ignore map, which is already accounted for elsewhere,then air density correction(%) = (70 + 459.7) x 100 ------------------ ((mat/10) + 459.7) For Celsius, with a 20 deg C reference and mat in deg C x 10 air density correction(%) = (20 + 273.2) x 100 ------------------ ((mat/10) + 273.2) */ if(flash4.Temp_Units == 0) { // deg F return((int)(529700 / (mat + 4597))); } else { // deg C return((int)(293200 / (mat + 2732))); } }
Where a value of 100 equals normal correction and zero is no correction. We could then try to "turn down" the airden correction.Code: Select all
* airden_scaling / 100
From the above equation, the actual values for Philip's table (at those temps) should be:
-7.3
-3.5
0.0
+3.3
+9.3
+14.6
James
EDIT: changed baro to airden.
Hello James,
I did read with a lot of interest this topic.
May i suggest something (maybe am i wrong, but...) ?
If the correction is based on the temperature of the fresh air only, and i take for instance air @ 20°c and air @ 40°c
The mass ratio would be
(273+20) / (273 +40) = 93,6%
BUT don't we have to take into account the fact that the air will be warmed up once it is in the cylinder head ?
Take for instance a very raw rule saying that it will receive 50 extra degrees from the head (i know it is not excatly how it works, it is just to start somewhere)
In that case the mass ratio would be
(273+20+50)/ (273+40+50) = 94,5%
Which makes a difference of 16% !!!
What do you think about it ?
Pierre
Re: Baro and MAT correction.
Mote than 1%, sorry...PierreL wrote:jsmcortina wrote:How about a code feature to scale down the correction?
MS1 used a hardcoded air density lookup table, so that was hard, but MS2 does this:
Could be changed to include something like. (This variable doesn't exist but would be EASY to add.)Code: Select all
int aircor_eq(int mat) { // returns air density correction from equation in % (100 is no correction) // mat in deg F x10 or deg Cx10 /* Following equations used: For Farenheit, air density (pounds/cubic feet) = .0391568 x map (kPa x 10) ------------------------- ((mat(degFx10)/10) + 459.7) If we take 70 degF as standard temperature (no density correction) and ignore map, which is already accounted for elsewhere,then air density correction(%) = (70 + 459.7) x 100 ------------------ ((mat/10) + 459.7) For Celsius, with a 20 deg C reference and mat in deg C x 10 air density correction(%) = (20 + 273.2) x 100 ------------------ ((mat/10) + 273.2) */ if(flash4.Temp_Units == 0) { // deg F return((int)(529700 / (mat + 4597))); } else { // deg C return((int)(293200 / (mat + 2732))); } }
Where a value of 100 equals normal correction and zero is no correction. We could then try to "turn down" the airden correction.Code: Select all
* airden_scaling / 100
From the above equation, the actual values for Philip's table (at those temps) should be:
-7.3
-3.5
0.0
+3.3
+9.3
+14.6
James
EDIT: changed baro to airden.
Hello James,
I did read with a lot of interest this topic.
May i suggest something (maybe am i wrong, but...) ?
If the correction is based on the temperature of the fresh air only, and i take for instance air @ 20°c and air @ 40°c
The mass ratio would be
(273+20) / (273 +40) = 93,6%
BUT don't we have to take into account the fact that the air will be warmed up once it is in the cylinder head ?
Take for instance a very raw rule saying that it will receive 50 extra degrees from the head (i know it is not excatly how it works, it is just to start somewhere)
In that case the mass ratio would be
(273+20+50)/ (273+40+50) = 94,5%
Which makes a difference of 16% !!!
What do you think about it ?
Pierre
-
- Super MS/Extra'er
- Posts: 1004
- Joined: Thu Feb 16, 2006 6:18 am
- Location: George, South Africa
Re: Baro and MAT ( IAT ) correction.
For those interested in barometric timing advance correction for normally aspirated high compression engines, have a read on this:
http://www.msextra.com/forums/viewtopic ... 01&t=46373
http://www.msextra.com/forums/viewtopic ... 01&t=46373
Kind regards
Philip
'74 Jensen Interceptor 440ci (EFI'ed with MS2 and wasted spark + GM 4L60e GPIO controlled - both on Extra FW)
Philip
'74 Jensen Interceptor 440ci (EFI'ed with MS2 and wasted spark + GM 4L60e GPIO controlled - both on Extra FW)