MAT table starting place?

This is a forum for discussing the development and testing of alpha MS2/Extra code. Documentation
(Runs on MS2 and Microsquirt)

Moderators: jsmcortina, muythaibxr

AbeFM
Super MS/Extra'er
Posts: 875
Joined: Wed Dec 05, 2007 1:40 pm
Location: San Diego, CA
Contact:

MAT table starting place?

Post by AbeFM »

I just looked at my AIT table, and noticed it's 0's (well, 100's) all the way across. Maybe instead of richening by coolant temps, I could do it the right way, using MAT. Could someone give me a good idea of what values to try first???

Thanks!
-Abe.
2000 VVT Miata turbo, MS3Pro

Contact me if interested in a MS-II 2nd gen NB Miata PnP board.
Philip Lochner
Super MS/Extra'er
Posts: 1004
Joined: Thu Feb 16, 2006 6:18 am
Location: George, South Africa

Post by Philip Lochner »

If you have stim (or some other way of changing the IAT value) on the bench, set it at values that result in the same PW at all values of IAT. Then, once your VE table is tuned (it will have to be done with IAT not changing or very little), then you can tune the IAT table such that your AFR at a particular operating point (say 2000rpm, 50kpa) does not change as a result of changing IAT.

But first you you will have to tune your WUE table properly as per the manual.
Kind regards
Philip
'74 Jensen Interceptor 440ci (EFI'ed with MS2 and wasted spark + GM 4L60e GPIO controlled - both on Extra FW)
jsmcortina
Site Admin
Posts: 39587
Joined: Mon May 03, 2004 1:34 am
Location: Birmingham, UK
Contact:

Post by jsmcortina »

Having the same pulse width at all MAT values defeats the purpose of the speed-density equation as you have nullified the density term.

Using the table to tune the MAT factor correction is fine, but I cannot think of any circumstances where it should be ignored.

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".
AbeFM
Super MS/Extra'er
Posts: 875
Joined: Wed Dec 05, 2007 1:40 pm
Location: San Diego, CA
Contact:

Post by AbeFM »

I'm still a bit confused here. My first guess would be by PV=nRT that this correction would be the same for all cars.

Could someone at least tell me what there's looks like? It's only ten points, you could just type it or do a screen shot?

50 130
80 120
110 90

ETC. At least to give me a place to start.
2000 VVT Miata turbo, MS3Pro

Contact me if interested in a MS-II 2nd gen NB Miata PnP board.
Philip Lochner
Super MS/Extra'er
Posts: 1004
Joined: Thu Feb 16, 2006 6:18 am
Location: George, South Africa

Post by Philip Lochner »

jsmcortina wrote:Having the same pulse width at all MAT values defeats the purpose of the speed-density equation as you have nullified the density term.
I am aware of that James, but I feel that its better to start with a MAT table that does not lean the mixtures out as much as it does when left at 0. So, the idea was to rather provide rich fuel mixes (protecting the engine) with an untuned MAT table and then tune it downwards, than having to tune it upwards.

Taking this approach also demonstrates to the squirter how much the PW can change as a result of MAT - which I think is a good thing. Many squirters seem to underestimate the importance of MAT and tuning for it.

IMHO, MS leans out too much with its internal MAT compensation.
Last edited by Philip Lochner on Mon Mar 03, 2008 4:37 am, edited 1 time in total.
Kind regards
Philip
'74 Jensen Interceptor 440ci (EFI'ed with MS2 and wasted spark + GM 4L60e GPIO controlled - both on Extra FW)
Philip Lochner
Super MS/Extra'er
Posts: 1004
Joined: Thu Feb 16, 2006 6:18 am
Location: George, South Africa

Post by Philip Lochner »

AbeFM wrote: At least to give me a place to start.
That might not be as straightforward as you think. The % by which you must enrich is dependant on MANY factors, even the injection scheme you are using.

But if you really need to start somewhere, then start with 150% in the coldest bin reducing 5% per bin. Note, however, these figures may prove either way too much or way too little.

Your WBO will (should) tell you.
Kind regards
Philip
'74 Jensen Interceptor 440ci (EFI'ed with MS2 and wasted spark + GM 4L60e GPIO controlled - both on Extra FW)
spyro
Experienced MS/Extra'er
Posts: 229
Joined: Thu Jan 12, 2006 1:32 pm

Post by spyro »

just a question
100% means no correction ? or 0
Philip Lochner
Super MS/Extra'er
Posts: 1004
Joined: Thu Feb 16, 2006 6:18 am
Location: George, South Africa

Post by Philip Lochner »

In the WUE tables 100% means no correction. Your VE table should be tuned while WUE = 100%, baro, and MAT is constant.

In the MAT tables 10% means that 10% enrichment will be added to whatever enrichment results from Baro, WUE.

Eg if BARO = 90%, WUE = 140% and MAT = 10%, total enrichment should be 140% (90 + 40 +10).
Kind regards
Philip
'74 Jensen Interceptor 440ci (EFI'ed with MS2 and wasted spark + GM 4L60e GPIO controlled - both on Extra FW)
md95
Master MS/Extra'er
Posts: 721
Joined: Tue Mar 15, 2005 6:44 pm
Location: Grand Rapids, MI
Contact:

Post by md95 »

I tune the MAT correction tables using Ego correction referencing how much fuel is being added/subtracted at various MAT temps. From my experience, the MAT correction factor seems too agressive. 85 degrees F seems to be the "middle ground", and temps lower than this, add too much fuel, and temps over this, take away too much fuel. So I basically use 85-90 degrees as the start point, with no change in fueling. By watching the EGO correction amount, referencing the MAT temps, i was able to adjust the MAT table accordingly. I found out that the standard MAT correction factor seemed to be off almost in a linear fashion. I'd have to look, but i believe i used 5% adjustments for every 15-20 degree MAT temps. So for example, at 105 degrees, i would have a bin set for that, and the adjustment would be 5% fuel added, and so on. My EGO correction stays fairly consistant with all ranges of MAT temps(ego within +/- 5%). It's easiest to do this at idle. Only downfall to using the MAT correction table this way, is you really need true MAT temps, so adjusting the really cold temps, well, the air needs to be that cold. Basically this means that you need to adjust hot temps when its really hot out, and cold temps when its cold out. It's something that is going to take seasonal adjustments, and considerable amount of time to get perfect.
-Matt-
1998 Mitsubishi Eclipse RS 420A Turbo- MS2 Extra- 2.1.0 Release
1992 Plymouth Laser RS 4g63 AWD Turbo MS2 Extra- 3.0.3s
hassmaschine
Super MS/Extra'er
Posts: 1331
Joined: Mon May 21, 2007 8:36 am

Post by hassmaschine »

md95 wrote:I tune the MAT correction tables using Ego correction referencing how much fuel is being added/subtracted at various MAT temps. From my experience, the MAT correction factor seems too agressive. 85 degrees F seems to be the "middle ground", and temps lower than this, add too much fuel, and temps over this, take away too much fuel. So I basically use 85-90 degrees as the start point, with no change in fueling. By watching the EGO correction amount, referencing the MAT temps, i was able to adjust the MAT table accordingly. I found out that the standard MAT correction factor seemed to be off almost in a linear fashion. I'd have to look, but i believe i used 5% adjustments for every 15-20 degree MAT temps. So for example, at 105 degrees, i would have a bin set for that, and the adjustment would be 5% fuel added, and so on. My EGO correction stays fairly consistant with all ranges of MAT temps(ego within +/- 5%). It's easiest to do this at idle. Only downfall to using the MAT correction table this way, is you really need true MAT temps, so adjusting the really cold temps, well, the air needs to be that cold. Basically this means that you need to adjust hot temps when its really hot out, and cold temps when its cold out. It's something that is going to take seasonal adjustments, and considerable amount of time to get perfect.
I agree, the standard MAT correction is way off. the problem for me is I need huge corrections for high temps (or it gets very lean) - and because negative values are not allowed, it's still too rich when it's cold out.
6040solder
Experienced MS/Extra'er
Posts: 307
Joined: Mon Oct 22, 2007 7:15 am
Location: Auckland, New Zealand

Post by 6040solder »

hassmaschine wrote:I agree, the standard MAT correction is way off. the problem for me is I need huge corrections for high temps (or it gets very lean) - and because negative values are not allowed, it's still too rich when it's cold out.
Where is your sensor placed? The OEMs don't use the sensor the same way we do, so their stock placement usually isn't all that good for our purpose. You have to place it to avoid heat soak if you want it to be fairly accurate and useful.

On a boosted car, immediately after the intercooler behind the bumper outside the engine bay is best.

On an NA car, as close to the intake of the air from the outside as possible without it getting wet etc.

Lots of people have been having issues with this for years, usually careful placement alleviates the problem for them somewhere between totally and a lot.
md95
Master MS/Extra'er
Posts: 721
Joined: Tue Mar 15, 2005 6:44 pm
Location: Grand Rapids, MI
Contact:

Post by md95 »

The most accurate placement for a air temp sensor is in the intake manifold plenum itself. This measures the exact air temp of the air, before it's ingested into the cylinders. If you place the sensor after an IC, or any other farther away location, it's almost certain that actual air temp will change before it reaches the cylinders. It's not a big deal at higher rpm, but at idle it can be significant. But the problem is two fold. Having the sensor in the plenum, can cause heat soak, but this is only a problem with a hot engine, not running. Once the engine starts, air flow cools the sensor down, very, very quickly, and it's not much to worry about.(true for an "open element" sensor). If i had the choice, i'd put it as close to the intake throttle body or plenum as possible, mostly for accuracy. Mine is right before the Throttle body, in the IC outlet pipe. Never had a heat soak issue...if it did, after start airflow cooled it down almost instantly. I think some have problems with heat soak when they use a closed element temp sensor. They dont react nearly as fast.
-Matt-
1998 Mitsubishi Eclipse RS 420A Turbo- MS2 Extra- 2.1.0 Release
1992 Plymouth Laser RS 4g63 AWD Turbo MS2 Extra- 3.0.3s
6040solder
Experienced MS/Extra'er
Posts: 307
Joined: Mon Oct 22, 2007 7:15 am
Location: Auckland, New Zealand

Post by 6040solder »

md95 wrote:The most accurate placement for a air temp sensor is in the intake manifold plenum itself.
Only if you want to measure the temperature of the metalwork where you mount it and the radiant heat from those surfaces.
This measures the exact air temp of the air, before it's ingested into the cylinders.
Disagree.
If you place the sensor after an IC, or any other farther away location, it's almost certain that actual air temp will change before it reaches the cylinders.
See fluid dynamics 101 where the air only flows down the middle of the pipe and the air touching the pipes does not move at all. If radiant heat hit the air, we would all be very very cold.
It's not a big deal at higher rpm, but at idle it can be significant.
See below.
But the problem is two fold. Having the sensor in the plenum, can cause heat soak, but this is only a problem with a hot engine, not running.
Bingo, we have a winner. This is a MUCH greater problem than the heat absorbed while flowing through the pipe in a laminar state. MUCH.
Once the engine starts, air flow cools the sensor down, very, very quickly, and it's not much to worry about.(true for an "open element" sensor).
I agree, but that contradicts your "idle air temp" statement. If there is enough air flowing to cool down the sensor then there is enough air flowing that the heat absorbed by it per volume is negligible.
I think some have problems with heat soak when they use a closed element temp sensor. They dont react nearly as fast.
Those are as good as useless. People have these problems with open sensors too.

Reality is that the flow even at idle is significant enough to not change in temperature very much at all. It enters the engine much like it enters the car.

Consider this : If the code is not broken (and I have some faith in our two heros) and your sensor is calibrated correctly and setup correctly with NO heatsoak, you do not need any correction. That feature was added because of heatsoak.

Consider also just how much heat gets conducted into the sensor head by the vacuum present when at idle with the throttle closed. Not a lot. Pure vacuum does not conduct at all by definition.
md95
Master MS/Extra'er
Posts: 721
Joined: Tue Mar 15, 2005 6:44 pm
Location: Grand Rapids, MI
Contact:

Post by md95 »

The sensing element in an open element temp sensor does not conduct to it's body. Break one apart, you will see.

There is a lot of misconception about heat soak on the MAT sensor. I guess i wasnt clear enough. For example, if the sensor is in the manifold plenum, and reads 150 degrees, engine off, stagnant air flow, then the air temp REALLY IS that temperature. There is no way for it not to be. The heat from the engine has heated the air to that temperature. The sensor may be "soaked with heat" but it doesnt matter, since thats the air temperature. When you go to crank the engine over, that hot air is getting into the cylinders, and needs the appropriate amount of fuel to match the density. As air travels over the sensor, it cools down very quickly, within seconds. The inlet air temp has changed, and so has the sensor's readings. This is all normal, and is not considered heat soak.

Lets also take into account that we are not flowing air through a STRAIGHT metal pipe(in most installations anyway), there are curves, and angles of the piping, where the air stream will most certainly hit, and absorb heat. So the "fluid dynamics" theory doesnt hold much water there. In a real world installation, heat from intake piping WILL effect the incoming air temperatures, as will intake manifold walls, runners, ect.

The BEST place to measure incoming engine air temperature will be the closest place before it's ingested into the cylinder. This is the real world air temp that the cylinder will see. Even still, there will be temp changes from intake valve opening, to actual point in time of spark plug firing, but that kind of accuracy isnt really needed, we can tune around that, not to mention that there isnt a really economical, feasible way to measure cylinder air temperature before compression.

Take for example my issue i had with lean hot restarts, before i had used the MAT correction table. On a hot restart, the engine would run a bit lean, for maybe 10 seconds or so, until MAT temps cooled off a bit. This is what led me to believe that the MAT correction was off. It had to be, because it ran lean, and didnt run smooth. If it were true heat soak, the engine most likely would not have run that lean, since the actual air temp would have not been that hot to begin with, it was only the sensor saying it was.

I've looked at the mat factor, and indeed it's not broken, nor incorrectly written. It matches what the scientific notations say it's supposed to be for air temperature density. For some reason, in real world use, it's too agressive. Many, upon many have seen this problem, and its not "heat soak" that is the cause. The air temp really is that hot, and the MAT factor is pulling out too much fuel, causing the engine to run lean. These are my theories, and seem to match what others are finding out as well, and agree with. Scientifically, it doesnt make sense, but it's what's happening.

I tried alternate sensor placement, to NO effect. This pretty much proves my theory. This just wasnt done on one of my installs, this was with at least 4 or so installs i've done, and sensor placement had no meaningful effect. A few needed more correction than others, but all needed correction.

So to Abe- Start out with 85 degrees being the start point with no correction. Use the bins in the table to adjust 5% of fuel per 15-20 degrees in MAT temps. Lower by 5% under 85 degrees, and raise by 5% above 85 degrees. The table values are interpolated, so you dont need tons of bins anyway. Thats a good starting point, from all of my installs and tuning, and they all seem to center around these perameters.
-Matt-
1998 Mitsubishi Eclipse RS 420A Turbo- MS2 Extra- 2.1.0 Release
1992 Plymouth Laser RS 4g63 AWD Turbo MS2 Extra- 3.0.3s
6040solder
Experienced MS/Extra'er
Posts: 307
Joined: Mon Oct 22, 2007 7:15 am
Location: Auckland, New Zealand

Post by 6040solder »

md95 wrote:The sensing element in an open element temp sensor does not conduct to it's body. Break one apart, you will see.
Wow, levitating sensor elements, cool, can you explain how they work?
For example, if the sensor is in the manifold plenum, and reads 150 degrees, engine off, stagnant air flow, then the air temp REALLY IS that temperature. There is no way for it not to be. The heat from the engine has heated the air to that temperature.
Yes, it is, but, do you care to calculate precisely how long that air is going into the engine? That's right, its all gone in a few revolutions...

I'm out. Last time was somewhat futile too.
AbeFM
Super MS/Extra'er
Posts: 875
Joined: Wed Dec 05, 2007 1:40 pm
Location: San Diego, CA
Contact:

Post by AbeFM »

Philip Lochner wrote: In the MAT tables 10% means that 10% enrichment will be added to whatever enrichment results from Baro, WUE.

Eg if BARO = 90%, WUE = 140% and MAT = 10%, total enrichment should be 140% (90 + 40 +10).
What you're saying is that my 100%-100%-100% table means I'm getting TWICE the normal fuel? That doesn't seem right.

Also, you are saying they are additive and not multiplicative?

This seems really strange, and doens't make much sense to me. Could someone verify this?
2000 VVT Miata turbo, MS3Pro

Contact me if interested in a MS-II 2nd gen NB Miata PnP board.
AbeFM
Super MS/Extra'er
Posts: 875
Joined: Wed Dec 05, 2007 1:40 pm
Location: San Diego, CA
Contact:

Post by AbeFM »

md95 wrote:Take for example my issue i had with lean hot restarts, before i had used the MAT correction table. On a hot restart, the engine would run a bit lean, for maybe 10 seconds or so, until MAT temps cooled off a bit. This is what led me to believe that the MAT correction was off. It had to be, because it ran lean, and didnt run smooth. If it were true heat soak, the engine most likely would not have run that lean, since the actual air temp would have not been that hot to begin with, it was only the sensor saying it was.

So to Abe- Start out with 85 degrees being the start point with no correction. Use the bins in the table to adjust 5% of fuel per 15-20 degrees in MAT temps. Lower by 5% under 85 degrees, and raise by 5% above 85 degrees. The table values are interpolated, so you dont need tons of bins anyway. Thats a good starting point, from all of my installs and tuning, and they all seem to center around these perameters.
THANKS!!! Yeah, theory is nice, but I want to start with "what should I do that other folks are doing and works", then argue semantics. Since it's the closest thing I've gotten to a number, I'll try it.
Consider this : If the code is not broken (and I have some faith in our two heros) and your sensor is calibrated correctly and setup correctly with NO heatsoak, you do not need any correction. That feature was added because of heatsoak.
Oh oh oh! So the MAT *is* being watched. This is an additional contribution/correction? OMG I feel better. I was wondering how the motor was running at all as well as it was. I don't notice huge issues with temp as is.
2000 VVT Miata turbo, MS3Pro

Contact me if interested in a MS-II 2nd gen NB Miata PnP board.
6040solder
Experienced MS/Extra'er
Posts: 307
Joined: Mon Oct 22, 2007 7:15 am
Location: Auckland, New Zealand

Post by 6040solder »

AbeFM wrote:Oh oh oh! So the MAT *is* being watched. This is an additional contribution/correction? OMG I feel better.
Precisely :-)
Philip Lochner
Super MS/Extra'er
Posts: 1004
Joined: Thu Feb 16, 2006 6:18 am
Location: George, South Africa

Post by Philip Lochner »

AbeFM wrote:What you're saying is that my 100%-100%-100% table means I'm getting TWICE the normal fuel? That doesn't seem right.

Also, you are saying they are additive and not multiplicative?
The tables are additive but the result is multiplicative as follows:

If you have 110% 120% 130% the total enrichment will be 160% (100% + 10% + 20% + 30%) meaning the open time of the pulsewidth, if they were all 100%, will be multiplied by 1.6.

I find it odd that you refer to MAT enrichments as 100% because in my MS2/extra, nil MAT enrichment on the table will be 0% and not 100%.

See: http://www.msextra.com/ms2extra/MS2-Ext ... htm#matcor
Kind regards
Philip
'74 Jensen Interceptor 440ci (EFI'ed with MS2 and wasted spark + GM 4L60e GPIO controlled - both on Extra FW)
hassmaschine
Super MS/Extra'er
Posts: 1331
Joined: Mon May 21, 2007 8:36 am

Post by hassmaschine »

6040solder wrote:
hassmaschine wrote:I agree, the standard MAT correction is way off. the problem for me is I need huge corrections for high temps (or it gets very lean) - and because negative values are not allowed, it's still too rich when it's cold out.
Where is your sensor placed? The OEMs don't use the sensor the same way we do, so their stock placement usually isn't all that good for our purpose. You have to place it to avoid heat soak if you want it to be fairly accurate and useful.

On a boosted car, immediately after the intercooler behind the bumper outside the engine bay is best.

On an NA car, as close to the intake of the air from the outside as possible without it getting wet etc.

Lots of people have been having issues with this for years, usually careful placement alleviates the problem for them somewhere between totally and a lot.
it's actually in the same place as the stock sensor - open element as well (which is what the stock sensor was).

my problem isn't really heatsoak, but when it's freezing outside the car is rich, and when it's warmer it gets lean. so I have to tune my car when it's freezing and use lots of MAT correction to compensate for when it's warmer out. if I could input negative values, I could tune it to stoich when it's warmer and tune it to run the same when it's colder out.
Post Reply