ULTM8Z

So I started playing around with the MAT tables and did a lot of reading on the previous threads and stickies, but I'm not quite getting it I think, including some that I started and posted in.

What's changed since the earlier threads I was involved in is that I now have a 12160244 LT1 MAT sensor in my cold air duct right now and the set up is nearly identical to the LT1's. The Enable Inv MAT table switch is enabled.

I had been running the factory ANHT MAT tables and then started looking at BDZL $DA3 (1993 LT1 Trans Am automatic trans) MAT tables.

When I plug in both tables verbatim, the car runs pretty choppy... hesitation and what not.

When I only plug in the Inv MAT look up table and keep the ANHT MAT compensation counts as is, it runs much smoother, but not much different than the AUJP tables in terms of power... not what I was expecting. I would have thought a more accurate accounting of air temperature by the ECM would have given some more power.

So what gives on the MAT compensation counts table? My MAT sensor and location is pretty much identical to the LT1. Why wouldn't running both tables be the way to go if my MAT set up is identical to the to the LT1. When do you have to change the MAT compensation counts table?

84Elky

Have also been playing around with this lately, so a couple of thoughts and questions.
What mask/bin are you using? What switch and bit are you referring to here? Is it 0x018 b0? If that's it, in AUJP and ANHT that switch only appears to affect whether MAT or Coolant is used to determine the adjustment of idle IAC steps during startup delay, so should have no effect once running. The switch that affects the calculation of gm/sec airflow is 0x018 b4. Stock AUJP and ANHT have b4=1 which says use the MAT Inv Lookup table at 0x80E to compute gm/sec air flow. if b4=0, that MAT table is never used and calculated gm/sec air flow will be much lower than if the table is used. Just something to check.


This seems strange because when comparing ANHT to AUJP, there appear to be no differences in the factory Comp counts tables at 0x7FD, and the only differences in the factory Inv MAT lookup table at 0x80E are:
---------------AJUP------ANHT
16 gm/sec----0.63------0.75
32 gm/sec----0.40------0.50
These difference would seem to have minimal effect on the calculated air flow. Is there possibly something else going on unrelated to the MAT tables?


The values in the 2 MAT tables for BDZL are radically different than AUJP and ANHT. That's because of the way the BDZL code apparently uses the values to calculate air flow. Can't just use BDZL values because the sensor is positioned similar to a LT1.


Most never change these table values (0x7FD, 0x80E) from factory settings despite running quite radical setups. The reason is because of the complicated code logic. Bottom line, the 2 MAT tables are used to compute a number of variables based on cylinder volume (part of the the air component), RPM, MAP, VE, EGR partial pressure, etc. All the computed values are then used for one and only one purpose -- to compute an injector PW based on the engine state at that moment. And PW rules! So mathematically, adjusting the MAT table values just adds one more variable to the PW calculation, which can more easily be affected by merely adjusting VE tables.


Power will generally only be affected if the calculated PW is correct resulting in the proper AFR.

ULTM8Z

Thanks for the detailed reply.

To answer your questions:

I'm running 7730 ANHT. However, I'm using Tunercat, so I'm not sure about the hex addresses.

There's a switch that enables the Inv MAT Lookup table. I have that enabled.

I mistakenly said AUJP in that one paragraph, when I meant ANHT. The comparison I meant to draw was between the BDZL Inv MAT look up and the ANHT Inv MAT Look up.

So BDZL $DA3 is different in that regard... it's strange because everything I use verbatim from BDZL seems to work great (particularly the AE).

At some point I'll need to break down and buy a WB O2 so I can check actual AFR. I've been tuning by seat of the pants up to now and have gotten it pretty good thus far... just trying to see if there's anything else to be gained.

84Elky

Hex addresses are as in my original post (AUJP and ANHT = same addresses). They merely represent where the item is located in the bin if open the bin in a hex editor.

Looking at ANHT, that switch is 0x018 b4 referenced in my post. It should be set.


Interesting the BDZL data works because I suspect the BDZL code that deals with the data is somewhat different than AUJP/ANHT. Maybe someone who is familiar with BDLZ code can comment. But again, those MAT tables appear to only be used to ultimately compute a PW. Proper VE settings will do the same thing without the agony of changing many things. The only other significant variables in the PW calculation are MAP and RPM which cannot be controlled. The only controllable items are VE and cylinder volume. Is cyl vol correctly calculated and entered for your setup? That value affects the air flow factor used in other PW calculations.

WBo2 certainly beneficial, but a lot can be determined without one. INTs and BLMs are INTs and BLMs with or without a WBo2.

Dominic Sorresso

I moved my MAT further upstream in the air intake tract. As a consequence, had to modify the existing MAT v CTS table. My objective was to maintain a consistent BLM even as the IATemps dropped or increased. I got pretty close to that particularly as I sat at idle where intake air temps vary more.

RBob

iTrader: (1)

Doesn't the LT1 use an intake air temperature (IAT) sensor, while the TPI is a manifold air temperature (MAT) sensor? This would explain the difference in table values.

Note this thread:

https://www.thirdgen.org/techbb2/sho...hreadid=265869

RBob.

Dominic Sorresso

RBob,

On the subject of MAT, what are Timed IAC steps v MAT and where does it help in the tune?

ULTM8Z

Yep. I have that sensor and I've printed out that thread and have re-read it many times trying to understand it. Some of this stuff simply goes over my head unfortunately...

ULTM8Z

Dominic,

In the link RBob posted, you had posted your MAT compensation tables, modded vs stock. But when I look at them, they look exactly the same... am I missing something?

Are what I'm seeing your modded tables? Or your stock tables?

Curious to see what you're actually running given your satisfaction at the response from the car.

Dominic Sorresso

ULT,

Look carefully, they are slightly different. However, the values I am using now are significantly different. I'll try posting values. Basically, where the stock table has 50/50 values in the 70-85C area, when I moved the MAT TO AN IAT, I needed to move those values down into the area of the table that the ECM would "see" most of the time. Not likely it'll see "MAT" temps above 45C for example.

ULTM8Z

So I did quite a bit of experimentation today...

I gave the LT1 Inverse MAT Lookup table another try and spent quite a bit of time really dialing in the VE's. They didn't change that much from when I was running the ANHT table, but tuning with the integrator, I must have smoothed out some lean spots or something because the car is much stronger now than when I was running the ANHT table (very noticeable improvement). There were areas where the integrator was hitting the mid 140's... so I got those down to low 120's... the engine seemed to like that.

I found and old thread of mine where I had put in the $DA3 LT1 table before and apparently noticed an improvement. But one thing and another over the years.... I changed cold air intake tubes, moved the MAT sensor around, matched my CCP system to factory third gen, etc... restarted my calibration from scratch a few times and I had apparently lost the recipe. So I couldn't figure out this time why I initially didn't see the same improvement.

But again, after getting the VE's dialed in via the integrator, I'm getting the improvement again.

I keep saying this... the closer I get to the $DA3 BDZL LT1 calibration, the better the car runs. The numbers seem to just port over verbatim

The only exception so far is the MAT compensation counts table. But things like Aceel Enrichment, throttle follower, LT4 Vette spark map (though modified to be more aggressive), and a host of other settings, transfer over remakably well.

Dominic, I'm still curious to see your current MAT settings though.

Dominic Sorresso

Here's the values I am using now. So this is a % MAT blend w CTS.

ULTM8Z

See, this makes sense the way it's set up. You tell the computer how to weigh the values coming in from the MAT and the CTS.

I'm still not sure I understand the MAT compensation counts table though.

Dominic Sorresso

So u can see how I moved the operating range down into more realistic temps for the IAT. However, the values for 80/92/110C are still the stock numbers. Didn't bother to change since I don't expect to see 180F intake air temps anytime soon. At least not for very long anyway. ��

84Elky

MAT Compensation counts and the Coolant/MAT Delta Tables -----

Looking at RBob's excellent explanation in the post above and what the code does, here's why IMO it's impossible to change these values and understand what's being affected without a lot of trial and error. Bear with me for a second in the example below as it's meant to show at a high level how complicated the use of these tables really is (assumes: Coolant = 190*F, MAT = 72*F, Air Flow = 32 gm/sec):

• A multiplier is first obtained from the Coolant/MAT Delta Table based on Air Flow (at 32 gm/sec, multiplier is .0.5)
• Inv MAT voltage from the sensor is next "scaled" (adjusted to a value consistent with coolant)
• The scaled MAT value is subtracted from coolant temp and the difference is multiplied by the 0.5 factor
• That result is multiplied by 2 and then another complicated hex multiplication and an addition to arrive at a final value used as input to get a value from the MAT counts table.
• In this example, the calculated input value to the counts table is 81 hex (129d) and the output is 97 counts. But looking at the factory counts table with that input you can't get that output (129 in would yields 128 counts out). That's because as RBob stated, the table label description of "MAT *C" (the input) is incorrect (and thus the values in that column are incorrect). Based on the gyrations above, what the heck is it actually? RBob calls the "T term" in his post, which is a massaged coolant/MAT differential -- NOT MAT.
• The 97 comp count value is then used with cylinder volume to compute a delta temp based/cylinder volume value for later PW calculation.

What???
That's what happens. So looking at all this, bottom line, higher values in the comp counts table will result in higher delta temp based/cylinder volume values, and thus higher PW.

Personally, I like Dominic's approach where there are 52-53 counts in the 15*C-36*C MAT range. But we know this is not really MAT; rather it's a massaged coolant/MAT differential. But the values are constant in a given range which will result in a constant delta temp based/cylinder volume value, which will result in a consistent input to the PW calculation. Are the values constant in the right range? I believe only trial and error can tell.

Dominic Sorresso

Trial and error is pretty much the process I used. A bit like playing "Battleship" with the tune. Once I got the VE pretty much where I wanted it, I then would look to see what direction overall BLMs went just watching the IAT temp, for instance, sitting at a light.
If they went higher, I would drop the MAT/CTS value emphasizing the "MAT" input more. I figured I was going to get close to the 50 number the stock cal was using in the operating range when the sensor was positioned in the air horn. The main observation was discovering what the typical temps would be for Intake Air between cruise and idle situations. Basically, that's what my table represents.

ULTM8Z

I seem to be having really good luck with the LT1 Inv MAT look up table. When I get everything dialed in around it, I'm going to revise my Miniram start bin.

msm z28

If the formula for the MAT compensation counts multiplies and divides by 128... is it doing so because of the code or because it's the value at zero airflow in the other MAT Table

Consequently when plugging in the LT1 MAT Comp tables which have values over 128 does this throw the math off if the 128 number is a constant in the code??

In other words are we just playing with the scale and going up or down?

Since we cant data trace the MAT Compensation Counts table we can't really be sure where the combination of coolant temp and air temp "BLEND" is effecting the change in VE.. we're only speculating

It's why last year I asked can we data trace the MAT Compensation table... My intake is similar to a miniram, but its a one off custom modified tunnel ram. Coolant temps go up and down depending on lots of variables along with air temps that go up or down... SO a one size fits all doesn't cut it... especially when people have different intakes with varying runner sizes and lengths

Thanks for anyone's input and clarity,
Sal

Vanilla Ice

Yes, we can trace any table.

84Elky

Sal --
Take a look at Post #7 in the link below and see if it answers your question or clarifies things. If not LMK.
https://www.thirdgen.org/forums/diy-...d-gm-make.html