[Cougar5.0]

Usually if OL Fuel calcs are off (yours seem to be too rich), it is due to injector slopes/offsets or MAF curve issues.

[Davin]

Patrick, I am about to make some changes, I have been looking at the differences between T4M0 and J4J1. It seems that the only multipler that is not "1s" in my tune is the Open Loop Fuel multiplier vs ACT. I can see that being a benefit since increasing it before had some benefit. I am also guessing that I will need to compensate by using the appropriate base fuel tables.

Also I though this was interesting..I was not aware of the "maf_temp_compensation" function..here are the stock tunes:

J4J1
-256 90 100 130 150 254
0 0 0.019531 0.039063 0.044922 0.044922

T4M0
-256 -256 -256 -256 -256 254
0 0 0 0 0 0

As you can see, my tune does not modify for temp vs maf in this function. Anyone know how this compensation works? Could having zero values be adding to my richness?

[Cougar5.0]

Everything else being the same, only slopes/offsets & MAF can mess up open loop fuel.

You can likley fix it by messing with those functions, but there is a fundamental issue if your OL calcs are off so much as to make the car stall.

[sailorbob]

Just gone back and re-read the start of this thread. Davin, are you still using the PRO-M transfer for your MAF? Someone shout out if I'm wrong but aren't the PRO-M transfer tables the old 'fudge' factor ones that are meant to be used with stock injector and MAF transfer table settings? If this is so then this is where a lot of your problems lie, you need to run a true MAF transfer table.

[patrickmx2]

See my comments mixed in below:

there is no doubt that the car is in OL when it stalls. If I make it to CL the LAMBSE goes up to around 14+ and everything is fine. It stalls after around 30 seconds the first time and the sooner with each subsequent attemt. Again, logging shows me that its definitly OL. I leaned the Open Loop Fuel Multiplier vs ACT function and made the hot startup LAMBSE around 12.1 instead of 11.1. This made it so that I could start and idle some of the time (instead of never)
Can you elaborate on turning adaptive control off?

Set the following to:
AFACT2 Adaptive Control Maximum ACT 254
AFACT1 Adaptive Control Minimum ACT 254

That math effectively turns it off.

Also, when you reference modifiers, are you referring to the open loop fuel multipliers in the functions? Yes

You said to use the J4J1 MAF transfer..are those based on 30# injectors?
First things first. If you are running a PRO-M meter, well that is probably your problem. See, those things lie to the EEC. They are manufactured with the assumption you don't have access to the transfer, aka, Tweecer. So they have to be calibrated. This is what I did with my car and Pro-M Meter. I run a STOCK electronics in my Pro-M housing. I can't speak for everyone but I happened to have two Pro-M meters. One was 24, the other was for 30's. Both housings were identical in size and sample tubing. I do know C&L uses different sample tubes. So I stuck my stock meter on the housing and used the factory MAF curve.

Think of it this way. It's just math and numbers. A given voltage is a given amount of air. The only way to get it right is to start with a base curve of numbers and let the EA watch your KAM's and it will aid you in recurving your transfer. For WOT you can curve the upper end with a wide band.

SO... If you use a J4J1 curve keep in mind you need a COBRA or equivalent sensor. If you only have a GT sensor that your car came with, use THAT sensor and factory curve for that car. I think I missed something when I recommended the J4J1 curve before. I was thinking this was a Cobra and this is a GT (looked back at your sig) Needless to say I do like the J4J1 tables over the U4P0.



The base fuel decay is also interesting, nearly double of my stock readings for the area of hot start...those values are subtracted from the base fuel LAMBSE right? Perhaps they are higher because the other multipliers are shut down... Could be but I'd have to sit down with a paper, pencil, and calculator probably because my brain is starting to hurt...

I'll try those other things you mentioned...

[patrickmx2]

Patrick, I am about to make some changes, I have been looking at the differences between T4M0 and J4J1. It seems that the only multipler that is not "1s" in my tune is the Open Loop Fuel multiplier vs ACT. I can see that being a benefit since increasing it before had some benefit. I am also guessing that I will need to compensate by using the appropriate base fuel tables.

Also I though this was interesting..I was not aware of the "maf_temp_compensation" function..here are the stock tunes:

J4J1
-256 90 100 130 150 254
0 0 0.019531 0.039063 0.044922 0.044922

T4M0
-256 -256 -256 -256 -256 254
0 0 0 0 0 0

As you can see, my tune does not modify for temp vs maf in this function. Anyone know how this compensation works? Could having zero values be adding to my richness?

Davin, are you familiar with the term LAMBDA? Let me see if I can help you understand some of those numbers.

14.64 = a Lambda of 1.

So 1=14.64

Now a modifier or multiplier of say .98 would mean this.

.98 x 14.64=14.3472

So .98 made it richer. Numbers below 1 are richer. (except for when you are looking at KAM's this sort of gets reversed... but lets not go there yet).

.88x14.64=12.8832 or basically an AFR of 12.88

likewise 1.06x14.64=15.518 So that is leaner. And that will be your commanded lambda, or Lambse.

BUT, BUT... remember 14.64 is stioch. Just a reference point.

Here is where the rubber meets the road. Take your fuel tables. If you have 13.53 in location that your load and temps line up on (row/column) you then take your Multiplier and apply it to THAT number.

.98X13.53= 13.26 a commanded AFR of 13.2594 rounded up.

So that is why I said null out your OL functions / multipliers. Because otherwise what ever you see in the tables will be different depending on what multiplier gets applied.

Last comment on this I hope. See, really you want the tables to be your foundation. Your BASIC information. Get that lined out first, then go about factoring multipliers back in. Temp, ACT/ ECT,,, various load points and so on.

I hope this helps.

[Davin]

I went ahead and change over to the Open Loop Fuel mult vs ACT and base fuel tables from J4J1 and the hot startup went really well (though it was only 85* out). This was probably b/c the LAMBSE was over 13:1 which is about +2 of what my T4M0 tables/multi were commanding. I get the concept of LAMBSE and multipliers, what I was unsure of is whether that much of a change was okay considering Ford seemed to think 11:1 was a better strategy. End result is it seems this might be fixed so thank you.

I think I have the stock MAF around somewhere. You're saying it would be better to use that one and the stock T4M0 MAF curve than the ProM w/ ProM curve right? It looks like they aren't taking new orders till July 27, so I guess I'll have to wait :(

[patrickmx2]

Great work everyone!

Hmm, lets clarify that Pro-M thing. See I'm running a Pro-M Housing but stock electronics. As usual I end up with parts all over my shop when friends come around working on their cars. So here is a for instance:

guy runs 19lb injectors and he wants a Pro-M meter. It will come in a box with housing and electronic meter attached to it already. It will be calibrated for 19lb injectors. Essentially they tweak the meter electronics due to the difference in air flow to better match a STOCK application. Problem: Cam, heads, intake are not stock now. All the load values will now be effected by what that meter sees.

Enter the Tweecer.

You can internally access the MAF transfer now. If you leave that PRO-M meter on it's supplied housing you are dealing with two things. The factory transfer (1) with Pro-M's modification (2) laid over the top of math and you will have to attempt to deal with the result. All that math is applied to your entire strategy. And the Pro-M math is unknown unless you attempt to enter in it's transfer to correct the issue back.

So... why not just put the factory sensor back on, but with Pro-M's larger housing and use the transfer it was calibrated for? Then you can recurve that single transfer and not have to deal with what I could only describe as two transfers blended together.

I attempted multiple combinations across many hours of tuning. Then it hit me. Hey, it's just an air pump. I just need a voltage. Most people only think of matching their injector size to a MAF. They forget about the transfer function in the EEC. Since you can access the injector size, the MAF transfer directly, you no longer need Pro-M's sensor and their efforts to blend it's voltages into your EEC's transfer. You have DIRECT access to that transfer with a tweecer so, change it, and your injector size.

This is before I ever fired the engine. Notice the Pro-M sensor still on the housing. I now am running that same intake system with a stock 19lb factory meter. Just the electronics were replaced with two screws. And yes, use the T4M0 curve as a foundation to build on.

[Davin]

I'm excited about trying out that strategy and letting the EA work for me. I bought it today, I'll just be waiting for the registration code. Unfortunatly while I kept the stock electronics, I didn't take good care and one of the wires is torn off a post. Last thing I'm wondering, can EA do just as good a job (or nearly) of perfecting a curve with my ProM 30# electronics or should I hit the junk yards and try to find a "used" T4M0 meter (and shell out the clams for it)?

[cgrey8]

You can tune a MAF and curve and injector slopes without a stock MAF. But it just takes longer and you do have to know what you are looking for when determining if you need to adjust the MAF or the injector slopes.

[patrickmx2]

Someone correct me if I'm wrong. I've gone to www.advanceautoparts.com

Click the below link. Then click the vehicles. I got this by searching for a 95 Mustang. One would get the idea that all of these cars share a few common things. Probably the MAF transfer, MAF, and injectors. Even spread across 4.6 and 5.0 platforms.

This reasoning is why I chose the route I took.

http://www.partsamerica.com/ProductGuid ... mber=29078

This is for a 1990 Mustang but it even fits V6 Taurus, 5.0 and 3.8 T-birds as well.

http://www.partsamerica.com/ProductGuid ... mber=29083

[Davin]

That's great information Patrick! I have a Ford-specific junk yard in my city that is very well organized. I'm guessing from what you posted that any MAF from a 94-95 would be the same. I guess the differences are in the strategies and not the electronics. Thanks

[2Shaker]

Davin,

After re-reading this thread I feel like we are still dealing with something more fundamental. I can't see where we ever got clarity on the location of the ACT sensor being eliminated as the culprit, and whether or not you determined the sensor is working right. Your ACT's are too high after the car sits a few minutes. You can tune around this with the ACT vs OL multiplier function and changes to the start up table, but I would be concerned that you are optimizing for this particular condition and some of the values you are putting in to fix it may adversely affect your tune under other scenarios.

[Davin]

2Shaker, yeah that's a good point. While the cobra tables have imporved things, there are still stalling issues. I was noticing that 45 seconds into a hot start, my ACT temp is still 146*. I did a resistance test on the sensor and it tested okay. It saw one place in my log where my temp while driving was around 110* but it raised to 132* while idleing at a light for 35 seconds. Shortly after that it went up to 150* before ending the log. I would have bought another just to see the difference but they run around $95. I bought a stock MAF sensor from the junk yard and I've been working on getting that working like Patrick suggested. I suppose I should take the time to seal the intake tube and see how what the sensor reads on regular underhood temps. Any other thoughts?

[2Shaker]

2Shaker, yeah that's a good point. While the cobra tables have imporved things, there are still stalling issues. I was noticing that 45 seconds into a hot start, my ACT temp is still 146*. I did a resistance test on the sensor and it tested okay. I would have bought another just to see the difference but they run around $95. I bought a stock MAF sensor from the junk yard and I've been working on getting that working like Patrick suggested. I suppose I should take the time to seal the intake tube and see how what the sensor reads on regular underhood temps. Any other thoughts?

Davin,

You want it to be exposed to air that is not from under the hood. I know the sensor is under there, and after a few minutes of being turned off the sensor and surrounding plumbing will become heat soaked. But as soon as you start cranking it will draw cool air from outside and the sensor should cool quickly so when it starts you should be reading temps close to ambient, unless you are also drawing the air from under the hood. And can you go back to the wrecking yard and get a sensor for cheap too?

[patrickmx2]

I'm excited about trying out that strategy and letting the EA work for me. I bought it today, I'll just be waiting for the registration code. Unfortunatly while I kept the stock electronics, I didn't take good care and one of the wires is torn off a post. Last thing I'm wondering, can EA do just as good a job (or nearly) of perfecting a curve with my ProM 30# electronics or should I hit the junk yards and try to find a "used" T4M0 meter (and shell out the clams for it)?

Since Clint returned from vacation, did you get the EA working? If so we need to get your MAF curve lined out before anything else. Also, injector slopes, etc... Let me know and we can dial that in first.

I have my ACT in the fender well area. Not in the actual air tubing.

[Davin]

I'll try relocating the sensor first and doing some logging.. if things don't change I'll look into another sensor. As I said above (edited)"It saw one place in my log where my temp while driving was around 110* but it raised to 132* while idleing at a light for 35 seconds. Shortly after that it went up to 150* before ending the log." I think heat soak is likely.

As far as the EA goes, yep, got it working and I've done about 7 logs/revisions on the curve since. The sensor I got is from a 92 crown vic but they said it was the same part #. I currently am using 30.499 and 37.999 for my high/low slopes. Not sure about my breakpoint and min pulse width..since I switched to v1.30B4 they read out as "0.00". I think in my old bin they were around ".00001975" and ".00000500". Load Scaling Switch is set at "2".

[patrickmx2]

Hey, have you seen this out there anywhere? I have a bad habit of cut and pasting but don't always grab the author. SO... I did not write this but it contributes to what I know about LOAD and LOADX.

[EDIT] I found the author. Dan Nicoson 1-1-06. I have provided a link to the original document below in this thread.

LOAD is used by the EEC to represent the power being used versus the power available at a given RPM. While this sounds an awful lot like Volumetric Efficiency (VE), in the purest sense LOAD is just a variable within the EEC algorithm needed for calculations during operation of the EEC. Please don't get caught up on weather or not LOAD is really VE, it doesn't matter.

It must be noted that LOAD can be calculated three different ways within the
T4M0 series of processors (and many others). There is a Load Scaling Switch in the scalars of most calibrations. On my 1994 Mustang GT the stock setting was "2".

LOAD Scaling Switch=2: This setting allows LOAD to be calculated with Throttle Position Sensor (TPS) and RPM. Possibly a temperature input to this calculation but essentially this situation is has very little ability to adapt to different parts combinations. Not a good selection other than for a near stock parts combination.

LOAD Scaling Switch=1: This setting calculates LOAD based on the airflow into the engine. Assuming accurate MAF calibration, the EEC looks at the mass of air flowing in based on the MAF, RPM, CID and temp and correlates this to the theoretical amount of air that could flow into the cylinder under these conditions. So even with wide open throttle, you might only see 75-80% LOAD at certain RPM.

LOAD Scaling Switch=0: This setting calculates LOAD exactly as in Load Scale Switch=1, but then correlates the LOAD to the VE function for the engine and results in LOADX. Here's an example (numbers are fictitious): At cruise & 2000 RPM on my 302 CID the EEC calculates a LOAD of 35%. Looking at the VE function the maximum LOAD for this parts combination is 70% of theoretical cylinder filling. So in this case LOADX = 35/70=50% LOADX. The reason for LOADX is to give you a wider range of numbers to scale your fuel and spark tables (more on scaling later). In theory if you have your VE function properly figured out you should get exactly 100% LOADX at every RPM at WOT with a naturally aspirated setup.

How is LOAD and LOADX affected by boost?

LOAD Scaling Switch=2: In this case you are screwed because the EEC will not know you have boost. Don't use this option with boost.

LOAD Scaling Switch=1: In this case you will see larger LOAD numbers as boost climbs. At a given RPM with 15 psi boost you should see twice the LOAD percentage than when NA. It is OK if LOAD is larger than 100%. The EEC recognizes LOAD values from 0 to 199%.

LOAD Scaling Switch=0: This case is ideal for boost. If you leave the VE
function set as it was for the naturally aspirated case (LOADX=100% at WOT,
every RPM) then 15 psi boost should get you right up to 199% LOADX. In
practice we would play with the VE function to keep the LOADX lower than the
max if you were running more than 15 psi boost.

[Davin]

yeah, I can't remember where I saw that...I think I have a link to it as a web page on my laptop...I had switched it to "0" but figured I should go back to "2" since I had to work out my MAF and didn't want that to be a factor...I wasn't sure if this used the MAF readings or the Tx too.

[patrickmx2]

Just my personal experience here, it has been easier for me to dial in my MAF while in LOAD=0. I think if I was dialing in WOT and the wide band and I was having trouble with my commanded AFR (OL fuel tables) I would go to LOAD=1 so that I would always know the row I was working with.

But you can't really scale load with a MAF that isn't correct because it depends on the MAF. So LOAD=1 might work better for other tuning issues while you were collecting data to curve the MAF transfer with. It would really depend on the situation but this is while a tune is in the infancy stage for sure. Not one you'd go lay down 1/4 mile passes with until the MAF curve was good.

I don't think I would ever see a need for LOAD=2, even on my stock SSP car.

[cgrey8]

If what you are referring to is Load Scaling Switch (PRLDSW), then I thought it could only be 0 or 1. I didn't think a larger value was valid. If it is, what is the significance of 2?

[Davin]

cgrey8 - check Patricks post on this page...

[sailorbob]

patrickmx2, I don't know where you got that quote from but it's wrong (badly). As far as the eec is concerned LOAD = VE and the 'Percent Load Switch' (aka PRLDSW) parameter determines the value of LOAD% where;

0 results in LOAD% = LOAD / peak LOAD
1 results in LOAD% = LOAD
2 results in LOAD% = inferred LOAD%

cgrey8, early eec's (e.g. GUF*) only have a value of 0 or 1, later eec's (e.g. CBAZA) have the three options.

EDIT: changed "inferred LOAD" to "inferred LOAD%"

[cgrey8]

...and the 'Percent Load Switch' (aka PRLDSW) parameter determines the value of LOAD% where;

0 results in LOAD% = LOAD / peak LOAD
1 results in LOAD% = LOAD
2 results in LOAD% = inferred LOAD
...

What is inferred LOAD?

...cgrey8, early eec's (e.g. GUF*) only have a value of 0 or 1, later eec's (e.g. CBAZA) have the three options.

That makes more sense. Thanks.

[sailorbob]

What is inferred LOAD?

Something else to confuse us

'Inferred LOAD' is a variation of LOAD that is determined from RPM, throttle angle and EGR% alone (i.e. without using the MAF or BP sensors).

'Inferred LOAD%' is the equivalent to LOAD% for the 'Inferred LOAD' value.

[cgrey8]

So is there any benefit in datalogging Inferred LOAD or tweaking to get Inferred LOAD to be close to actual? Or is it only used during a MAF failure?

[sailorbob]

Got me on both those questions, I don't know. I prefer to have LOAD% = LOAD, that way I always know where I am with things. I hate the idea of having different load ranges for different tables, if the engine hits 150% load I want both my fuel and spark tables to be roughly the same.

[patrickmx2]

I've got one car on LOAD=1, the other on LOAD=0. Both of those are GUFC. My CBAZA car is on LOAD=0.

On naturally aspirated setups like mine, I'm finding LOAD=1 easier to work with. I'm never ever over 100% LOAD. So I know exactly where the EEC is looking in the table at all times. If I'm at low loads, and hit open loop, I don't need it hitting too high in the tables. I've tried adjusting my FN035 but that is still difficult to pin down what row it might land on.

LOAD is largely determined by air flow. If your cubic inch has been correctly entered, and a given engine only has so much VE, meaning there will be a gap between peak load and 100% cylinder filling, then one way to look at it is where ever the engine hits a peak load is where you can safely assume it will be when you want to put in your commanded AFR in the table.

The problem is with boosted applications where you are as much as doubling displacement (or more) within the same confined physical cubic inch. Thus producing higher loads due to more air flow.

For instance a 5.0 with 14lbs of boost could be viewed as a 10.0 liter. You have two atmospheres in the combustion chamber. In simple terms there is only one way to get the EEC to get the AFR right. Help the MAF be truthful in the air volume, and give it a reference to that volume for fuel to go with it. It would seem to me that setting LOAD=1 and using a wide band on the DYNO that one could just richen the MAF curve to nail the AFR.

Then add or adjust the OL table to display loads over 100%. The problem is due to VE, if a given engine can only muster 80% load, and boosted with 14lbs it can do say 130%, what about that 20% of the VE curve below 100? Would it go to 160% at 14lbs (double the original 80%).

If you have two engines that are identical in displacement, but one with better heads, it should show a higher load due to moving more air at a given rpm. So it has a higher VE.

Ok, I'm getting off on several tangents here.... time for some aspirin.

[patrickmx2]

patrickmx2, I don't know where you got that quote from but it's wrong (badly). As far as the eec is concerned LOAD = VE and the 'Percent Load Switch' (aka PRLDSW) parameter determines the value of LOAD% where;

0 results in LOAD% = LOAD / peak LOAD
1 results in LOAD% = LOAD
2 results in LOAD% = inferred LOAD%

cgrey8, early eec's (e.g. GUF*) only have a value of 0 or 1, later eec's (e.g. CBAZA) have the three options.

EDIT: changed "inferred LOAD" to "inferred LOAD%"

FYI: I found it here. I had clipped part of it for a text file which basically just had the 3 forms of LOAD=

http://autos.groups.yahoo.com/group/twe ... 20Scaling/


sailorbob let me be sure to let you know I deeply respect your opinions and everyone here. I always want all of you to hear my typed words in a pleasant tone. Can you review that document and help with corrections or clear up some confusion? It was authored by Dan Nicoson 1-1-06

[sailorbob]

Dan Nicoson's usually pretty good when it comes to the eec, I don't know what's happened on this one (though I see a big disclaimer in there).

There's quite a bit of amendment required to that document and, TBH, I'm not sure if I'm the right person to tackle it (my inability to explain things well and bluntness being the reasons why :) ).

I consider that most of the problem with LOAD% is that people don't know what it's for (myself included). I *think* that it is there to indicate that maximum cylinder filling at a particular rpm is being achieved, i.e. VE at that rpm is 100%. If you alter the VE characteristics of an engine you cannot hope to get a correct LOAD% value unless you know how to adjust 'Peak Load at Sea Level vs RPM' (aka FN035) to achieve what I've theorised (most people just use it to get LOAD% down to 100% as a maximum and not throughout the rev range). Now on an engine that's capable of VE's greater than 100% this makes the concept of aiming for a LOAD% equal to 100% rather daft and, for example, on a forced induction ecu I have Ford didn't try to do this and set the 'Percent Load Switch' (aka PRLDSW) to 1.