PDA

View Full Version : More E38 weirdness...



gmtech16450yz
February 7th, 2009, 06:10 AM
Ok, this is interesting...

I just looked at an EFILive tuned '08 GMC Sierra 6.0 VMax ECM (E38) with both EFILive and HP Tuners programs and noticed a few weird things, A/C disable/enable was one of them and DFCO was another.

In EFILive the A/C disable rpm is set to 4000 rpm, the re-enable is showing 3800 rpm. If you look at the same tune in HP Tuners it shows the opposite- disable shows 3800 rpm and re-enable shows 4000.

In DFCO B2409 and B2410 in EFILive do not seem to make any difference in this OS (12617631), putting in any number does not change the vehicle at all. I noticed in EFILive the DFCO disable (B2409) shows 1000 rpm and enable (B2410) shows 1300 rpm. In HP Tuners disable shows 1050 rpm and enable shows 1700 rpm. Funny thing is the DFCO on the truck turns off when decelerating and it goes below exactly 1700 rpm, so I'm thinking the HP Tuners may be correct? (I haven't tried reflashing with different numbers in HP Tuners yet.)

Anybody know what's up?

gmh308
February 7th, 2009, 04:23 PM
Ok, this is interesting...

I just looked at an EFILive tuned '08 GMC Sierra 6.0 VMax ECM (E38) with both EFILive and HP Tuners programs and noticed a few weird things, A/C disable/enable was one of them and DFCO was another.

In EFILive the A/C disable rpm is set to 4000 rpm, the re-enable is showing 3800 rpm. If you look at the same tune in HP Tuners it shows the opposite- disable shows 3800 rpm and re-enable shows 4000.

In DFCO B2409 and B2410 in EFILive do not seem to make any difference in this OS (12617631), putting in any number does not change the vehicle at all. I noticed in EFILive the DFCO disable (B2409) shows 1000 rpm and enable (B2410) shows 1300 rpm. In HP Tuners disable shows 1050 rpm and enable shows 1700 rpm. Funny thing is the DFCO on the truck turns off when decelerating and it goes below exactly 1700 rpm, so I'm thinking the HP Tuners may be correct? (I haven't tried reflashing with different numbers in HP Tuners yet.)

Anybody know what's up?

EFILive looks like it is correct for the A/C on/off.

And with DFCO, do you mean DFCO disables > goes from DFCO is enabled and fuel is cut, to fuel flow is = on/DFCO disabled and engine operating "normally", as the rpm passes through 1700rpm and is decreasing?

Reason for clarification: When DFCO is "on", "it" the fuel, is off, and DFCO should be "on" when decel anyway until it gets to disable rpm.

Have seen this with other parameters too, between EFI and HPT. Like speedo. The number in the ECM was the same, they both scaled it out differently. On one, the cal number was presented incorrectly......guess wHicH?

gmtech16450yz
February 7th, 2009, 05:17 PM
And with DFCO, do you mean DFCO disables > goes from DFCO is enabled and fuel is cut, to fuel flow is = on/DFCO disabled and engine operating "normally", as the rpm passes through 1700rpm and is decreasing?

Yes, that's what I meant.

On the A/C, in the stock tune the disable and enable rpms are both 0, TPS is 100. Looks like all they're working off of is shutting it off at full throttle only.

Thanks for your reply.

gmh308
February 7th, 2009, 11:55 PM
Yes, that's what I meant.

On the A/C, in the stock tune the disable and enable rpms are both 0, TPS is 100. Looks like all they're working off of is shutting it off at full throttle only.

Thanks for your reply.

You're welcome. Which trans in this truck?

GMPX
February 8th, 2009, 10:26 AM
In EFILive the A/C disable rpm is set to 4000 rpm, the re-enable is showing 3800 rpm. If you look at the same tune in HP Tuners it shows the opposite- disable shows 3800 rpm and re-enable shows 4000.
To make sense of this you have to think about what the purpose of the function is and it is to protect the A/C compressor and give a little more power at High RPM.
So looking at EFILive's numbers, you want to shut the compressor off at 4,000 RPM and only allow it to come back on once the RPM falls below 3,800. That makes sense right?


In DFCO B2409 and B2410 in EFILive do not seem to make any difference in this OS (12617631), putting in any number does not change the vehicle at all. I noticed in EFILive the DFCO disable (B2409) shows 1000 rpm and enable (B2410) shows 1300 rpm. In HP Tuners disable shows 1050 rpm and enable shows 1700 rpm. Funny thing is the DFCO on the truck turns off when decelerating and it goes below exactly 1700 rpm, so I'm thinking the HP Tuners may be correct? (I haven't tried reflashing with different numbers in HP Tuners yet.)

Anybody know what's up?
This one can be a little more confusing, B2409 is used if DFCO is active then it will be disabled once it falls below that RPM value (1000RPM), to allow it to be re-activated it must rise above B2410 (1300RPM), I think you will find that the ECM is not using RPM as the main qualifier to have DFCO active, it will be an airflow thing, keep an eye on the airflow values and compare that to B2404.

Cheers,
Ross

Redline Motorsports
February 8th, 2009, 03:28 PM
What I don't understand, and see there points, is that how can ANY software that is designed to read the same parameters be expressed in different values?

Why should IFR tables be differently scaled same year same OS? Just as an example. Is it avoid copy/paste issues????

Howard

gmh308
February 8th, 2009, 06:33 PM
What I don't understand, and see there points, is that how can ANY software that is designed to read the same parameters be expressed in different values?

Why should IFR tables be differently scaled same year same OS? Just as an example. Is it avoid copy/paste issues????

Howard

Just not a perfect science mapping "black boxes" from the outside. :shock:

GMPX
February 9th, 2009, 09:16 AM
Why should IFR tables be differently scaled same year same OS? Just as an example. Is it avoid copy/paste issues????
GM have changed the IFR table structure I think 3 times since the E38 hit the streets, in table size and scaling, that has nothing to do with EFILive, apart from annoying us :)

Cheers,
Ross

gmtech16450yz
February 9th, 2009, 02:12 PM
Ok, first off this is NOT an EFI vs. HPT situation, I'm just trying to figure out some things that haven't made sense and I couldn't really care less which program gives me the answers. Truth is using BOTH programs has given me the answers.

I had a chance to play around with the A/C and DFCO, here's what I learned so far...

On the A/C, EFI reads the enable and disable correctly as far as which is which, but the rpm is only an enabler for the tps check. If you put the rpm at 0, the A/C will never cut out no matter what the tps is. It only cuts out at the tps setting if the rpm is BELOW the set numbers. If you set the rpm at 7000, it will then cut out at whatever tps you set, at whatever rpm.

On DFCO, EFI is not reading the numbers correctly. I can change the enable and disable in HPT and the truck definitely changes to exactly what I set the rpms at. If I read the HPT tune in EFI it still shows the same numbers it always does, they don't change even though they definitely have been changed by HPT. If you change the numbers in EFI, they have no effect on the truck. (I'm talking about B2409, B2410 and B2415.)

Yes I understand how and why the A/C cutout and DFCO work, that's not the problem. And yes, I've already figured out the airflow/DFCO correlation, stock settings would only allow DFCO if you chopped the throttle at about 80mph! There are a few parameters I've tried to work with and found that even going from one extreme to the other on numbers, they make no difference what so ever to the actual vehicle. Using HPT has shown me what I suspected, there are parameters that aren't working properly. At least in this particular OS on this particular vehicle. And yes, I understand reverse engineering this stuff is extremely difficult and I am SO happy you all have been able to figure how to get into GM's programming and be able to modify it.

Redline Motorsports
February 9th, 2009, 03:48 PM
Nobody was making it a debate but the fact that both softwares should be reading the same info and don't was the point. I'm not a computer hacker so pardon my ignorance but the tables weather read buy either one should be the same. I have seen this myself on many occasions.

gmtech16450yz
February 9th, 2009, 04:22 PM
Nobody was making it a debate but the fact that both softwares should be reading the same info and don't was the point. I'm not a computer hacker so pardon my ignorance but the tables weather read buy either one should be the same.

Exactly!

I know nobody was making it a debate, I was just mentioning both programs quite a bit and I wanted to make sure nobody thought I was bashing one program or the other. Just trying to figure things out.

Thanks.

gmh308
February 10th, 2009, 09:12 AM
Nobody was making it a debate but the fact that both softwares should be reading the same info and don't was the point. I'm not a computer hacker so pardon my ignorance but the tables weather read buy either one should be the same. I have seen this myself on many occasions.

It would be great if they did. :) Unfortunately, in many/most/all(?) cases the ECM does not store data in plain old real world numbers. Scaling the data out into both metric and imperial (two types of imperial?) from the native format, often in unknown units (even though the ECM's internals are purportedly all metric), must be a real challenge to get perfect all the time. :shock:

GMPX
February 10th, 2009, 12:22 PM
Nobody was making it a debate but the fact that both softwares should be reading the same info and don't was the point. I'm not a computer hacker so pardon my ignorance but the tables weather read buy either one should be the same. I have seen this myself on many occasions.
Ok, look at the picture below, EFILive turns that in to a nice 3D map for you all to tune with, how we turn that in to a map is where things can go wrong, I can't call GM software engineering and ask them how much do I multiply the numbers in that table by to figure out the real value, we have to try to figure that out by ourselves.
This is where you see some differences between all after market software companies, EFILive, HPT, LS2Edit, we are all in the same boat. Reverse engineering is not an exact science, but I seriously doubt that the car companies would ever release tuning data to the after market for obvious reasons.
I am sure if you asked HPT or LS2Edit they would not provide you with any sort of guarantee that every parameter / table is 100% as GM know it.
There is also the human factor, I have been guilty of fat fingers entering incorrect table address's, it happens. As an example, for the E38 / E67 we now support 74 operating systems for them with approx 600 parameters/tables per OS, that is 44,400 address's that we need to define, find me a person that could type 44,400 24bit address's without making one mistake and he/she can have a job right away.

Ian is also correct on the Metric issue, the ECM works in Metric, sometimes GM even change the scaling between OS's which also confuses things. EFILive is native Metric too, so all development is done in metric, sometimes the conversion back to imperial is not set up correctly so you might see errors there too. We have worked hard to try to fix that situation on the upcoming Version 8 software.

Cheers,
Ross

gmh308
February 10th, 2009, 12:39 PM
As an example, for the E38 / E67 we now support 74 operating systems for them with approx 600 parameters/tables per OS, that is 44,400 address's that we need to define, find me a person that could type 44,400 24bit address's without making one mistake and he/she can have a job right away.

Cheers,
Ross

44.4k addresses for E38/E67! Stunning. A single E38 has ~17k data points in timing alone. Multiply that by 74 too! I knew it was a bigger number than I could work on fingers and toes.....:shock:

With 24 bit data, thats over a "megapoint" of data. If you got "only" 1,000 wrong, that is a failure rate of less than 0.1%, which is an exceptional outcome in quality terms for "hand made" stuff!

And then there is the 3D diags and the PID linking and all that stuff. Aiyah! :shock:

I am going cross eyed just thinking about it! :Eyecrazy:

:cheers:

gmtech16450yz
February 13th, 2009, 07:42 AM
Here's an update on this...
Oh btw, the trans is a 4L70

Ok, sorry it took me awhile to get back to you guys, little busy this week, probably worked on 30 cars since monday.
I had a chance to look at the logs I did when I was trying to figure out the A/C cutout thing, when I thought it was turning off from tps it looked like it was more because of simply pressure cycling. It was a little hard to assess A/C changes when it's so cold you can't keep the compressor on long enough! This would be easier to figure out in the summer.

I figured I'd try again the other day and ended up going through about a dozen reflashes, in both EFILive and HP Tuners. I tried every combination of rpm and tps settings, and guess what, IT DOESN'T WORK! No combination of settings in EFI or HPT would get the compressor to shut off, even at 5600 rpm/ 100% tps.

I also took a look at probably 8 or 10 stock truck and car calibrations to see what the factory settings were, I found everything from 0 to 170% tps and between 0 and 6800 on the rpm settings. Some rpm settings were high disable and lower enable, and some were reversed. I also looked in SI to see what GM had to say, they say A/C will cut out above 5000 rpm in most of the cars and trucks I looked at, which tells me that's just a generic description because they all couldn't be the same, and my truck definitely does not turn the A/C off at 5000 rpm. It also says it will shut off at full throttle, but I have logs that show 100% tps with the A/C still on.

BTW there is a high pressure cut-out, SI says it's 429psi, again might not be exactly that on all cars and trucks, but I know they all have a high pres cut out at some point.

So I got the DFCO working like I want it to, but I can confirm that the A/C settings on E38's, and others obviously, DO NOT WORK. It would be nice if it did, but I'd much rather you EFILive guys work on support for at least the T43 in my '09 CTS! (hint hint) This A/C thing just seems to be yet another E38 table that has no effect on the actual vehicle, no matter what numbers you put in it.

And yes, I know enough (almost nothing actually) about programming to know it's a ton of work to reverse engineer this stuff, and again, I'm thrilled you guys have done the hard work so we have these tools to fix what GM (and the EPA) have given us. Thanks!