E35B Digital MAF Scaling

[kidturbo]

Hi all,

Would it be possible to get a cax file to modify the E35B Digital MAF tables?

I'm looking to run a new style GM digital MAF with a higher a 12k HZ frequency range. I verified the E35B ECM will read up to 14818 Hz, but table B0205 currently ends at 8696Hz. If supplied a frequency above 9200Hz, the G/s reading falls to zero. So it appears to be running out of resolution in that table. The new style sensors should work fine if table values can be aligned.

Thanks

-K

21083

21085

21084

[SASDakota]

It may actually be a protection. There is likely some sort of threshold that when hit or crossed over it defaults to 0 g/s thinking the sensor has failed. That would be where you need to look. IMO

[kidturbo]

I think it's just running off the end of the scale. But like you said, it would set an out of range code for certain. Reason I would like to adjust the scale and test it. Any help with finding that line would be greatly appreciated. Just bench testing at the moment, then I have a half dozen ECM's to upgrade.

[SASDakota]

Ross, What was the deciding factor for EFI live to not map the axis for MAF digital {B0205}? Is it hardcoded somewhere? Similar to mapping counts in the Cummins stuff.

[kidturbo]

Don't think he's stumbled onto this thread yet. Maybe we should PM him a link to it?

[kidturbo]

Anyone have the A2L file for this OS to share?

[GMPX]

Don't think he's stumbled onto this thread yet. Maybe we should PM him a link to it?
Yeah sorry I do tend to miss a lot of threads on here.


Ross, What was the deciding factor for EFI live to not map the axis for MAF digital {B0205}? Is it hardcoded somewhere? Similar to mapping counts in the Cummins stuff.
I don't remember, it was 10 years ago :shock:

I have a feeling we actually hard coded those axis values in to the software from whatever was in the stock files (I know that is not ideal), but maybe we hadn't sorted out the axis displays back then and it never got revisited.

Can either of you do .cax files? I'm happy to tell you where the axis is for a given OS and you can add it in yourselves (I'll just post it up here). I am a bit reluctant to go and add this in to EFILive then regenerate the E35B .calz files again. I did do them back in Feb this year and it was a big pain to have to check over everything.

Cheers,
Ross

[kidturbo]

I've added a cax file for MAP scaling to this OS before, so probably could handle it with a little reading.

Thanks

-K

[GMPX]

Let me know the OS and I'll save you the trouble (I'll PM the details).

[kidturbo]

12628594

Since I all ready use this build with a MAP scaling mod, please let me know if any special tricks to using multiple cax files in same tune file.

-K

[SASDakota]

kidturbo, If you send me your current CAX I'll just add in the stuff Ross has for you then you can see how to stack multiple parameters onto one CAX file.

It isn't rocket science, just a bit of a learning curve figuring out the syntax for CAX files.

Shoot me a PM if you would like some help with it.

[GMPX]

I've just had a look at this and the axis is defined in 'uS', we converted those values to Hz because for most people that would make more sense. I think that is why we went for the fixed values because we couldn't do the uS > Hz conversion on the axis values. But there is a BIG problem, the axis is defined using 16 bit signed values (no idea why they used signed), so the maximum positive value is 0x7FFF, the end value at the top of the axis from factory is 0x7F9E, meaning there is nothing left, you can't go over 0x8000 or the numbers become negative. I think you are stuck with what is there unfortunately.

[SASDakota]

uS for a maf scale?

What if we do something like the Subaru guys and half the scale, then half the table axis' that reference the scale?

[GMPX]

uS for a maf scale?
Don't shoot the messenger.....it's Bosch :shock:


What if we do something like the Subaru guys and half the scale, then half the table axis' that reference the scale?
I guess if you can figure out where in the OS code they set that up it might be possible, but I don't know what flow on effect that might have elsewhere in the system.

[kidturbo]

That's interesting. Leave it to Bosch to over complicate things..

What about on the Analog option? Looks like half the cells in that table axis are set to 5.00 for a reason. Could we improve on the resolution by adjusting that axis to fit better with a 910gps hard limit?

While not my ideal solution, adding two 5v signals together is actually easier than adding dual Hz inputs. That's basically all I'm trying to accomplish, tally up two MAF signal values into a usable total the ECM can work with. Do they even still make Analog MAF's...

[GMPX]

Can I throw my hands in the air on this one? :grin:
I don't know what the ECM's internal limits are with air flow though or even if the same limits apply to both frequency and analogue MAF inputs.

Also I'm not sure what you mean by adding two 5V signals together, to be fed in to the same input?

[kidturbo]

Also I'm not sure what you mean by adding two 5V signals together, to be fed in to the same input?

It's on a twin turbo setup. Was considering combining two separate MAF Hz sensor signals into a single output using a micro controller. However those output totals could exceed the ECM's 15khz max limit shown in my first post from bench testing. So I would need to place limits on the Hz output from the controller, no matter what the actual total. As where with voltage MAF setup, limit would be standard 5v max. Easy.

Anyways, if we can't increase axis values to up around 14-15khz, then there isn't much to be gained by this exercise. Resolution can't be improved, so will just remap within existing scale. Not ideal, but what are the other options?

[joecar]

What are you trying to do again (your original objective)...?

[ScarabEpic22]

Not sure if it's completely applicable here, but I know gas guys simply increase the tube size the LS3/7 card-style MAF sits in to change the resolution. Going from a 3" -> 4" tube means the overall speed of the air decreases (lowering MAF values) while keeping the same mass of air flowing into the engine. Downside is you lose resolution at idle and part-throttle.

[kidturbo]

What are you trying to do again (your original objective)...?

Change the digital MAF axis from 8696hz to around 14khz so we can run the new style GM digital MAF's with 12khz frequency range. In first post pics you can see the ECM understands a 15khz input signal from an emulator.

[GMPX]

But that doesn't fix the problem of the scaling in the ECM where the airflow is currently set to 0x7F9E and you can't go over 0x8000 in value.

[kidturbo]

Now he should be caught up.. :hihi:

Still leaves me wondering why the ECM can understand a 15khz signal but only 9khz is usable. Hardware and software guys must not chat much..

[koolky]

I've just had a look at this and the axis is defined in 'uS', we converted those values to Hz because for most people that would make more sense. I think that is why we went for the fixed values because we couldn't do the uS > Hz conversion on the axis values. But there is a BIG problem, the axis is defined using 16 bit signed values (no idea why they used signed), so the maximum positive value is 0x7FFF, the end value at the top of the axis from factory is 0x7F9E, meaning there is nothing left, you can't go over 0x8000 or the numbers become negative. I think you are stuck with what is there unfortunately.

Doesn't uS increase in value as Hz decrease? eg. 918 Hz = 1890 uS (0x7F9E) and on the other end: 8696Hz = 115 uS (0x0D89). Therefor if we want to see 15kHz aka 67 uS would we not be moving down in HEX (~0x07D0)?

21661

[koolky]

Also, if I try to write a CAX for the AXIS I get the following:

-----------------------------------------------------------
Address Overlap Report
The following calibrations have been excluded:

- User calibration: 191E76 overlaps one or more EFILive calibration(s)

User defined definitions may overlap other user defined definitions (to support interleaved axis/data definitions) by checking the option: View->Properties->[Options]->"Allow *.cax calibrations to overlap other *.cax calibrations."


I have that option checked off.

[GMC-2002-Dmax]

Any CAX you create cannot overlap an existing EFI-Live table.

Disable the OS CALZ by temporarily renaming it.

Then reopen the CTZ File, only the CAX you created will open up any table in the V7.5 Editor, resave the CTZ with the new values you altered in the file.

Re-Enable the CALZ by properly naming it back to what it was.

I do that for some of the other stuff I work with when the definition I want to change is already defined within EFI-Live's calz.

[koolky]

kidturbo - try this CAX file and do what GMC-2002-Dmax mentioned in the last post. I guess you'll want to copy the axis into Excel to keep track of your row labels because the EFILive label won't change when you re-scale and re-enable the .CALZ

21662

[kidturbo]

Thanks, I'll give that a try. Gotta order some more autocal's first. Sold last one with the licensed E35B I was testing on. :thumb_yello:

[koolky]

Thanks, I'll give that a try. Gotta order some more autocal's first. Sold last one with the licensed E35B I was testing on. :thumb_yello:
21828