A 3.4 DOHC Build then... F40 Turbo (Page 24/57)
RobertISaar JUL 02, 09:01 AM
in 8F:

byte 4C is raw MAT counts(which is the MAT you see in the datastream). at ED07, it's the only place written to, immediately after an A/D read of channel 8, so no modification is made to it at all. what you see is what the sensor is reporting.

after that, typical MAT offset and then normalization and then DTC checking. either an estimated(when DTC present) or normalized MAT value is stored to byte C6. it's the only place it is stored, so no filtering is applied to it either.

looking for other places in the code where either are referenced, noticed that 864E, bit 5 is actually not enabled in the code... normally used to switch between MAT and IRT for the alpha-n idle table. MAT is always used.

8653, bit 7 determines if normalized MAT is directly used apparently nowhere in the code(no stores are made after calc is done and next subroutine ignores generated values) or if IRT is calculated. that does have some filtering built in but it's not actually used in the base pulse calculation either(like it is in later code)... normalized MAT is at D059.
after looking at all of this, i'm reminded of why i don't particularly like 8F, too buggy.
Will JUL 02, 09:17 AM

quote
Originally posted by Slowbuild:

A little off topic, but what do you think the response time is for those air temp sensors. I've found them to be really slow. Unless I'm in boost for a long time there is no change in temp according to the logs.

Is there a type of sensor that's faster or something?

Chay



The Fiero open-element IAT sensor responds fairly quickly. However, some applications, like the L98 tuned port 350, use a coolant temp sensor for the IAT. This sensor has the element enclosed in a brass shroud and is MUCH slower to respond than the Fiero-type IAT.
Fierobsessed JUL 02, 02:30 PM

quote
Originally posted by RobertISaar:
after looking at all of this, i'm reminded of why i don't particularly like 8F, too buggy.



It is pretty unique, almost seems unfinished. It kind of makes me want to use manual $DF, and focus on adding in real support for boost. If done carefully, really not a whole heck of a lot needs to be done. Think about modding the section that makes the map variables, have it create normal map variables based on 2-bar value *2, capped at 255. Then, use the 2 bar variable only to reference the main fueling table and spark, extend those tables with some boost columns. Then you could borrow the waste gate controls from $58 or $8F. Ok this part is a bit hairy. but its doable, IF, you have the ram space. There's always plenty of program space though!!

But in all honesty, Using 8F and dealing with its difficiencies and crappy workarounds is easier then creating a highly custom OS like above. I think I'll look into the feasibility of DF, or maybe A1? Either way, I'm going with 8F for now, if that turns out to be just fine and drivability is up to my standard, then it is what it is!
Slowbuild JUL 02, 02:47 PM
I'm using a stock sensor, no brass, there is a plastic protected element that looks like glass kinda. I tested it in hot water etc and it seemed to function ok, but laggy. Maybe I need to freshen it up a bit.

Do you see changes in air temps in your logs soon after going on boost?

Thanks,

Chay
RobertISaar JUL 02, 02:51 PM
that is NEARLY identical to the route that i went with nAst1 to add boost support....

i did run out of program space though(i have less than 200 bytes of usable space left, some of it not continuous, either). using a 9396, have TONS of leftover RAM, i need to build a custom MEMCAL adapter to open up portions of the addressing space to be used with the PROM, since the way GM connected what would be pin A15 on the PROM is actually connected to VCC. then dealing with CS, OE... it's been interesting to come up with a solution that isn't overcomplex.
Fierobsessed JUL 03, 08:25 AM
I took some time today and read up on the nAst1 project some more. It was quite some time ago that I took a look at it. I dismissed it as it wasn't particularly useful to me a after a few pages in because without boost support, 8F really is a better choice to start tweaking from...

Well, I read up about 20 pages of the 54, and the more I read, the more interested I am! I didn't realize that it had gained boost support. That's kind of a big deal. I'll certainly entertain test bedding for you with this project.

Only one thing I wanted to mention. Without reading up on 537 posts, It's really hard to understand the list of features, changes, uses, and Pinout's of your project. It would be very helpful if you could include that data with your uploaded package, or in the post that the link that you have in your signature directs to.

That being said. You have one seriously ambitious project, and I am blown away with your intricate knowledge of code to function. I find it difficult to follow the disassemblies in the more complicated routines. I know I can make routines, but they won't have interrupts, stack pushes/pulls, Fdiv's or anything like that. Kind of like working with the tools you know best, and working around the ones you're not comfortable with. I'm comfortable with port accesses, table lookup mods, ALDL mods, and a decent number of other stuff. But I've always had trouble correlating a performance issue to a specific table/value.

This is a bit O/T, but is mostly directed to you.

Quick story. My Van (LT4 Hot cam Vortec 355 with a BBC TBI Running '427/$0D) had a severe surge at idle in gear. I ported the flags for AE, PE, DE, and DFCO to the ALDL to help figure it out.
The engine was getting a AE pulse with each surge due to an unstable MAP sensor reading, IE my cam is a bit aggressive. So I raised the Delta MAP threshold for AE. This fixed the aggressive surging. But to this day it still surges, and all I can think is: Is it an error in my VE table? Spark table? Is it an Idle setting? Is it a closed loop fuel tuning parameter that is out of control due to a moved O2 sensor? I can never figure all that stuff out. Far too many possible reasons. ALDL isn't telling a clear enough story, all its saying is that the IACV is going nuts trying to quench the surge, O2 is going crazy, and so is the fueling. I gave up figuring it out years ago. It drives ok enough for me. There is too many settings for me to figure out what is causing the issue. I find that so frustrating, I gave up trying.

And my opinion is that the 16197427/$0D might be the most flexible ECM. It does everything from 4L60E's TH350's, Manual Transmissions, L4's to V8's... It really is a huge improvement over the 7730. It has some flaws though. Not enough PWM/Digital Outs, or Analog Ins, and No boost support. It sort of has a built in TCM that runs semi independently of the ECM too. It's a great ECM for 2.8 Fiero's, and Especially 4.9 Caddy cars, but no one on here has done that yet

As for my project... Still waiting on a pressure plate. Hope it comes in soon! In the meantime, Ill keep playing with the FMHE and its plumbing.
RobertISaar JUL 03, 04:08 PM
if you think the thread is bad, the directory i have everything stored in is probably a little worse... i'll see what i can do to somewhat organize it and perhaps document some things that i may have forgotten to do before.

the 7427 is would be an interesting place to start with if you wanted to run full custom code seeing how it's a P6(so the 6811 pink book is a lot more relevant). the lack of I/O(and non underhood case) is why i chose to go with probably the most common P4 family units still in existance. the 2.5K of RAM is also nice with the 9396, since the MPFI 60V6 factory code already runs really close to hitting stack overflow.
FieroWannaBe JUL 04, 12:26 AM
OSE 12p.
Read up, and see what you think of that.
FieroWannaBe JUL 04, 12:27 AM
Dopple post

[This message has been edited by FieroWannaBe (edited 07-04-2013).]

sleevePAPA JUL 04, 03:21 AM
for $0D, does it have IDLE over/underspeed ignition error tables? Not familiar with $0D but $8D uses ignition timing to finely control idle and those tables allow ECM more/less control. Usually a choppy cam will throw off the synergy of the calibration and idle is the first major tuning hurdle.

Also, 12P might be worth a look if interested as stated above.

http://delcohacking.net/for...topic.php?f=27&t=356