Stock Mazda Fuel & Ignition Maps
 

I dont want to get too deep into this but here is what they look like from the N3A7 ROM..

I've attached the Primary Fuel map, Leading & Trailing Ignition maps.

Damn man! What are you, some kind of sleepless tinkering junkie?

;)

you're one to talk ;-)

Can you upload this in table form? What are you using to dump the chip? and can you explain the units used here

WOW u has haxed the Motorola chip ?

Wow!

So the OEM maps are 12 (rpm units) x 14 (load units) so the PFC's 20 x 20 is a finer matrix

Buy what is 12? 8000rpm? And what is 14? Can't be 10psi because fuel cut is higher.

I would expect stock maps to go past 8000 rpm; because for years I drove on stock ecu to 8500. I would expect rpm to ~9000 (750 increments); and from x? inches to 13-14 psi?

^ unlikely it is in psi... even GM does all their turbo speed density maps (GMC Syclone) in kilopascals

is the fuel map a volumetric efficiency table like GM uses? And are there boost modifier tables... again I am pretty familiar with the Syclone setup so I wonder how these compare

I don't want to to get into it too much because I'm still trying to figure things out (read: learning).

Currently I don't know the scalars but I'm making highly uneducated guesses at the scales (500 - 8000 RPM & -700mmHg to +700mmHg).

At this point, I believe the maps are Load vs. RPM based. There is most likely Throttle vs. RPM maps too (tip-in,etc) but I've yet to locate them.

In conjunction with some help, I've managed to bang out an IDA Pro Processor module to aid in disassembling the firmware from the external memory chips. My goal is to build a reflashable stock "single turbo" ecu but this is a long ways off from where i stand now.

You will have to get some idea of how load is calculated then. On MAF based systems, it is usually measured airflow divided by rpm or measured airflow divided by some "ideal" airflow.

On the Syclones (MAP based just like the FD), fuel is calculated like this:



BPW = BPC * MAPP * T' * A/F' * VE * F33C * BLM * DFCO * DE * CLT * F77

Where:

BPW = Base Pulse Width

BPC = Base Pulse Constant Term

MAPP = Manifold Pressure Term

T' = Inverse Temperature Term

A/F' = Inverse Air Fuel Ratio Term

VE = Volumetric Efficiency Term

F33C = Battery Voltage Correction Term

BLM = Block Learn Correction Term

DFCO = Decel Fuel Cutoff Term

DE = Decel Enleanment Term

CLT = Closed Loop Correction Term

F77 = Turbo Boost Multiplier

see http://www.nwstp.com/forum/P4_section_9.asp

Factory ECU's are designed to model the engine and its potential for emissions, which explains some of the complication. Here is a list of some of the most important tables in the Syclone ECU: http://www.nwstp.com/forum/chiptable.asp

Here are the two most important fuel tables in the Syclone ECU:

https://www.rx7club.com/attachment.p...1&d=1259792721

This is the basic volumetric efficiency % vs rpm vs map table. But when the Syclone goes into boost, the boost modifier trim kicks in:

https://www.rx7club.com/attachment.p...1&d=1259792721

This is a two dimensional table of MAP (kilopascals absolute pressure) vs I believe a volumetric efficiency correction factor. The two tables work together along with all the other corrections. I'll see if I can dig up some stuff on the Honda internal ECU architecture. It may be closer to the FD design than the GM architecture.

Again, can you post those maps in a more readable 2 dimensional table form?

Why do I need to do this again?
Knowing what the numbers mean wont really help. Case & point would be with 90% of the "tuners" out tuning cars probably couldn't tell us how the computer figures up the fuel at the end of the routine but they "know" how to add/subtract fuel in various increments.

If you really want to get some kind of Hondata/Chrome style chipped ECU going, you better do your best to understand how the ECU works. Otherwise you'll end up with nothing more than a Pettit Unlimited style ECU modification, and my understanding is that they just dumped fuel in. Remember that there's injector staging and all sorts of other things going on.

Have you ever done any flash tuning or significant chip tuning? It's not the same as tuning an SAFC or a Power FC or a Haltech. Stock ECU's are very complicated and they put a lot of things in random Hexadecimal type of values or power of two. Only programs like Hondata have managed to simplify them some. The most expensive standalones are still pretty dumbed down. When you reflash an SR20 or KA24 ECU there is an injector constant you have to adjust and all sorts other crap. And timing is even more important... are those degrees BTDC? What are they? Is it on actual RPM or some value from the Ne sensor?

On Karman vortex MAF based ECU's, if you didn't know that the tables were based on Karmann Hz you'd never understand what you were dealing with. Hertz/Karmann signals are not the same as hotwire signals.

check out other sites like romraider.com, pgmfi.org, etc

I was smart enough to yank the maps off and identify them :-)


I'm leaving work, when I get home I'll post a few images that might help you better understand where I'm going with this (after dinner of course).

subscribed

This might give you a better idea of what I'm doing...

interested

looks like assembly language?

Yup. Brings back memories of the late '80s!

8bit (dis)Assembly Motorola baby!

I actually ended up having to write the disassembler plug-in for IDA.. It's loosely based on the 68k series but still remains a private semi-custom Denso chip.

..no hex editing done here. recompile or bust! :-)

look like nuclear missile codes to me :lol:

To the OP:
The 2nd Gen Ford Probe uses the same micro. and a couple of guys on probetalk have done quite a bit of disassembly on it and are very familiar with it. Look for Mike 94PGT or ASword , they may be able to give some info you might not already have.

Which processor?

HC11G5 or D8x? ...FD ecu has 3 in it.

For this project to really move any further.. it'll take an EE to decipher the Denso/Toshiba processor.. My BDM isn't working with the G5 proc.. mfgr isn't really any help either.

so how about putting those maps in a chart rather than a 3d rendering? I'd like to see the trailing split progression

its less than you might think. it's 0* split in vac areas and max 4* split in boost. that's really it.

ether you guys are ridiculously smart or you all need a vacation.

I have not clue what you guys are talking about but it sounds cool.

Info like this makes me think of the good old days, back when the signal-to-noise ratio of the internet was nice and high... before myspace, facebook, and iTunes.

i really cant understand this.. but good work for doing what you do

Yeah over my head too, I generally submerge myself with the information until I understand it. But I would just drown, trying to understand all this. Explain it in simple terms lol JUST KIDDING

There is still plenty of that. Usually not on car forums though. :)

I have a programmer board/adapter coming for the Motorola G5 processor using the M68SPGMR11 programmer. :nod:

Was a bit rough finding this one.

I think the guys talking geek speak here are electrical hardware and/or software engineers by training or they have too much spare time on their hands... there is a lot of jargon in any discipline, and they are just speaking to that. Pretty neat stuff regardless. I love threads like this :)

^^ 4* split in boost? I thought Mazda spec was at at LEAST 14* split in boost. I'm no expert but that's what I understand :dunno:

I haven't ever see any evidence proving numbers.

If you don't know the units used in the Mazda tables, how do you know it's 4 degrees split? serious question

yeah back in college we used the motorola hc11 chip lol brings back memorys haha

I don't think he was implying absolute values but rather relative. some areas of L/T maps look the same i.e. zero split; other areas have up to 4 x's magnitude difference.so maybe 4 x 3.5 degs == 14 deg total split

5k rpm, 0 mmHg = 36.9* (176) of ign... on the trailing map it -4.6 (170) of the primary map.

you're right - the entire trailing map (in vacuum) is of the same value... it fades from -4* degree to -14.8*

in fact, 5K Revs column is 0* trailing until atmospheric pressure which is then -4.9*, -9.5* then 14.8*
it continues like that all over the map.

Only way to prove numbers in the real world would be to use a Mazda tester/monitor tool. Also, the service highlights manual indicates that timing can vary up to 7 degrees based on knock sensor feedback alone.

https://www.rx7club.com/attachment.p...1&d=1260059720

for reference, here are the early model PFC timing maps (current PFC default maps are different). Supposedly these are close to stock.

https://www.rx7club.com/attachment.p...1&d=1260059720

https://www.rx7club.com/attachment.p...1&d=1260059720

here is the final injector pulsewidth map (hidden basemap * Commander INJ map correction) for that same model PFC. this does not include air temp etc corrections

https://www.rx7club.com/attachment.p...1&d=1260060264

10000 PIM units = atmospheric pressure
20000 PIM units = 1.0 kg/cm^2 boost or 14.22 psi

-- leading timing is in degrees BTDC
-- trailing timing is the number of degrees of separation between firing of the trailing and leading plugs
-- fuel map is in milliseconds of injector pulsewidth

subscribed buddy ;) (not that i understand it all, but i do understand enough to know that its pimp ish lol)

updates?

What kind of updates were you looking for?

I've since moved on to other projects to keep me busy.

I think I know the answer, but have you found any of the other fuel/timing maps in the ECU or figured out what units the maps are actually representing?

the cosmo new model training book has the timing map in it, clearly split is small at low load, and big at high load.

the rx8 can be plugged into the dealership computer, it also runs more split @wot than it does at say half throttle.

i do not specifically know about the FD of course...

Of course there is more information... OMP, correction tables, etc..

There are a few people here who I've shared that information with but for something more specific you could always log the FD and build your own map for these or run the ECU on a workbench.

I didn't have any help extracting this information and dont feel the need to release it. Sorry.

^ ouch

I can respect that you're protective of work that you've put in but what was the point of posting the images of the maps in the first place, to gloat? Poor form.

^_^