those images give a lot of insight. anything outside of the stock engine and those maps aren't of any real major help but they are much better than the "stock map"(s) floating around for the stock 13B.

To address your "gloat" remark - if i wanted to gloat, I could have very well posted everything I found inside the ROM's. I have a lot invested to make those maps see the light of day.

So, there you have it. I didn't want to share but the 3D Maps to settle some arguments that have gone on for a while. Now that its settled and I've managed to achieve other goals with the data I found I've moved on to another project. The short version of my goals in extracting this data was to build the best calibration for my RX-7. I haven't seen any published data showing how the OMP works - but now that I know I'll surely use it on my FD and I'll have a better understanding of how to configure the computer to control it.

I'll gladly share some hardware info which is aid someone else to extract the ROM's. I have a lot invested in extracting it and do not want to give that away. Call me an ass but thats how I feel about it. Whats really amazing is how the 20B maps look compared to the FD!

hwnd,
Thankls for posting this info for all of us to see. I work with/for a company who's main product is "chip" tunning for a very well known make. I appreciate your hard work and for all those who think he should openly give out all his info; go out and take 4+ years of programing classes, invest thousand of dollars in hardware, spend countless hours weeding through computer code, and then post it yourself.

Thank you very much for understanding. I'm not sure if this will ring a bell but IDA Pro Advance wasn't cheap and having to write the proc-module was a little beyond my scope but I managed to get it done and produced damn near re-compilable results! :-)

Please dont take this the wrong way, but unless you stand to profit from the information your collecting I cant understand why you wouldnt release everything. The worst part of the RX7 community the way I see it is to many people hold onto their "secrets"... I mean I get that you put a lot of effort/money into getting all that you could from the factory computer, and if you use that information to develop and sell a product that people want/need then I totally understand, but if its just to have some token knowledge that the next guy doesnt have then its just another blow to the community.

I do have interests in this data from a monetary gain perspective.
in the end - I'll provide a tunable stock ECU with the options of changing map sensors, etc;

If I only said "hey this is all i could uncover" it would be lying which i'm not in favor of.

There are plenty of car communities with open source tuning of the factory ECU. Hondas, Subarus, and Mitsus all have people using their knowledge and hardware for everyone's benefit.

But the Mazda ECU doesn't... at all.

I've been able to flash my own box - so naturally I suppose I have to give that information away.
I see your point.

Out of curiosity, is the FD ecu re-flashable? I know many of the mazda ecu's of the time period aren't, which has essentially killed the idea of making a re-flashable ecu for other Mazdas.

I guess the point Im getting at is - You may want to investigate the validity of the business plan if it includes having people send their ecu to you to install a programmable add on board with a flashable chip.

Also lastly... I'm amazed at how much time / money you must have put into this. I'd be focusing on a newer, more common ecu with more potential buyers myself :D

not in "stock" trim, it isn't.
now you're catching on - this is my plan.

well be that as it may - i'd have competition - as where in this area I wouldn't.
besides, id have to be interested in newer cars. Maybe I'll work on the 930 ECU next.
I've been toying with the idea of buying another 911 turbo and stepping away from the rx-7 stuff for a while. this might be a good reason.


another point as to why i even bothered.. i wanted to know how the OMP works (what controls it, etc).. so that i can successfully & correctly calibrate my own ecu in the RX-7.
haven't you ever wondered how it worked? wouldn't it be nice if your standalone was built on a legit "base map"? these were my ideas and i ran with 'em.

I even swapped over the RZ/Spirit-R ABS unit into my FD (the abs pump with the EBD option)... I just tinker with stuff too much I guess.

Care to elaborate? Did you pull maps from the stock cosmo computer? I'm wondering what differences there are in the maps. I always assumed it was somewhat more conservative and plus one rotor...

Some of you may find this interesting. These are the factory timing maps on a series 4 T2, disassembled from the stock N332/N333 ECU. These engines were rated at 182hp and about the same amount of torque. Max boost was a little over 6psi and they ran on 87 octane in stock form.

https://www.rx7club.com/attachment.p...1&d=1297919843

https://www.rx7club.com/attachment.p...1&d=1297919844

https://www.rx7club.com/attachment.p...1&d=1297919844
Credit to RotaryRocket88 for extracting this out of the Rtek software.

The X axis is rpm and the Y axis is load. Load is calculated from the airflow meter signal instead of a MAP sensor. We don't know the exact formula for calculating load, but it is proportional to engine airflow divided by rpm. More airflow per engine revolution will result in higher load. So here we we have confirmed for sure that Mazda really likes to run around 15 split at full load. The Rx-8 and the T2 did it. The early-release FD PFC basemaps ran this type of split as well. You can see that this engine has a significant amount of timing over 5000rpm even running 87 octane. You can credit the 8.5:1 rotors, small ports, and small factory turbo for that.

^ I should point out that the last table isn't exactly showing the load/timing for an engine at WOT. I highlighted those cells in green because they marked the peak load levels where timing went constant. These two graphs show load vs. boost vs. timing much better:

https://www.rx7club.com/attachment.p...5&d=1297836996