MyRxBad

iTrader: (1)

I know someone who has a shop manual for the 8 IIRC. I'll ask him for the ECU part# and get back to you.

OverLOAD

I want to ge the Japan ECU part, and do a drop in replacement.. Depending on the prohibitiveness of the cost...

However I am also interested in a device that could modify my ECU output to match the JDM spec..

OverLOAD

canzoomer

I got a price last night ( daytime in Tokyo) of around $1100 US dollars..
Gasp!

Smoker

Good work Canzoomer. Good to see you are putting your passion into good use.

Just a note, I remember there's this Japanese magazine which tested the 5-sp version and the 6-sp version and have them clocked for almost the same time. So don't be so sure that the JDM spec 6-Sp ECU is the one that will get you the full 247hp.

As far as we know, we have YET to see a dyno for the 6-sp JDM version.So really, we don't know whether the Official hp number is not lowered at Japan because the engine IS really producing 247 hp or just that there's isn't a lot of people who actually dyno'ed their car over there.

OverLOAD

If someone has any code for the JDM spec RX-8 ECU's, please PM me, I'd like to see what I can unearth...


OverLOAD

canzoomer

Thanks. I already obtained that info, and more, so no longer needed. Appreciate the help however!

Some news:
Today we pulled an ECU from an RX-8 at the shop, and did a bunch of work.

Interesting stuff:
The ECU board is inside a cast metal case, and the metal case is surrounded by a hollow plastic shroud. The shroud connects to two small air ducts that draw cool air in from the front of the grille, so as to cool the ECU housing.
VERY neat trick, but complicated.

The brackets that attach the shroud to the ECU housing use tamperproof screws. We had the correct driver tips on hand, so no problem.

The board has 3 flash programmable PROMs on it and one single burn ( one time programmable) PROM on it.
All components are surface mounted, and the whole board is coated in a protective waterproof applicate.

The programming points are a grid of via test holes on the back side of the board beneath each PROM.

Changing the chips would require a decent surface mount station and patience. Reprogramming in place requires a special probe device with pins to match the board holes. Very specialized tool.

Also, I received news that the JSpec ECU part is indeed the same as the one here, but with different code on it at time of supply and installation.
When purchased new in Japan or USA the part has NO ECU map code on it, and is supposed to be programmed by the dsitribution support centre in the respective country. They do this for the dealer upon receipt of the VIN and the service request.
So, for a JSpec ECU to be useful to us, it has to be prepared and installed into a Japan car first, then can be pulled and sent to us. However, when requesting a new ECU board, teh dealer has to send the old (failed) one back, so this means no spares with live code can be easily obtained, unless a car is temporarily disabled by lack of an ECU.

That avenue in general appears to be very difficult. One would need the dump of each of the 3 programmable chips, and they would have to either come up with the programmer tool (not available to dealers) or strip off the surface mount chips and apply them to a special programmer with contact base for this surface mount package type. Also not commonly available.

The most practical way of adjusting ECU behaviour appears to be to intercept the ECU data and signals at the connectors to the ECU packace with a add-in computer with over-ride maps and the ability to isolate and pass through signals or replace them as needed. In preliminary testing today we determined the pinouts for the sensors, and the signal type and levels to send data to the fuel and ignition controllers.

We now know that the Apexi and other control computers whould be able to over-ride the ECU to allow us to insert custom maps and responses to the system, bypassing the ECU when applicable.

When replacing the ECU, and re-attaching the battery I found the blinking traction system light on my panel, and was able to reset it by the method of turning the wheel to far left, far right, and back to centre, which alowed it to calibrate and return to normal state with the blinking light going off and the DSC working again.

The CPU is a 16 bit, 12 register device, semi custom, with a 4MHz clock, and there is a set of gate array logic on a big GAL ASIC chip, also surface mount, with about 220 contacts. A LOT of hardwired logic on that part!

Board is 4 layer, and nicely made.
The price tag on it is not too far out of line considering it's nature, probable quantity of manufacture, and level of tech..

More news to follow.
Next up:
Figuring out board/connectors to intercept the plugs to the ECU
Programming some stating map to the Apexi.
Testing, more testing, and some more testing.

Evolv

iTrader: (1)

holy.. interesting read, but a little over my head.

What's the bottom line?

When you are all done with the ECU modification, how difficlult will this be to duplicate?

I'm hoping that there will be solutions, answers and fixes to the power/fuel issues at hand prior to the spring, which is when I take delivery of my black beauty.

canzoomer

Bottom line:
1) It is possible to intercept data from sensors, and to provide alternate commands to fuel and ignition systems.

2) By doing this it is possible to over-ride the ECU, and send signals to systems which will result in leaner burn, and better power output all along the rpm curve, and specifically above 7300rpm.
It is IMPOSSIBLE to know what gains will be made as of yet.
Once we are at a point that we feel it is possible to build and supply, we will run dyno tests, and co-operate with others interested in their own dyno testing.
At that time actual figures will be stated.


I can GUESS ( NOTE THAT I SAID "GUESS") that we should easily get back the 247HP, and I FEEL that we can probably pick up around 10 to 15 HP from 4000rpm up to 7300rpm, and about 30HP or more at 8500rpm.

If and when we complete this our goal is to produce a device that connects into the ECU connector blocks, and which will sit in the cab of the car. This device can be set to do nothing, or to insert modified response data to the ECU and system controls.

Thus it can be turned off to leave stock settings, or on to produce more power and better mileage.


Laymans terms:
Hey Billy-Bob, wanna chip your Mazda?

Warning: If used it is very possible that the device would make the car not pass emmissions regulations in some locales. If so the device could only be legally used on a race track, in sanctioned events and practice.

There, just did my best to cover my butt.

Evolv

iTrader: (1)

Say Betty-Lou, if we was to get a divorce, would u still be ma sister?

---------------------------------

In order to prevent warrantee loss, would this system be easily removable, or somehow undetectable by Mazda at times of servicing?

canzoomer

To insall or remove the device one would:
1) Open the ECU cover at the front right, under the hood. 3 bolts and 2 snap posts.

2) Remove the battery cover ( 3 clips).
Disconnect the battery.

3) Disconnect 4 electrical connector plugs.
Unplug the ECU plugs ( 6 of them, all in a row).

4) Plug the interface board into the ECU sockets where we just removed the plugs.

5) Re-insert the ECU lugs into the interface board.

6) Re-assemble the covers and connectors. 3 bolts, 4 plugs, 2 clips. Reconnect the battery.

7) Start the car and calibrate the traction control system ( turn wheel fully left, right, back to centre)

It took me about 15 minutes to disassemble it yesterday, but could now do it in 5, as I now know what to do.
It took me about 10 minutes to put it all back together yesterday.

Once done it is indetectable.

Naturally one needs to be reasonably careful. If you were to break a connector the dealer might have some questions why you were in there in the first place. Of course one could claim curiousity, and admit to clumsiness..

"I was looking for a plug for my radar detector"

<grin>

MyRxBad

iTrader: (1)

Good write up canzoomer.
Sounds like they really went hi tech on the ECU build. Especially with the 2 flash roms.

Keep up the good work.

Evolv

iTrader: (1)

As long as the final Dyno results are satisfactory, and no possibility of engine damage, then you can count me in :D

Thanks for the great work canzoomer

bgparsons3

This is fascinating. I'm not a car guy - I mean, I love my 8, been waiting for it since '96 when I saw the last RX-7's sold in Houston. But I really don't understand that much of what goes on under the hood (beyond the basics that I understand as a mechanical engineer).

However, I am an engineer and would love to hack my 8. If the price is right, and the risk is low, I'm in.

I do have one problem. I understand that in North America, the 8's ECU is programmed with a pig-rich fuel/air ratio. I also understand that this was done to comply with strict emissions regulations. Further, I understand that emissions controls are designed to protect the environment.

So, our cars are all burning more hydrocarbons than is necessary so we can protect the environment? I don't understand.

Smoker

Few Interesting Things came up on the forum today:

Paul Yaw discovered that this car cannot be properly dyno'ed at its full potential with the front wheel not moving.
http://www.rx8forum.com/showthread.p...&highlight=YAW


One magazine dyno'ed the J-Spec RX-8 and got 217.0ps @ 9000RPM
http://www.rx8forum.com/showthread.p...ghlight=Option


take a look and enjoy.

[Edit] - Fixed the broken links

mdmaclean

Smoker,

Your links are not working for me.

Smoker

-Fixed. Please Click again. :D :D

canzoomer

Yeah, that is the trade-off.

Here is how it basically works:
We use catalytic converters to change the unburned hydrocarbons and carbon monoxide in the exhaust into carbon dioxide and water vapor. The converter use chemical catalysis to create this change.
A catalytic process has to be powered by something. In this case heat.
So, we burn more fuel, to make more heat, to make the cats work better.
Mazda used to use a device called a thermal reactor on rotaries.
They are also a device to burn the products of incomplete combustion, but they do it directly, by adding more fuel and burning the resulting gas mixture. Kind of an "afterburner" but no thrust is produced. They have been abandoned as they use even more fuel than a cat.

This makes less "bad" pollutants, releases MORE CO2, and uses more non-replaceable fossil fuels.

That is considered to be the lesser of two evils.
We are burning MORE hydrocarbons, but burning them more completely.

canzoomer

Well, Paul is a bright guy, and he makes some good points, but he does not prove anything. I agree that there is a possibility that the car goes rich and down on power if the front wheels are not turning, but there IS NO speed sensor for the front wheels in action ONCE YOU TURN OFF THE DSC!

And the GTech numbers people have shown are amazingly similar to the chassis dyno results so far.

On the Japanese thread the writer (x28) claims that the JSpec RX-8 makes 217HP. That is perfectly in line with a car making 247HP at the engine, and losing about 30hp to driveline loss.

Comparing that with the numbers we have seen here, we see that the cars in North America that make (at best) 187HP, and often less, are down on power by at least 30HP from the Japanese model.

This is completely consistent with everything we have seen here so far.

And here is the final statement to all those who consider me to be a "whiner":

That is a heck of a lot more than "9HP" that Mazda claims.

Oh, and in Japan they got a 0-62.7 in 7.04 seconds ( that is 0-60 in 6.74) and a 1/4 mile in 14.97

So, the cars that Mazda North America ran to make the slips submitted to Rotary Power are apparently a good half second faster than the Japan cars?
I smell bullshit.. I don't know if it is coming from Mazda, Rotary Power, or Artex in Japan, but it still smells like bullshit.

canzoomer

One can not guarantee "no possibility of engine damage"

Examples:
1) If an engine makes more power it is subject to more stress, and will probably not last as long.
Of course how you drive it is probably the bigger factor in this.
Do 8K clutch drops all week and I suspect you might hurt your car.

2) If you are handling the ECU and touch a connector, and are in dry air, with wool and polyester clothing, there is a very good chance you might cause static electricity damage to the ECU.
If you wear a ground strap to the chassis then this is very unlikely to happen. Normal electronics precautions are advised!

3) If you run crap 85 octane gas, with low anti-knock properties, and a fixed fuel air curve that is leaner than stock, there is a very good chance of engine knock ( I can almost guarnatee it!).
We are trying to see if we can keep the anti-knock sensor in circuit, but I can not say at this point if this will be feasible.

4) If you run 40 more horsepower, turn off the DSC, crank it and loop the car, and hit a Buick, then there is a very good chance of engine damage ( not to mention fender, bumper, and human damage)

There will always be SOME risk!

Smoker

wait, you cannot just assume that. There is a good chance by disabling the DSC, you are simply telingl the computer NOT to do DSC for you BUT that doesn't mean that the computer will stop receiving and processing the information from the front wheel. Hmmm.... kind of like turning off the ringer off your cell phone but your phone is still receiving the incoming calls. Just that the ringing function is off. Hope you know what I'm trying to say.

And you are right, Paul Yaw didn't exactly post any "real proof" on the thread but given the fact that he carefully picking apart the engine and the ECU and figuring all fine details out with the help of the Shop manual. I think we should have some faith in the guy and give him some time to show his findings to us. Given the fact that this is his and his own business, I don't think he is just making things up over there. Be Patient guys !

Broker73

I applaud you for all the work you have done, but why are you always trying to find something wrong??..........do you realize that the times they posted ( 0-60 in 6.74 secs) all depend on the launch from start.....I read an atricle on the G-35 from a smaller mag, and it stated a 0-60 time in 6.6secs........yet from a hard launch (4000rpm) R & T got a time 0f 6.1secs.....same thing for the 8.............it all works out....hard launch 5.9secs.......but the 5-60 time really widens here, and from start it makes a huge difference on your time......like maybe .5secs !! ...................and yes I figured something was wrong with all these dyno numbers, there is no cover-up here.........:o ............I have driven the car, and been a passanger many times......it didn't take a wizard to figure out that the car was putting down more than 180whp...................

canzoomer

The hell with patience. We can test things now!

We went to the simple procedure of hooking up a meter to the wheel sensor today.

When the DSC is on there is a 12 volt potential andabout 120 milliamp current to the sensor, and it is modulated to a pulse by the wheel turning. With the car on a stand, you can turn the front wheel and verify this.

With the DSC fully disabled, as in by pressing and holding the button for 5 seconds, there is NO current on the sensor line. There is NO pulse from the wheel sensor.

The method used on the wheel sensor is similar to a fan for a computer: A tach pulse is generated when the wheel spins.

You can verify this for yourself:
Raise the front of the car on a jack stand.
Start the car.
Put a multimeter on the connector to the wheel sensor.
Spin the wheel. See the meter jump on the blue wire connection.
See the 12 volts on the black and grey lines.

Now turn off the DSC. Repeat meter tests.

However, there is one other possible test we need to do:
The car has some air flow sensors that I have not yet found. They are on the schematic. But it does not show their physical location.
They are not MAFs, and are simple flow sensors.
If they can use ducts to cool the shroud around the ECU, I do not put it past them to put a flow sensor or two in places to detect the air, and if no pressure, de-tune the car to prevent heat problems or something.


They MIGHT be there to detect air flow in the engine compartment.

Or, theymight juist be pressure sensors in some vacuum or air line. Got to find them first to tell..

Anybody mention that this damned car is COMPLICATED?

The electrical manual is about 100+ pages. And costs about $125

That damned computer and sensor array is way more complicated than the PC I am sitting in front of..

Speed Racer

What Broker73 is saying about the launch techniques used during test makes sense and match my own observations.

I made three runs and lauched at a different RPM each time. You can see the results below. The first run I tried to start out at the lowest possible RPMs and quickly feather the clutch. On the second run I dumped the clutch at around 6k RPMs and the third was at 8k RPMs. If you look at the graph it is pretty easy to tell when the tires hook up or in the case of the 1st run it was when the clutch was fully engaged. Just look for the point where the RPMs start climbing in a straight line. The second and third run are prety interesting because it actually took me longer to get to 60' (due to lots of wheel spin) on the third run but the rest of the run was faster. Like 0.538 seconds faster in the 1/4 mile and that supports what Broker73 said.

Bottom line is that higher RPM starts dramatically improve your times with this car.

P.S. But you won't catch me doing them very often because I know that some thing is going to break sooner rather than later.

Broker73

great info, and it looks like you made some good 1/4mile times.......that is why I have been amazed on here at the people that cry about performance......no doubt there are faster cars out there, but lets face it, on the street if you drive it hard enough (not that I would), you will blow the doors off most cars next to you........all depends on how hard to rev it, and that is what makes this car fun!.......never been one to do modes to a car, other than throw in a K&N filter, and after 6000k on the new cars I have owned, they really seem to open up.........I can't wait to feel that with the 8............test drove a 350z twice, and it just made me feel more happy about my choice.....it's a nice car, and fast, but not that much faster, and not the refined ride and handling of the 8...........we are so lucky!..........

aussie77

Great work canzoomer. Good to see you are finally channeling that anger and frustration into something positive. I hope this trend continues and we can maybe 'get along' a little better now

It has never been about burying our heads in the sand (well for some it has, but for most of us it hasn't). For me at least it is about patience. I am not one to go running around screaming at people without all the facts. Even now it is obvious that we do not have all the facts - but finally some of the pieces are starting to come together (Yaw, these posts from canzoomer).

Canzoomer the reason you frustrated so many of us before wasn't that you were talking 'truth', and we were too stubborn to hear it. As many of us have posted over and over, we are aware that presently there is still some disparity between the listed hp and the rwhp. Like everyone else we want answers. Some of us unfortunately don't have the resources to truly test the situation and GET answers. So we were willing to wait, enjoy our still fun-to-drive cars, and find out the truth. You ran around posting negative things in EVERY thread on this damn board for a while. You posted a LOT of assumptions (not facts), and had a habit of doing so in completely inappropriate places. Someone could start a thread asking what oil to use, and I'd be waiting for canzoomer to reply with "It doesn't matter what oil you use you still won't be getting 238 hp!". That got old very quickly.

That said, I personally applaud what you are doing now. Thanks for being one of the people with the resources to get some answers, and for pursuing them. Hopefully everything will come out sooner or later - if you guys can use all of this info to PROVE that the car isn't getting even 238 hp, then we can take the PROOF to Mazda and demand something be done about it. That would be an even better situation than having to buy stuff to fix it ourselves

Here's hoping we get answers, one way or the other.