I guess I should have clarified, I didn't mean piggyback as in a signal interceptor type of unit. I was referring to keeping the stock PCM to control all the canbus functions besides engine control. Though from the looks of it that seems like the thing to do if I choose to go standalone.

Do you have a link to a working Adaptronics Turblown setup? My friend was pretty much lied to by Elliot when he was sold a Turblown S2 Turbokit with the Adaptronics.

You've got to do all kinds of crazy shiat to get the Adaptronic to work with the S2. Not worth the hassle IMHO.

The post wasn't really directed at you as much as anyone who used unichip's terrible piggyback connector.

I feel like you can get quite a few of the Can functions to work without the stock PCM, but here is a happy alternative I thought of. The Car's ignition system/etc wants to see that PCM, it doesn't care WHERE it is! So if your Aftermarket, Haltech is what I would use, was in the engine bay, there isn't a reason that you can't relocate the stock PCM somewhere else and use the 2wire Canbus (4 wires +/- H L), to keep those factory modules happy.

Another idea I have explored, is Factory PCM tuning, but anything requiring a module would use existing tables and logic to compensate. IE the IAT table sets certain things based on the air temp and barometric temp. The IAT is resistance based, so with a digital poteniometer you could create the voltage drop and compensate based on your tuning table & scale. This would be a piggyback of sorts, but the kind you can control. Or just keep it simple and use the Canbus triple to mod the packets and handle anything the stock PCM can't. For example adding a 3 bar MAP sensor that is Canbus, the PCM wouldn't know what to do with that code, but the Canbus triple would see those packets and execute changing PCM packets to compensate for whatever is happening with the stock logic.

Example

00x000000x000 (Canpacket-Haltech MAP 14.7psi) CBT Mod packet +/- 0.0 volts MAF/IAT (at atmo)
00x00000x005 (Canpacket-Haltech Map 20psi) CBT mod packet -.5 volts MAF 3.5v-.5 resistance = higher voltage lower
PCM "o'Shucks, I need to add more fuel! It is bloody cold in here mate."
CBT "Haha tricked you again douchebag"

^I'm not sure what you are trying to achieve. If you want to tune an s2, just get Mazdaedit or Versatuner and be done with it. No hardware hacking required.

If you are referring to me, that is my plan, I am just a massive pain in everyone's ass including my own trying to be a engineer when I should just shut up and do what I am told. Sometimes it yields cool results that are new, but most of the time it just makes people go WTF???


I am actually really interested in learning about CAN at the heart of it because even the Nissan Leaf and Prius being electric cars use CAN to operate the Battery Management Systems and etc. The RX8 is truly a marvel when it comes to sensors and communication, and most of it goes unknown to the driver how much that piece of engineering is working to put a smile on your face. I mean, YAW, Gyro, and torque sensors? Mazda really did design it to move sideways.

CAN has absolutely nothing to do with actual engine management... Monitoring and silly stuff like manipulating gauges, yes, but you're not going to find a bus capable of actual EMS.

Well, if Haltech uses CAN-attached MAP sensors, then the CAN does have something to do with engine management. I had thought the CAN bus would be too slow for that kind of thing, but I guess not, now that we're past dialup speeds.

The idea about piggybacking a unit onto the PCM to interpret data from an engine sensor on the CAN is completely sound. It may not be strictly necessary in the case of going FI (I don't know enough about that process), but it isn't invalid and could have other applications.

Other applications, certainly.

Reading a CAN sensor and then doing something else, like take control of drive by wire is not a CAN exclusive activity. My point is that you cannot simply expect to write a different register and it magically work for all the functions you are trying to control.

There will also be scenarios where you cannot get the existing message off the bus and it will be read anyway. Let's add this to a use case such as changing fuel based on something like a MAF or MAP value and the results are potentially disastrous.

Want to rewrite your LCD readout or freak out your gauges? Knock yourself out.

See what I'm getting at? If everything was managed over the HS CAN network then there would be 4 wires going into your ECU -- 12V+, Ground, CAN Lo, and CAN Hi. That simply is not the case on this platform.

Sorry man, I have respect for you, but it isn't just yes or no here.

http://www.normalexception.net/docum...l%20System.pdf

http://www.normalexception.net/docum...0Connector.pdf

Normalexception.net - Documents

http://www.normalexception.net/docum...highlights.pdf
(chapter 9)


If you go through the electrical diagrams, CAN touches ALOT of everything, on the series II there is two buses, not just one. HS CAN,(PCM), and the instrument cluster can. There is also a direct to ECU manipulation bus on the OBDII connector that could let you override any PCM function based on whatever Mazda specific register it uses.

Anything on that bus can be manipulated. The TCM/EPS is a big one. It sends everything from the steering wheel angle (it's own CAN register), to how much torque the car is supposed to product at X RPM.

Who transmits and who receives can vary indefinitely. I could program a Atmel chip to take over the Canbus if I wanted to and run the show. There is no "squelch" function on the car, so I can spam a PID 100 times and bog down everything with updates. (when you have a short in the clock spring it does this) The PCM Harness has most of the PCM to Sensor functions, but it can be manipulated via CAN. That "torque demand" table can set a lot of how the car behaves. The point being, how the bus behaves is set by whatever is connected, not by 1 device. You can also re-terminate with a 120ohm resistor and break the bus up.


It is possible, and easier than you would think. They added the Can functionality to make the car better, and it clearly worked with how smooth the series II can be at it's best. They more than likely had to keep most of the sensors with the PCM directly because no BCM exists and can transceivers were not as widely used in 2007 when the series II was being designed. Now most 10 dollar microcontrollers can understand CAN with a voltage shifter.


On any other car of this time period, I would agree CAN-data is useless, but on a 2009+ car, it controls everything because it cuts the harness runs in 1/25. Code is cheap, wiring, maintenance, service bulletins, extra-parts, labor, etc isn't. Can really comes into it's own with BMS on the Leaf and new generation electric cars.

If the M-MDS can understand it and communicate with it, then anything can with the right protocol. Keep in mind, it isn't magic bullshit in a plastic box, it a microcontroller with a set of instructions and if thens. That means you can reprogram it, replace it, manipulate it, trick it, etc.

I am not going to arque that it is EASY, but you can use the PCM's terminations to feed other devices and pull registers to execute functions on external units. Example
TCM: Speed 75mph (block to PCM)
PCM: Maf 210@5v (thru from PCM)
Haltech: WB02: 15:01 (Block to PCM)
MAP: 2.1BAR (Block to PCM)
Haltech: I=75DC BCV (Write to registerbus00)

Getting the OE PCM to do what I think ShellDude is saying is difficult, but CAN is adaptable. I am just talking about requesting data from it, and using a termination bus to take action. So like a Haltech RE interfaced into a hybrid factory CAN, with a CAN MAP, CAN Wideband, CAN Termination Point (CAN to I/O), etc) with the CBT only passing relative data.

Not really related to this thread, but have been doing some extensive boning up on all the current Skyactiv models (including ND MX-5), WOW...

Particularly the full blown models with EVERYTHING in them...
DTC (Diagnostic Trouble Codes), well there are 9 times yes 9 times the number of codes and sub codes to the last S2 RX-8.

58 different warning lights (total) in dash cluster, same icon can change amber, red, green and some blue.

Control Modules front and back (just for body, radar cruise, LDW, SCBS, AH, on and on and on), then engine, then gearbox, then there is the infotainment stuff, plus all the usual stuff we are used to.

These control modules are so sensitive that re-connecting the negative battery terminal is shorting some modules out in new CX-3 and Mazda 2 if the operator is not clean and not sparking post as they re-install terminal clamp.

It appears they may have to install (from factory) built in manual circuit breaker switches on the battery.

Jesus, warranty claims much? They need to keep it simple. RGB LEDs in the Dash, 4 Wire CAN with a relay for a "soft start" when connecting the battery. I am glad you said something, that really turns me off to the new MX-5. According to most dealers here, a USB to 5V USB charger can blow up your body control module. (Yes they are idiots)

I never said they are idiots, on the contrary.

Things change, just because it is not set up or made to how you like them to be, does mean to say they are wanting.

I was saying our Dealers in Texas are mostly idiots. Mazda is pretty damn genius, but they delegate the diagnostics to the techs and HOPE they paid attention during Mazda's amazing training process. I absolutely love how detailed Mazda gets with their cars, just hate how the techs pick the first thing to blame even if it isn't on the list. I doubt Mazda has anything to do with that, it is a FORD dealer practice generally. If it was up to Mazda they would sell Mazdaspeed parts to you and have the techs install them.

Hey... if you think you can pull off a full CAN managed solution then all the power to you. I have neither the time or the patience to reverse engineer an entire vehicle and cook up solutions for all the compensating controls you'd have to implement to mimic what the manufacturer spent years in R&D to bring to market.

It's an industry with very little standardization so anything you do develop will also be a one off.

Anyone ever got this error in Mazdaedit? I am trying to figure out what PID it is so I can get it to stop.

error: unable to read response at code 0x0108

Anyone disable the OMP or change the steps on the Series II? I want to see about turning it down at the top end. I am running 8oz premix and E15-E30.


https://cimg1.ibsrv.net/gimg/www.rx8...ddf621223e.png


I want to avoid using the 4cycle oil if I can as much as possible, but not make it unsafe if the oil doesn't mix at a lower tank level.

Same thing here. I was premixing 6oz a tank last track day.
At high rpm close to redline only. I had some hesitation.
I'm thinking just maybe to much oil with that High rpm/load with what I
was adding. Same car, 2010 Rx8 R3. Mine is bone stock right now.
But will be putting on the full racingbeat exhaust in the spring. And I have
Mazdaedit ready to go.
They are adding a lot at wide open high load, high rpm. I was
sucking down the oil. Used about half a quart to 3/4 quart in
3 day weekend track days.

Half a quart to 3/4 a quart for a 3 day track weekend is nothing. Unless you only ran a few laps.

seemed a lot to me comparing to my older rotaries. But I have heard the 8 does use more.
I'd still like mostly premix in there if I can get it that way. Though I hadn't been on track
that much with my older ones. But I've premixed for a long time to. It's been awhile since
I was keeping track of oil used in the engine.

I was on track quit a bit all weekend. Lost count actually. About 8 times each day

I top up my oil and gas every 2 sessions. It will depend on track and driver, but I'm using half to 2/3 a quart of oil and 10gallons of gas every 50-60 minutes on track.

I typically premix 1oz/2 gallons, no issues with hesitation or anything else. I am thinking of going lower on premix though. This is an s2 and I'm running about 13:1 AFR wot.

So, I don't think premix is the reason for your hesitation. I would check ignition (plugs, coils, leads) and AFR at wot (could be an issue if it is very rich, or lean).

It is possible to disable the omp and only run premix. One thread on here with someone (magnusracing) doing that, running on track, and the enging holding up. So it is possible and probably won't kill your engine right off the bat. Is there conclusive data on what the safest, or best way to run is? Nope... Your other option is to lower the amount of premix, and trust that 3rd injector in the s2 to do it's thing. That's the direction I am leaning in.

My cat is toast! I get 6 in the city and 21 on the highway! I think I am going to increase the OMP after dealing with the cat problem and just add in a little transmission fluid to the 4 cycle oil... I will just use the best oil I can that doesn't make garbage.

Any data on the weight of the oil effecting the OMP? How about Zinc content?

I have heard 15W-40W is a better fit the 5W-20/5W-30 for the rotaries. Been using Rotella T.

:scratchhe

OK, sorry to hear that. Are you sure it is the cat? ie have you pulled it and visually checked?
Low mileage could still be ignition or something else. Be aware that track time puts more stress on cat, coils and plugs, so all will have reduced lifespan. Also, once the ignition start to go, that puts more stress on the cat. It could be you have to replace the lot... Removing the cat when on track and then changing plugs every year and coils every 2 seems to work for me. I do about 12k miles plus 12 track days/yr. I also left the OMP tables stock, but add premix on track. Have not generally done premix on the street, I don't drive that hard day to day.

I have not heard any concerns of the oil affecting the OMP, assuming it is suitable for the engine in general I think you will be fine. Also have not heard why you would want to mix in some transmission fluid?
All kinds of debate about oil across the internet... A fully synthetic 30 weight seems to work well for me. Folks are using 40 weight with success as well. There's a good thread here on this: https://www.rx8club.com/series-i-tec...m-here-184241/

Where do people get this myth that ATF is good as a two-cycle oil???

It's a good hydraulic oil, full of additives specific to making it a good hydraulic oil, and none of which make it burnable in an engine.

Specifically, the better the ATF, the more dimethylpolysiloxane it has, to prevent foaming inside the transmission, the worst thing an ATF can do.

Silicon is poison to the zirconia coating in lambda sensors - tiny amounts will kill them quickly; not just the cat one, but the vital and expensive air/fuel ratio controller. Even a small amount of silicone RTF inside an engine can destroy these sensors, so purposely adding ATF to your fuel is just plain dumb.

To paraphrase the great Dick Cheney - "You don't know what you don't know."
If you have no idea about ATF additives, don't just guess.....

S




ref: SAE technical paper 860478