kickerfox

I opened up the RX-8 wiring harness last night and started eliminating the connectors I won't need like the coils, injectors and various solenoids. I'm leaving the coolant temp, oil pressure, oil level, ESS, and throttle body for now. I've found all the wires I'll need for power to the Isuzu ECU and engine so at least it'll run. That should be fairly motivating when it does.

Hopefully my friend with an RX-8 lets me disconnect things from his car to see what it takes to keep it out of limp mode. He has a Cob so I'll get a chance to see what programing options there are for the stock ECU.

I finally found a deal on a crankshaft. I'll get my hand on that next week and have it spec'd so I can finally get the bottom end together.

Has anyone heard of issues with ACL bearings having tolerances all over the place? King bearings are cheap and I've heard good things about them. There's also Clevite bearings which I don't know much about in a performance application. Any advice here?

Paco664

i have always heard good things about clevite...

ABdragonVX

I've got good reviews on clevite, mixed on acl and never heard of kings

kickerfox

King has been around awhile. They claim to have 0.00010 tolerance.

I was reading about someone who had bearing variance of .005-.014" using ACLs and when he switched to Kings he had .008" across the board.

Kings are 2-layer that have a thicker coating that's supposed to increase "particle embedability" (great for engine life) but if that coating is soft I'd worry about it crushing over time in a performance/turbo application.

Here's a good video explaining bearing constructions.


I'm undecided but leaning towards tri-metal. That's what the OEM bearings were.

kickerfox

RX8 parts are good for something. Making cordless leaf blowers! :D

Wait... don't I have an engine swap to be doing?

Paco664

ok... that is funny...

kickerfox

I'm doing what I can to clean things up before assembly and no that's not paint although I will paint the block before assembly. Color suggestions? I should paint it pink just for you guys. :P

I'm not looking forward to stripping the heads though I am looking forward to the end product. The car's electrical harness is stripped of unnecessary wiring but I can't start splicing until the engine is in.

Paco664

update??

kickerfox

I cut a flange off the intake manifold to send to the shop for cloning. Ripped out unneeded wiring from the 8's body harness. Also been going over access port options to see what I can change in the 8's ECU like fan temps. I've been talking to a few guys on the Isuzu forums who turbo'd the v6. They are both going to use the FIC and I'm helping them find injectors. Looks like the 2012 Mustang injectors will work for them. And got the flange made for the 8's throttle body and pulled the front end to size up for the FMIC.

Anyone have a '04 trigger wheel?

I'm working on the intake manifold now. I'm not very good with 3d CAD and time is money at a CNC shop especially if they have to do the design. I took the cut off flange and milled it so the rough cut side is parallel with the head flange. Then I'll scan both sides, scale them to proper dimensions, import them into CAD and have it plot the area between them. The Isuzu intake runners aren't round so I'm going to edit them to be round and also the same size as the tubing I'm using for the new runners. The reason for going through this much trouble is because the fuel injectors fit nicely into the factory flange right at the head. I won't have to worry about welding in injector bungs. I'll just need to weld on tabs to hold the fuel rail.

kickerfox

My CAD program sucks. I can't import scans in PDF format.

I have to redraw flanges from scratch. I'm placing the openings first then I'll add the radiuses around them to pretty it up. Everything measures up within .25mm. Close enough. Slowly but surely these parts will get done and then put it all together.

rx7roller02

Maybe it is just me, but I'm partial to the Ford 3.8L V6 from the Thunderbird Supercoupes....1994-95 engines were detuned to make 230 hp at the flywheel...and 330 ft. lbs of torque at 2500 rpms...stock. Get rid of the ridiculously bad exhaust, free up the intake and fuel, and drop an overdrive pulley on the M90 and youre in business. I have no idea if it would fit in the 8, but just a thought. Ford actually put a right angle into the exhaust to drop the power down on these so that the Mustang GT would still be faster....no kidding. You could make massive amounts of torque with one of these.

kickerfox

As soon as it comes DOHC with an aluminum block I'll look into it. It's a little too late to turn around but thanks anyways.

I really like what I found with the Isuzu engine now that I got my hands inside of one.

Paco664

i have been considering that.. also thinking about a chevy v6...


just thinking tho...

kickerfox

Careful what engine you choose Paco. There could be... disagreements.

Paco664

lol... the immensity of the **** that i do not give as to others opinions is so large as to not be able to be contained by the world wide web in totality...

*(in other words if i gave less of a **** about their opinions the universe would collapse from the vacuum that it created.... lol..)

kickerfox

This is fun! My friend and I recreated the pulse train from the RX8's ESS using audio software and injected it into the RX-8 ESS connector. We made the frequency of the signal at 1200hz which is 2000rpm. Playing the audio sample shot the tach straight to 2000rpm but within one second it dropped to zero BUT the throttle pedal was now responding to user input up to about 10%. Without the signal present, the throttle pedal would only fully open the throttle body 100% to de-flood the engine, I assume. I'm guessing the ECU does indeed think the engine is running but until I fake the sensors it's stuck in limp mode.

The waveform was based off a sine-wave but it's roughly what a CAS (or ESS) sounds like for those who are curious.

http://www4.zippyshare.com/v/69630200/file.html

EDIT - Plugged in the oil injection pump and the MAF sensor and had throttle control for about 10 seconds before it entered limp mode. Getting there.

kickerfox

I jumped all the ECU's solenoid leads with a 1k series resistor to power and that seems to have eliminated all the solenoid codes. The missing injectors and coils do not throw any codes. The only codes I have now are for no ABS communication (ABS motor is currently removed) and one for the ESS which is likely because my signal isn't perfect or because the sensor is missing. I'll put a resistor parallel with the sensor input and see if it helps.

I now have full control of the throttle body from idle to WOT so it seems I'm keeping the 8's ECU out of limp mode. As it is now, this gives me throttle control of the Isuzu engine and full instrument panel function. WIN!

kickerfox

So far so good. I haven't found a way to fake out the oil pump but I can retain throttle control from idle until the tach is pegged (over 10k rpm).

I have a ton of codes (most I can trick) but the following sensors/relays/solenoids are totally disconnected and I still have throttle control.

-All the solenoids/relays*
*Secondary injection solenoid
*Variable dynamic intake effect solenoid
*Secondary shutter valve solenoid
*Air intake duct solenoid
*Purge solenoid
*Air pump relay
-APV motor/switch
-SSV switch
-Fuel injectors (6)
-Ignition coils (4)
-Front and rear O2 sensor
-Airflow/Intake temp sensor
-Knock sensor
-Electronic steering control module and steering rack

Now that the throttle and tach are working, I know I can wire in the Isuzu ECU/engine and have drivability. Next will be tricking the Isuzu's idle air control motor since the RX-8's ECU has control of idle. My DSC won't function without the ability to calibrate it. I'm guessing this also disables traction control. That's fine though.

j9fd3s

cool!

this is probably obvious, but the solenoid and relay codes will go away, once they are plugged in.

the oil pump is traditionally tricky, as it has the output stepper motor, and an input position signal, so you can't just add a resistor, it needs to see certain positions vs rpm and load, so it might be easier just to tuck it away somewhere.

kickerfox

I'm THAT board at work today. Think the ECU can track it?

Video - TinyPic - Free Image Hosting, Photo Sharing & Video Hosting

kickerfox

I can trick all the relays/solenoids with a 1k pull-up resistor to 12v. The ECU thinks the coils are still in place. It works for keeping codes clear but I don't need to do it apparently.

The oil pump doesn't have positional feedback other then "closed". The sensor on the back looks like a TPS but I believe it's just a switch. The pump's stepper motor cycles to the closed position at idle. It looks like the ECU sends it pulses to position the stepper motor but doesn't know where it is except by counting the pulses it sent. I'll be looking deeper into it but yes your right it could be tucked away somewhere. If I can simulate it easy enough I might just try.

kickerfox

Yep. It can track it.

j9fd3s

lmao! your whole project was worth doing just for that

Paco664

update??

kickerfox

Waiting on parts and tied up negotiating a building purchase. Haven't had time to do anything other then fiddle with the wiring.

Anyone have an '04 trigger wheel? I can't find one.