I just separated the housings and placed the rotors to the side.

As I expected, nothing going on with the rotors or apex seals.

I did see some suspicious things with the housings.

There is a little notch/flaw in the front iron about 1mm in diameter and 0.5mm deep.



The bottom half of the sides is discolored on both the front and rear covers. I'm not sure if this is normal, but there isn't any deformation in either iron.

Front and rear rotor housings have no apparent wear. I'm eyeballing everything using a micrometer for anything that looks less than perfect.

I noticed this when I took off the rear iron:



Though, that might be some leftover coolant coming up when I removed the tension bolts.

Does this look like a breach?



The inner and outer coolant seals seem like they changed in size, and they won't fit back in the grooves right. I'm replacing them with the Pineapple-racing heavy duty water seal kit.

 

My car computer components showed up.

Stuff:

mini-ITX VIA 800Mhz/512MB RAM
30GB HDD (spare)
7" mimo USB monitor
DC-to-DC ATX power supply
Windows XP SP3

Unfortunately, the monitor only supports Windows XP/Vista (I wanted to use Linux), but it is a cost-effective solution for a portable screen with real VGA.

This system will talk directly to the PowerFC via i2c.

My Japanese is pretty bad but this site gives me a good enough idea on what to look for. http://kaele.com/~kashima/car/pfcadp/

 

I'm reusing a component I put together before for a robotics project. The system (not just this board) made use of the serial port as a bus, so that I could talk to multiple PIC microcontrollers at once.

This should be helpful for creating analog inputs in addition to the data received from the PFC. The circuit is pretty generic. I did this so that I don't have to rebuild it everytime for serial port interfaces.

 

Here's something interesting I found the first time I took a shot at this interface a few years ago.

Left = PS2 standard input
Right = PFC commander pinout



If you connected the two, the Clock, Data, Power, Ground lines all hook up (the backlight would not have power). While I would never suggest trying to jam a PS/2 cable from your PC to the PFC, it does make the figuring out the communication protocol much easier.

 

Prototype circuit. Not finished, but this is pretty close to what I need to sneak-a-peek at the PowerFC.



After wiring it together, realized that the 20-pin PIC16Fxxx is just enough chip to handle this task. Operating at 4Mhz, it can reliably talk to the PC at a 19.2K, but the i2c chatter can easily outrun this. So, I added a 32KB memory chip for buffer memory.

 

Turns out I did not need the 32K sram for some simple sampling. It will be added after I determine some specifics.

The PIC was programmed to monitor SBO/CLK lines of the PFC at 100KHz and send 80byte bursts to the PC. This represents: (4 samples/byte) * (10 us/sample) * (80 bytes) = 3200 us or 3.2 milliseconds per block.

First, I tested the handheld by itself, and it repeatedly sends the bytes at 19.2K baud (8-bit even-parity): 0xD7 0x02 0x26. I'm pretty sure this is a greeting, but it is in a RS-232 mode.

I removed the PowerFC from the car and hooked up power and ground in my office then placed the PIC in-line. Now, I'm getting i2c behavior.

I'm still trying to decode the waveform, but once I figure out the pattern, I can write some code to disassemble the remainder of packet data.

 

Haltech ECU.

 

Sweet

 

Rotor side sides aren't here yet.

This is exciting stuff.

Apparently, there are some differences between how the PC talks to Datalogit vs. directly to the PowerFC.

Items of note:

0xD7 - The Commander is sending menu selection information to the PowerFC whenever a screen change occurs. This makes sense. It hit me when I powered up the system today and the first screen the Commander went to was the 1-channel battery voltage watch (where I left it when I powered off). There's quite a bit of data in this packet, so it will be a little while before I can completely decode it.

0xD7 0x02 0x26 - The Commander is asking the PowerFC to tell it what screen to show next.

0xE1, 0xE2, 0xE3, 0xE4 packets manipulate the IGL map. 0xE0 sends a write command for a cell.

0xC1, 0xC2, 0xC3, 0xC4 packets manipulate the IGT map. 0xC0 sends a write command.

0xC6, 0xC7, 0xC8, 0xC9 packets manipulate the Injector map. 0xC5 sends a write command.

More to come...

 

I've decided to create a thread specifically for the PowerFC work.

https://www.rx7club.com/power-fc-forum-47/bluecu-powerfc-tuning-software-840225/

 

Replaced some side seals and cleaned the remainder of the rotors. Tension bolts are back in.

I turned the e-shaft to listen for leaks in the rotor cambers. To my encouragement, the sounds were more pronounced now than before I took the block apart. This is by no means a legitimate test, but it is nice to hear that no air is leaking through the coolant channels.

 

Flywheel nut is back on. E-shaft bolt is back on. Replaced some bolts for the front cover. Eliminated the OMP.

Apparently, the last person to put the water pump on cross threaded the crap out of the housing rods, so they'll need to be replaced.

Hopefully, I can find a hardware store which stocks M8x1.25 120mm bolts.

 

Bolt threading tools, ftw.

I was able to save the front housing bolts. I also replaced a bunch of worn nuts and bolts.

The engine is coming together nicely. I made the mistake of breaking off a bit one in one of the UIM holes, so working that out is going to suck up most of tomorrow.