Email me your session and log file.

A big thanks to the guys offering to help. Andrew at Rasant suggested switching the vr + and - wires (for the crank sensor). That did the trick. Also had to change the Crank Missing Tooth detection threshold to 2 since I have a 60 -2 crank wheel. For reference, the Clewett three wire (looks like two wires plus a shield) crank sensor is mag and the Clewett cam sensor is a hall sensor.

Unfortunately my hose making skills were not up to par (had some leaks) so I will remake them and test them before mounting them again. At least I could verify that the fuel pump works fine.

Gentlemen, we have engine start!

I had all of the hardware done by my schedule but those pesky fuel lines kept delaying me. At first it leaked at the hose/connector interfaces. My bad. Redone correctly. Then I decided I wanted some other fittings to remove some of the hoses. Wait for it to come. Decide I want another fitting, got it, but it was the wrong one; order the new one, etc. Finally I got to the point where the hoses were all good but the leak was at the fitting interface! Rather than buy other fittings, I found some washers that go inside and bingo, problem solved. Except that another fitting leaked where it has an internal seal. Since I had extra fittings, I replaced the offending one, tested and all was dry.

After getting the engine up to temp, I noticed that the fuel pressure was low so I adjusted the regulator. Checked the timing and it was good.

Now the tuning fun begins...

Here are pics of the washers. Used in aviation. Got them from Aircraft Spruce.

http://forums.pelicanparts.com/uploa...1504313390.jpg


I tested the decklid after fixing my bracket mishap and found I will need to cut into it to get clearance for the AC pulley. Early on I documented how I had to raise the compressor to clear the fuel rail.

Its been a while since I updated so here goes. At this point I am calling the project a succes. Still refining VE but it is completely street useable and conservative until I get the dyno tune. I have one lead on a tuner and am waiting to hear back. Idle was a bit of a pain but is relatively under control. Todays idle was great. I can't wait get it properly timed as I currently have a map along the lines of the factory distributer. Also, I need to work on the accel enrichment.

I ran into a series of videos by Adaptronic that are great for those who want to learn to tune efis. They are found in youtube in the AdaptronicECU channel.

It is currently normally aspirated but I am now working on rerouting the bypass valve on the supercharger so as to not interfere with the fuel rails. Should be boosting pretty soon.

BTW is the factory 5 deg BTDC at idle still a good target now that I have the flexibility to modify it?

I use 8 on my 3.4 at idle. Sure, 12-15 makes for better throttle response, but the engine sounds like it's too advanced tone wise. 8 on my motor produces a nice deep rumble.

It's been a while since I updated so here goes. So far I am very happy with the way the car works and responds. It essentially has a street tune and will eventually get dynoed when I reinstall the supercharger. here are a few things I learned.

On the VE table:
1) It's a shame that the infinity does not have an auto tune function like the Megasquirt. They already give you what the VE should be but does not automate the update.
2) the AEMdata software is great, but, again, I had to spend hours looking at my data logs to estimate VEs.
3) then came this awesome tool from a fellow Pelicanite
http://forums.pelicanparts.com/porsche-911-technical-forum/941314-tuning-aem-infinity-check-out-helpful-tool.html
Much more automated but I had to preprocess the data to remove points where there are transients, wallWetting, etc. A lot of work up front. Then I figured that it would be better to use the median rather than the mean so...
4) I developed my own Excel to give me the mean values rather than averages. Although not as neat as #3, it gave me the data I wanted. Then had it finally look at number of hits on each table entry and remove those with few hits and finally...
5) I found MegaLogViewer. OMG. This thing lets you create your own filters, load the data as a csv from AEM data and produces the corrected VE table, along with many other visualization tools. Finally, what I was looking for! Also, the thing is blazing fast even on my old computer.
Get this. It is what you need.

Idle:
1) I decided I wanted to use the three pin IACV used on the 3.2. It takes 12v in the center pin and I use two lowside outputs to PWM the remaining pins. When one is 75% voltage, the other is at 25%. Fully open one way is 0% to 100%, etc.
2) had a bit of problem getting the valve to reliably respond in the idle position table. A little search showed that these things can be very nonlinear. After reading this, Bosch rotary idle valve info - FrankenCIS - FrankenCIS Community
I did an experiment. Without starting the car, look inside the IACV, and change the position value from 0 to 100% in steps of 10%. Well it seems that below 25% and above 75% the values are nonlinear so I changed my lowside tables so that 0 is 25% and 100 is 75%. Now its much better. I am much happier with the idle response. Using the idle screw, I adjusted the idle to where the response of the IACV is reasonable.
3) once his is set up right and the VE at idle is close, there are less hassles, particularly with feedback. On the IACV PID, I am exploring different values but it works fine with P=D=0. I was set at 0.02. For the spark feedback I'm at 0.002 for P, 0.001 for D and 0 for I.
4) At the advice of others on the board, I set idle AFR to 13.5. I dont have an ITB setup so I explored leaner mixtures. Lo and behold, it works just fine at 14.0 and even leaner. The EFI makes for a smooth engine. I left it at 14. Why not save gas?
5) above about position 60 or so, the IACV is sufficiently open that you can hear an annoying sucking sound- need to avoid having it open it so much.

Well, I will keep posting as I continue to learn and improve the tuning!

Good read Dr J. Glad MLV branched out. Great tool.

I wouldn't fear runner a little fatter at idle as idle doesn't use that much fuel. Makes transients much easier when you're at mid-13's, at least in my case using batch fire

Use of sensors for engine protection
 

Thanks for the suggestion ^^

On another note:
Has anybody used the OEM sensors for engine protection? For example: connect the temp sensor to the infinity to use for engine protection strategy while using a lowside output to move the gages?

Presumably you would put a voltage signal through a resistor to the gage and tap the signal at the resistor to gage connection.

While we are at it, anyone connect the transmission odometer sender to get Vss?

Dr. J,

For VSS, Porsche uses a plate with 8 magnets mounted to it that is installed inside the transmission. Each axle revolution gives 8 pulses. There is a sensor (actually a reed switch) that is mounted to the outside of the transmission case that senses when a magnet passes by to create a 12vdc pulse to ground signal. The sensor is connected to the speedometer in the tunnel where the shift coupling is located. Use a meter to determine which lead goes to ground, then use the other lead to tee off and hook up to your AEM input. If AEM has an input that is looking for a ground (pulsed input) then there would have to be some parameters setup for tire size, and final gear ratio in the software. That is the way it works for MS3, not sure about AEM. If you do get it work with AEM be aware that this speed reading is much more accurate than the Porsche speedometer. My speedo read 5-7 mph fast, which I knew. I installed a speedometer calibrator to adjust the signal going into the speedometer to correct mine.

Rasant has identified new rail plugs that are 1/3 to 1/2 the thickness of this version. Using those eliminated my A/C clearance issues.

That's good to know. However I already did all the mods and cut the decklid at the appropriate location. Hopefully the next person along can get the smaller rail plugs.

On to the supercharger. I realized that being speed density I can use the bypass valve as a blow off valve and not recirculate the air, so... I sealed the part of the supercharger that accepted the recirculated air, took off all of the piping, and mounted it in.

I was done around midnight but had to try to start it. Woe, was that thing loud. All of the air coming out at idle made too much noise. Turned it off and called it a day.

Decided I need to recirculate the air, so I'm having the pipe back to the supercharger modified so that it does not hit the fuel rail. It will come out the back, up and over the top. As soon as I have it together I will post a picture.

I have Oil Temp and Oil Pressure sensors going into the Infinity then use 2 lowsides to control the gauges on the dash. PM me for details on how to set this up. Works really well, no resistors or anything wonky required. It's nice knowing that what the gauges are displaying is what the Infinity is seeing.

A short while ago, I noticed that when sitting at idle, the reading on my wideband sensor would "blip"...
My AFR at idle is approx 14.5 - 14.7.. but it would blip to 16, 17, 18 etc and it is VERY cyclical. It is just under 2 seconds every time.

Is this an indication of a failing wideband? or is something failing on my ecu? (Infinity 506)

Mike, I sent my observations via email. I saw the blip in your logs. At times leaning a lot and at others enriching periodically. I could not match it to wall wetting or anything else though I suggested a few additional parameters to log that may help us out.

I've just ordered a new wideband sensor.. I don't know it if is failing or whether something is causing the 'blips'.. but we will find out soon enough!