Corvair Fuel Injection Blog
What it took to get from there to here
Goals:
The use reliable, cost effective and off-the-shelf domestic parts available at major auto parts suppliers (adapters, linkage, etc)
Develop fuel injection to use REGULAR unleaded gas, improve performance and economy.
Improve engine durability and drivability through use of modern techniques (Electronics, Fresh air PVC) in conjunction with Fuel Injection
All additions/modifications are BOLT-ON and could be removed if stock is desired.
First.... I'm not against Megasquirt, but at $400-500 you must spend for the assembled computer and harness... the "stick it to the man" system was not for me after I found out I was "The Man". After researching the DIY sites and holding to the goals above, the GM 1227747 computer was chosen as it is inexpensive ($35 in Junk Yard/$70.00 at O'Reilly Auto) and is easy to program. The 2732 Chips are $2.00 each and the software to create your bin files are free or next to free. A Chip Emulator (Romulator by xtronics) was used to enable live changes to the 7747 computer while driving and this project would not have been successful without it This is my daily driver for my business and I drive about 200 miles a day making on-site visits. I'm under the assumption that if I can make a 68 w/dome pistons not knock and decent power on regular gas, the project would be better suited for other Corvair engines.
Engine/Trans: 1968 164cid, 110 HP (dome pistons and heads), PG, 3.55
TBI Phase I - Development and Testing Stage
Jan 08-Jun 08 (Mileage on Car 78223)
A center mount TBI system was chosen to test the computer and develop the knowledge for controlling air-fuel mixture and timing with an air-cooled engine. Testing started 29 Apr 08 using a laptop with tuner pro and a GM 1227747 ECM. A 2.0/2.5 Throttle body and a converted GM 4.3 distributor (7 pin) for EST (provided by Ed Kelly). Testing was stopped 29 Jun 08.after 2,000 miles and converted to dual TBI.
Top end performance is unbelievable compared to the Carbs
Controlling timing/knocking was major challenge using ESC. ESC was disconnected after I found it was making timing erratic.
Stalling was a problem on quick stops (No VSS, so the engine had to rely on RPM to correct for quick decel)
Real fuel mileage increase of only 2 MPG. (I did keep my foot in it a lot though)
Included Fresh Air PVC with internal orifice. (IAC circuit did not like fully open PVC)
Ignore dates on Photos
TBI Phase II (Dual Throttle Bodies) July 08 to Jan 09
Dual Throttle bodies were added with 2" adapter plates. The Carb bolts had to be removed, but was a straight on bolt-on. Linkage is now controlled by cable to the transmission rod.
Chip was burned to get the laptop out of the front seat and enjoy driving. Interstate driving at 75 mph is non-labored and does not struggle to keep that speed.
Top end performance is still unbelievable, easy start, good transition and actually a joy to drive....... I love this freaking car
Stalling was corrected after adding 2nd Throttle body and increased power and smoother.
Linkage was the hardest thing, but slick now using $5 lawnmower part., Temp Sensor fits into the top head perfectly and registered about 190 f on a warm day.
Fresh Air PVC internal orifice removed. (almost no oil dilution)
GM ECM mounted behind back seat and I did drill a 3" hole between the engine compartment for the ECM wiring. (Picture has chip out)
TBI Phase III Final Setup for Fuel Injection
Changes that are being made now are cosmetic
Currently a Daily Driver with over 11,000 miles on this setup as of 20 Dec 08
In Tank Fuel Pump w/ Modified 3 Prong Oil Safety Switch
The noise of the external pump is driving me crazy so we are using a standard Bosch Pump for a 89 Chevy Truck and a Oil Safety Switch for a 76 Vega, we now have the following
A pump so quiet that you don't know if it is running (with the engine shut off and the key during the TBI 2 sec pre-start)
Safety Switch that gives you the Idiot light when no pressure is present and ground the Fuel Pump Relay when pressure is present. (Ground out the P Terminal)
Idle Control (IAC) and VSS (Variable Speed Sensor)
I broke down and purchased a VSS from JTR and it should mount to the back of the Speedo between the cable. The cable and the VSS or 2 different sizes and JTR never returned my email about it.
The latest bin has been created and I'm too happy. The car runs great and does not miss a lick.... I get in it in the morning, turn the key, put and gear and drive (just like a real car). It has good power through out and I burned a couple of chips just as a backup for the romulator as I won't need this thing much longer. The GM DIS really made a difference in smooth power and I will NOT go back to HEI.
COMPLETION: 23 Jan 09 (Mileage on car 93,480)
The project is now completed and only cosmetic changes will be made. All Idle issues were corrected by using spark advance. IAC problems were corrected by using maximum retraction steps and retarding timing at fall points. The final bin file for the Corvair has been written and will not be visited for awhile as its time to drive and enjoy this car. The timing is conservative and I'm employing a knock sensor the stop even an accidental knock. I love this car and is a joy to drive. This was a long and hard road, but was well worth the trouble.
Everyone want to know how much.... I spent well into $2000.00 in R&D, but it could be DIY for less than $500.00
Because no project is never perfect.......
UPDATED: 24 May 09
The Knock sensor and the IAC have been my latest project to perfect. The knock sensor would give false positives at around 3600 RPMs and was knocking the timing down and the IAC was never that dependable, but I think I have found the answer.
IAC: Airflow around the IAC is so intense inside the bore that it disrupts the airflow going into the intake which in turn causes the stall. I have rigged up an external IAC and Idle is now controlled without stalling. I have been programming around it so long, I might work it a little more.
Knock Sensor: I basically added a small piece of pipe to the end of the sensor and the false knocks stopped.
Speed Sensor: The VSS I ordered from JTR seemed to be a great product, but the VSS would not connect to the Corvair small (5/8 vs 7/8) speedo threads. I used a VR sensor from the DIS project, used a NPN transistor, pullup resistor and mounted the Sensor at the yoke of the shaft. The 2 yoke shaft creates a magnetic break field and generates a signal (AC Sine). The transistor take the AC Sine and acts as a relay and "relays from ground" the GM computer (same thing the VSS does) and it generates a square wave DC signal (ie: MPH). The only draw back right now is that it only reads at speeds above 22 miles per hour, which is fine for the computer.
Update 1 Jun 09: Built and installed a remote IAC motor on a block of alum. This work great off the crossover tube and makes this car idle great in warm weather, but lacks the volume of leak required for high idle.
Update 13 Jun 09: Turned over 100,000 Miles and back to zero. The injection system has a good 27,000 miles and the DIS has about 10000 miles and all with no failures (yet... knocking on wood). I do miss having something to do with my car though.... maybe I will wash it.
My latest Thoughts on the ECM and Spark Advance (Jul 09)
I'm happy with the system except at only one time. High ambient temp under load. I currently use oil as my coolant temp for the ECM and it does warm up slow, but no problem have been encountered as I set the Spark advance low at the low temps and really concentrate on the Engine at regular operating temp. (180-230). The ping is pretty much around 90-100 degrees ambient, which does not happen much, but it does happen down south which is what I want to stop completely without using the knock sensor. I found the sister (for the 89-92 Camaro V8) for the 1227747 computer which is the 1228746 ECM. This ECM uses the IAT sensor which uses the ambient temp along with the coolant temp to adjust spark advance.
I am getting ready to add the 1228746 ECM and the IAT sensor to the Corvair. I've still got a lot of work to do before I can actually run it and hoping I will find the time before summer ends.
Oct 09
The 1228746 ECM project (which I think is the last missing piece) is now underway and a DIY or an authorized installer Kit should be ready for summer. Its going to take a lot of work and more research, but it took many people asking to make me do what I should have done a long time ago. I have been concentrating on fuel mileage lately and got a surprising 23 MPG on a trip to Maggie Valley, NC today. I've added the highway mode to the ECM to lean the fuel and improve mileage... Looks like it may have worked. Bottom end power is still a problem with the Powerglide without pinging the engine to death. I'm hoping the 1228746 ECM will help with bottom-end power (2 degrees below ping) and keep the fuel mileage good. I disconnected the additional oil cooler noted above as I'm wanting to cut-out possible areas of failure..... I drive this car everyday
Dec 09 (Clubbing Baby Seals)
I've been concentrating on fuel mileage way to much and allowing the pinging to overtake my life. Running 14.7:1 AFR might be best for "planet earth" and pocket book, but who really cares about that.... ITS POWER. I went back to the 1227747 ECM (because its easy to get) and put in open loop. I have set the open loop at approx 12.5:1 AFR and 2 things happened, 1) More power at all RPMs 2) The richer mixture makes the fuel burn slower which reduces ping.... yea you use more gas, but I came to dance not sit on the sidelines. I also have started using the 87 Octane fuel and it gives a little extra buffer after my daily 60 mile run @ 80 MPH which puts the engine at max temp. I can't stop driving this car and drive it every chance I get. Im getting about 17 MPG now, but this 110/PG is feeling its oats.
Feb 2010 (40,000 miles on ECM/Fuel Injection)
We've proved the Fuel injection system is dependable as it been 40,000 miles without a single failure, but I did throw a belt. The 1227747 ECM development will soon some to an end and the next step should be made. The ECM idle strategy leaves a little to be desired and development work has already started on the new ECM. The current bin being used it suitable for Regular Gas during the winter time as summer has not been evaluated as of yet, but is a GREAT running bin and is my daily driver. Adapter blocks are going to be sent to a machine shop for replication and hopefully get a good price. I plan to stock pile and modify enough parts to equip 10 engines, but that is later this summer.
1 Mar 2010 - Idle Update on the 1227747 - IDLE PROBLEM SOLVED
We've got a great running car with the 1227747 ECM, but the idle had always been a problem during cold to warm. The 1227747 must have the Variable speed sensor (VSS) connected in order the the ECM to make idle calculations. Voltage jitter has also been a problem in the past, but that has been corrected with my latest bin file. I have made many runs from 30 degree to warm and it has behaved all times. The real cold idle is no higher than 1200 RPMs and settles down to 750 after warm and transitions well to between drive and neutral. I'm just about out of cold weather to do much more on it this year and got to get the summer calculations made also to make it bullet proof. We must also have both IAC motors connected to make the dual TBI idle correctly and the MAP sensor should be connected to the crossover pipe, not to the TBI. Also running midgrade fuel now instead of low 85 octane... this is mainly for pinging when head temps climb.
Second Generation ECM (PCM) is scheduled for development at the end of this year (2010)
I plan to use the 94-95 GM PCM which has tons of memory and uses a better spark knock theory. It also has better idle routines and has larger fuel and spark tables. RPM/speed limiters, Wideband monitoring are just some of the good things about this ECM. I have everything to install minus the wiring harness, but since I got the 1227747 to idle, I want to finish this project before I start another. Its a lot of work and cash to get these projects going..... over $1000 so far on the new ECM and I haven't even got it in the car. Extra parts, romulator (512) wideband O2 sensor, etc... the cost went up pretty quick.