1979-1993 & 94 Conv [Subscribe to Daily Digest] |
Hi guys,
Since I've determined that my AMM is defective and I have to order one tomorrow, I thought I'd check out if the AIC is good or not.
After examinging the output signals from teh ECU with a scope I incorrectly assume that the motor held its position with a 50% duty cycle; 5ms On for Pin 1 and then 5ms on for Pin 3. Moving the valve so that it would open or close seemed to be a simple matter of adding an extra 5mS pulse to pin 1 or 3 so that there would be a single 10ms long pulse to shift the valve and then it would stay stable at the new location.
Wrong! I used a Revision 1 surplus uController I had from my Aluminum casting furnace and whipped up some code to do the above algorithm. The motor immediately went to the 1/2 open point and vibrated nicely. I then addedthe 5mS pulses in either direction and the motor open or closed the valve for a moment and then returned to the same spot.
Hmmm. A bit of trace cutting and rewiring and some new software and I had a 112Hz PWM signal that I could adjust from the keyboard by tapping the comma and the period keys. I started it at 50% duty cycle.
Hooked that up, applied power and Yippee! AIC moved to 50% open point and after that I can send it comma or period characters to open or close the AIC. It moves smoothly throughout the entire range from full open to fully closed.
So for the folks reverse engineering an LH2.2 ECU, that's what the code is supposed to do. I measured 100Hz from the SAAB ECU but obviously 112Hz works too. I imagine the ECU will attempt to reach 850RPM using the Distributor Hall Effect Sensor signal and close the loop by adjusting the pulse width out to the AIC.
How well it does that depends on a large number of factors like Vacuum leaks, O2 sensor response and AMM response since there is obviously another control loop running that tries to maintain the O2 sensor at 0.5V.
When I designed the Honda VTEC controller, we removed all Automatic Idle Control because an aircraft just doesn't need something like that and I maintained a closed loop mixture control by monitoring the O2 sensor and adjusting Injector Pulse Width.
I will guess that the SAAB controller will first adjust idle to its target value based on Airco on or off and throttle switch and a separate control loop will adjsut the mixture to maintain emissions by trying for 14.7:1 Air:Fuel ratio (0.5V O2 sensor).
The vacuum leaks will change the mixture which change the O2 which change the Injector pulse width which changes the idle speed which is compensated by the AIC. But the the new idle speed changes the vacuum leak flow which changes the mixture which changes the O2 ....
Hence the puppy hunts all over the place. When the engine is cold the Bank Fire Injection system atmoizes and distributes the fuel so much poorer plus with the cold engine enrichment (overly rich mixture) that the hunting is far worse. Muli-port sequential is less sensitive to that but a heck of a lot harder to write good code for and wouldn't have fit in the older technology processor used back in teh early 80's.
So that's how the AIC works. Given enough prompting I could probably be convinced to develop a small AIC Motor tester but the cost would be in the range of a spare brand new AIC so it's probably not worth it.
Cheers All
John
posted by 64.180.68...
No Site Registration is Required to Post - Site Membership is optional (Member Features List), but helps to keep the site online
for all Saabers. If the site helps you, please consider helping the site by becoming a member.