Here is an article that appeared in Hemmings Motor News- more than you ever wanted to know about the CCC Q-jet and applies to the Dual-jet as well.
When General Motors' line of 1981 model-year automobiles reached the showrooms in the autumn of 1980, many car enthusiasts were surprised when they looked under the hood. They believed the company had gone mad. Electrical wires were routed into a carburetor full of gasoline. How can that be? According to conventional wisdom, it made no sense. But with those cars, the electronic series of Rochester carburetors made their debut; from that day on, carburetor function was never the same.
Rethinking an old concept
With the federal government mandating stricter emission control standards, a more aggressive approach to fuel delivery was required. The catalytic converter, which had been cleaning up exhaust since 1975, was very sensitive to the air/fuel ratio emitted from the engine. It was only efficient at a mixture strength of 14.7:1. If the amount of fuel and air skewed much beyond that, whether rich or lean, the rate of conversion was greatly reduced and tailpipe emissions increased exponentially.
Thus, the electronic carburetor--technically speaking, the proper term is "feedback carburetor"-- was not actually an emission-control device, but really a way to ensure that the catalytic converter did its job properly. The goal was to keep the air/fuel ratio at a stoichiometric ratio (the aforementioned 14.7:1) during as many operating conditions as possible. The oxygen sensor acted as an auditor of the results.
General Motors introduced four different electronic carburetors--three different two-barrel designs and a four-barrel model. The two venturi carburetors were the DualJet and VariJet; a unique two-barrel application for the four-cylinder Chevy Chevette was produced by Holley. The four-barrel model was a new version of the old QuadraJet.
If the prefix "E" was placed in front of the model number on a carburetor, it indicated an electronic version. The last letter identified the style of choke. Again, an "E" was used for an electric choke, while a hot-air style was cataloged with a letter "C." For example, an electronic QuadraJet with an electric choke would be model E4ME. The same carburetor with a hot air choke would be identified as E4MC.
The two-barrel DualJet (E2M series) was basically a QuadraJet that was cut off behind the primary barrels, and it looked like just that: a shortened four-barrel that only had two barrels. In contrast, the VariJet (E2S) was a QuadraJet cut longitudinally. It had both a primary and secondary bore and, thus, was a progressive two-barrel design. The VariJet did not resemble the QuadraJet it was based on; it had a more rectangular shape. The GM/Holley Chevette carburetor was also a progressive two-barrel.
The Rochester-produced E4M and E2M feedback carburetors shared basic operating components, while the E2S VariJet used the same theory of mixture control but with a unique fuel control solenoid that also incorporated the air bleed. The Holley/Chevette design used a mixture control circuit similar--but not identical--to the E2S.
On a standard Rochester carburetor, a vacuum piston that sensed engine load controlled metering rods, which were responsible for enrichment. During high vacuum operations (light load), the metering rods would be pulled down into the main jets (the rods were stepped), limiting fuel flow. As load increased, the throttle opened and the vacuum signal weakened. The spring-loaded metering rods then moved up out of the main jets and increased fuel flow as a function of signal strength. Rochester engineers believed that, if modified, the metering rods could be used at all times to tailor the mixture strength for emission control.
The carburetor was modified to eliminate the vacuum-operated power piston; this was replaced by a duty-cycle solenoid controlled by a small microprocessor located in the passenger compartment. Part of a complete engine management system that GM called Computer Command Control, or CCC, that microprocessor was identified as an ECM, or Engine Control Module.
The metering rods and main jets on the new carburetors were redesigned to accommodate the mixture control solenoid (M/C solenoid). The traditional jets were very small and shallow, while the new design was tall--a little more than one inch in height. The M/C solenoid was a fast-acting electromagnet that had a small plunger in its center with a lightweight spring under it; the ECM provided both 12 volts and a ground circuit for the solenoid.
When the ground was applied (internally in the ECM through a driver circuit), a magnetic field was created and the plunger was pulled down into the solenoid. The plunger had two wings on it that sat over the top of the stepped metering rods. When the metering rods were pulled into the main jets, the mixture would be driven lean. On command from the oxygen sensor, if the air/fuel ratio went too lean, the ground circuit would be removed and the springs under the metering rods and the plunger would now push them up out of the jets, supplying more fuel.
The movement (or cycling, as it's technically known) of the M/C solenoid occurred at a rate of 10 hertz, or cycles, per second. This often led to confusion because the duty cycle was fixed: Only the length of time in each position was altered, with the lean command (ground circuit applied, metering rods fully seated) as the determining factor.
With the ignition on and engine off, the cycling of the M/C solenoid could be heard as a constant clicking sound.
To understand the action of the metering rods, think of a dump truck that travels from an excavation site to a dump site. Instead of cycles per second, consider the truck's trips per hour. The truck driver is required to make 10 trips per hour from the work to the dump site. If he were to spend equal time at each site, then in one hour he would spend 30 minutes at each location. If he liked to talk to someone at the dump site, he could still make 10 trips per hour, but he would need to shorten his time at the loading site. If the inverse were true, he would need to hasten his time at the dump site.
Now replace the driver with the M/C solenoid and shorten the time from one hour to 10 seconds. If the oxygen sensor told the ECM the mixture was too rich, the metering rods would still cycle at 10 times per second, but would spend more time in the main jets (driving the mixture lean) than out of the jets. If the mixture was determined to be too lean, then a greater amount of time would be spent out of the main jets (going rich) and less time spent in the jets. Ten trips per second would still happen, but the time rich versus lean would be different.
GM provided a way to monitor the trend of the metering rods with an ignition dwell meter set on the six-cylinder scale and connected to a pigtail wire that was factory-spliced into the ground side of the M/C solenoid. Since full deflection of the dwell meter in that setting was 60 degrees, the dwell or duty-cycle activity of the metering rods was likened to that value.
An M/C dwell of 30 degrees was considered perfect: The metering rods were making 10 trips per second and spending an equal amount of time in and out of the main jets. As the M/C dwell went higher than 30 degrees, the metering rods were spending more time in the main jets (shutting off fuel) to try and lean the mixture back out to 14.7:1. Conversely, if the dwell was less than 30 degrees, the ECM was trying to richen the mixture. The dwell period was not only a function of the oxygen sensor's determination of the mixture strength, but also of internal adjustments in the carburetor and the condition of the engine.
Rochester's E2M and E4M carburetors had two travel adjustments on the metering rods. The screw that held the M/C solenoid into the float bowl casting was also the lean stop. It determined how far into the main jets the metering rods traveled when the magnetic field was energized. In the air horn was the rich stop that controlled the amount of travel out of the main jets when commanded to add fuel. In addition, a sophisticated air bleed circuit was designed to follow the M/C solenoid plunger; it would either add or subtract air. The air bleed circuit was adjustable for position. The last area of tuning was the idle mixture screws.
The E2S did not have a lean and rich stop, or an adjustable air bleed. The M/C solenoid was designed as one unit and had no internal or external calibration.
When calibrating an electronic Rochester carburetor, the rich and lean stops, along with the air bleed, were fixed gauge settings. A special tool was required to set the metering rod travel at 1.304 inches from rich to lean stop. Another tool, called a shepherd's hook, was used to set the air bleed height only after the proper metering-rod travel was set. Then the idle mixture screws were used to fine-tune the idle quality and set the M/C dwell to the proper 30 degrees at idle with the transmission in drive.
The Rochester feedback carburetor was an excellent and trouble-free design, especially if a knowledgeable mechanic worked on it. Sadly, technicians often took the easy way out when working on these units. Often, a rough/poor idle complaint, a problem requiring internal adjustments to the unit, was instead addressed by closing off the idle air bleed and choking the circuit for oxygen. Another common mistake was trying to achieve proper performance by only adjusting the idle mixture screws. In either case, poor performance and customer dissatisfaction was the result.
As emission standards became more stringent, the feedback carburetor was replaced by electronic fuel injection. Hopefully, the marvel that was the Rochester electronic carburetor will now be viewed with the respect that it deserves.
