My check engine light came on and the diagnostic.

Check the no.2 cylinder compression if its within tolerances

A different mechanic. Swap the number two injector with one of the other ones, then see if the code switches to that other cylinder.

Injectors should be replaced in sets with rebuilt units so they are flow-matched. GM owners are very familiar with misfires in high-mileage engines caused by mismatched injectors right from the factory. Chrysler buys their injectors in matched sets and rarely has an injector problem. Replacing just one injector on your car can introduce a new problem to add to the current one. That makes diagnosis even more difficult. Engine Computers are not commonly the cause of intermittent misfires. Do a compression test before spending the money on an unneeded computer.

A faulty injector can be properly diagnosed by a competent tech using a lab scope or performing and injector balance test so it doesn't have to be guess work. I don't agree with having to replace all of them though. The difference in injectors would be minimal and replacing them all would not guarantee that they all would come from the same batch anyway.

That's why you buy flow-matched sets. Linder Technical Services in Indianapolis puts on very expensive, high-level classes all over the country. They have solved a lot of problems by providing replacement injector sets after they have found no defects with those sent to them. They just flow different rates leaving one or two cylinders lean compared to the others. Their biggest sellers are by far for GM products and they claim many of the car owners are surprised by how smoothly the engine runs with the matched injectors. Chrysler buys their injectors from Bosch in flow-matched sets. GM grabs a handful out of a big bin. That is the word from the experts, not me. To be fair, there's a lot more GM products on the road so of course they're going to sell more injectors for GM vehicles. No mention was made that I recall about other car brands. Check out their web site and their "Dirty Dozen" most common failures.

It seems too many mechanics rely on resistance readings to determine if an injector is bad and they condemn way more than necessary. I don't have an issue with replacing just one defective injector but when you're looking for the cause of an intermittent problem, I'd look at the ignition system first and compression problems last. If one cylinder is running a little lean, you are less likely to feel that compared to a loss of spark, but the Engine Computer can pick that up. That's the most common cause of misfire codes that you can't feel.

A lab scope can show if an injector is opening at the right time but it won't show a restriction. If you're good with a scope you can also look at the oxygen sensor waveform to tell if a problem is due to ignition, fuel, or compression. Most of us will never be that good. It's one thing to watch a slide presentation in a class but quite different to try to do it on a customer's car when you're charging by the hour. Even those instructors have to fiddle around for a long time to show what they want us to see, ... BUT, in my case the instructor owns a specialty shop that only solves the one in a hundred no one else can solve. Their customers are other shops, and they have plenty of time to experiment and put together a presentation on their findings. We don't have that luxury.

For the injector balance test, are you going to cancel one at a time with a scanner? How would the results of a restricted injector be different than from low compression? Are you watching the drop in engine speed or are you doing something else?

Given that this is an intermittent problem, measuring the injector's resistance is useless unless you can do it at the instant the Check Engine light turns on. That's not practical. You COULD replace the suspect injector and hope you didn't waste $290.00, or you could just swap two of them and let the Engine Computer tell you if you're on the right track.

The computer is the last thing I would suspect. If one injector is not firing due to lack of current flow, the diagnostic fault code should reflect that. I'd suspect a corroded connector terminal first, especially for an intermittent problem. As it is now, it's just detecting the misfire, not the cause. On engines that batch-fire the injectors, two or three would quit at the same time and you would definitely feel that. Too many people, it seems, jump on the computer right away as the cure-all for problems that haven't been properly diagnosed. I'd want to rule out every other possibility before I bought a computer.

[quote]For the injector balance test, are you going to cancel one at a time with a scanner? How would the results of a restricted injector be different than from low compression? Are you watching the drop in engine speed or are you doing something else?[/Quote]

That's not how an injector balance test is done.
It's done using a fuel pressure gauge with the engine off. The fuel pressure is brought up to spec and the scan tool turns the injector on for a specific measured time and the pressure drop is recorded. This is done for each injector and any restricted injectors will show considerably less pressure drop.

Dandy. That makes more sense. My guess is that takes five or ten minutes to complete if you have the tools at hand.