UPDATE 22

Injectors: For lack of the test harness, used the paper clip method to get to the signals on connector B-10 at the firewall. It was really hard to get them in the back of the connector and I wasn't quite sure they were making proper contact. Got scope traces for what I think were injectors 1, 2, 3 and 4. The results were not uniform. The first 2 traces are injectors 1 and 2, the second set is injectors 3 and 4.





Next I tried making a test harness out of old individual wires from an RS-232 set of connectors. This time it was really hard to position the wires and keep them from disconnecting. The following 4 traces are what I believe were injectors 1 and 2, and 3 and 4. Again, completely inconsistent traces.






So, this is not normal, presumably, right? Does anyone know what traces are correct and which ones are not? What I've read is that 12 V is normal, but I get 6 V and more than 20. Does this indicate bad injectors? or a bad computer that is sending bad control signals. The static test (resistance) indicates that they are all ok: 1 - 15.75, 2 - 15.63, 3 - 15.73, 4 - 15.66, 5 - 15.64, 6 - 15.56, (Ohms)

Next I'll try with a set of wires with alligator clips. This will make for better and more reliable connections. I'll report back when this is done.

 

Everything you are describing sounds like a vacuum leak. There looks to be a vacuum line coming off the fuel regulator on the rail. If it's running lean, that could be where the leak is.

Easy way to check for a vacuum leak is while vehicle is running, spray some brake cleaner or throttle cleaner around the lines. When you hear rpms increase, you have found your leak.
Check all over that engine compartment, even if you don't think you have any vacuum lines in the area. Also worth checking to see if you have a pcv or ccv valve that can be replaced, should be low cost and in reality that would be a maintenance item anyway.

 

I tested all the hoses at the top of the engine with throttle cleaner (see UPDATES 7 and 21 above) but didn't see any changes in RPMs. At the time (UPDATE 21) I also measured the vacuum: 10 in Hg. Is this what it should be? I just tested again, spraying all over the top of the engine, even where there were no vacuum lines. No change in RPMs.

I don't have the EGR system. At the time this was only required for vehicles sold in California. The only mention I see in the manual to a PCV valve is a "Connection for PCV hose" but the drawing doesn't match what I see under the hood. I don't see anything that could be a PCV valve.

I suppose the relevant question re: vacuum is: what should it be under normal working conditions? and is the 10 in Hg I see normal?

Thanks for replying.

 

From google AI search:

For a 1990 Mitsubishi Montero with a gasoline engine, the target idle vacuum pressure is generally between 17 and 22 inches of mercury.

Checking online, I was able to find a manual for 1992 Montero and found the following:

 

Thanks. Just a few days ago, it occurred to me to ask the AI and I got the same numbers. Posing the question a slightly different way returned 18-22. Your 1992 manual page confirms it. I have a vacuum leak that I can't find. I'm taking it to the local mechanic next week. Hopefully he'll find it.

 

UPDATE 23

I found the leak! It's the brake booster. I unplugged the vacuum line that goes to it and put the vacuum gauge on it. Solid 20-22 in Hg. It doesn't change much when accelerating. Unfortunately, it didn't solve the low idle and stalling when hot or the misfires. Not sure if the computer will eventually adjust after driving around a bit, but I'll have to get the booster replaced first.

Is it conceivable that it could take some time for the computer to adjust after driving around and then the low idle, misfires and horrible gas mileage will all be solved? I have replaced every single part in the ignition system: plugs, wires, ignition control module, coil, distributor cap and rotor.

The signals to all the injectors are consistent. The following traces are for injectors 1 and 2, 3 and 4, and 5 and 6. All were done with the probes set at 10x, so the real voltages spike at about 60. Is this normal?

 

Yes, computer has to "re-learn" and remap the fuel mixture for the new no vacuum leak condition. It shouldn't take that long. Eliminating the vacuum leak will help with fuel mileage. However, I'm not sure if it will relolve the misfires and stalling. Only time will tell

 

UPDATE 24

Replaced the brake booster and the vacuum is just as bad as with the old one (about 10 in Hg. If I disconnect the brake booster hose and measure the vacuum there, it's 20, the supposedly normal vacuum. This leaves 3 possibilities:

1 - this loss of vacuum is normal for a well functioning brake booster, which means I wasted $300 replacing a perfectly good one, or

2 - the replacement booster is also bad, which means I have to do it all over again after returning it and losing the labor to replace it (the local mechanic did the replacing), or

3 - the local mechanic didn't install it properly or didn't adjust something right. If this is the case I can take it back to him and complain.

So, which one is it? What is the proper behavior of a brake booster? is is supposed to maintain 20 in Hg vacuum?

Of course, the original behavior is as bad as ever, maybe even worse as the vacuum is more erratic now: low idle, misfires and stalling when warm.

What can I do now? replace the damn thing again? Did I get a broken replacement?

Thanks for any further ideas.

 

Under normal idle conditions the vacuum pressure should not drop that low. Try replacing the hose and see if this resolves your issue. I know it may be odd, but doing this is much cheaper/easier than having the booster replaced again if that's all it was to begin with. (My apologies if you already replaced the hose).

 

Re-reading UPDATE 24 above I realize it wasn't clear that I disconnected the brake booster hose at the brake booster and placed the vacuum gauge there. So the measured pressure (20 in Hg) includes the brake booster hose. The leak has to be inside the brake booster. I wonder if there's some way that the old one that is now out could be tested, opened and fixed without sending it out to be rebuilt. Thanks for taking the time to reply.