Today was another day spent working on 1082/2000. I finally got the car running yesterday, but it was suffering from a misfire and was leaking fuel from a couple of the injectors. My first stop this morning was Bell Road Mitsubishi to pick up a set of new seals for the injectors. I could have picked up a kit at Autozone or something, but when it comes to something that keeps your car from catching fire, I don’t mind shelling out for OEM quality and fitment. Karma came correct, however, and they set me back a mere $12.

Before I was actually standing over the disconnected fuel rail again (for the sixth or seventh time in 48 hours), I spent some time at the shop. Since we’ve removed the 550cc injectors and swapped the ECU out for one with a stock EPROM, it makes sense that we just continue to restore 1082 to stock configuration and use the go-fast bits on one of the other Galants. With that in mind, I climbed up into the mezzanine at the shop and took down a spare TD05H-14b assembly, complete with O2 housing and janky, nobody-knows-where-it-came-from J-pipe.

I also removed the front bumper from The Flying Galant to get access to the stock intercooler and remainder of plumbing not already removed. My first thought was that it was really nice being able to work on the bumper with the car on a lift, but then my arms kinda got tired and I got a face full of dried up, crusty bugs, dirt, and road gnar with each piece of trim I touched. In the end, Dave gave me a hand pulling the hulking boat anchor off the car (no badge on this particular GVR4, actually). GVR4s can be a bit intimidating without their makeup.

Stock intercooler plumbing and turbo in hand, I headed back to the house to resume my work on 1082. First order of business was to pull the fuel rail and replace the hardened lower injector seals with the supple, new ones I got at the dealership. I’ve done this enough that it’s really a piece of cake. In fact, someone remind me on Monday to take more pictures of the actual steps involved in removing the injectors so I can do a simple write-up, okay?

Once the new seals were installed, I fired the engine. SURPRISE! It still sounds like a damn Subaru. The new seals managed to reduce the fuel leakage by 90% or so (still some leaking at the base of injector #1, I’ll manage), but the engine still sounds like crap. I decided it was time to do a proper load balance test and, with the engine idling, began pulling spark plug wires.

Pulled #1 – There was spark, and the engine wanted to die out.
Pulled #2 – There was spark, and the engine wanted to die out.
Pulled #3 – There was spark, but the idle didn’t change!
Pulled #4 – There was spark, but the idle didn’t change!

Considering my world has revolved around chasing various fuel supply issues on this car lately, I turned my attention to the injector clips.

Pulled #1 – Engine wanted to die.
Pulled #2 – Engine wanted to die.
Pulled #3 – No change in idle.
Pulled #4 – No change in idle.

What are the chances the RC550s were stuck from bad gas and half this other set of stock injectors off a running (albeit poorly) Galant were duds as well? I mean, that’s like saying 150% of the injectors I had were bad? C’mon! I went back over to 464 and pulled out the factory service manual (FSM).

There’s no oscilloscope in Keith’s yard. It’s also been long enough since I used one that I would probably use it wrong and open a wormhole to some other dimension, so I did what I could with my trusty multi-meter.

First thing I did was ohm out the injectors and compared the results to the FSM specs. All this involves is disconnecting the injector clip and touching the meter leads to the two pins inside the injector. They all reported the proper resistance across their terminals almost exactly. I don’t recall the exact spec, but all four injectors were identical. No dice there.

Next, I removed the resistor pack for the injectors (documented top right in the image above, actually) and ohm’d it out according to the FSM. It’s a 5-pin connector, with pins 1 and 3 at the top (the side with only two pins, if you’re reading this to chase down your own GVR4 misfire/no-start nightmare). One lead is touched to pin 3 and the other then touches pins 1, 4, 5, and 6. The resistance should be something like 6.5ohms. In this case, the values were spot on, ruling out the resistor pack.

Then I switched the key to the ON position and checked for 12VDC at the harness. To do this, the negative meter probe is grounded and the positive probe is touched to pin 1 on each clip. All four connectors were sending voltage to the injectors, so I was able to confirm that at least half of the harness was intact.

The final step in the FSM is to ground pins on the ECU harness connector and check for continuity to the other pins on the clips, but the leads on my meter aren’t long enough to reach that (nor do I have multiple jumpers to test out by other means), so I simply set my meter to measure duty cycle and touched the probes to both pins in the connectors with the engine running on the two, known good cylinders. I was getting a consistent 10.5Hz, which suggests (to me) that the MFI circuit was sound and that the ECU was switching something somewhere to fire those two injectors.

(NOTE: If you don’t have a meter – or a stethoscope with which to listen to the injectors while the engine is running, which can help detect if one doesn’t sound like the others – another test to perform would be to swap the suspected bad injector with a known good one. For example, in my case, I would swap the #2 and #3 injectors, then start the car again. Then I would pull the connector on each in turn. If disconnecting the #2 injector had resulted in the car wanting to die before, but did nothing now, and pulling the #3 injector previously had no affect, but now caused the engine to stumble, then it would be a good way to confirm that the injector was, in fact, the problem.)

With the harness pretty much ruled out as the problem, I headed back to the shop to score another set of injectors. These are already installed in a rail, so I’ll be installing them first thing Monday morning. Ordinarily, I’d be confident that this will solve the problem, but being overly confident can lead to making stupid mistakes, so I’ll just keep my fingers crossed.