The Saga of The Fuel Injectors...
Note: This is a story about basic troubleshooting and the dangers of f****** around with non-stock components in experimental ways. If my results, methods, tips and conclusions seems like common knowledge to you, remember this: there was a time in your life when you, too, didn't already know it all. Also, please use some common sense with some of the more flammable techniques described thereafter...don't set yourself or your Jeep on fire.
A few days ago, I began an experimental test to see whether or not a set of four-hole Ford fuel injectors would function correctly in the TJ 4.0, and if they would yield any sort of performance benefits. The plan was to swap them in at the same time I changed tire sizes, so that I could get a baseline reading with the new injectors. After a few tanks of gas I would swap my cleaned-up stock injectors back in and see what sort of change in performance I got.
From the start, we had issues. The Ford injectors were very hard to get into place in the fuel rail...we struggled and struggled and struggled with every aspect of swapping them. I'd like to say that there was some sort of nefarious cause at work, but honestly: it was just one of those days where nothing wanted to cooperate. Even the fuel rail was being difficult, and it's RIGHT THERE. It's right there! It's not even hard to get to!
Actually, you know what probably caused it? Greta being mad at me because I never took off that stupid K&N sticker that the previous owner put on there because he thought he did something smart. That has to be it. Ugh...party foul on my part. I gotta fix that.
In all seriousness I assumed the problems were due to stiffer O-rings but I found out a few days later that I was wrong: the O-rings were actually being distorted a bit due to a slight difference in the size of the ring grooves on the Ford injectors, and even with silicon grease on them they didn't want to seat properly in the fuel rail. I thought I had them solidly in-place when I first installed them but upon pressurizing the fuel rail I found that the #4 injector was leaking...in fact, it was spraying fuel directly onto the manifolds. This was an extremely
dangerous situation so we quickly shut everything off and got ready for a fire; I had the extinguisher pulled from its carrier in the front seat. However, the gas just boiled on the top of the engine and dissipated.
With the danger over we pulled the rail, changed out the O-ring in #4, and got it all buttoned up and the computer re-set. Then I had the same problem with #5. Same extinguisher drill as before, only slightly more rehearsed. To summarize: we pulled and re-installed everything about three times over a hot engine. I'd say we screwed around with it for at least two or three hours. This was my first clue that it wasn't a modification that was going to work, but I wanted to give it a fair shot.
When we finally got it all sealed up everything ran fine until about four days later, when I got the dreaded P0171 code.
Why is P0171 so dreaded? Because it can be immensely hard to track down and fix, at times. It results from an overly-lean condition being detected in the exhaust on the upstream O2 sensor in Bank 1. So...one of the front three cylinders was malfunctioning in terms of air/fuel. Nothing was detectable in the engine actually running
so there was no reason to panic over it. Serious problems usually have serious symptoms; you can usually tell when something is going horridly awry.
Even though I had a suspicion that the fuel injectors were the culprit, I wanted to do some systematic troubleshooting. With that in mind, P0171 comes from three things, usually:
- A dirty MAF sensor.
- A vacuum leak.
- A faulty O2 sensor.
Well, I don't have a MAF sensor, so that's out. Vacuum is harder to detect...it could come from cracked lines, cracked fittings, or - gods forbid - a cracked gasket or manifold. However, it's easy to search out:
Valuable Information: Starting Fluid is CHEAP, and when it gets into an engine that's running it will cause a momentary hesitation and then a surge in idle. To find a vaccum leak, introduce starting fluid systematically to different areas of a cold engine...using it on a hot engine could cause a fire in some cases, so don't do it. Spray it in your intake at first to see the change in idle that you're looking for, and then go over anything that looks even partially vacuum-related. Don't just hit the lines...soak fittings, canisters, sensors, manifolds, gaskets...soak EVERYTHING. If you go over the engine systematically and slowly, any sort of vacuum leak can usually be found. If you use less than about three cans, you've not used enough. If the engine gets hot while doing this, then stop, let it cool, and start again. Always keep a fire extinguisher on hand. I've never seen starting fluid light from hitting a hot surface, but it's still flammable. Use this technique at your own risk!
So I soaked the entire top half of the engine, and didn't get any results. That left only one normal culprit: the O2 sensors.
Valuable Information: When you're tracking down a code that has something to do with an O2 sensor and the sensor itself is suspect, you can try swapping the positions of the sensors themselves. In my case, I can swap either the uppers or the lowers, front to rear. If you do this and get the same code showing up with a different sensor in place, you know that you actually have an issue and that the sensors are good. If you get a different code that seems to follow the sensor, then your sensor is bad. It's a free diagnostic...and saving money is always good in the wasteland.
But wait, there's more!
Valuable Information: O2 sensors are surprisingly easy to get out. You don't need an O2 sensor socket to do it. All you really need is a 22mm wrench and some Liquid Wrench. Again, we start with a cold engine, and we spray penetrating oil on the sensor connections. If the spray begins to smoke then the exhaust is still hot: wait a few minutes and try again. When it hits and doesn't immediately smoke, you're good to go. Soak the sensor connections and then fire up the engine. Let it run for a few minutes and then shut it down: if you start seeing smoke coming from the sensor connections, immediately shut down the engine and wait until everything cools...the smoke is the oil beginning to burn off. Repeat this pattern a few times...the heat and cooling cycles will help the penetrating oil work into the threads of the sensors and then they'll come right out. Heavy rust may take more work. Again, use this method at your own risk, and keep the extinguisher handy. You DO have a fire extinguisher in your rig, don't you?
After the sensors are out they can change places with each other. However, in my case there's a slight issue...namely, the sensor plugs themselves are different. Here's the female part of the front sensor connection:
And here's the non-matching rear:
Those plugs are identical, except for the position of that stupid little alignment rail on the left side...it's dropped down on the rear plug. The inner connector is exactly the same, and it's only the outer housing that prevents swapping the sensors front-to-rear. For the life of me, I don't know why they're different. As an example: 2004 O2 sensors are the same part number, if I recall correctly...so there's no reason - electronically - that they can't be swapped front to rear for the free diagnostic in that year or any other. In this case it's a purely physical appendage that doesn't have any real reason for existing...sort of like the human appendix.
This is irritating as hell, but the good news is that we're not out of luck...all it takes is a slight modification to the male portions of the plugs, and they can be moved one to the other. To modify the male portion of the plug you simply file off the offending alignment rail, as indicated thusly.
And then you snap them together! Don't think it'll work and connect properly? Fine, here's a shot of the rear
O2 sensor plugged into the front
O2 sensor connection, and working just fine.
Keep in mind that with an O2 sensor connection, each side is a double connector...the female side is female on the inner connector and the outer shell, and the male is male in both portions respectively. What does this mean? Simple: even after filing off one guide rail, there are still two additional guide rails and an entire housing left to properly align, secure and weather-protect the connection. There's no danger in making this modification if you want to check sensor function.
After the sensors were swapped I cleared the computer, fired Greta back up, and waited about five minutes. Sure enough, P0171 showed back up...the O2 sensors were thereby perfectly fine, which meant that the injectors were the last thing that could possibly be causing a lean condition. I moved the O2 sensors back to their proper locations and removed the fuel rail for what seemed like the eightieth time. Several of the O-rings stuck in place, and had to be gently removed. What I found upon removing those stuck rings is a seriously disturbing image if you know what you're looking at:
Those are torn and distorted O-rings that were still holding pressure, somehow. Look to the center of the image...on the left, an O-ring with a large chunk missing from the surface. On the right, the telltale signs of a ring that was pinched. Neither of these caused an issue while installed, but at some point these rings would likely have failed catastrophically.
What caused the damage? As I said, the diameter of the Ford injector O-ring grooves and the placement of the rings on them was slightly different...but it was different enough that it caused the injectors to be ultra-difficult to install and remove, which in turn caused the crushing and tearing of the rings. When I replaced the stock Jeep injectors and O-rings they literally popped into place with no complaint. I was amazed at the effect of such a minute difference in sizing. When Greta was fired up with the stock injectors back in place, P0171 was nowhere to be found. No leaks, no hassle, no struggle.
Moral of the Story:
1) Sometimes, the listed causes of a code aren't the things that are responsible. Use your head. In this case, I knew what was likely the problem and a few minutes of work made sure of it. This saved me from spending extra time and money chasing my tail.
2) O2 sensors can be as easy as anything else to remove, if you prep them properly.
3) When you're doing anything experimental with an engine, USE EXTREME CAUTION. I glossed over the gas leak earlier, but the truth of it was that somewhere around a quarter cup of gas was shot over the top of a still-hot engine in the time that the rail was pressurized. That much fuel is enough to cause a serious fire, which could destroy your Jeep and/or cause someone serious injury or death. This happened because I wasn't paying enough attention. Learn from this mistake.