gptoyz

The Skinny:
So let me preface, this is one of many possible fixes for an unstable, rough or stumbling idle.

If your TPS is ok....
If your Air Idle Control Valve is functioning properly....
If you have already whacked your MAP...
If your crank isn't walking....
If your spark plugs are operating nominally...
If your O2 sensors are operating nominally....
If you have a stock flywheel...
If you don't have any vacuum leaks...
If your ECU doesn't have any issues....

Then you may have leaky injector syndrome!!!

Background:
Here's a video I originally made when we were first experienced the idling issue:

Modmonkey's 03 AP1 has been having an idling issue stemming from about 2013. After warmup, the idle would drop below the Honda spec of 800+/-50 rpm. If you revved the engine, the idle would drop, possibly stumble before catching itself, other times the car would just stall out. Coming to a violent or abrupt stop would result in the car stalling out. The worst part, EVERYONE could see and hear that there was an issue but no CEL. It is the issue many AP1 owners had been starting to exhibit more and more frequently and of course we turned to S2Ki for answers. Our main suspects were IAC Valve as that seemed to be the most prevalent fix. We tried the typical home remedies including:
- cleaning the IAC
- snake oil AKA fuel injector cleaner
- seafoam in the intake manifold
- ECU/Idle relearn procedure
However, none of the DIY solutions had an effect on the issue.

We enlisted the help of a local s2000 friendly shop. They tinkered, poked and prodded and suggested various solutions:
- new IAC Valve
- AP1 ECU swap
- crankwalk
- insufficient ground
- get aftermarket ECU and remap (did not even contemplate this approach!!!)
Finally, they settled on a bandaid which was to bump the mechanical Idle speed screw on the AP1 throttle body so that the vehicle wouldn't have a chance to stall out. Wow this alleviated the problem, and I temper somewhat, it was annoying to have an S2000 idle @ 12xx rpm when the spec is less and it also obviously increased fuel consumption.

A friend of his suggested maybe the fuel injectors, mod monkey wasn't convinced of the science but he procured 4 bench tested second hand stock fuel injectors. Unconvinced of the efficacy, the injectors lay dormant (nearly 2 years) in my tool box until I could justify why this would be a solution. So keep in mind a few things.
#1 we are hobbyists, so most of cars are in relatively stock form with minor bolt ons.
#2 We all had day jobs, so tinkering with this problem was becoming an after thought.
#3 Short of an OBD2 reader, we had limited visibility as to what was going on in the car
#4 Test equipment was limited to multimeter and other basic electronic test equipment

Rationalizing the injectors:
While Mod Monkey's AP1 is an 03, it had a paltry 80k miles on the clock and had been maintained in an excellent state, all things considered, but it was, nonetheless, still about 14 years old. Anything original from the factory, was at least that old. Most of us think of injectors from the performance aspect, that is @ 100% duty cycle can it keep up with fueling demands? In a performance application, fuel injectors would probably be one of the first mods when fueling demands aren't satiated as it is low hanging fruit. However, we weren't having any of the performance related to fueling issues: no flat spots, no power issues on track. Our issue was germane to off throttle, idle speed and braking, completely the opposite of performance demands. My epiphany came one day observing a leaky garden hose. What if the fuel injectors are leaking at low duty cycle demands? My brain began working though the logic:
What happens if the injectors are leaky @ low demands?
Then too much fuel. The ecu would demand X and the injectors would be delivering 1xx% of X.
What would happen if there's too much fuel?
The fueling would be too rich.
Why wouldn't the ecu be able to compensate for the rich condition?
It can't get enough air to compensate.
The throttle body is still a cable, so nothing except the accelerator pedal and the IAC valve can get more air into the combustion chamber.
The air requirements must be beyond the IAC valves capcacity.
...or...
Because it can't shut the injectors down more than 0% or 0 volts or whatever "off" state is.

But how did the injector go bad? Mod monkey has always used top grade fuel, had been on top of his maintenance and his AP1 had relatively low miles for it's vintage. I looked at the anatomy fuel injectors and the best theory I could justify is that with age and usage, the spring inside the fuel injector was probably weak and probably can't keep the injector shut, firmly.

I started racking my head to justify my theory and realized that my experience with carbureted motorcycle engines, a history I had buried and suppressed at the onset of fuel injected bikes, bubbled to the surface. This was no different than a bad jetting/carb setting or when some a-hole would twist the throttle on your parked bike and flood your carbs. When we first encountered the obvious problem, the spark plugs were completely fouled and the primary o2 was also fouled. Leaky injectors causing a rich running state was now fitting the bill of evidence and I was about 90% sure installing the newer fuel injectors would alleviate the problem. Unfortunately, implementing my theory was going to have to wait as it's not my car.

Success!!!:
Fast forward to May 2017, I finally get my first crack and the chance to tear into the AP1's fuel rail while we were addressing an air pump issue. I have limited experience messing with injectors, I've maybe tinkered with injector replacement 3x in my life, for most of my demands, stock fuel injectors have been sufficient. However, upon successfully removing the AP1's stock injectors, there was significant fouling and pooling brownish fuel around where the cushions seat on the engine side, still, other than that - no obvious signs that would indicate to our untrained eyes that the injectors were faulty or obviously defective compared to the replacement ones. We fired up the car and okay, so far so good, no engine fire, car seems to be idling normally high. Blipping the throttle, yielded good throttle response with no weird hesitation or stumbling. Ok, time to adjust the idle set screw back to a more normal position and launch our OBD2 diagnostic app.
Moment of truth...car started up....yay!!!
Idle is at about 8xx....awesome!!!
Blip throttle....crisp response....SWEEETTTT!!!!

We drove around a bit trying to induce the scenarios that would exacerbate the idling issue and so far, car was behaving like a normal s2000. 3/4 of a presidency, countless hours reading, google fu leveled up...the sweet taste of success @ 1:30 am!!!

Post mortem:
Ultimately, I would like to satiate my curiosity by sending in his original injectors and having them bench tested for flow so that I can unequivocally attribute fault to the injectors. As of the moment, I'm satisfied to have this off of my plate. If I have any volunteers that would like to test them or knows someone who can, let me know. I'm guessing, if we had a wide band o2 and an AF gauge installed, it would probably be glaringly obvious that the car was running way too rich @ idle, but I stress putting in unnecessary mods is not in our nature. If it was a dedicated race car or force inducted, totally different story.

LeonV

Great job! FWIW, you should be able to see a rich condition on a good OBD scanner that has a live data option. Fuel trims will be heavily biased (-25% or whatever the limit on the ECU is for max fuel trim) in an attempt to get AFR back to stoichiometry.

gptoyz

So the best tool I had at my disposal was an 04 ap2. I did an A to B comparison of the LTFT And STFT with an ELM327 obd2 scanners and iPhones.

Using the ap2 as the baseline, nothing about the ap1s measurable parameters were completely out of whack. That's what made diagnosing so difficult.

LeonV

Were they not heavily biased one way or the other? Typically in a good running engine, STFT will be a few percent one way or another and LTFT will be under 5%.

gptoyz

It's been over a year since I had done that AP1 AP2 comparison of LTFT/STFT. However, I was reading that trims under 10% and total trim under 10% were deemed "acceptable". At least that's what the internets had said. I recall, the values not setting off any alarms in my brain but I can't recall the exact value.

I wish i had known > 5% was the demarcation for worry.

vortamock

Congrats on getting it fixed!

HarryD

Nice writeup.

Please do get his injectors sent off to be bench/flow-tested (and then cleaned/replace o-rings) -- should cost $75-125 for the set and will 100% confirm if they were indeed leaking or not?

gptoyz

http://fuelinjectorclinic.com/flow-testing

Thinking of sending it to this place!

Anybody had any experience?

reyes2k

Having the exact same issue, where did you buy the injectors you used.

gptoyz

I believe my friend purchased them from ballade, so that would be:
https://www.balladesports.com/produc...0-injector-set

Definitely make sure you also get an OBD2 scanner an APP that can get you visibility into what's going on with the Long Term and Short term Fuel trim values, too