Correction - hard start not fixed, leaky fuel inj?
Correction guys - my fuel pump "fix" must have been placebo effect. To answer the question - I found the check valve in the old assembly when I tore it apart for learning. It looked fine and it was somewhere near the outlet line but I'm sorry I don't remember exactly where. It was a simple ball in crimped cylinder like you'd see in a coolant thermostat. It certainly could have been leaking slightly, but it wasn't all gummed up or anything and no obvious damage.
I've continued to have hard starting issues and this weekend I found descriptions on this site and others that match my symptoms. Leaky fuel injectors! In the past couple of years I've already done coils, plugs, valve cover gaskets, main fuel pump, fuel filter and the performance after start-up across the entire speed/load map is wonderful, implying I have not compression or other chamber issues. I was slightly wondering about the cam and crank position sensors but not high on those as a problem for a couple of reasons. 1. all posts I've seen about it talk about "ok when warm" or not working at all. My hard starts are also when warm, so it's not that. 2. without crank sensor the ECU won't be able to time the fuel or spark at all. Fuel timing isn't so important for running but spark is critical, so my crank sensor must be fine. Without a cam sensor, an ECU could theoretically go waste fuel and spark (fires every 360 degrees) and then when it sees a combustion event represented in an acceleration of the crank shaft, it could infer cam position. I don't think this is likely because it would almost certainly set a code, and I'm not sure the production ECUs actually would implement such a limp-home strategy, but possible.
So, by process of elimination it's highly improbable that my spark, induction/valve/chamber or total fuel flow rate capability are a problem. Since the injectors obviously can flow fuel, and since no codes are being set, it's not like I have a non-functional injector. But if I don't have the expected pressure, then the pulse-width the ECU commands will deliver an entirely wrong quantity of fuel, the A/F will be way off the small window of combustibility and viola - hard start. Eventually the wall-wetting will accumulate enough fuel and it'll run, then the pressure control will be more active and the ECU will take over. The leak must be small when compared to normal operating flow rates, but large when compared with the flow rate required for startup - b/c it isn't running, then needs to step on.
I have been meaning to buy a fuel pressure tester to confirm all this and noticed that Harbor Freight's got one for $16 so I'll pick one up for fun. I expect to see my pressure build then seep down after the pump disengages. The seep down is likely the result of my injectors being varnished up and not seating in the off position fully, thus bleeding pressure out of the rail. I've noticed that if I try to crank immediately or if I let the pump run for a long time, stop and wait, then in both cases I have the problem. There seems to be a sweet spot in delay between key-on and cranking to avoid the hard start. That would be consistent with the pump running until the ECU sees a high pressure satisfied, then bleeding down. The ECU doesn't expect it to bleed, so may not be controlling pressure as actively as it could be. The other posts I've read fit my symptoms well and like I said, by process of elimination this is about all I'm left with.
Now I need to decide on how to fix it. I've run a total of one bottle of Tecron through the system over the entire vehicle's life of 149k, about 5k ago. I see three options 1. on-car cleaning (direct solvent to rail) 2. off-car cleaning (injectors in ultra-sonic bath) or 3. new or remanuf injectors. I'm waiting on a price from Max on new and see approx $35-40 for remans on the web. The rails look pretty easy to pull, but the solvent would be simplest - if it works.
prob not.