inconsistent starting

[micbrody]

I am close to my wits end!!! Every time I think I have figured out why I have intermittent starting problems, I discover that I have not fixed the problem. When the car does start, it runs great! So I have focused on the starting sequence.
Things that I have checked and/or adjusted:
1) CSI and Thermo switch that's in series with it. I have twice removed and cranked to see if it fires when cold. It does; and shuts off after a few seconds (depending on temp of coolant)
2)I have checked coolant T sensor. It's internal resistance is appropriate at hot vs cold; and the resistance from connector to ground and ECU plug is zero.
3) I had a small leak in brake booster. I changed it. Brakes slightly better; idle is now rock steady (idle would vary when vacuum would pop off at 15cm h20 pressure)
4) adjusted idle using procedure with idle adjustment screw, along with checking throttle position sensor; also checked resistance of plug from TPS to ECU, all zero ohms. When started and warm, car idles around 900 and steady
5) checked coil resistances; both the low and high resistance sides. Changed old plugs and wires. Changed magnetic pickup in distributor. Last year, replaced flaky ICM
6) checked grounds to everything. Checked the grounds going to ECU 5,16,17; the intake manifold grounding. the exhaust manifold ground. I added additional heavy gauge ground to exhaust manifold
7) I have (three times!!!) checked the double relay. Everything (on bench testing )works perfectly. Relay easily fires with a small 9 volt battery; and there is continuity without any resistance when appropriate... in every way I can test it. I also checked the resistance from the plug that goes to ignition switch. I also replaced ignition switch. I also checked resistance of plug wires to ECU plug. I have deoxyed every connection I can think of.
8 )I have removed AFM; adjusted copper swing arm so it is on a new part of resistor board. Everything works as it should. (Very interesting integrated circuit resistor board; the engineers actually measure two resistances on board (one from each side of swing arm; this way it can cancel out any variability (caused by oxidation, age, humidity,etc...) in conduction resistance of copper swing arm contact to board.) And fuel pump switch definitely works
9)twice checked AAV. opens and closes appropriately with temperature

The only system I think I have not fully explored is the fuel delivery system. When car runs, it runs great; so I would think that the fuel pump and pressure regulator is fine. However, in my readings on the subject, I can theorize a few situations that could explain my starting issue.....though really not sure.......so any knowledgeable board member help would be greatly appreciated:

My last theory of what's going on: I believe when car is off, there should be residual pressure in fuel rail. (true or false?) So when I immediately start car and injector opens, fuel is immediately delivered to intake of cylinder. Maybe I have a slow leak in : 1) injector or 2) fuel pressure regulator ....allowing fuel to flow back to gas tank or 3) internal check valve of fuel pump.

If any of this were to happen (and they were slow leaks), could this explain why when sitting overnight, it take 3-15 seconds to start car (needed time to repressurize fuel rail? and/or uneven lean/rich fuel mixtures in individual cylinders if injector leaking ); yet car starts right away after 5 minutes to 3 hours of car being off.

I guess I need to check fuel pressure in fuel rail, etc..... Does anyone know if advanced auto or auto zone loans out fuel pressure gauges?

[124JOE]

yes i think its the fuel press reg
if its a slow leek then try "when cold"turn the key on for five seconds then turn it over
still hard to start then its not a slow leek

#8 if you put the 8 just before the ) then thats the symble for cool.so put a space after the 8 or a#

hope this helps.joe

[micbrody]

Thanks for quick reply

Question: your statement "if its a slow leek then try "when cold"turn the key on for five seconds". Do you mean "start" or "on". If "on" but not cranking, why would anything change in fuel delivery system. The fuel pump is not energized in "on" position unless car already started

[124JOE]

yes just to the on position
i thought the fuel pump would run up the pressure then stop

[micbrody]

The fuel pump is not suppose to run with engine off but ignition in "on". Via the double relay and AFM contact switch, when the car is running and ignition in 'on', the fuel pump works. As a safety feature, if you are in an accident where engine stops, then the contact in AFM will shut off (no air intake moving flap), and that would shut off fuel pump. This way, pump won't be spilling gas in engine compartment if bad accident, and occupant (unconscious or disabled) is unable to turn off ignition switch.

My car initially was rewired and bypassed in a way that fuel pump was always running in "on" position. I cleaned up ignition contacts and double relay, and then have reversed it.

[RRoller123]

Mine is wired that way too; as an alternative to rewiring it, I put in a inertial shutoff switch in line with the pump feed wire. But rewiring it as stock would be the overall preferred method.

[micbrody]

Last time I started car was about 18 hrs ago.
Today I slowly (and with towels) disconnected the fuel line to CSI. There was no pressure in line.

I removed fuel rail with injectors still attached. I reconnected electrical; stuck a screw driver in AFM flap; and turned ignition on to "run". Fuel pump energized. I saw no gross leaks; and only sound I heard was turbulent noise of fuel returning from pressure regulator to line going back to tank.

Injector tips were damp. I wiped them dry. Within a few minutes, they were glistening again. I repeated process a few times (wiping injectors dry and turning on fuel pump for 20 seconds). Every time the tips got wet in a minute or two; I waited 30/60/90 minutes. No visible drops formed; but that could be because of evaporation. (CSI was dry each time)

The injector hoses look original.

2 Question:

1)Should injectors be totally sealed when pressurized? I think they should be

2)Should pressure remain in fuel rail after 18 hours?

Plan:
1) about a year ago I bought fuel injection replacement line kit. ( the short injector lines are shorter than what's on injectors now. Is that normal?) I will replace all the old lines in engine compartment.

2) clean injectors. Do you think the cleaning will solve the weeping problem? I read through some ways of cleaning them. Anyone have a preference? Any 3rd party injectors that are reasonably priced that people have had luck?

3) borrow a fuel pressure gauge from Autozone, so I can check on all other aspects of fuel lines/regulators

[RRoller123]

I am no expert, but I think the answers to your 2 questions are #1 yes, #2 no. 18 hours is a lot of time for the pressure to bleed off back to the tank. I sent my injectors (including the CSI but not the rail; I pressure tested it and it was fine) to OK Injectors and they tested and serviced them (internal and external seals, clean, test, provide flow data). What a difference! Money well spent. The test data showed enormous improvement. The dripping injectors may be causing an over rich situation contributing to the inconsistent starting?

[micbrody]

Thanks for quick reply.
I am HOPING that leaky injectors can explain the chaotic nature of the starts.
For instance, when I initially got car, it would sometimes start easier when cold than when warm; there were many other reasons for that; but maybe the artificially rich mixture from car sitting overnight during cold weather helped it start; but if overnight was warm, then it was too rich;

In any event, I definitely need to do something about injectors. Like most of this car, I like the challenge of trying to fix things myself , so I will try some techniques I saw on Internet for cleaning injectors. If the "weeping of fuel" stops, then I will just re-install; if not, then off to a injector reconditioning service .
How much do your injector service cost?

[RRoller123]

It was a couple of years ago, but I seem to remember something like 25 bucks an injector?

[micbrody]

Did you notice a difference in performance after conditioned injectors re-installed?

[RRoller123]

Oh yes, absolutely! Ran MUCH better after injector service. The car was a complete running mess when I got it. It is now only a partial running mess....

[micbrody]

update on fuel diagnostics:

1) this morning, no pressure in fuel rail

2) went to autozone and borrowed fuel pressure gauge. I inserted at CSI fuel line. energized fuel pump. Gauge reads 38 and constant. repeated a few times; 38 each time. I hooked up mityvac to regulator. when I pumped it went down a few dashes... (36).

3) re-energized pump; turned off. Pressure held steady at 38. I then (with 9 volt battery, hooked up leads to each injector and pulsed. They all work; no easy way to quantitatively assess how well they flowed.

Plan:
I will quantify how fast system loses pressure (pump to 38; turn off; and check gauge every 10-15 minutes

I will disconnect and clean injectors. While in the neighborhood, I will replace all the fuel lines.

[micbrody]

cleaned injectors
reinstalled all new FI fuel lines

started fuel pump:
1) sounds fine
2) able to pulse injectors and get spray
3) CSI injector not spraying when tested
4) Fuel pressure gauge started at 10 but now stuck on zero



??????????????????????????????????????????????????

I re-energized pump and pinched the fuel return. I could hear fuel pump strain; but with outflow restricted, the CSI worked when pulsed. At this same time, the pressure gauge still stuck on zero.

I "think" I have two coincidental things going on

1) vapor lock in rail system/pressure regulator, not allowing pressure to rise
2) pressure gauge is not working

SO: how do I proceed?

Is there a trick to get vapor lock out?
or should I just re-install injectors to engine, and try to run. Maybe after a few minutes the vapor lock will get dissipated via the injectors

[RRoller123]

I would expect that the air in the rail would be pushed back intto the tank very quickly, and then it shouldn't take long for the injectors to pass the air in them. ?