Fuel injected: good cold start ; poor hot start

[RRoller123]

Also could be failure of the fuel pressure regulator that hangs on the fuel rail. Pressure should be in the mid 30's. If you pinch off the return line, does the pressure rise up to spec?

[Daveje]

In my case yes, pinching off the return line brings the pressure up to spec, and beyond. Is this an indication that the fuel pressure regulator
is definitely bad? I replaced it a few years ago and it doesn't have many miles on it. I was going with the fuel pump first since the one I have is a non Bosch aftermarket and definitely seems to not hold the pressure in the rail.

[RRoller123]

Sounds like that is the case, the pumps usually are capable of higher than the regulated pressure. The data supports the idea that the Pressure Regulator is bad.

[Daveje]

Hmm, OK, that's what I was afraid of but I want to change the pump anyway. I'll be putting it in this weekend and I'll post results, good or bad!

[RRoller123]

If the old pump is not noisy, keep it as a spare because it is likely fine, given the pressures that it is giving with the return pinched off. And watch the wiring of the new pump, it is ridiculously easy to wire it backwards, as I found out, even with some care. On my car, the Green wire was hot, which threw me off, so even though the two ring terminals were different sizes , I got it backwards, then for awhile I presumed that the pump was not priming.

[Daveje]

So just an update, got the new pump in and I now my fuel rail holds pressure after shut off. Hot starts are good now so I'm happy about that.
Still a low idle after hot starting but that is improved also. I have a fuel pressure regulator on order and I hope that solves that.
There are some good deals out there if you shop around. If you buy the Bosch 39133 pump be warned it does not come with the 2 different sized nuts so expect a trip to the hardware store.

[micbrody]

I just had another thought.
PO had temperature sender in coolant T coming from the upper screw outlet (not lower screw).
I assumed the tip would be in the coolant in any hole. But may be there is air "stuck" in the top of T; but still allows coolant flow.
At cold start, the sender is cold and sends a high resistance signal to ECU telling it to enrich the fuel mixture.
When car is hot, and then shut off, maybe the sender cools off faster than the coolant fluid/engine. When hot engine starts, fuel is "flooding" the start conditions because ECU is seeing a "colder" engine than in reality.

[spider2081]

PO had temperature sender in coolant T coming from the upper screw outlet (not lower screw).

I think its possible to install the coolant "T" upside down. The side with the restrictor hole in it is suppose to point to the thermostat. If it is upside down the the temp sensor would be on top and the restrictor would be pointing toward the radiator. Not sure what this would cause but the water flow through out the system would not be working as designed.

[micbrody]

Today I removed the T-coolant thermo sender. Even though it was mounted in top hole, the T was filled to top with coolant.
With the temp probe removed, I tested the resistance at different water temperatures. All resistances seemed good.

I drained some radiater fluid; inspected coolant T. It was installed correctly with restricter side on thermostat side.
I re-installed the temp probe in bottom hole; filled and burped system.

The car started ; but lots of variability at start eventually it warmed up and remained constant at RPMs about 900. When I gunned the engine, it revved fine; but when throttle released, it again dipped below 900 (to 500), before recovering to 900.

I sprayed all potential leak sites with carb cleaner; pinched AAV when car warm to see if maybe it was still open.... pinched vacuum line to booster.....,etc... etc.. etc.. Nothing changed RPM. My conclusion: AFM might be flaky.

I removed; tested the resistance compared to an extra AFM I had; definitely the resistance was different than extra AFM; more importantly: resistance would jump from 120 ohms to 1500 ohms to 200 ohms with very tiny movement of AFM door; the extra afm had variable resistance; but from 120-300 over same type of movement. I opened up cover. Tested the resistor board. It just is not conducting much in the low end. I had previously (about 1.5 yrs ago) refurbished (moved the swing arm; cleaned (de-oxyed) the resistor board ; It worked fine; but now very little conduction left on the black portion).
I installed the extra AFM. Car started immediately. Idle seemed appropriate over the warm up. I then gunned the throttle and released. RPMs came down to original idle RPM........problem solved.

As for the bad resistor board AFM: for kicks, I ordered some conductive paint ($10 on amazon). I will recoat board. If it works, then this might be away for others to fix their bad AFMs.

[RRoller123]

That thick film resistive material (TFM) on the wiper board in the AFM is carefully calibrated by its formula (basically it is a low resistance semiconductor) to provide the proper resistance through the wiper that the ECU wants to see. It is also formulated to resist excessive wear throughout tens of thousands of wipe cycles. It is very unlikely that the addition of a conductive paint will do anything but ruin the board, but you may get lucky! That would be quite a find if it works out, so please keep us up to date on what you find!

Have you tried the bending procedure to get the wiper to run on a new section of the original track? There are many posts her on the Forum on how to do this. You basically bend the wiper arm to track either inside or outside of the original position, to get it to ride on fresh TFM.

[micbrody]

I did do the bend procedure. It tracked well when I did it; resistance was good and gradual. I tested with slowly opening door; but also with test probes lightly dragging across black surface. When close to each other, very little reistance. As probes separate, a gradual increase in resistance.
That was over a year ago.

Yesterday, many parts of black has very high resistance..... When I moved AFM , I would find a "dead" spot, and then manually press harder on the copper contacts(thinking maybe my bending procedure distorted so it was not making connection). Even with pressure, no change in resistance (1500 ohms range). I then removed the swing arm; checked with ohmmeter probes. Very little conductivity in the black area. Conclusion: resistive surface is bad. Not sure why my "rehab" only lasted a year. I wonder if the "de-oxy" spray was ultimately a bad idea. Maybe it made the black stuff change to more non-conductive over time....

In any event, this discussion board/community has been so valuable, I figured I would try the re-coat. If it works, that IMHO would be an easy and inexpensive way to fix many of the AFMs

[RRoller123]

I will be the first one praying that it does work, that would be quite a breakthrough on this common problem!