93 V6 fuel pressure release.

[sgould]

I think I'll try the ECU once it's refuelled.  I got the new battery fitted.  Now it's raining again...

[carrera]

Thinking a little more, it might be worth taking readings from the ECU before you fill up.

You'll have two data points that way

When it stops raining 

[sgould]

It's sunny now, but too late to start anything.  I may have a bit of time tomorrow, but she needs the car at some point.

One point I realised was that the CarScanner report said that the level sensor failure was "high output", which implies that there is less resistance than expected.  So the sensor must reduce the current as the level drops.  But it also implies that the sensor has a short circuit, not a break as I had imagined.

[carrera]

There's what looks ike a decent diagnostic in WIS, but it does assume you have access to the tank

Buried away I see that for EU empty is supposed to be 250 ohms, full 40

You could try applying 12v in the appropriate places to measure the voltage the ECU would see - seems to be operating on 5v due to a 487 ohm resistor though. An idea anyway

[sgould]

OK, so that points to a break in the sensor resistance at high level.  The resistance for a break being infinity, it would give a reading of empty.

[sgould]

Nothing today.  It's cold and wet.  We had a frost overnight.

[aerojon]

9/10 the P0089 is caused by the pump.

[sgould]

That's the consensus on the internet, but I've run a long record of pump pressure and it hasn't dropped at all.  There's no suggestion of hesitation at full throttle, so when it first happened, I changed the sensor and it seemed to solve the problem.  Six months later it was back and I heard a theory that the filter allowed very fine particles through which, over time, sat in the far end of the fuel rail.  It was suggested that I filled the fuel rail with fuel treatment, which i did and the light went off again for almost a year  Now it's back.

I filled the fuel rail again with the intention of leaving it standing overnight, but the failed battery meant that the treatment stayed in for 36 hours.  It's started and I've run it a few miles and the light is still off.  I felt the engine was running a bit smoother, but it was maybe my imagination.  The SWMBO took the car out and when she came home, said "What did you do to my car? It's running a lot smoother than is was".  Maybe the fuel treatment has cleared an injector, or maybe the new battery is helping.  She's out on a much longer drive this evening.  We shall see if the light has stayed off.

I'm hoping it stays dry enough to check the level sensors tomorrow, if the rain stays away, but it isn't looking promising.

[sgould]

Oh well!  She's home and the engine light is back on...

If it's coming on more regularly, I'll clear it, run with the pressure recorder and see if I can pick up the instant it triggers.

[sgould]

Oh! ****

In a gap in the rain I went out, cleared the code, set the Car Scanner to show the fuel pressure.  Went out and retraced the route she did last night, then came home via the motorway and the supermarket.  The light stayed off.  So no data to compare.

[carrera]

Did you pick up from WIS that the P00089 code is also triggered by running out of fuel, or at least the car thinking you have run out of fuel...

[sgould]

No, I didn't.   I'll double check the circuitry when it's warm enough for my fingers to grip the multimeter probes...

I have been thinking that the P0089 fuel pressure fault and the level sensor fault may be linked,  but I think that will depend on the fault being in the sensor attached to the pump, rather than the separate sensor in the other half of the tank.

I was also pondering if there's a pressure drop or spike when the pump starts up.  From my hydraulics background, I wondered that if there's a small blockage in the feed.  If it stops the flow briefly the pressure will be lower than expected at the sensor and could trigger a fault.  If the blockage releases suddenly, the pressure would surge and also trigger a fault.  That sort of blockage is typical of a sticking non-return valve.   Further reading required...

[carrera]

Can you put the pump into a permanent live situation ?

That way you can circulate your treated fuel without actually consuming any, and perhaps even graph the fuel pressure at the same time.

I would have the charger onto the battery, to maintain state & not drag your new one down...

[sgould]

The V6 fuel system does not have a return pipe through a pressure relief valve.  The pump is electronically controlled and varies the fuel pressure under the control of the ECU.

I have cans of Forte engine and injector cleaner coming tomorrow. 

The P0089 fault in WIS does mention it being triggered by no fuel.  It also mentions a blockage in the fuel line.  I'm also thinking about a loose electrical connection.  The level sensor may have a a fault but separate to that, if one of the level sensor signals to the ECU is broken for a moment or two, the ECU may think the tank is empty and trigger P0089, it won't show on the gauge because that has a damped response.  Perhaps I can read the instant fuel level with Tech2...

She went out to a singing engagement tonight.  No engine light on the way there, but it came on as she started to drive home. 

[carrera]

Ah, fair point, it's the first to move to move away from traditional FI type design to the newer direct injection, like the NG 9-5's

You're right about the damped fuel gauge response, you can see that by filling up a car with fuel while its idling because you have grave concerns about the battery or something else. Seen that.

Question is, does Tech 2 give access to both sensors, or is it the combined damped output response. One way to find out.