Fuel gauge issue (was...Anyone know of a circuit diagram for the MIU?)

[phoenix]

All Saab's fuel sensors are resistive up to the point they get to the ECU. Some pumps use PWM to control the flow rate though.

As you say, there's two sensors for the AWD version but it's not clear how the Motronic ECU handles the information. However, given that the sensors are high impedance when empty, I would speculate that one of them has an intermittent open circuit.

[carrera]

Or perhaps sticking in the low position, whether through friction on the pivot, physical obstruction eg a loose baffle or simple elctrical non-connection

[sgould]

I think that there's a fault in the level sensor. it doesn't appear to be a random loose connection.  It's done three full tanks since the fault was noticed.  Each time the gauge reads empty (not completely dead) after filling to full.  It then stays on empty until about 60 miles have been driven, when it wakes up and reads around ¾ full.

[carrera]

Sounds very similar to the 9000 I rescued. In that case it was effectively fuel varnish on the pivot point. Your car is much newer, but perhaps the low mileage of a couple of years ago has contributed here?

Comparative resistance readings may be a useful indicator here ?

[sgould]

I have finally braved the cold weather with my arthritic hands and connected the Tech2.  It's showing a few faults, but only one for the fuel level sensor.  P0463 which is a short in the wiring. Since the sensor is a variable resistance item, I assume that the "wiring" will include the variable resistor.

However, the Tech2 info is lacking.  It only shows one level reading, so I cannot find out which sensor is faulty.   That's annoying as the two sensors are wired to the ECU separately. 

When it gets warm enough to safely use my fingers outdoors, I will have a closer look, but I think it will be the level sensor itself rather than a loose connection, as the level reading returns to the gauge at the same point each time.  Dropping the tank will be too big a job for me these days, so it will probably be off to the garage.

[sgould]

Investigating parts availability...

It looks as if the second fuel level sensor is unobtanium.  I can't find one listed in  the UK/EU.  Maybe the part number has changed.   If it is in need of a new sender, I hope the O ring seal stays intact.  A new one is available from Saab at £48

Saabits do an alternative pump with sender for £208, and list the official Saab "OE" version, if available, at £638 - another from me.

[carrera]

How about this ?

https://www.esaabparts.com/saab/parts/12847003

Bill has it listed as special order... I just had a delivery of special order stuff, so may be available on this part number

If it is the sender, it's probably repairable, but then you'll be paying labour rates, rather than new part prices, so not necessarily cheaper

[sgould]

I’ll have another look.  The parts catalogue I have seems to be increasingly out of date. 

There’s a fair bit of investigation to do yet. And I need to wait for a warm day.

[carrera]

I was tipped off that esaabparts has an up to date catalogue, so updated part numbers are available from there 

[phoenix]

Bit of an assumption here. If what the ECU is doing is taking the lowest value and using hat for the fuel gauge, if you were to substitute a resistor of the same value as full on the faulty one, then the ECU would still defer to the lower reading from the other sensor. Maybe not quite as accurate, but it would be better than what you have at the moment.

[sgould]

I'll look at the resistance idea.  It might work, but it would always show that side full.  So may not work if it's the pump side. 

I've thought about using the esaab site for part numbers, but I've also looked at saabwisonline, but they are both US based and may not include EU part numbers, so caution needed.

[carrera]

If you search on the EU part number, then I would have thought esaab would only find equivalent new part numbers