Recalibrating the fuel gauge using a Spiyda Gauge Wizard

[John Vine]

Following my unsuccessful attempts to get my fuel gauge recalibrated, I decided to apply a little science and fit a Spiyda Gauge Wizard.

I wired it in without problem, using a short extra wire to link the sender input to the Wizard, and piggybacking the power source off the gauge 12v supply (switched).  I then syphoned out the fuel and collected a set of sender resistances:

Given that around 6L of fuel is unreachable and therefore unuseable, my next move is to load the Wizard with the resistances corresponding to the five gauge positions between Empty and Full.

The snag is that the very range where I wanted greater precision (1/4 to Empty) is going to escape me.  As you can see, the sender resistance is already maxed out at 205 throughout that range.  It seems that the only answer is to remove the pump and then bend the float wire downwards.

I’ve never removed a fuel pump before, let alone modify the float.  What should I be looking out for?  Are there gaskets that will need replacing?

JV

[johnnyseven]

Hi John- I’ve mailed my findings

[StevehS3]

Thinking about the bottom quarter area and a possible explanation. Do you think the gauge was stuck until the 14th litre was added then it jumped from 205 to 176 ohms and in practice with all the vibration and slushing around of fuel when driving it wouldn’t have done that? Apart from that anomaly it seems as good as can be expected.

[John Vine]

Thanks, Johnny!

Thinking about the bottom quarter area and a possible explanation...

Hi Stephen -- I don't think it was stuck as I rocked the car back and forth and the resistance didn't change.  I'm going to remove the pump and examine exactly the sweep of the float arm to see whether I can bend it downwards.  If I can get the float to reach the 6- litre zone, I'll be able to take a fresh set of resistance readings and then recalibrate the Spiyda to match.  Wish me luck!

JV

[Jonathan Kay]

Is it possible that the float is on the bottom of the tank rather than the resistor at the end of its travel?

Good luck.

Jonathan

PS: If there any old or tired rubber hoses nearby it might be a convenient time to replace them.

[Mark w]

Hi John 

I thought about using a Spiyda but opted instead to bend the float support first .

I am assuming that you have the pump in tank set up as your car is a duratec ? .If so remove the boot floor and you can then access the top of the tank .Unscrew the big black " nut " that holds the pump in place .I did this by tapping it round using a piece of wood and a hammer , it has projecting ribs on the outside  .Once the resistance is overcome mine unscrewed by hand . You can then remove the whole assemply and bend the float arm . 

I did mine by trial and error .I had two attempts to get it right .Basically i was looking for a more accurate reading at the lower end of the scale . My guage now shows full for " ages " but i can now safely run down to empty knowing i dont have a huge amount of fuel left unused in the tank .

I didnt have to replace any rubbers 

PS dont forget to bend it the right way !!

[John Vine]

Is it possible that the float is on the bottom of the tank rather than the resistor at the end of its travel?

Hi Jonathan:  I don't think so, because if it were the resistance would change as soon as I put the first litre in.  My guess is that the float was hanging in the air until I'd added 13-14 litres.  It would be useful to know the full resistance range of the sender.  The part number of the Ford pump is, I think, 3M51-9H307 P1697156AT, but I've had no luck searching.

I am assuming that you have the pump in tank set up as your car is a duratec ?

Hi Mark:  Thanks for that.  Yes, it's an R400D, and the pump sits in the tank. I plan to bend the float arm downwards.  I hope that's right!  I also plan to leave the fuel line attached, although others have suggested it's better to detach it.  Any thoughts?

JV

[Mark w]

John , i didnt disconnect anything but i did manage to splash a bit of fuel about ! .

I seem to remember when you pull the pump out it has a plastic " bucket " assembly on the bottom which holds a fair bit of fuel ( presumably for anti surge protection ) which i then tipped on the floor as i withdrew the pump !!

Downwards ? Now you've got me thinking .....

So you bend the arm so the float goes lower , except it cannot stay there because the level of the fuel is the same so the float rises to its original position but giving a lower resistance reading now and moving the needle higher up the gauge .

Yep that seems to make sense and must have been what i did .

 

[John Vine]

Thanks, Mark, and for confirming the direction of bending!  Also, I intend to order a new gasket, just in case.

It's worth saying that the Spiyda is a doddle to set up.  Fiddly at times, but very straightforward.  I reckon that, with the float behaving properly, I'll be able to get the gauge spot on.  (Very satisfying too, as I dislike intensely anything on the 7 not working properly.)

JV

[StevehS3]

Hi John, firstly here’s the good luck message!

I wondered if your gauge actually moves in one step from 1/4 to E in practice? My needle does seem to gradually move between 1/4 and E albeit rather more quickly than expected which ties in with the non-linearity of the sensor/measurement. Having said that the gauge is heavily smoothed so I expect the needle wouldn’t immediately jump from 1/4 to E anyway.

I expect you are right and the float is hanging in air below 1/4 full.

[Ivaan]

It's strange that for a car with "added lightness" we're all carrying around a dead 6Ltrs ( 6Kgs ) of fuel, that could be fairly easily removed by re-profiling the bottom of the tank.

[John Vine]

I wondered if your gauge actually moves in one step from 1/4 to E in practice?

Interesting thought.  One of the Wizard's functions is to allow you to move the needle slowly across the gauge so as to set the five calibration points (E, 1/4, 1/2, etc).  The needle moves in a steady and smooth fashion when doing this (it's well damped, of course), which suggests that it's reflecting small changes in sender resistance faithfully.

JV

[Mark w]

Hi John

How did you establish that the first 6L were unreachable ? did you put a litre in at a time and try to start the engine ?

I ask because i ran my car pretty low today ( to refresh my memory as to how the gauge behaved near the empty reading ! )  and i managed to get 29.8L in on a fill .( needle was reading empty ! ) 

This either means i got within 0.2L of getting dumped at the side of the road or the unusable amount is actually less than 6L .

I guess i am just trying to establish how far i can trust the message from the gauge now ive altered the float position !

[Garth]

Watching this with interest John as one of my winter jobs is to fit the wizard and new fuel gauge!

Garth

[John Vine]

Hi Mark,

Nothing so empirical, I'm afraid.  I relied on hearsay from members, plus a bit of experience.  When I first had my current 7, I attempted to find the point at which the engine would start missing.  This turned out to be around 5L remaining in the tank.  The general consensus was more like 6L, so I adopted that.

JV

[Strangely]

Mark, it must depend upon the fuel tank configuration, because on my 1996 K my record is getting 36.58 litres in on a fill-up, and before I changed the tank I ran the engine until it stopped and there were literally just drops left in the tank. One thing I have noticed is that when the level gets very low, the car won't stand going up even the slightest slope without conking out with fuel starvation.

[John Vine]

One thing I have noticed is that when the level gets very low, the car won't stand going up even the slightest slope without conking out with fuel starvation.

That happened to me.  I'd returned to my B&B after a Brands t/d, and parked on the (sloping) lawn.  Next morning, the car just would NOT start.  I enjoyed a long walk to the nearest petrol station to buy petrol and a container to put it in!

JV

[Mark w]

I had a K series car and the pick up outlet was in the end of the tank , My Duratec has the pump sat in the top ,

The pump in the latter doesn't reach the bottom of the tank which might explain the difference between a 30L fill and a 36L fill .

JOHN 

Can you just clarify how you have wired the Spiyda please , my electrical knowledge is poor !

Did you cut the sensor wire near the gauge and connect  the relevant wires from the Spiyda  to the two cut ends ? ( i am OK re the live and earth ) . How/ where  did you connect to take the resistance readings ?  

Having gone to the trouble of bending my float arm i am minded to recalibrate the gauge now . 

[John Vine]

Hi Mark,

I've just removed the pump, so I'll shortly be measuring the resistance range of the sender.  Then I plan to measure just how much higher the float sits above the bottom of the empty tank.   If I then add, say, three litres of fuel, I should be able to work out how far to bend the float downwards so that it just reaches the fuel.  I should then have sufficient range to set the E-1/4 segment more accurately.

This is how the Wizard fly-leads are set up:

The minimum connections are red, black, blue and green.  The other two are for optional features.

The fuel gauge connector on my car looks like this (edited 17Aug18 : looking in the direction of cable entry):

This is how I made the connections:

1. Detached the 3-way plastic connector from the back of the sender, and pulled out the green/black sender wire. (In fact, I pulled all three wires and removed the plastic plug completely.)
2. Crimped a female spade to the end of the blue Wizard wire, and connected this to the gauge sender terminal (GB 148 above).
3. Crimped a ring connector to the end of the Wizard black wire, and attached this to the (nearby) chassis earthing point.
4. Crimped a male spade to the end of the Wizard green wire, and connected this to the female spade on the (now loose) sender wire.
5. Piggybacked the red Wizard wire onto the gauge green wire (12v switched supply) using (horror of horrors!) a Scotchlok.

Doing it this way, I avoided cutting any of the original wiring, so could reinstate the standard set-up if necessary.

To measure sender resistances, touch the multimeter probes to the green/black and black gauge wires (or, if you've already hooked up the Wizard, to the green and black terminals in the Wizard connector block).

Hope that helps!

JV 

[John Vine]

Mark (and others),

Just a thought....

My car has the 1/4-sweep fuel gauge:

I don't know whether any of the thread so far (or indeed the Wizard itself) applies to the more recent 3/4-sweep gauges.  Can anyone with a more recent 7 comment?

JV

[John Vine]

Re the gauge connector, the Feb 2015 Assembly Guide shows this (for Sigma variants with standard dash):

Note that the sender wire has changed to Yellow.

But for the 420 (standard dash), it shows this:

Which is the same as for my 2008 R400D.   Puzzling.

JV

[Mark w]

Hi John 

Thanks for the very informative info above .I attach below a couple of photos from the wiring and connection to the three quarter sweep gauge .

The second wire from the right is actually green and black, i am assuming the sender signal ? .

So the connector to the gauge is different from yours  . I am assuming we have live , neutral ,sender wire and a fourth wire for the illumination of the gauge . 

In my case the available slack in the loom at this point is NIL ! I think the easiest way to complete the wiring for the spiyda is to use male and female connectors to make an extension section of loom ( with the spiyda within its length ) and plug it in the gap between loom and gauge .

Can anyone suggest a source for the mini  connectors ?

[John Vine]

An update...

This is the float position on the standard pump:

 

 

The rear edge of the pump base rests on the tank floor, so the float isn't going to register anything for quite a while.  In the original setting (Focus?), I understand that the floor of the tank has a recess to accommodate the bottom of the pump.  This would mean that all the tank contents would be useable.  I suspect that, at “empty”, the float sits on the bottom of the tank.

I bent the float arm like this:

I left a gap between the base of the float and the bottom of the pump, so as not to overdo things first time.  I also reversed the mounting of the float, so that it pointed down the slope of the tank floor (that is, towards the front of the car). I refitted the pump into the (empty) tank and added six litres of fuel, but the sender still registered max resistance (203).  I removed the pump again and bent the float further, so that the bottom edge was level with the bottom of the pump.   The resistance with the tank empty now dropped to 163.  I then added fuel a litre at a time and recorded the sender resistances:

For info, the resistance range of the sender is 9 (full) to 203 (empty).

So, it now looks as though I have the desired precision where I wanted it -- in the E-1/4 range.  The next step is to recalibrate the Wizard with my new calibration points.

If all this proves successful, I plan to make use of one of the Wizard’s optional features: a “low fuel” warning lamp.

JV

[Jonathan Kay]

Nice work. I'm convinced.

What would that be like on the road without the Wizard?

Jonathan

[John Vine]

Hi Mark,

Yes, I'd say the green/black is the sender wire.  You can always check by measuring resistance between that wire and a chassis earth.  You should get a value between 9-203 if the pump/sender is the same as mine.

As for connectors, maybe you'll need to ask either CC or Caerbont.  But it might just be one of these.  I agree that a little mini-loom would do the job.

JV