24v Lightweight LR fuel tanks reading 1/8 when empty!

[ScarboroughSeadog]

This has become a niggle.....

Late model (1979) 24V- no voltage regulator on these models. Just replaced fuel gauge and one of the lucas switches and now I have a fuel reading at least. BUT both tanks read 1/8 full when empty (they read full when full). Its not to do with the position of the sender float (which is at the bottom of the tank). The gauge returns to empty when the ignition is switched off.

Any thoughts on what the prob could be? Both tanks are well earthed.

Thanks

 

 

[fv1609]

I would be suspicious that your fuel gauge is incorrect or not earthed. The reason that 24v gauges require no stabilising voltage is that these gauges have two windings. One winding to pull the needle to empty & another winding to pull needle to full. These work in opposition depending on the fuel level but as both are energised by the same supply both windings are subject to the same voltage fluctuations.

12v gauges have a single winding so are voltage dependant & require that 10v stabiliser. Although the stabiliser is merely a thermal switch going on & off the erratic output is smoothed by the inertia of thermal gauge which responds slowly.

If you do have the correct 24v gauge bear in mind it has three connections. One to supply, one to the sensor & earth.

 

[ruxy]

Reading 1/8th full when emempty  - stop complaining , this is amed at auto electricians not instrument mechanics - it's one step up from having a dip-stick.  The FFR has a moving iron gauge which has the advantage of being independant of supply voltage , the gauge coils are 'cut' at manufacture , the float variable resistance is fixed (don't try re-calibration by bending the float arm.  Measuring tank content when nearly empty , just when you need good accuracy - will it's down to luck , the luck that by some accident you have good calibration. (where means to calibrate were not provided).  a problem is that the float weight is fixed (unless you have a part sinker) , the float swings on a arc from a  fixed fulcrum point - so as the tank gets to empty then the distance from fulcrum to c of g of float automatically reduces - basic engineering has the advantage of cheapness (value engineering) .  At least with a MOD spec. Rover  - you have twin tanks , so don't let either get below  1/4 full. 

[ScarboroughSeadog]

It’s the correct 24V part- direct part number replacement for the old one. 
It has (and should have I believe) only 2 connections to the gauge itself- the cluster housing is earthed.
I’ve just noticed that the gauge reads full when the tank switch is in the middle (selecting neither tank). Pretty sure it should read zero. Any thoughts if that points to a particular fault?
pic shows the gauge

[fv1609]

Thre are two coils there at 90 degrees to each other. To work correctly it should be earthed at the case (red arrow)

Fuel tank sensor resistor on the right.

[ruxy]

Both fuel tank sender units & dash level gauges for 24 volt are different to 12 volt.  You should have 560757 senders , No.. should be stampd on and  552616 gauges.  

Hhmm   -  tap in neutral position , somewhere in my office I have a tap/bracket with switches assembly but I can't find it.  IIRC the switches are N.C.  (normally closed).    Fuel tap lever in neutral - you would only do that as a gratis extra anti-theft (expect running for only 1/2 mile downhill - don't ask , I caught the wife out once).  The fused supply is fed white (Ignition control) & you don't leave Ign. on without engine running.  So - yes , I think you would have both switch contacts bridged  & point (cinch) between the control coil & deflecting coil feeding both tank units , not a normal situation - you are not supposed to do this & I would expect a spurious reading on the gauge.

 

[ScarboroughSeadog]

Thanks for this detailed response. Looks like the red arrow point is electrically connected to the right bolt (as it appears in the photo). There are 3 earths attached to the cluster securing bolts- perhaps one of them should be connected to the right bolt of the fuel gauge (it hasn’t been for many years- 20, probably). 

The oil and water temp gauges aren’t earthed directly.

[ScarboroughSeadog]

7 minutes ago, ruxy said:

Both fuel tank sender units & dash level gauges for 24 volt are different to 12 volt.  You should have 560757 senders , No.. should be stampd on and  552616 gauges.  

Hhmm   -  tap in neutral position , somewhere in my office I have a tap/bracket with switches assembly but I can't find it.  IIRC the switches are N.C.  (normally closed).    Fuel tap lever in neutral - you would only do that as a gratis extra anti-theft (expect running for only 1/2 mile downhill - don't ask , I caught the wife out once).  The fused supply is fed white (Ignition control) & you don't leave Ign. on without engine running.  So - yes , I think you would have both switch contacts bridged  & point (cinch) between the control coil & deflecting coil feeding both tank units , not a normal situation - you are not supposed to do this & I would expect a spurious reading on the gauge.

 

Late model (1979) 24v models with enclosed cluster gauges can use the 12v sender units- checked this with Blanchards. I have the correct 24v gauge.

Thanks for the detailed response which I will try and decipher tomorrow. 
 
As you say, mid point on the tank lever is useful anti theft measure, but can catch you out!

I did read (on hmvf) the gauge should read empty at the tank lever ‘neutral’ position- hence the query.

[fv1609]

[ScarboroughSeadog]

2 minutes ago, fv1609 said:

How should the earth point shown be correctly earthed? With direct earth from the ‘to fuel tank’ bolt (in addition  to the fuel tank connection)?

[fv1609]

It presses up on the inside of the metal case onto an extrusion that is the weld point for the earth thread the other side. So it is important there is no internal or external corosion and there is a good earth connection

These are resistance readings I measured on a new fuel gauge.

TEST 1: Terminal 1 to metal case  - 320 ohms

TEST 2: Terminal 2 to metal case – 750 ohms

TEST 3: Terminal 1 to Terminal 2 – 450 ohms

If in TEST 2 and TEST 3 both are high or both open circuit, it suggests that the internal earth strip is not effectively contacting the earth stud on the inside of the case. This will probably be due to corrosion.

A reading of appx 600 ohms on TEST 1 suggests a failure on the “Full” winding, but an open circuit suggests a failure of the “Empty” winding.

Open circuit on TEST 2 suggests a failure of the “Empty” winding. Coil failure would be most likely where the end windings are terminated as the wire is very thin.

On TEST 3 a reading above 750 ohms suggests a failure of the “Full” winding. An open circuit suggests a failure of the 220 ohm resistor or its spot weld.

The water temperature gauge works on a similar two-coil principle, except that the float and variable resistor are replaced by a temperature sensitive resistor in the engine’s water jacket.

[ruxy]

Yousay your LWT is  1979,  but what is the VRM  - is it a HF  ?      I have some very accurate Solihull records for build up but nearest FFR is  GJ  (1977)  I do have records for most Contracts of  1979 and 1980 - unfortunately these are only for 12 volt GS.

So - I have to go to the next best 'definative'  Fiche   RTC9968FA   dated  June  1993  ,   I have found a few errors but at least it is official  Land Rover Parts Ltd.

fiche 1   page  O8

PRC1789   Fuel gauge   24 volt

PRC1788   Fuel gauge   12 volt   Printing is a bit blurred but  think correct.

So  -  supercession ,  I would say change in contractor supply not change in specification.

 

 

[ScarboroughSeadog]

12 hours ago, ruxy said:

Yousay your LWT is  1979,  but what is the VRM  - is it a HF  ?      I have some very accurate Solihull records for build up but nearest FFR is  GJ  (1977)  I do have records for most Contracts of  1979 and 1980 - unfortunately these are only for 12 volt GS.

So - I have to go to the next best 'definative'  Fiche   RTC9968FA   dated  June  1993  ,   I have found a few errors but at least it is official  Land Rover Parts Ltd.

fiche 1   page  O8

PRC1789   Fuel gauge   24 volt

PRC1788   Fuel gauge   12 volt   Printing is a bit blurred but  think correct.

So  -  supercession ,  I would say change in contractor supply not change in specification.

 

 

Thanks- yes correct fuel gauge is PRC1789.

It's FFR  20HF63......

[ScarboroughSeadog]

12 hours ago, fv1609 said:

It presses up on the inside of the metal case onto an extrusion that is the weld point for the earth thread the other side. So it is important there is no internal or external corosion and there is a good earth connection

Thanks Clive- my cluster is a later part I think. The gauge fits directly into a common cluster housing.

[fv1609]

Yes fair enough, the pictures I was using come from the gauge on my Wolf. I was using those as guide to the commonality of the two-coil system, I was forgetting you would have a common cluster.

[ruxy]

20 minutes ago, ScarboroughSeadog said:

Thanks- yes correct fuel gauge is PRC1789.

It's FFR  20HF63......

Then with PRC1789 - you must use tank sender unit   90560612   - the only one suitable  (and in fact suitable for 12 volt vehicles - but only with gauge  (panel)   PRC1768

[ScarboroughSeadog]

8 minutes ago, ruxy said:

Then with PRC1789 - you must use tank sender unit   90560612   - the only one suitable  (and in fact suitable for 12 volt vehicles - but only with gauge  (panel)   PRC1768

Understood- that's the sender I have.

[ruxy]

1 hour ago, ScarboroughSeadog said:

Understood- that's the sender I have.

You need to know for absolute certain.   Starting 1977 - there were lots of electrical & mechanical  'change points'  , only a few such as axles are given to/from Serial numbers.  Defender was complex with civvy - but at least normally the change points were stated with chassis/VIN  number. 

I think you will have a single ammeter on the shunt box between seat-back squabs  ?

--------

 Your   20HF63  -  I think built very early 1979  ,   Contract  FVE 22A/87  Item 1 

I have  23HF63   but not much help because it is  12 volt.  FVE22A/87  Item 3   - and I know this Item were built between March and June 1979  ,   having said that - they could have gone down thje lines alongside each other , because as daft as it may seem  Item 1 and 2 vehickles  were built as late as  November  1979.

[ScarboroughSeadog]

1 hour ago, ruxy said:

You need to know for absolute certain.   Starting 1977 - there were lots of electrical & mechanical  'change points'  , only a few such as axles are given to/from Serial numbers.  Defender was complex with civvy - but at least normally the change points were stated with chassis/VIN  number. 

I think you will have a single ammeter on the shunt box between seat-back squabs  ?

--------

 Your   20HF63  -  I think built very early 1979  ,   Contract  FVE 22A/87  Item 1 

I have  23HF63   but not much help because it is  12 volt.  FVE22A/87  Item 3   - and I know this Item were built between March and June 1979  ,   having said that - they could have gone down thje lines alongside each other , because as daft as it may seem  Item 1 and 2 vehickles  were built as late as  November  1979.

Yes, I have a single ammeter on the shunt box- which is located in the rear tub behind the drivers seat. Correct, contract FVE22A/87. First entry on the VHC is 27/11/79. Chassis 95105646A.

[ruxy]

14 minutes ago, ScarboroughSeadog said:

Yes, I have a single ammeter on the shunt box- which is located in the rear tub behind the drivers seat. Correct, contract FVE22A/87. First entry on the VHC is 27/11/79. Chassis 95105646A.

Shunt-box in tub  , that's first location of ammeter/shunt which indicates with Late genny & control box , it then was located with controller  between seat squabs, finally the controller was mounted in a stupid position under the tub on small brackets they welded to the chassis.  

Then the requirement is a final check - that you have PRC1789 cluster mounted  level gauge along with 90560612 tank sender unit. The above two parts will have been calibration checked  & if new / NOS they should be OK.   The only other thing is , because of common fault water entry from vent-seals getting in the binacle box - look for any green verdigris on Lucar connectios (it produces a high resistance that eventually stops current flow).  If you are realy keen then disconnect & check every wire for soundness of insulation & good continuity.  A good AVO and Wee Megger comes in handy.  1/4 indicated full when empty - that is still sort of acceptable.  . 

[ScarboroughSeadog]

Yes- PRC1789 and 90560612 fitted.

Have cleaned the various connections. I'm going to replace the second Lucas switch on the changeover tap. This should be NC (normally closed), but if it has deteriorated and is remaining partially open when not activated, I wonder if that could explain my bogus reading......

[ruxy]

These switches on the change over tap - for automotive switches they are quite good quality (and now £ expensive) , I would check that the plungers are not sticking in any way and have a good mechanical/electrical make/break.  They probably match Burgess cheaper industrail micro-switches so not Mickey Mouse  ,  however ther are far better switches used on machine tools and lifts etc.   It's one of those things keeping warm - study of catalogues for a direct swop if possible , or make a different bracket rather than butcher original to mount the better switches.  There is plenty of thread adjustment for the switches - I normally use a spare pair/bracket/tap  and bench set-up & test ,  avoid the plungers being hammered in further than required - it just damages the switch.   The casings can corrode as exposed - when I remember I just borrow my sons mountain bike type chain sprays and give a blast.  Most LWT now are infrequent use - that does not help - switches are always first suspect.

[ScarboroughSeadog]

New Lucas switch about £30. Mine are over 40 years old so don't mind renewing speculatively. Chain spray probably good preventative idea.

[osiford]

I've just scanned this thread, and from what I can gather, the gauge in question reads OK, except at the critical empty end of the scale, so the basic system seems to be working OK. I've had to fit a number of new sender units and gauges on various vehicles, and none of them are exactly precision calibrated instruments. I have dealt with most as follows. Remove sender unit, and drain tank. Next put in a known quantity of fuel, say a gallon. This is, if you like, a "reserve", as it should have this amount available still, when the gauge reads empty.  If your gauge was reading too high a level when the tank ran out, bend the float arm slightly upwards. Reinstall and check reading on gauge. If still not showing empty, remove sender and re tweak. Some folk may not agree with this approach, but I can't see any problem, and I've always been happy with my fuel level readings on the road. Why make things complicated, if a simple solution is available? Just my thoughts, for what they are worth.

[ruxy]

You know  ,  many bread& butter car manufacturers went off this style of sender unit because they are useless for accuracy.   Take  Fiat Group  ,  small cars  £ cheap   Uno Mk1  ,  Uno Mk. 2  Punto  - they had a plastic tube down the tank from the top (a little larger on dia. than a toilet roll tube )  inside is a tubular float (giving variable resistance)  about the size of said toilet roll card tube , this has a small coil spring at each end.  This design probably went back to the 500D of the 1960's  + earlier  ?     Land Rover could have introduced similar but didn't, you have to work around what you have if you want originality,  probably ndividuals would had done better operating modifications if it were not for fuel safety and insurance cover.