Jump to content


Fuel Sender Diagnosis/Repair


  • Please log in to reply
47 replies to this topic

#1 _Yella SLuR_

_Yella SLuR_
  • Guests

Posted 29 August 2006 - 12:40 AM

Since there ain't nothing on it in the manuals.

1. Diagnose if it is the sender or the guage. Mod Edited.. Please see Dangerous's post below - Tiny.

2. Attach an earth wire to the sender body, and other end to an earth on the body. If the gauge works, it's inadequate earth/insulation between the tank and body. Run earth wire.

3. Still not working, it's in the sender itself. pull the sender apart, undo metal tab at the back only, and slide out from two folded metal tabs at the front. Grab multi-meter. Check resistance of resistor wire end to end. Should read 60 ohms. If no reading, resistor wire is broken somewhere along it's length. Move probe along to locate, may be able to solder, if not new resistor wire required.

4. If resistor wire checks out. Check connection between resistor wire, and power supply attachment. If broken, solder to restore power (my problem).

5. still no worky, check tension of copper arm and contactor to resistor wire.

6. still no worky, get professional help.

Hope this helps. Cost me $50 to find out. Damn tradesman, it's all so simple when you know what you are doing. Hope it saves you a buck.

Edited by Tiny, 29 August 2006 - 06:05 PM.


#2 TerrA LX

TerrA LX

    Fulcrum Fixture

  • Members
  • PipPipPipPipPipPipPip
  • 14,241 posts
  • Location:Sid 'n' knee
  • Joined: 31-May 06

Posted 29 August 2006 - 12:55 AM

Thanx Yella SLuR :spoton:
, can we stick this?

#3 _Yella SLuR_

_Yella SLuR_
  • Guests

Posted 29 August 2006 - 01:01 AM

If you want. Step 3 is a bit confusing. You need to remove the sender from the tank and remove the cover from the actual sender mechanism on the fuel pick-up. Where the bendy stick/float connects into.

#4 rodomo

rodomo

    To advertise here, call 13TORANA

  • Members
  • PipPipPipPipPipPipPip
  • 18,026 posts
  • Name:R - O - B Dammit!
  • Location:Way out west of Melbourne Awstraylya
  • Joined: 10-December 05

Posted 29 August 2006 - 01:14 AM

Prepare to be "adjusted" in your diagnosis as you haven't stipulated the make, model and year of the vehicle you have repaired. :tease:

Edited by rodomo, 29 August 2006 - 01:16 AM.


#5 _Yella SLuR_

_Yella SLuR_
  • Guests

Posted 29 August 2006 - 01:16 AM

Need you ask??? LH of course, but think it's actually and LX sender (solid black plastic float).

#6 _keith1962_

_keith1962_
  • Guests

Posted 29 August 2006 - 06:17 AM

Its ok Pat i knew what car you were talking about.

cheers Keith

#7 _devilsadvocate_

_devilsadvocate_
  • Guests

Posted 29 August 2006 - 06:46 AM

Since there ain't nothing on it in the manuals.

1. Diagnose if it is the sender or the guage.  Pop the wire off the sender and earth it out on the body.  If person in cabin with ignition on sees guage go all the way, it's the sender, not the guage.  Don't do this too long, you are running zero resistance, which Torrie guages don't like.  Result will be non functioning guage (need 2nd person so it doesn't sit at full for too long, preferably not at all).

I wouldnt recommend this method for a general how to fix it. If you sit there long enough for the gauge to move to max or to move at all, (they dont move to fast since they are damped), that can be plenty of time to administer the last rights to a gauge that may have been overdriven on previous occasions.

Much safer to run a dumby load as suggested by Rodomo, better still visit your electronics shop and for a couple of dollars invest in a 0-500 Ohm potentiometer(variable resistor). Using your multimeter set the potentiometer to the min resistance value for your gauge(10 OHM for a LH), generally setting to the min resistance value will see the gauge go full scale on most Holdens.

Edited by devilsadvocate, 29 August 2006 - 06:47 AM.


#8 _Yella SLuR_

_Yella SLuR_
  • Guests

Posted 29 August 2006 - 08:51 AM

Just out of interest, is an LH sender different to an LX? I know UC is different. Fatter power pin, and float is plastic ball rather than the floating solid plastic. UC sender also alot more compact than LH/LX. Maybe I should take some pics tonight.

Agreed, not the best method, but quick and easy. Potentiometer sounds good, but probably beyond the requirements of most peeps. Gauge takes a while to get to the top, even with zero resistance. Personally, I'd just be looking for some sort of movement in the guage, then stop the test. It doesn't have to go all the way to full.

Edited by Yella SLuR, 29 August 2006 - 08:53 AM.


#9 Dangerous

Dangerous

    Forum Fixture

  • Members
  • PipPipPipPipPip
  • 948 posts
  • Location:Adelaide, SA
  • Joined: 14-November 05

Posted 29 August 2006 - 03:08 PM

Yella, please please PLEASE change your original post, and stop this dangerous practice! :fool: :fool: You should NEVER short out an LH/LX gauge. LH to UC and HJ (or is it HX?) to HZ (and early Commodore) gauges are a bimetallic strip gauge. They have two metal strips with dis-similar expansion coefficients, with a heating wire wrapped around the strips. The reading from the sender unit denotes how much current can flow through the heating wire, to heat the two strips. Because one strip expands at a faster rate than the other, the strip bends, causing a deflection on the gauge.

Now, LH/LX sender units have a minimum resistance of 10 ohms, with the voltage to all the gauges (except the voltage gauge) limited to around 10 volts by the voltage regulator. This means that the absolute flat out maximum current that can flow through the gauge is less than one amp, as the heater wire has some resistance.

If you short the sender wire to earth, the only resistance in the circuit is that of the heater wire, so a MUCH higher current than 1 amp can flow. This can do one of two things almost instantly, neither of them good. -

Firstly, it can cause a more rapid increase in temperature than the bimetallic strip is capable of handling, which can permanently warp the strip. This will cause the gauge to then permanently read high or low and/or slow. You've just stuffed the accuracy of the gauge.

Secondly, it can either burn out the heater wire, or break one of the bimetallic strips, either of which will cause the gauge to die permanently.

SL/R gauges are getting too rare to destroy them this easily. Would you flatten the throttle on your engine while it was running and in neutral (even if only for a short time) just to see if the engine was still working? I hope not, but that's the equivalent of what you're doing to the gauge.

Re the potentiometer, they can be used, but most don't have the current capacity to test the guage near its upper limit, and the potentiometer will burn out. Much better to use some power resistors, or get a 10 ohm 5 or 10 watt resistor, and solder in series with the potentiometer. Minimum settign one the pot will tehn correspond to minimum reading on the gauge, and effectively limit the current flow.

LH and LX sender units are the same in principle, although Holden did tend to change around the connector design in most Torana models.

Edited by Dangerous, 29 August 2006 - 03:09 PM.


#10 _devilsadvocate_

_devilsadvocate_
  • Guests

Posted 29 August 2006 - 03:37 PM

Yes, I stand corrected. Most potentiometers wont cope with a 10 Ohm load at 10V for too long. Fine for use with a UC where one would be setting the min resistance to 40 Ohm, with a 3W pot or if testing at the other end of the scale. Potentiometer can come in handy, especially if you are trying to work increase or decrease the resistance one needs to add to a sender to get the gauge(be it temp or oil etc) to read exactly how youd like it too.

#11 _Yella SLuR_

_Yella SLuR_
  • Guests

Posted 29 August 2006 - 04:24 PM

Mmm, can't. Oh well, at least we is all the wiser now for having discussed it. Will just count myself lucky I suppose (beginner's luck). Only reading I'm really interested in is the empty reading.

This is the first fuel guage problem I've ever had, so will count it as a once in a life time experience.

Good info as always Dangerous. Much appreciated, I'm that little bit wiser.

#12 Tiny

Tiny

    Oh My, Don't you post alot

  • Administrators
  • PipPipPipPipPipPipPip
  • 10,018 posts
  • Name:Tiny
  • Location:Sydney
  • Joined: 04-February 07

Posted 29 August 2006 - 06:10 PM

Hey Guys.

I've modified the above post as per dangerous's comments! Hope thats ok pat!!

Dangerous: Can you reccomend any way of diagnosis by a torry lover with minimal equipment? ( especially an in-car diagnosis?)

I know that guage repair people have a "dummy load" box wehre they hook up the guage and wind up the resistance checking the guage-read to a known resistance value. Are these available cheaply or is there another way?
Cheers for the good advise all round!

#13 _Yella SLuR_

_Yella SLuR_
  • Guests

Posted 29 August 2006 - 09:54 PM

Don't have a problem with that. As always, I'm interested in seeing the info on here is as correct as it can be, the future stock of our Torries depend on it. Change away. Great explanation of the reasons there Dangerous.

#14 _Yella SLuR_

_Yella SLuR_
  • Guests

Posted 29 August 2006 - 10:00 PM

1. Test the Gauge.  Don't short the sender wire to earth, the only resistance in the circuit is that of the heater wire, so a MUCH higher current than 1 amp can flow. This can do one of two things almost instantly-

Firstly, it can cause a more rapid increase in temperature than the bimetallic strip is capable of handling, which can permanently warp the strip. This will cause the gauge to then permanently read high or low and/or slow. You've just stuffed the accuracy of the gauge.

Secondly, it can either burn out the heater wire, or break one of the bimetallic strips, either of which will cause the gauge to die permanently.

SL/R gauges are getting too rare to destroy them this easily.

Potentiometers can be used, but most don't have the current capacity to test the guage near its upper limit, and the potentiometer will burn out. Much better to use some power resistors, or get a 10 ohm 5 or 10 watt resistor, and solder in series with the potentiometer. Minimum setting on the pot will then correspond to minimum reading on the gauge, and effectively limit the current flow.

^ I believe that is the relevant change for Point 1, unless Dangerous has an alternative test method. I've changed some text, and fixed the spellos.

Edited by Yella SLuR, 29 August 2006 - 10:01 PM.


#15 Tiny

Tiny

    Oh My, Don't you post alot

  • Administrators
  • PipPipPipPipPipPipPip
  • 10,018 posts
  • Name:Tiny
  • Location:Sydney
  • Joined: 04-February 07

Posted 29 August 2006 - 10:20 PM

Great stuff! I'll wait to see if dangerous has any more information, and then add that into the first post!

Cheers!

#16 _devilsadvocate_

_devilsadvocate_
  • Guests

Posted 29 August 2006 - 11:49 PM

I know that guage repair people have a "dummy load" box wehre they hook up the guage and wind up the resistance checking the guage-read to a known resistance value. Are these available cheaply or is there another way?
Cheers for the good advise all round!

I dont think there is another way around to check the end points of the gauge(unless you have a replacement sender lying around) other than what has been suggested above. One needs to know the resistance end points of the sensor and have some dummy loads to apply to check with the gauge.
If the aim is to simply check that the whether the gauge is operational on a holden, then Rodomo's suggestion of the test lamp is probably the easiest for those that arent familiar with multimeters and component resistors, with the proviso that that the rating of the lamp is <5W and is not used for too long on a post 78 Holden, those with the 40-280Ohm fuel senders.

#17 _devilsadvocate_

_devilsadvocate_
  • Guests

Posted 30 August 2006 - 07:13 AM

If you short the sender wire to earth, the only resistance in the circuit is that of the heater wire, so a MUCH higher current than 1 amp can flow.

Are you sure about this Dangerous? You are most likely right (as you normally are) but just raising the point for consideration
Giving this topic some more thought, popped out to the UC and established by way of voltage drop across the sender that the gauge/ in the UC has a internalresistance of ~45 Ohms which together with a 40 Ohm min sender would give a max current reading of 0.12A, which explains why Ive never cooked a pot testing these gauges


Won't the current that flows be dependent on the resistance of the heating wire in the gauge, you are implying that the gauge itself will have a resistance of less than 10 Ohm (by the current exceeding 1A if shorted)
I dont have any LH/X gauges to try out myself, but what is the resistance of the gauge itself?, or what is the maxium current that flows through a LH gauge/sender, those with LH's could measure the voltage across the sender when the tank is full to get an answer for this?

#18 Dr Terry

Dr Terry

    Technical + Numbers Guru + Moderator

  • Moderators
  • PipPipPipPipPipPip
  • 3,315 posts
  • Location:Eastwood (Sydney) NSW
  • Joined: 13-November 05

Posted 30 August 2006 - 09:07 AM

Hi Guys.

I thought is was time to throw in my 2c worth.

The easiest way to check LH & LX gauges (& yes, they are the same, but UC is different) is to use substitute resistors (or dummy loads) as mention earlier. This works for all 3 gauges on LH/LX, by the way. Go to Jaycar, Dick smith or similar & buy three 1/4W resistors, a 75, 25 & 10 ohm. Just put these, one at a time between the disconnected sender wire & a good earth & switch the ignition on & wait for the gauges to warm-up & stabilise. The 75 ohm will make the gauge read to the empty mark (or cold or low for the temp & oil gauges), the 25 will move the gauge to 1/2 & the 10 ohm to full scale. That's it, if the gauge doesn't do this you have a gauge or voltage reg fault. You can also use these ohm readings with a mulitmeter to check a fuel sender out of the car. AKAIK the current thru the sender is less than 100 mA.

Dr Terry.

#19 Dangerous

Dangerous

    Forum Fixture

  • Members
  • PipPipPipPipPip
  • 948 posts
  • Location:Adelaide, SA
  • Joined: 14-November 05

Posted 30 August 2006 - 01:39 PM

I think I need to clarify what I said vs what I meant. DA, you're right - what you have quoted me as saying about more than 1 amp being able to flow is not correct. I was using the '1 amp' as a hypothetical maximum current , and then trying to say that if you short the gauge to earth, a much higher current is allowed to flow than what the gauge was designed for. It certainly wouldn't need a current as big as 1 amp for the gauge to operate. Offhand, I do not know the 'hot' resistance of the heater wire element - the maximum current flow will depend on the cold resistance of this bit of wire, and any other internal resuistance in the circuit. It might be as little as 100 mA for normal operation, but the maximum design current will still be exceeded if the gauge is shorted to earth on an LH/LX.

The test method that Dr Terry has described is OK, except that you can't easily get a 75 or a 25 ohm resistor, but you can get a 68 and a 22 or 27 ohm resistor - the mid scale resistor should be roughly half way between the 10 ohm and 75 ohm range of the gauge, and the nearest preferred value would be a 33 ohm resistor. I'd also recommend at least 1/2 watt or 1 watt - not only because they can absorb more power, but they are also larger with larger leads, which make them easier to use. Resistors are only a few cents each.

If it were me (and it was, cos I've bought them :D ) you could buy a 72 ohm, 10 ohm and 33 ohm resistor. 72 ohm bridged from the gauge wire to earth will give empty/cold/low, the 10 ohm will give full/hot/high, and the 33 ohm will give a mid range reading.

You can just hold the resistors in place with your hand and you won't get electrocuted as the voltage is way too small, or you could buy some miniature jumper lead sets when you buy the resistors.

For LH/LX, it's the only Torana where the three gauges (oil, temp and fuel level) are identical in their operation with all three sender units using the same resistance range for operation, so any sender unit will cause an identical reading if attached to any one of the three gauges.

As far as in car diagnosis of LH/LX gauges goes;

(1) If all three gauges (not the voltage gauge, as it's hooked straight into the 12 volt rail, bypassing the voltage stabiliser) suddenly head towards full scale for no apparent reason, turn off the ignition immediately, as this is a sign that the instrument panel voltage stabiliser has failed and is short circuiting, and is allowing too much voltage to be fed to the gauges. The car can still be driven, but I'd recommend disconnection the oil, temp and fuel gauge sender wires until you're sure that the voltage stabiliser is replaced or working OK. Removing the wires will protect the gauges.

(2) If all three gauges fall to zero for no apparent reason, the voltage stabiliser has probably failed in an open circuit - no immediate danger to the instruments. Unfortunately the voltage stabiliser rarely fails like this - more often like described above.

(3) if all three gauges move down a bit when the park or headlights are turned on, it's a sign that the instrument cluster earth is not good enough (it's barely good enough from factory). Run an extra earth lead from the screw holding on the voltage stabiliser on the back of the cluster to one of the four 7/16" UNF bolts holding the steel steering column bracket onto the column. DO NOT use any of the larger three silver bolts, as they form a part of the collapsible steering column assembly.

(4) If the fuel gauge dies, or reads low, it's usually that the tank isn't properly earthed (once again, no real earthing from factory). Run a piece of wire from a small hose clamp attached to the metal fuel pickup line on the fuel pickup/sender unit to a good earth on the body.

(5) If one gauge dies, eg the oil pressure gauge, and you think you've still got oil pressure, swap over the oil pressure sender wire with the temperature sender wire (you might need a bit of wire to extend it to reach). The Oil pressure gauge should then read engine temperature, and the temperature gauge should then read oil pressure. If the temperature gauge then shows accurate oil pressure but the oil pressure gauge does not show accurate temperature, then your oil pressure gauge is probably faulty. If the temperature gauge does not show accurate oil pressure, then you probably have a faulty oil pressure sender unit. The same goes to test the fuel tank sender unit.

(6) As I mentioned, having a known working spare fuel tank sender unit is handy for checking any of the fuel, oil or temp gauges. Connect the sender unit connector on the spare fuel sender unit up to the sender wire for the gauge to be tested, earth the body of the spare fuel sender unit, and as you raise and lower the float, the gauge under test should raise and lower with it (but slowly).

(7) Dr Terry's method, with resistors as outlined.

For other early holdens, HK to HQ and LC and LJ gauges are a different design. The test method is as above, except in this case, for the fuel gauge the resistance ranges are 0 to 30 ohms (ie you can short the fuel gauge sender cable for the low reading), temperature is 315 to 22 ohms, and oil pressure is 10 to 190 ohms.

For UCs, as DA has mentioned, fuel and temp resistance ranges are 282-40 ohms, but oil is 10 to 180 ohms. Not sure if these changed through the life of the UC or not.

Sorry if my previous posts caused confusion.

#20 _devilsadvocate_

_devilsadvocate_
  • Guests

Posted 30 August 2006 - 02:16 PM

According to my friendly instrument repair man, LX/H fuel gauge winding is 12 Ohm, ie max current <0.46A

#21 Dr Terry

Dr Terry

    Technical + Numbers Guru + Moderator

  • Moderators
  • PipPipPipPipPipPip
  • 3,315 posts
  • Location:Eastwood (Sydney) NSW
  • Joined: 13-November 05

Posted 30 August 2006 - 02:17 PM

Hi Guys.

Dangerous said:- "the mid scale resistor should be roughly half way between the 10 ohm and 75 ohm range of the gauge, and the nearest preferred value would be a 33 ohm resistor. "

This is not correct, the thermister (sender) is non-linear. A 33 ohm resistor will result in a gauge reading closer to a 1/3 scale.

You are correct about the commonly available resistors, but you can make a 75 ohm by using two 150 ohm ones in parallel & you can make a 25 ohm by using a 47 ohm & a 56 ohm in parallel.

Dr Terry.

#22 _devilsadvocate_

_devilsadvocate_
  • Guests

Posted 30 August 2006 - 07:58 PM

This is not correct, the thermister (sender) is non-linear. A 33 ohm resistor will result in a gauge reading closer to a 1/3 scale.

Dr Terry:Thermistor? we are talking fuel gauge? Typically temp gauges arent linear because of the phenomenon you mention(the behaviour of the thermistor is exponential), for the correct LH temp sender the resistance can be calculated by the formula
Res = 382.98e^(-0.0286Temp) (this is from measurements supplied by hb1200)

Dangerous, I dont think the UC fuel and temp gauge are identical. The temp gauge starts way off scale whereas the fuel gauge starts basically at empty.
Ive used my temp gauge as the fuel gauge(after a fautly lpg sender/convertor burnt it out) and found it would be driven off scale by the correct sendor( which was the correct resistance value) and had to put a 10Ohm resistor in series with it to stop going overscale. Repairing the fuel gauge to specs sees the needle nicely on the full mark with the same sender.

#23 Dangerous

Dangerous

    Forum Fixture

  • Members
  • PipPipPipPipPip
  • 948 posts
  • Location:Adelaide, SA
  • Joined: 14-November 05

Posted 31 August 2006 - 10:54 AM

^ ahh, OK. I was going off the specified VDO sender units listed for the UC. Maybe they've made another mistake?

#24 _Yella SLuR_

_Yella SLuR_
  • Guests

Posted 31 August 2006 - 11:40 AM

Plenty of stuff there for peeps to make their own mind up on how to tackle it now.

One step missing from my explanation, was to do the multi-meter test of the sender unit for 10-70(??) ohm resistance prior to pulling it apart which verifies that the problem is indeed in the sender.

Wish I could join the discussion, but my knowledge of electronics can only be described as "Pre-Neanderthal".

Edited by Yella SLuR, 31 August 2006 - 11:41 AM.


#25 Dr Terry

Dr Terry

    Technical + Numbers Guru + Moderator

  • Moderators
  • PipPipPipPipPipPip
  • 3,315 posts
  • Location:Eastwood (Sydney) NSW
  • Joined: 13-November 05

Posted 31 August 2006 - 12:00 PM

Hi Guys.

Devils, I was using the temp sender (thermister) as an example. The 3 gauges senders are all non-linear, look at the way the fuel float arc swings, it too is non-linear.

In answer to Yella's query, it is easier to substitute for the sender, it saves removing it from the tank. But having said that, if the sender is out check it using the ohms scale on a multi-meter. The advantage of using dummy loads (substitute resistors) is that it not only checks the operation of the gauge, regulator & wiring etc. it also checks the gauge's accuracy, which is difficult if you're just checking the sender.

Dr Terry




3 user(s) are reading this topic

0 members, 3 guests, 0 anonymous users