improved guages

[john_d]

does anyone thing there will there ever be a time when we will get decent lpg guages? similar to petrol perhaps? they would need to change the multivalve obviously which would push the price up but its it likely to happen?

[Gilbertd]

The sender (when it works) does send a linear(ish) output so could go to a standard type gauge. I think it's just that kit manufacturers consider that a few LEDS to give an indication is enough. If you've got a 0-90 Ohm sender, you could use a bargraph LED voltmeter module to give more steps. How accurate the tank gauge sender is though is anyones guess (extrememly inaccurate on all of my cars!).

[john_d]

my gas guage light has come on, despite the fact i put 17 litres in on sunday & my petrol guage is a dodgy as a £2.50 coin.

[Gilbertd]

You should be able to set what it reads in the software (yes, I know, damn computer again). Fill it up and then set the reading at that point to full, when it's empty, set it again and set the others somewhere in between. I think the gauge on my recently acquired P38 Range Rover works but is a touch on the pessimistic side. It was empty when I got the car and bunged 20 quids worth of gas in it. Now as this was at the BP just down the road would have been around 30 litres in a 90 litre tank. The red light stayed on, changed to one green as I pulled off the forecourt and was back to red within a mile. As I've had the engine in tiny pieces since then (but got it all put back together today and it's running as sweet as they come!), haven't filled it up yet to see what it reads then. When I do I'll set the levels in the software so it reads something sensible.

[Clifford Pope]

It seems to me the green lights are as accurate as a petrol gauge anyway. Basically they just say Full, 3/4, half, 1/4, in the red, the same as a gauge does. Anything else is spurious accuracy.

[DaveyD]

Had our system fitted 850 miles ago. Can do 240 miles on a tank but 130 miles are in the red. Total waste of time. Hopefully they can tweek with software on the 1000 mile service. Or can they ..........

Kind Regards, David

[classicswede]

You can normaly adjust the position of the gauge on the multi valve to give a more relistic reserve reading

[junkcatcher]

The system used in lpg can never be anything like as accurate as petrol gauge even with a cylinder tank.
Without even considering tank shape and lack of baffles.

The lpg level gauge is pivoted near the top tank, this gives a smalle angular movement fot a large change in level and a very no linear response.

A petrol/diesel gauge system usually has the the pivot for the float at about half way height of the tank this gives the best sesnsitivity and a more linear response.

[SimonHobson]

The system used in lpg can never be anything like as accurate as petrol gauge even with a cylinder tank.
Without even considering tank shape and lack of baffles.

The lpg level gauge is pivoted near the top tank, this gives a smalle angular movement fot a large change in level and a very no linear response.

I disagree there. On the 4hole tanks on my old Disco, the float was pivoted in the middle of the tank and geared to the indicator/sender. On the Emer multivalves I'm familiar with, although the pivot is offset somewhat, the linkage is designed to correct somewhat for that.

The real problem is that the sender for a petrol gauge typically has a LOT of steps - easily 20 or more, and they are carefully staged to correct for odd tank shapes. On the LPG senders I've had apart*, they have only had about 10 and so are very coarse.

* I've converted two to drive the dashboard gauge by swapping the resistors. On my old Disco it worked quite well and worked the gauge really nicely - that was a 'large' sender on a 4 hole tank. When I did my mates Series III it was a different matter - it was a small sender (AEB 1090 IIRC), and it has a wonderful design 'feature' of not having any provision for a signal below 25% That's right, when the tank gets down to 25% (of nominal, ie 16l for an 80l tank filled to 64l) it just goes open circuit and the gauge drops off the bottom of the scale, but above that it was 'good enough'.

I reckon there are two types of sender these days. Some (ie 0-90R) are designed to give a fairly linear voltage when fed with a constant current. Others are designed to give a fairly linear current when fed from a fixed voltage via a resistor. In the case of the AEB 1090, I vaguely recall that when I've measured it, the Leonardo feeds it with 5V through a 12k resistor, and the sensor varies from very little resistance when full to open circuit when empty.
I've a circuit drawn out to take several of these sensors, add the currents, and multiply them up to drive the dashboard gauge. Now if I can find someone selling 8 day weeks I might find time to build it

[mat_fenwick]

I've converted two to drive the dashboard gauge by swapping the resistors

Now this I would be very interested in! I'd certainly be willing to exchange some beer tokens for a circuit diagram, particularly the resistor values for your Disco (we have a 1996 V8 and I would assume they would be the same as yours). Did you wire it so the gauge would display whichever fuel it was running on?
Obviously I would have to check the resistance of our senders currently, and the fact it has twin tanks may mean the resistor values would need to be different anyway...

I had better get the LPG system working again anyway before playing with such niceties!

[Sc00by]

Hi all,
it should be fairly simple to drive the dashboard/OE fuel gauge from the LPG tank providing the LPG sensor works in the same direction as the OE sensor eg if the OE sensor is low resistance full and high resistance empty, the LPG sensor would need to behave in the same way. Presumably if it behave the opposite way, it could be turned upside in the tank but i'm not sure on this. If the sensors behave in the same way, a few other factors are now needed.

Empty resistance of OE sensor
Full resistance of OE sensor
Empty Resistance of LPG sensor
Full resistance of LPG sensor

If by concidence the values are the same or even within a small margin, a simple changeover relay would be needed. These relays are also known as SPCO (single pole changeover). Take the common contact from the relay to the fuel gauge using the original sender wire. Now take the normally closed (NC) contact from the relay to the OE fuel sender and the normally open (NO) contact to the LPG sensor. Wire one side of the relay coil to earth and feed the other side of the relay coil from the solenoid valve feed on the LPG tank. In this way, when the LPG is off, the OE fuel gauge will read the contents of the petrol tank. When the LPG system is energised, the solenoid valve on the LPG tank will be energised. This will energise the relay coil, switching the contacts from NC to NO and hence changing the OE gauge feed from petrol to LPG.
If the sensors are different values, some calculation will be needed but the formula will be different depending whether the LPG sensor has a greater maximum resistance then the OE sensor. Quite happy to help out anyone who wants to do this with calculations etc. Some lineaarity will be lost by adding resistors to alter the overall resistance of the LPG sensor but if it's done right, that should be minimised.

Cheers,
Dave

[SimonHobson]

it should be fairly simple to drive the dashboard/OE fuel gauge from the LPG tank providing the LPG sensor works in the same direction as the OE sensor eg if the OE sensor is low resistance full and high resistance empty, the LPG sensor would need to behave in the same way. Presumably if it behave the opposite way, it could be turned upside in the tank but i'm not sure on this.

In general they can't be turned over, but more on that later ...

If the sensors behave in the same way, a few other factors are now needed.

Empty resistance of OE sensor
Full resistance of OE sensor
Empty Resistance of LPG sensor
Full resistance of LPG sensor

If by concidence the values are the same or even within a small margin

That's the problem, I've not seen anything for LPG that's even remotely close. I don't have any figures to hand, but IIRC the resistance required for the gauge was nothing like any standard LPG sender. I simply took the sender apart and replaced the resistors with the values needed.

The first one was relatively easy, it was one of the large senders (over an inch in diameter) used on 4-hole tanks. Also, this had a large number of steps, and pads to connect the wires to have the resistance work either way - so I could convert to the required "resistance decreases as fuel fill increases".

The second one was harder as it was a small one (about 1/2 inch diameter) on a multivalve. It had less steps available, and used smaller resistors - both use surface mount which are "fiddly" to work with if you're not used to it. I also did the calculations of power dissipation and determined that for certain fuel quantities, the resistors were actually on (or slightly over) their power dissipation rating - about 1/8W IIRC. This was with the tank nearly full so only a small number of resistors were in circuit and the gauge current was at it's highest.

The hard part was working out what the values needed to be at each step.

, a simple changeover relay would be needed. These relays are also known as SPCO (single pole changeover). Take the common contact from the relay to the fuel gauge using the original sender wire. Now take the normally closed (NC) contact from the relay to the OE fuel sender and the normally open (NO) contact to the LPG sensor. Wire one side of the relay coil to earth and feed the other side of the relay coil from the solenoid valve feed on the LPG tank. In this way, when the LPG is off, the OE fuel gauge will read the contents of the petrol tank. When the LPG system is energised, the solenoid valve on the LPG tank will be energised. This will energise the relay coil, switching the contacts from NC to NO and hence changing the OE gauge feed from petrol to LPG.

Correct, that's exactly how to do it and how it would work. Though the two I've done had manual switches (both carb engines), and so I just used a double pole switch with one pole switching fuels and the other switching the gauge.

me lineaarity will be lost by adding resistors to alter the overall resistance of the LPG sensor but if it's done right, that should be minimised

I doubt you could do it successfully by adding resistors externally - my experience suggests the correction would be too much. The result would be, as you say, non-linear operation and I suspect you'd struggle to even get a full range of movement on the gauge, let alone anything vaguely linear !


Unfortunately, I no longer have any of my original measurements or calculations. In both cases they went as part of the documentation pack when the vehicles were sold. I know where one vehicle is as another friend now owns it, but it's been via ownership of a complete idiot who "lost" all of that.

[Gilbertd]

It isn't quite that simple. The figures for a P38 Range Rover fuel gauge sender (as I've got the infor for that but I suspect all are going to be roughly similar) are 19 ohms at full and 270 ohms at empty. The low fuel warning comes on if the resistance is in excess of 175 ohms which says that it is far from linear otherwise it would be coming on at about half full. Petrol senders use a rheostat so the resistance is constantly variable but LPG gauges, as Simon has already hinted at, use pads and switch different resistors in so go in steps. Common resistances for LPG senders are 0-90 or 0-1050 so neither are even remotely close to petrol sender units. There are no baffles in an LPG tank and the sender is not damped either, so the gauge would probably only remain stable when you are parked.

I'm not saying it can't be done, but it hardly seems worth the effort.

[Sc00by]

Hi, the DPDT switch would work well in that application, however for the addition of a simple relay to function as the second pole of the switch, you save having to run the gauge wire from the tanks area to the dash.

From what you're saying about your calculations, i suspect you've neglected to take into account the internal resistance of the gauge. Checking through the workshop manual for my Rover (genuine Rover manual on CD ROM), the sender unit for mine gives the following :-

Full 19.6R (R= Ohms)
Empty 240R
167R triggers low fuel lamp.

This suggests to me the internal resistance of the gauge must be reasonably high, as if it was low, let's say 4R to keep the figures simple, the gauge would dissipate 12V/24R = 0.5A. In turn this would give 12V x 0.5A = 6W which would get rather warm in use and eventually overheat in terminal manner! Although 6W doesn't sound much, try holding onto a 5W sidelight bulb after it's been on for a few minutes!!

If i can find a 20R resistor later today, i will perform the following experiment to determine the internal resistance of the gauge. Replace the sender unit with the 20R resistor, switch the ignition on and measure the voltage across the resistor. From this i can determine the curent flowing in the fuel gauge circuit (V/20R = I) and from that, the internal resistance of the gauge ( [12-V]/I = iR). Let's say for the purposes of illustrating this that the voltage across the resistor is 6V - this gives 6/20 = 0.3A. Therefore, [6-6=6]/0.3A = 20R so the gauge would have an internal resistance of 20R.

Assuming your Landy has a similar resistance on the tank sender, you will first of all need to fit a 20R resistor (in series) with the LPG 0-90R sender. I know there are other senders available for LPG tanks, but i don't know the resistance for them so i'll work with what i know! Personally i would suggest fitting the 20R resistor between the earth teminal on the LPG sender and earth. This would then give a total resistance of 110R when empty so would give an OE gauge reading of around half full. If you can live with the OE gauge reading half full when the LPG is empty and full when it's full, this is all you need to do. You can "shift" the gauge down a bit by adding in a 63R resistor which should (if i've got my calcs right) put the OE low fuel lamp off above 1/4 tank and give a full reading of about 3/4 on the OE gauge.

Just seen Gilberts post about the other type of sender giving 0 - 1050R. This type of sender is much easier to alter with external resistors. You'll still need the 20R resistor between the sender earth terminal and earth, but a 363R resistor across the sender terminals.
This will give (1050 x 363) / (1050 + 363) = 381150 / 1413 = 269.75R. Straightaway there's an obvious glitch - i've neglected to include the 20R resistor used to limit the current on full scale so re-working this to give 250R gives 328R for the parallel resistor - (1050 x 330) / (1050 + 330) = 346500 / 1380 = 251R. The nearest value to 328R is 330R, hence the different value used in the calculations. As can be seen, 250+20 is 270R for a Rangie OE sender and knowing how the parts bin got raided when under the auspices of BL, there's a very good chance the Rangie uses a Landy sender unit!

So with a 0-90R LPG sensor it's possible but you'll lose about half the OE gauges movement, with a 0-1050R sender, you can keep the OE gauge functioning fully including the low fuel lamp. However, if the low fuel lamp is switched by a separate switch in the tank, this will still operate independently of the LPG tank.

Hope this all makes sense and is of use to someone!!!

[mat_fenwick]

Yes - this is useful thanks! I'll probably be best getting my head around it before a hard day at work though...

It's made slightly more complicated on ours by the twin LPG tanks, but I'm sure I can tweak things to suit.