I've just bought a 1938 Ruby and I'm chasing some minor faults. This concerns the fuel gauge which didn't work at all. Checked out Malcolm Watts excellent article at http://www.austin7.org/Technical%20Articles/Fuel%20Gauge%20Working/ and with some other advice checked out the sender. The arm was seized solid and I thought the problem solved. Unfortunately it wasn't.
The gauge passed the test given in the article ie. it goes full scale when the tank terminal is grounded and now the sender appears to be ok.
So now I'm chasing down what can be done when the gauge works but doesn't work properly.
With the sender on the bench the resistance range is o to 25 ohms. With the tank about half full the resistance is about 12 ohms. Does this sound like the correct sender for the Ruby ?
Another question is (if the sender is the correct one) has anyone ever measured the range of actual resistances found between tank empty and tank full ?
I've plenty of other observations but will stop at this point. Roger

Location: A little East of Sandy

Roger, much confusion around this, the earlier units measured approx. 25 ohms tank empty and approx 1 ohm tank full.

The later units were the reverse of this, approx. 1 ohm tank empty and approx. 25 ohms tank full.

Faulty operation with apparently good units usually comes down to bad earths at either end of the circuit. Either the tank unit or the guage.

Ian Mc.

Location: Shropshire

Thanks for that information Ian. I have made sure all of the connections, especially the earths, are all ok. My sender is minimum resistance at max fuel. So that makes it an early type then - although it's in a 1938 Ruby. At least the gauge must be the correct one for the sender.
I now think that the gauge must be faulty. Today I received a 25ohm potentiometer from eBay and tested the gauge on the bench. It will go full scale (5gals) with a minimum resistance but it will not lift of the "Empty" end pin until the resistance has fallen to 13 ohms. Hence my half full tank was showing empty.
I have temporarily put a 47ohm resistance across the tank sender which makes it show about 1 Gal.
I guess a 25 ohm resistor placed there will allow the gauge to read across it's whole range (empty to full) but the intermediate readings will be meaningless.
When I get a suitable resistor I'll try it. Maybe better than nothing.
But perhaps I've missed something ?

Location: A little East of Sandy

Roger, if you want to check your gauge, between the "B" terminal and the "T" terminal it should read 17 ohms (all external wiring disconnected).
Between the earth terminal and the "T" terminal it should read 25 ohms (again, all external wiring disconnected).
If they have been subjected to long periods of damp there is a tendency for the windings of one or both of those 2 coils to suffer a breakdown in insulation and read lower values.

Once we know what these readings are, maybe we can figure out a stratagem to overcome the error?

Ian Mc.

Location: Shropshire

Thanks for that info Ian. I'll check those resistances over the weekend and report back.
So far I've only measured the B terminal to earth to check the value of the resistor needed for 12v operation. That was 30 ohm as I believe it should be.

Location: A little East of Sandy

One of the greater mysteries in this life is how you can have a sliding electrical contact in very close association with the petrol vapour of a tank. In some circumstances of age and wear sparking must occur, and the shaft hole can become enlarged, but not well known as a source of interesting developments.

Any explanations?

Bob Culver

Location: Auckland

Bob..Perhaps because in a low voltage, low current circuit where the impedance is entirely resistive no sparking occurs ?

Location: A little East of Sandy

Location: Shropshire

Agree with Stuart. No explosions as it is all fuel vapour, with very little air.And only a fifth of air is oxygen.

Location: United Kingdom

Hi Stuart and Andrew

Yes, I have heard that theory, but it is not hard to envisage exceptions; a tank run dry and the ignition switched on hours after, ignition on with cap off on a windy day, lost in motion etc. Perhaps when putting cars back on the road should wait a while after filling the tank. Maybe it is why we used to be told to switch off whilst filling!
I have read that the minimum energy to fire an ideal mixture is a very tiny fraction of that in an ignition spark(even a Seven magneto one!)

On reflection I suppose it is equally odd that static sparks do not cause more trouble.

Bob Culver

Location: Auckland

Petrol is only ignited by a spark if the air/fuel mixture is almost exactly 11:1 by weight. This goes some way to explain why petrol engined cars can be so difficult to start unless everything is exactly right. Even with a strong spark unless the mixture is very close to 11:1 it won't go.
It also is the reason why carburation was such a difficult problem in the early days of motoring, with a lot of different solutions attempted to keep the mixture within the parameters needed. The big problem being that air is compressible and petrol isn't. Unless the fuel feed is closely compensated there are variations in mixture depending on the vacuum in the inlet tract. As vacuum increases the density of the air decreases but the not the fuel, causing the mixture to weaken. If it strays from 11:1 the engine dies. This happens especially on throttle opening, hence compensating jets and pump jets. This is all now largely overcome by electronics and accurately metered fuel injection. Mass flow meters measure the weight of air going into the engine and can therefore allow the ECU to adjust the mixture dynamically.
It is widely thought that petrol is the major cause of car fires, but other fluids are much more dangerous, especially non-synthetic brake fluid. Petrol spilled on a hot exhaust evaporates and disappears. It does not ignite unless there is a spark and the mixture is right. Brake fluid sticks and it's flash point is low so it bursts into flame and mixture is irrelevant. This is often the cause of fires resulting from crashes, especially on cars where the brake fluid reservoirs were not positively fixed. Engine oil can do the same.
If you have a car that has the brake fluid reservoir merely pushed into the top of the master cylinder or servo then put some cable ties round it. It can dislodge in an accident a cover the engine in brake fluid.
These days a fuel pump is usually located within the fuel tank and this demonstrates just how unlikely it is that a fuel sender unit could cause an explosion. A motor must be more more likely to cause sparks.
We just don't hear of fuel tanks exploding because of sender units or fuel pumps which almost all cars have.

Location: Melrose, Scottish Borders

I think that the general public, the majority of whom know nothing of critical gas/oxygen ratios, get a misleading impression of the likely hood of explosions as a result of watching TV dramas and especially films where invariably car crashes have an immediate dramatic explosion through the use of napalm or something !

Back to the case of the possibly faulty gauge. The reading I get are:
B to Gnd: 29.5 ohms (42)
T to Gnd: 12.5 ohms (25)
B to T: 16 ohms (16)

...Ian's figures in brackets.

So it would appear that there are shorted turns in the Tank section resistance ?

Location: A little East of Sandy

That would be my conclusion, Roger.
Obviously, the best solution would be a replacement, good, gauge. Not least because the failing winding might continue to degenerate negating any attempt to regularise the readings by tinkering with the tank units resistance.

Ian Mc.

Location: Shropshire

Thanks again for your info Ian. Eventually I will get a replacement gauge but I don't suppose that will be easy or cheap. Meantime I've found a 25 ohm resistor and connected it across the sender. I will soon learn to interpret the readings and I'll watch out for any further deterioration!

Location: A little East of Sandy

That was indeed, Roger to be my suggested plan "B" - if you can live with it.

A replacement gauge might be cheaper and easier than you think. Talk to people like Tony Betts at 7 County Austins.

Ian Mc.

Location: Shropshire

Hi Roger. I looked up the calibration for my 1936 gauge, which is of the G33 "Sender is 25 ohms when empty" type. The required sender resistances and corresponding sender currents with a 6v battery connected were as folllows: MAX is 0.3 ohms/297 mA, 5 gall is 1.6 ohms/247 mA, 4 gall is 3.4 ohms/206 mA, 3 gall is 7 ohms/155 mA, 2 gall is 9.3 ohms/134 mA, 1 gall is 14 ohms/105 mA, 0 gall is 20.7 ohms/80 mA and minimum is 26.8 ohms/66 mA. The gauge internals are remarkably ingenious, with the needle pulled in two opposite directions by a tug-of-war team comprising two coils with magnetic pole pieces. As the sender resistance drops, the current in the "T to ground" coil goes down and the current in the "B to T" coil goes up. I guess this ratio arrangement makes it immune to a certain extent to battery voltage (but not for use directly on 12v !). From memory the coil winding wire is silk or cotton insulated, so shorted turns are not impossible after nearly 80 years.

Location: New Forest

John. Thank you for those useful figures. It's too soon for me to know how the bodged original gauge will read after adding the 25 ohm resistor but early signs are that it will fall quickly from the "Full" and head for empty faster than it should. I can probably live with that kind of error as it's safer than the other way around.

However, being a heretic as far as total originality is concerned I've already bench tested a simple circuit that takes a steady 140ma and uses a modern 1 ma meter movement. I think It will fit inside the old case if it becomes necessary. With those resistance readings I can roughly predict the linearity when connected to the sender. Probably too much detail but I found a VU meter which displays upside down (ideal) and has a red section around what would be the low fuel end.

The original fuel gauge of that period is indeed ingenious. Like the equally ingenious cutout it was designed when electronics was in its infancy. The gauge copes with the range of voltages that a 6v battery exhibits from discharge to being charged. Roughly 6 to 7 volts but probably it betters that.

My replacement, if it ever gets used, uses a voltage stabiliser as I believe all post war cars did. I took the tiny pcb from inside an old mobile phone in-car charger (the size of a 10p piece)to get 7.5v. The car has to be 12v though to make use of those. With apologies to the mechanics for talking dirty
getting good with the captcha..

Location: A little East of Sandy