How to test for bad earth?

[anthonym]

Worky no-worky is easy, but poor worky?

anthony

[Jonathan Kay]

How many circuits are suspect?

And how many points of access to the suspect circuits do you have?

...

And can you follow revilla's *demystification article? That's on the feed side but some of the electrical and diagnostic concepts are similar.

Jonathan

* Or first demystification article, as I hope.

[revilla]

Poor earths (or any other connections for that matter) come in two flavours. Sometimes they have a higher electrical resistance than they should, usually due to corrosion forming insulating layers in joints. Sometimes they are intermittent, working fine one minute and not the next, usually where something is physically loose and moving due to vibrations. The second one is hard to test for, as it will pass any test you throw at it one minute and fail the next. With increased resistance, the problem is that in a lot of circuits, the current is high enough that what constitutes a "high" resistance may still be too low to sensibly measure with a multimeter. Ohm's law says that the voltage drop across a circuit is given by the resistance of the circuit multiplied by the current flowing through the circuit. So a fairly low resistance can still give a significant voltage drop if the current is high enough. But that's the key to testing, look for the voltage drop induced when the circuit is under load. For a simple wired connection you would normally expect to drop only a fraction of a volt when the circuit is passing the maximum current expected of it. In the case of your fuel pump, going by what we saw in John Vine's car he was dropping about 0.15V measured between ground terminal at the pump and the battery negative terminal (or a more convenient good ground point that isn't passing current). That seems about right. If the voltage across these points is a lot higher than that, you may have a problem.

[anthonym]

Thank you - I may be some time digesting

[anthonym]

Ok here is what my Autel PS100 probe instructions say about:

Checking for Bad Ground Contacts

Probe the suspected ground wire with the probe tip. 

Observe the green LED. Depress the power switch forward then release. If the green LED went out and the red LED came on, this is not a true ground.

If the circuit breaker tripped, this circuit is more than likely a good ground. Keep in mind that high current components such as starter motors will also trip the circuit breaker. (From pages 16/17)

I haven’t tried it yet. 

[anthonym]

And about voltage drops: 

text starts (same instruction book as above) page 18

Red/Green Polarity LED

The Red/Green Polarity LED lights up when the probe tip voltage matches the battery voltage within +/- 0.8 volts. It is added information that could be valuable to the techncian.

If the circuit you are testing is not within +/- 0.8 volt of supply voltage, you will see the voltage reading on the LCD but you will not hear a tone or see a red or green LED. This tells you either you have a voltage drop in excess of 0.8 volt from battery voltage or you are probing a circuit than has an increase of 0.8 volt or more over battery voltage.

To determine battery voltage, simply remove the tip from the circuit and press the power switch forward. Battery voltage will then be displayed on the LCD. The difference between the battery voltage and what is read on the circuit is either voltage drop or voltage increase. This allows you to determine a voltage drop without running back to the battery. It is just another one of time saving feature the tool has.

end

Ok so the English could be better. Green means an earth connection, red means a supply connection. This all sounds to me like if I probe an earth or supply connection, it will tell me the voltage drop immediately.

[anthonym]

sooo... a bad earth will show a voltage drop ? Presumably the circuit must be "live"/switched on. So with the probe I would be measuring actual voltage drops (or not) as opposed to measures of resistance. 

I found the book online https://www.auteltech.com/u/cms/www/201209/07015346yrdh.pdf

[revilla]

Anthony,

It sounds from your description as though that is for detecting voltage drops on the supply side but not the earth side. You're not looking for a supply side voltage dropping below a certain threshold below the battery voltage, you're looking for an earth side voltage rising above a certain threshold above the battery earth voltage.

Hope you don't mind me saying so but my honest opinion is: You would be better off with cheap and simply multimeter and learning to understand what it's telling you than using a probe that tries to do the thinking for you. Once you get the hang of it, it's not complicated and you will understand a lot more of what is going on.

People on here will be happy to help, advise and explain.

Testing a good earth by deliberately tripping a circuit breaker doesn't sound like a sound way to test anything to me.

Andrew

[anthonym]

It detects both ways. Can do multi meter functions too. It is saying earth side voltage drop compared to battery voltage : isn’t that what we are looking for? This gives the actual voltage drop without doing calculations. 

However, it seems the tool is not so well known outside garages. My Ford agent head of the workshop introduced me to it. 

Anthony

[revilla]

You don't need to do any calculations to find the problem with a £5 multimeter. I'll try to draw a diagram on a piece of paper that will show you what I mean. Give me a bit to think about the simplest way to explain it. I'll probably have to start off with some maths and calculations but you'll see that they all drop out in the end and what comes out of it is really simple.

[Ivaan]

Thanks guys, this is really useful stuff for me too.

 

[rj]

I'll second Revillas thoughts on probes vs multimeters. 

Gear that make the conclusions for you rarely come to the right.

[revilla]

OK so let's get the mathematical concepts sorted and out of the way first.

In this diagram I've drawn a rectangular symbol that I'll use to represent an electrical resistance. Any electrical component, connection, cable etc. has some electrical resistance. It doesn't pass currently perfectly easily, it takes a little bit of voltage to push current through it. The unit of resistance is the "ohm". It takes one volt to push a current of one amp through a resistance of one ohm. Extending this a little bit we get what is called Ohm's Law:

The voltage across any piece of a circuit equals the current flowing through it, multiplied by its resistance.

This is a simple relationship but you can think about it in different ways, and it has some important consequences.

• You can either think of putting a voltage across a circuit and this drives a current through it.
• You can also think of it as passing a current through a circuit and getting a voltage drop (i.e. a difference in voltage between the two ends).
• They're just two ways of thinking of the same thing, and the voltage will be the current multiplied by the resistance in both cases.
• Since anything multiplied by zero is zero, if there's no current flowing an a piece of a circuit, there won't be any voltage across it, i.e. the voltage at both ends will be the same.

OK so now let's look a simplified version of the fuel pump circuit (it could be any other circuit such as headlights etc.).

So we have a battery driving the fuel pump. The whole supply side of the circuit I've lumped together into one piece. This will include the fuse, relay, wiring, joints in the wiring etc. But for now we'll just think of it as "the supply wiring" and let's imagine that it is healthy and has a very low resistance of 0.01 ohms (by the way, that's far too low to read directly with a multimeter). I've done the same for the the earth wiring. In reality this will be made up of the fuel pump plug, a wire running up to a joint somewhere by the gearbox, a wire from there up to the wiper motor bolt, then the chassis itself across to the wire that hops to the engine block, through the engine block and the wire back to the battery. But for now we can just think of that whole piece of the circuit as "the earth wiring".

Two more important concepts to introduce here:

• In a closed piece of circuit (i.e. with no other external connections other than the two ends), the current is the same everywhere. Think of it like water flowing in a hosepipe. So long as there are no leaks (it's closed), the amount of water going in one end is the same as the amount of the water flowing through the middle and the amount of water coming out of the other end. 
• If you add up the voltages across pieces of a circuit going from one point (call it A) and another point (call it B), you will get the same answer whatever route you follow through the circuit. So in the circuit above, we could take two routes from A to B. We could either go across the battery, and in this case we know the voltage will drop by (for the sake of argument) 12.5V. Or we could go the long way around through the wiring and fuel pump. This rule tells us that the voltages across each of the pieces of the long route round the circuit must also add up to 12.5V.

OK so now I need to do the last bit of maths to explain this, and then we'll get onto the simple answer that comes out of it:

We assumed that the supply wiring and earth wiring had a resistance of 0.01 ohms each. Let's also assume that the fuel pump has a resistance of 1 ohm. So the total resistance around the circuit is (0.01 + 0.01 + 1) = 1.02 ohms. We can apply Ohm's Law to the whole circuit to find out the current flowing, which is voltage divided by resistance, so 12.5 / 1.02 = 12.25 amps.

We also know that this current must be the same at each point around the circuit, so we can apply Ohm's Law to each section of the circuit to calculate the voltages (current multiplied by resistance in each case) we would see putting a multimeter across each bit of the circuit:

• Supply wiring, 12.25 * 0.01 = 0.12 volts.
• Fuel pump, 12.25 * 1 = 12.25 volts.
• Earth wiring 12.25 * 0.01 = 0.12 volts.

If you add these up, they do indeed add up to 12.5V (to the accuracy of the number of decimals I've worked it out to) as expected. All is healthy, the wiring is doing its job of delivering most of the battery voltage (12.5V) to the fuel pump (12.25V).

But imagine we repeated the exercise assuming that the earth wiring has a total resistance of 0.5 ohms, not 0.01 ohms. This still seems like a very low resistance, and the "continuity check" on a multimeter would almost certainly still show this as a good connection.

The total resistance would then be (0.01 + 0.5 + 1) = 1.51 ohms. The current flowing would then be 12.5 / 1.51 = 8.28 amps. The voltages would works out to be:

• Supply wiring, 8.28 * 0.01 = 0.08 volts.
• Fuel pump, 8.28 * 1 = 8.28 volts.
• Earth wiring 8.28 * 0.5 = 4.14 volts.

Again they add up to 12.5V. This time though there is a lot less current flowing (8.28 amps instead of 12.25 amps) and the voltage at the fuel pump is a lot less than the battery voltage (reduced from 12.25 volts to 8.28 volts). Clearly nowhere near as healthy.

BUT I promised you that you wouldn't need to do all of this maths to use the results. Any here's why ...

In each of the cases above, we can look at the way the resistance is split around the circuit and the way in which the battery voltage is split around the circuit:

In the first case:

• Supply wiring, 0.01 / 1.02 * 100 = 0.98% of the total resistance, 0.12 / 12.5 * 100 = 0.98% of the battery voltage.
• Fuel pump, 1 / 1.02 * 100 = 98% of the total resistance, 12.25 / 12.5 * 100 = 98% of the battery voltage.
• Earth wiring, 0.01 / 1.02 * 100 = 0.98% of the total resistance, 0.12 / 12.5 * 100 = 0.98% of the battery voltage.

In the second case:

• Supply wiring, 0.01 / 1.51 * 100 = 0.66% of the total resistance, 0.08 / 12.5 * 100 = 0.66% of the battery voltage.
• Fuel pump, 1 / 1.51 * 100 = 66% of the total resistance, 8.28 / 12.5 * 100 = 66% of the battery voltage.
• Earth wiring, 0.5 / 1.51 * 100 = 33% of the total resistance, 4.14 / 12.5 * 100 = 33% of the battery voltage.

So the percentage of the battery voltage dropped across any part of the circuit also tells you what percentage of the total resistance is in that part of the circuit.

This is the really useful result, because instead of trying to work out resistances, or doing any maths, or even thinking about resistances at all, we know that "bad bits of the circuit" will have large chunks of the battery voltage across them. All we need to do is find where the voltage is being dropped and that tells us directly where there is high resistance and a problem with the circuit.

In the first case above:

• Measuring with a multimeter between the battery positive terminal and the supply terminal of the fuel pump we would have read 0.12V, this is only a tiny proportion of the battery voltage voltage, around one hundredth of the total, so no major problem there.
• Measuring with a multimeter between the earth terminal of the fuel pump and the battery negative terminal we would also have read  0.12V and come to the same conclusion about the earth wiring.
• Most of the voltage is being delivered to the pump.

In the second case above:

• Measuring with a multimeter between the battery positive terminal and the supply terminal of the fuel pump we would have read 0.08V, this is only a tiny proportion of the battery voltage voltage, so no major problem there.
• Measuring with a multimeter between the earth terminal of the fuel pump and the battery negative terminal we would have read 4.14V, which is about a third of the battery voltage! There is clearly a problem here.
• Only two thirds of the voltage is being delivered to the pump.

So straight away, find the voltage drop, you've found the problem. In this case the 0.5 ohms we assumed was in the earth wiring somewhere has straightaway showed itself as over 4 volts being lost between the fuel pump earth terminal and the battery negative, which should ideally be at the same voltage.

Of course in reality the circuit is more complicated, as I've started to show below:

What I had previously drawn as "the supply wiring" actually has lots of components, such as the fuse, inertia switch, MFRU relay, lots of wires etc. What I had previously drawn as "the earth wiring" actually has lots of components too. But the same principle as above applies. All of the components form a closed loop, the current will be the same around the loop and the voltages around the loop will still add up to the battery voltage. You are still looking for say 98% of the voltage to reach the pump, and if there's somewhere bad in the circuit it will show up as a voltage across that bit of the circuit.

All you need to do is:

• Measure the battery voltage, and the voltage delivered across the pump, to see if there's a problem.
• Follow the circuit around, measuring the voltage across each part. The part where most of the voltage is lost is where most of the problem lies.

You can still treat big chunks of the circuit as single pieces as I did at the start. So if you measure the voltage across "the supply wiring" and you only find 0.12V, then there's no point in looking at the individual bits of the supply wiring as you know it is all healthy. Similarly if the measure the voltage across "the earth" wiring and find 4.14V, then you know the problem lies in there somewhere and you can start measuring the voltages across the individual bits.

One useful thing to notice. We can divide the "the earth" wiring into two convenient chunks. Firstly the wiring from the battery through the engine block to the chassis, and secondly the wiring from the chassis point at the wiper motor bolt down to the fuel pump. If, with the fuel pump running, there is only a tiny voltage between the battery negative and the chassis, then any earth fault must be somewhere between where the wiring is bolted to the chassis and the fuel pump.

It is pretty inconvenient to try to string a multimeter between the battery negative at the front of the car and the fuel pump earth terminal at the back. But if you look at the above diagram, if you know the fault lies between the chassis and the fuel pump, you can measure the voltage from any point on the chassis - there won't be any significant voltage differences between different points on the chassis as the currents flowing though it will be small in the context of the tiny tiny resistances of the metal components. And we won't be passing any significant current through the connection where the multimeter probe touches the chassis, so even if we don't get a perfect connection the voltage drop across it will be tiny.

One important point to note:

All of this is based on a closed loop of circuit. As soon as you disconnect anything, you no longer have a closed loop. This includes for example where you turn off the ignition switch, or the ECU turns off the relay in the MFRU. You need to do these tests with the circuit operating. So in this case, you need to measure all of these voltages with the fuel pump running, otherwise the results will be meaningless (there will be no current flowing and all of the resistance and voltage drop will be in the switch that is turned off!).

So to summarise: The simple way to find the points of high resistance in a circuit is just to look for where there a voltage drop around the circuit, with the circuit powered up and connected.

Anthony,

In your case, if you want to test the earth connection to your fuel pump, all you need to know is, with the engine running (so fuel pump operating), what is the voltage between the earth connection on the fuel pump and a decent convenient earth point on the chassis. Nothing more complicated than that. No maths to do. Whilst helping John Vine recently we measured this on his car and got around 0.15V. Use this as a guide. If you get something around this, all is normal. If you get less you're doing well. If you get significantly more, there's a problem to look at. My guess is all will be fine and you can get out and enjoy the car!

Hope that's useful.

Andrew

[anthonym]

Thank you; I will be some time. Does this belong in the tech wiki? Looks like a labour of love, so thanks again. 

I am surprised by the apparent anti-powerscanner reactions. It’s just another tool and served me well last week. Of course I have several multi meters of varying price, including Fluke. Including a £5 one in the car. Seems to me no tool can do one’s thinking and the more the merrier. Like my gearbox.

Can we not understand and use both? I am.

1972 I studied electronics, but it’s hard to remember now.

Anthony

[Jim 123]

Andrew, a really great explanation.

[revilla]

Anthony, yes of course you can use whatever tools you like. I think the point that people are making is that one general purpose multimeter is really all you need. If you learn to use it properly will tell you pretty much everything that the other tools will tell you. The difference is, when you learn how to use the multimeter you are learning general principles that can then be applied to test and diagnose any circuit. You don't need a specific tool for each specific purpose. There will always be a special case for which there isn't a special tool, but a multimeter and an understanding of the principles can be used everywhere.

I used to do a lot of electronics as a hobby. I've designed and built small computer systems (I mean from indicl usual chips, not the plug and play systems available today). I've diagnosed and repaired radios and TVs. Remote control systems, radio transmitters ... and lately I've reverse engineered the EU3 K loom to correct the wiring diagram, built a number of looms, analysed and documented a lot of the systems and circuits on the car and designed numerous fixes for numerous problems. For all of those diverse areas of electronics over the years taken fairly seriously, the only measuring tools I've ever owned and used (or needed) are a multimeter and for fast time-based circuits an oscilloscope (which in principle is just a multimeter that takes many readings per second).

There's no problem if you want to continue using your probe, but in terms of support in here most people will be able to help you in terms of basic voltage and current measurements as they are the terms in which most people work and understand and give you the underlying information needed to properly understand and diagnose what is going on.

[anthonym]

All sounds good to me. I was wondering about an oscilloscope. I’m not clear about difference between a multimeter and the powerscan. Aside from the 12v supply. Seems to me they do or offer much the same thing.

[anthonym]

So, istr ? my pump needs 8 amps. My battery is currently at 11.7 volts so circuit resistance greater than 

12.5 / 1.02 = 12.25 amps.

11.7 / R = 8 amps

11.8 / 8 = 1.475 ohms max resistance before pump is compromised if it needs 8 amps.

Sensitivity:

13.8 / 8 = 1.725

12.8 / 8 = 1.600 (0.125 less)

11.8 / 8 = 1.475 (0.125 less) etc

7.8 / 8 = 0.975 (0.75 down from 13.8 ) 

So with a healthy alternator there is a point where it will run if started, but can’t start if battery voltage is low.

Will insufficient current mean no pump operation, or slow operation (which could be worse judging by other comments in another thread). 

Anthony

 

 

 

[revilla]

I'll try to work through the numbers for you later. But in the meantime ... Your battery is 11.7V? What kind of battery have you got, it looks like a standard Banner in the photograph? If it's a normal lead acid battery, 11.7V is very very sick indeed! Even a battery which is to all intents and purposes flat as a pancake will normally read well over 12V when it's not being asked to drive anything.

[Ivaan]

Andrew. Thank you for a great clear explanation. This definitely needs to go somewhere for future reference.

Perhaps a Low Flying article ? 

Clive

[revilla]

Anthony,

I'm not sure I follow your numbers.

I think you may be over thinking this.

Firstly, if the pump normally operates at 8A which sounds entirely reasonable, it will have an effective resistance of about 1.7 ohms. This will result in it drawing about 8A at an operating voltage of around 13.5V, which is about where the supply voltage should be with the engine running and the alternator charging. ANY additional resistance in the circuit will increase the total resistance and therefore reduce the current, compromising the pump performance. There just isn't a threshold at which it starts to affect things as you describe.

I think you need to concentrate on what a wire's basic job is - to deliver the current required by the load without losing too much voltage along the way. If it's doing that, it's basically OK. If you're losing significant voltage along the way, resistance is too high and there's a problem to look at.

You need to concentrate on making sure that the wiring is right. I wouldn't worry about or try to think through what will happen at various different levels of "wrongness". Nothing in the system was designed to work with high resistance wiring, poor connections, reduced voltage at the pump etc. It was designed and tested to work properly when all the components are working properly. It's just guesswork at what point the pump will stop delivering enough pressure. You need to make sure everything is as it should be and then it will be fine.

As you say, running the fuel pump poorly is probably worse than not running it at all, the engine will probably run a low fuel pressure, lean mixture and end up detonating and doing untold damage. So don't worry about thresholds above and which it should be OK. You need to make sure that pretty much all of the battery voltage is available at the pump. 

So I would go back to the measurement I suggested before. Just check the voltage between the fuel pump earth terminal and a good convenient earth point on the chassis and make sure that it's a small fraction of a volt with the pump running and all will be OK. If the supply side is also OK now, measuring the voltage across the pump terminals should give you something pretty close to the battery voltage.

Speaking of battery voltage ...

State of Charge	Sealed or Flooded Lead Acid battery voltage	Gel battery voltage	AGM battery voltage
100%	12.70+	12.85+	12.80+
75%	12.40	12.65	12.60
50%	12.20	12.35	12.30
25%	12.00	12.00	12.00
0%	11.80	11.80	11.80

 

This table gives the approximate voltage you would expect to measure across the battery terminals when the battery is unloaded, i.e. disconnected, master switch off (or pretty close with just everything turned off). As you will see, 1.7V is beyond flat.

Can you measure the three voltages that Jonathan is fond of asking for:

• Battery voltage at rest.
• Battery voltage while cranking.
• Battery voltage at 3000rpm.

These will enable us to work out what is going on.

Andrew

[anthonym]

One of the things I am looking for is the reason why the pump failed to operate only after having engine stopped and then wishing to start again, whether after a short stop (an hour or minutes) or two days later. Every time has been after (apparently) faultless running for hours, and stop for however long and then fail to start (no pump priming). 

She's starting fine at the moment and all connections between the battery to the final loom connect (but not including that) are new, inertia switch and fusebox being cut out. I will replace the inertia switch (on order), but not the fusebox connection as that is now a short inline fuse connect.

Have to figure out how to measure at 3,000 

Anthony

[Jonathan Kay]

Have to figure out how to measure at 3,000 

Is the problem the lack of an assistant?

Jonathan

[anthonym]

Yes Banner. 11.7 was before I charged it again - bearing in mind constant starts and no miles, it's "ok" but no longer new and recently hammered with start failures. I have several new ones (acid not yet added) at home so no problem, maybe last season for this one. 

[anthonym]

@jk yes! one that can operate an R500 pedal.. maybe I can do something with longer wires.