Test Instruments & Measurement

Chapter 3 said your meter is an instrument with characteristics, not an oracle. This chapter is about which instrument, and how much of its answer you are entitled to believe.

Every measurement you take is a claim you are about to act on — that a conductor is dead, that a motor is drawing normal current, that an insulation fault is real. The instrument's specification is what backs the claim. Choosing the wrong instrument does not usually produce an obvious error; it produces a confident, stable, plausible number that is wrong, and you then reason carefully from it.

The instruments, and what each is for

| Instrument | The claim it backs | Where it misleads | | --- | --- | --- | | Two-pole voltage tester | This conductor is live / dead | No fine resolution; not for measuring a value | | Digital multimeter | This point is at this many volts | Ghost voltage on a high-impedance input | | Clamp meter | This conductor is carrying this many amps | Jaw around the wrong number of conductors | | Insulation tester | This winding or cable is sound to earth | Destroys electronics left connected | | Process / loop calibrator | This 4–20 mA loop reads what the PLC sees | Sourcing when you meant to measure | | Thermal camera | This joint is hotter than its neighbours | Load-dependent; emissivity; reflections | | Oscilloscope | This signal has this shape over time | Overkill for most faults, and slow to set up |

The column that matters is the third one. Knowing what an instrument cannot tell you is what stops you over-reading its answer.

CAT ratings: where you are standing, not what you are reading

This is the specification people misread most often, and the one with the worst consequences.

A CAT rating describes the transient overvoltage the instrument can survive at a given point in the installation. It is not a statement about the steady voltage it can display.

The reason locations differ is impedance. At a socket outlet there is a building's worth of cable between you and the transformer, and that cable limits how much energy a fault can deliver. At the incomer there is almost nothing in the way. The same nominal 400 V is a far more dangerous 400 V at the main panel than at a plug.

That produces the fact worth memorising:

A CAT III 600 V meter is the safer instrument in a control panel than a CAT II 1000 V meter. The bigger voltage number is the weaker instrument for that location. If you take one thing from this chapter, take that.

Proving dead: the right tool is not the multimeter

For proving a conductor dead — the Chapter 2 verification step — a two-pole voltage tester is the correct instrument, not a DMM.

A two-pole tester loads the circuit as it measures. That does two useful things: it collapses ghost voltage, so you do not chase induced readings on a dead cable, and it fails in a more honest direction. A DMM's ~10 MΩ input barely loads anything, which is exactly what makes it show 60 V on an isolated conductor.

A DMM is the right instrument when you need a value. A two-pole tester is the right instrument when you need a verdict. Proving dead is a verdict.

Clamp meters, and the reading that looks like a dead circuit

A clamp meter measures the magnetic field around a conductor, which means it does not have to break the circuit — the single most useful property of any instrument in this trade.

Two kinds exist and the difference decides whether you can use it at all:

• Current-transformer clamps read AC only. They are cheap, accurate and the common type.
• Hall-effect clamps read AC and DC. If you need DC current — a 24 V supply's load, a battery circuit, a drive's DC link — a CT clamp will read nothing at all and give you no warning that it cannot.

Other things the jaw cares about: centre the conductor, because accuracy falls off near the jaw tips; close it fully, because an air gap wrecks the reading; and at low currents on a large jaw, expect poor resolution — a jaw sized for 600 A is not the tool for 0.4 A. For very large or awkward conductors, a flexible coil that wraps rather than clamps is easier and usually more accurate than forcing a rigid jaw into place.

Insulation testers

An insulation tester applies a high DC voltage — typically 250, 500 or 1000 V — and measures the resulting leakage as a resistance, usually in megohms. It is how you find the winding or cable that is breaking down but has not shorted yet, which is the failure a voltage measurement cannot see.

Three rules carry most of the value:

Never on a live circuit, for the reasons Chapter 3 gave for ordinary resistance measurement, only more so.

Disconnect the electronics first. A drive, a PLC card, a soft starter or a power supply left connected will see 500 V DC across components rated for a fraction of that. This is one of the most expensive self-inflicted faults in maintenance, and the damage is often not immediate — the device fails a week later and nobody connects the two events.

Choose the test voltage for the equipment, not for the biggest number the instrument offers. Testing a 24 V circuit at 1000 V proves nothing useful and risks damage.

Interpretation is comparative more than absolute. A single reading of 200 MΩ means little on its own; the same motor reading 200 MΩ last quarter and 8 MΩ today is telling you something urgent. Trend it, and note the temperature and humidity, because both move the number.

Process and loop calibrators

For 4–20 mA instrumentation, a loop calibrator does two distinct jobs and confusing them is the classic error.

• Measure — read the loop current as it is, without disturbing it.
• Source / simulate — inject a current you choose, to prove what the receiving end does with it.

Sourcing 12 mA into a live control loop makes the PLC believe the tank is half full. If that value drives anything, it will act on it. Sourcing is a powerful diagnostic — it separates a field fault from a receiving-end fault in one move — but it is an intervention, and it belongs to Chapter 7 alongside the loops themselves.

Thermal cameras

Thermal imaging finds loose connections, overloaded conductors and failing bearings before they fail, and it does it without contact. Three cautions:

It is comparative. The useful reading is one phase against its neighbours, or one joint against an identical joint. Absolute temperatures depend on emissivity, and shiny metal — busbar, a clean terminal — has low emissivity and will under-read badly.

It reflects. A polished surface will show you the temperature of whatever is in front of it, including you.

It is load-dependent. A high-resistance joint dissipates power in proportion to current squared. Survey a line at 20 % load and the joint that will glow at full production looks unremarkable. Thermal surveys are done under real load or they are decoration.

Calibration, and what it is worth

A calibration certificate says the instrument was correct on a bench on a date. It says nothing about the drop it took last Tuesday, or the leads that have been trapped in a panel door.

So calibration status is necessary for a measurement anyone will rely on formally, and it is never sufficient. The live-dead-live check from Chapter 3 is what tells you the instrument is working now, and it takes twenty seconds.

Selecting and trusting an instrument, in the order that keeps you out of trouble.

1. Say what claim you need to make

   "Is this dead", "how much current is it drawing", "is this winding breaking down" and "what does the PLC see" are four different questions needing four different instruments. Naming the claim first stops you reaching for the meter in your pocket and then bending the question to fit it.

2. Check the CAT rating against where you will be standing

   Safety

   Not against the voltage you expect to read. Panel and motor circuits are CAT III; incomers and anything outdoors are CAT IV. Then check the leads and probes separately, because the assembly is only as good as its weakest part.

3. Inspect the instrument before you trust it

   Safety

   Leads for cracks, nicks and stiffness; probe tips for damage; the case for impact; the battery for life. A meter that has been dropped and looks fine is a meter with an unknown specification.

4. Prove it live-dead-live

   Safety

   On a known source, then the circuit, then the known source again. Twenty seconds, and it is the only evidence you will ever have that the instrument was working at the moment it told you a circuit was dead.

5. Set the mode and range deliberately, then look at it again

   Particularly after any current measurement — a meter left in amps and applied across a voltage is the single most destructive habit in this chapter. Confirm AC or DC, and confirm which jacks the leads are actually in.

6. Get the technique right, not just the instrument

   One conductor in the jaw, centred, closed fully. Probes on the terminals you intend and not bridging a neighbour. For insulation testing, everything sensitive disconnected first. Most bad readings are technique, not equipment.

7. Predict the number before you read it

   From Chapter 3: work out roughly what the value should be, then measure. A reading that matches your model confirms both. A reading that does not means either the circuit or your model is wrong — and either way you have learned something, which is not true of a number you had no expectation for.

8. Record it with its conditions

   The value, where it was taken, and what the machine was doing — load, speed, temperature. An insulation or thermal reading without its conditions cannot be compared with the next one, and comparison is where most of the value in these instruments lives.

• Choosing a meter by its voltage number

  Why

  1000 V looks like more capability than 600 V, and on the shelf it is more expensive, so it reads as the better instrument. But the CAT letter describes where it can survive and the voltage describes what it can display, and in a control panel the letter is the half that decides whether you are hurt. CAT III 600 V beats CAT II 1000 V there, every time.

• Proving dead with a multimeter

  Why

  It is in your hand, it reads volts, and it will show 0.0 V on a dead conductor — so it appears to do the job. What it also does is show 60 V on a dead conductor with induced voltage on it, and 0.0 V on a live one if a lead has failed. A two-pole tester loads the circuit and does neither.

• Leaving the meter in current mode

  Why

  The measurement before this one needed amps, and nothing about the meter's appearance objects when you then put the probes across a supply. In current mode the instrument is a short circuit between your two probe tips. At CAT III that is not a blown fuse.

• Insulation testing a circuit with the drive still connected

  Why

  The isolation is done, the machine is dead, and the test feels like the safe, thorough thing to do — which it is, for the cable. The drive on the other end sees 500 V DC across semiconductors that will not tolerate it. The reason this keeps happening is that the consequence is invisible today and arrives next week.

• Believing an absolute number from a thermal camera

  Why

  The display shows a temperature to one decimal place, which is a very convincing format. That number depends on emissivity, angle, distance, reflections and load. Compare like with like — this phase against the other two — and treat the absolute figure as an estimate.

• Treating a calibration sticker as proof the meter works

  Why

  The sticker is evidence about a bench on a date, and it is genuinely required for formal work. It is not evidence about the drop the meter took last week or the lead that is failing intermittently. Live-dead-live is what tells you about now.

Five questions. Each is answerable by reasoning about what an instrument can and cannot tell you, rather than by recalling a specification.