Multimeters - simple, possibly dumb question.

I am in the market for a new multimeter as I have decided to try and learn electronics. I think I will get a manual-ranged one as auto-ranging ones cost more and I see little benefit (but what do I know…). My question is, what generally happens if you select the wrong range? Back in the analogue days of my high school years if you connected it to 100V with the range set to 10V, the needle could fly across the scale with such force you could damage it and the physics teacher got mad. Can you damage a DMM somehow, or do they have protection built in?

While I have got your attention, any advice for what features are worth getting and which less so?

The most obviously useful:
[ul]
[li]DC voltage - duh[/li][li]AC voltage - ditto[/li][li]DC current - what range is useful?[/li][li]AC current - ditto[/li][li]Resistance - no brainer[/li][li]Audible continuity tester - has proven most useful even before I became properly interested.[/li][/ul]

and the less obvious:
[ul]
[li]Diode tester[/li][li]Capacitor tester[/li][li]Transistor tester[/li][li]Temperature - really?[/li][li]Frequency - sound like it should be useful, but how useful?[/li][li]Any good ones I have left off[/li][/ul]

Yes, you can damage one by selecting the wrong range, although I haven’t bought a high-end unit in quite a while and it’s possible they have breakers or fuses.

In practice, you’re most likely to damage a meter by setting it in current mode and then hooking it up in parallel to something. Damage in voltage mode is much less likely, even if you go outside the range, since voltmeters have very high (ideally infinite) internal resistance.

If you’re just trying to get into the electronics hobby, you really don’t need to worry about ranges, even a cheap one will have every range you will possibly need. When I get home I will post what model I have if you’d like a specific one that I know works good. But pretty much googling or ebaying “multimeter” and spending 50 bucks gets you one that will do more than you will ever need already.

Generally, the meter will just show -OL- or similar, and can safely handle up to 2,000 volts or so on the main volts/ohms inputs. At that point, take a good look at your test leads - chances are that they are not safe to 2,000 volts.

DMMs have gotten cheap as dirt - I picked up an auto-ranging unit for about $20 a couple of years ago. Just twist the dial to what you think you’re measuring (DC, AC or ohms) and put the leads on. The decimal point will then light up in the appropriate place, and either kV or mV will light, if you’re on either up to 1.999 kV or on millivolts.

The only thing you really need to be careful with is current. By selecting different jacks for the leads, you can read either 0-10 amps or 0-200 mA. The milliamp shunt is the one that you can blow pretty easily if you put it across voltage. The good (read: expensive) meters will have a circuit breaker for the current shunts. The cheap ones like mine probably cost less than a circuit breaker.

The one feature that hasn’t been mentioned, but can be very,very useful is a data output to a computer.

If you play around with a thermocouple or an RTD circuit and you want to calibrate your circuit to the real world, the temperature function can come in handy. Two easy temperature points that you can generate almost anywhere are your skin temperature and the temperature of a cup of ice water that is mostly ice.

I’ve used the frequency measuring capability of a meter at work where I deal with pulse input and output circuits. I’ve never had a reason to measure the frequency at home. I also have an oscilloscope at home though if I want to look at waveforms and frequencies.

Flukes are pretty good with fuse protection. El cheapo meters, well, not so much. YMMV (your meter may vary).

I’ve got a fairly high end meter at home, but most often I end up grabbing the little el-cheapo just because it is easier to use. You can get a simple, easy to use, auto ranging meter fairly cheaply these days. If you can find one where you don’t have to switch the lead jacks for current and voltage you have a much less chance of accidentally frying your meter.

I also have an old analog meter at home as well. There are some things you can tell by watching the needle swing that you can’t necessarily tell from a digital meter. If I only wanted one meter though it would be digital.

If you think you may be doing any troubleshooting of power inputs, a must have is MIN/MAX setting that will show the maximum or minimum voltage seen. That has been a real life saver for proving intermittent power issues.

It doesn’t even need to be mostly ice. As long as it’s at equilibrium (which it’ll reach quickly), any amount of ice will work.

When one of the junior engineers that works for me wanted to check out a thermocouple module to tell if it was reading correctly, I told her to grab the thermocouple with her hand to measure skin temperature then dunk the thermocouple in ice water to measure something close to 0 deg C. She stuck two ice cubes in a cup of water and wondered why she wasn’t getting anywhere close to 0 deg C. So now I always tell people make sure it’s mostly ice. :stuck_out_tongue:

It is often very nice to have a bar graph along the bottom of the display, so you can see transient things that are too fast to read numbers.

I like and have used all the obvious and less obvious features you mention, with the exception of frequency. I thought it was a great thing to include, but have yet to find a use for it. I’m curious who has, and what they use it for.

Here are the things I often find weak or lacking in a DMM:

Limited high resistance, often 20 megohms. If you want to verify proper lead resistance in RTDs for temperature measurement you have to go above 100 megohms. This kind of problem occurs in other measurement of measurement situations too. I eventually bought a special meter that goes to 10^14 or so ohms, which is another fascinating story.

Limited low resistance and 2 wire resistance measurement. If you want good precision for resistances of hundreds of ohms or less, or you want to be able to measure typical wire resistances with any useful precision, you need 4 wire resistance measurement (sometimes called Kelvin wiring). I bought a special meter for this, too, that can measure 1 milliohm with a couple significant digits.

Input impedance too low on voltage inputs. If you are trying to measure high impedance sources this can ruin the measurement. At work I measure some small and weak signals that I must amplify first, and my 100 megohm input impedance amplifier loads the signal too much to use, so I found a 10^12 ohm input impedance amplifier. This was also an example of having too high a minimum range for voltage; I also need these amplifiers because I want a drift and noise floor of ten or a hundred nanovolts.

We primarily use 7.5 and 8.5 digit DMM’s in our lab. These have all the-bells-and-whistles, and I will almost always use one in lieu of a handheld DVM when working in the lab. Still, there are times when a handheld DVM comes in handy, such as when I need something very portable or when I just need to make a quick-and-dirty measurement.

Here are a couple features I wish were included on all handheld DVMs:

  1. The ability to decrease the input impedance when measuring voltage, say to 10 kΩ. This can come in handy when troubling shooting AC power systems, as it eliminates problems with “ghost voltages” that arise when there’s a leakage path.

  2. The ability to turn off autoranging. (I hate autoranging with a passion.)

100 MΩ lead resistance in RTDs?? Never heard of that. Are you talking about an isolation measurement, perchance? At any rate, I think the reason handheld DVM don’t typically measure above 20 MΩ is twofold: 1) It would require a higher voltage, which could have safety issues. 2) More complexity, such as specialized (shielded/guarded) cables.

Any 7.5 or 8.5 digit DMM will do 4W resistance measurements. I really don’t see a reason to expect a handheld DVM to do it.

I think most modern DMMs have10 GΩ of input impedance. You’ll need something like a electrometer if you need higher impedance.

As much as the lab guys love their serious kit, back in the hobbyist world, especially when it is your own money, things are a bit more mundane.

Most of the DMMs add features because it is easy. Indeed many features are almost free, with the cost of adding the switch on the front probably exceeding the actual implementation.

Anyway, here is a hobbyist’s viewpoint. I’m not an EE, but i do do a lot of hobbyist stuff, mostly building HiFi, and general fun stuff, plus fixing things. I have been through a few DMMs over the years.

There are a lot of DMMs on the market. From the usual Flukes, right down to astonishingly cheap and nasty no-name stuff (typically rebranded by a chain reseller.) Just on features it is hard to work out where the money goes, there can be a ten to one difference in price for the same apparent feature set. But as usual you do eventually get what you pay for.

So, things I like.

A bar graph
True RMS AC reading
10A current
capacitance tester
inductance tester (however these are usually of rather limited utility on DMMs having a poor range)
peak hold
diode test
transistor test with Hfe readings (sort of useful for matching, but not as useful as a curve tracer.)

And some must haves:
Good leads - this means silicone insulated with proper shrouding. This is a serious safety issue. You may simply need to buy proper leads.

Safety rating of the meter. Cheap no-name meters will not be as well made, and in the event of some rare, but dangerous, fault conditions fail to properly protect you. Meters come in “Category” ratings. If the meter has no Cat rating, don’t buy it. Personally I would look for a Cat II rated meter for general messing about with. A Cat I is OK for general use, but a hobbyist will occasionally have need to look at things that are mains powered and live, and at this point Cat II makes sense. (Cat III and Cat IV are for serious, professional, work, if you need one for this you will already know.)

This is actually somewhat useful, as it measures the voltage drop across a diode at a given DC current (around 1 mA or so). As with any measurement, however, you may get erroneous results when trying to measure a diode in-circuit.

Not all that helpful for troubleshooting. Generally speaking, capacitors don’t fail simply because their capacitance is off. They usually fail because they short, open, have too much DC leakage, or an ESR that’s way too high. However, if your goal is to simply measure the capacitance of a capacitor, then I suppose this function can be useful. (Though there are much better instruments for measuring capacitance and inductance.)

Not useful.

Yea, you can connect a thermocouple or thermistor to some DVMs. I consider this to be a nice feature.

I rarely have a need for a portable frequency measuring device; when I measure frequency, I am almost always in the lab. Therefore I’ll use a laboratory grade (e.g. Agilent) frequency counter or digital oscilloscope for measuring frequency.

It’s a nice feature when doing troubleshooting, and I’ve used it quite a bit when ringing out cables and connectors.

I still have a VOM I bought at Radio Shack in the early 80’s. Does everything I need.

They haven’t changed much except for the digital displays.

Quoth Napier:

How do you use that? I’d think that at those resistances, the air or whatever’s holding the leads in place would effectively be a “short”.

The Keithley 6517A can measure up to 10[sup]16[/sup] ohms.

A modern high resistance meter uses a feedback ammeter to make the measurement. A special cable that has a “guard” conductor must also be used. More info here:

Yo, Crafter_Man, nice to see you again.

I said that incompletely. What I care about is the resistance from the leads to ground or cable sheath. So, yes, how isolated the leads are from the sheath. I think I usually hear this called “lead shunt resistance” implying shunting to ground, not just “lead resistance” as I said.
And, by the way, if I understood the signs along the way, the higher voltage thing is a red herring. One often tests insulation with a high range ohmeter, and one often cares about high voltage breakdown, so the two are often conflated, as in a megger. But they don’t have to be. If your interest is in the tiny error that a shunt to ground causes in, say, precision platinum resistance thermometry, then you want to measure that shunt at a low voltage. Creating a spark trail through some insulation porosity is testing the wrong thing about your leads - and possibly spoiling them, too.
But, hey, you’re my inside man when it comes to SPRT and standard resistor questions. How do you hear this problem referred to? And, when you check this isolation, you don’t use a megger, do you?

Napier:

You’re correct that you can use a relatively low voltage (say 10 V) to measure high resistance. And for your application that’s what I’d probably do, since isolation resistance is usually a function of voltage, and you would be more interested in the resistance near your operating voltage vs. 1000 V, for example. However, if a person is trying to evaluate the integrity of ground isolation, they usually measure the resistance at a higher voltage. We use 500 V. (Just about anything that is supposed to be electrically isolated from ground should easily be able to hold off 500 V.) A resistance above 10[sup]8[/sup] Ω 10[sup]9[/sup] Ω at 500 V is generally considered O.K. For most things I usually see it in the 10[sup]11[/sup] Ω or 10[sup]12[/sup] Ω range.

At work I use a HP 4329A high resistance meter to make the measurement. It’s an old unit, but it still works well. (And it is easier to use than our newer HP 4349A.) When performing a standard isolation test, I start at 10 V and step my way up to 500 V. In other words I set the voltage to 10 V and measure the resistance, set the voltage to 25 V and measure the resistance, set the voltage to 50 V and measure the resistance, set the voltage to 100 V and measure the resistance, set the voltage to 250 V and measure the resistance, and then finally set the voltage to 500 V and measure the resistance. If I see what I consider to be an abnormally low resistance reading at any voltage, I will assume there is a problem with the DUT’s insulation and will halt the test. If I am successfully able to get to 500 V while maintaining a “good” high resistance reading at each voltage - which occurs about 98% of the time - then I will assume the unit is well isolated from ground.

Recently we received 11 PRTs removed from an aircraft. We subjected them to many electrical tests, one of them being the aforementioned isolation test. Each passed with no problem at 500 V. I think the isolation resistance was around 10[sup]12[/sup] Ω for each PRT.