Analog meters - Getting the most out of your meter - Embedded.com

Analog meters – Getting the most out of your meter

Editor's note: If you're like us and can't bear to part with that crusty old (but magnificent) analog meter, David Ashton offers everything you want to know about these gems in this three-part series:


Testing meters
If you buy old meters or take them out of old equipment, you can’t assume that the markings on the scale on the meter is what it takes to actually deflect it. Max and I were discussing one he purchased that had a scale of 0-2000 V, yet had no internal resistors to make it indicate that — the multiplier resistance must have been external. So you’d need to measure it to find out the full scale deflection (FSD), and if possible its resistance as well, though this is not essential.

The easiest way to do this is with a variable power supply. Let’s say you have at your disposal a 0-20V DC power supply. The smallest current a meter will measure is likely to be 50 µA. To get a current of 50 µA at 10V DC you’ll need a resistance of 200 KO. So hook up your variable power supply in series with the meter and a 220 KO or 180 KO resistor, and your best DC current DMM like this:

Slowly turn up the voltage on the power supply and see what the meter does. If the meter does not deflect to FSD, reduce the resistor by a decade (to 22 KO) or even to 2 KO or 220 O if the initial deflection was very small. Once you get it to FSD, read the current off your DMM.

If you don’t have a variable supply, use for example a 5V supply and a 100 KO potentiometer in series with a 10 KO resistor, and then a 10K pot +1K resistor and so on. Make sure the pot is initially at maximum resistance, and decrease the resistance until you get FSD. This is basically what Max has been doing in his recent blog Oh, No! My Antique Analog Meter Has Twitched Its Last. He made up a nifty test box with 3 pots and a polarity reversal switch, driving it from his Arduino with PWM at full duty cycle. Although this does not tell Max directly what his meter’s FSD is, it tells him what series (multiplier) resistor he needs to get FSD — and seeing as he only wants to drive his meters from Arduinos, this is precisely what he needs to know.  For more general purposes, you could make the same sort of thing using a couple of sockets or clips to connect your DVM(s) as shown above.

You now know your meter’s FSD current. It’s also handy to know the meter’s resistance and/or the voltage across it at FSD — it will simplify working out shunts and multipliers later. Remove your DMM from the meter current path (or use a second one — all engineers have got at least two DMMs, right?), set it to voltage and measure the voltage across the meter at FSD — it will usually be less than a volt. From this you can work out the meter resistance (RMETER = VFSD / IFSD ).

A word of caution here. Don’t just stick your ohm-meter across a meter. If it’s a sensitive meter, the measuring current may be an order of magnitude or two greater than the meter is designed for, and you may bend the needle round the end-stop. If you know the meter sensitivity is 1 mA or more, by all means use a multimeter to measure it, you’re not likely to do much damage.

Protecting your meter
Meters are occasionallysubject to overloads and especially in the case of sensitive moving coiltypes, these may damage them. As noted above, most meters will haveunder 200mV across their terminals at FSD. So germanium diodes (whichstart conducting at around 200 mV) are ideal for protecting sensitivemeters but make sure the FSD voltage is less than 200mV! Put a couple ofdiodes back to back across the meter, and it will not get more thanabout a 50% overload, which it should cope with. Germanium diodes mayleak slightly, which will have the effect of shunting the meter, so makesure your calibration is still correct.

When you are dealing with current meters and shunts, the diodesshould be able to take more than the full shunt current. Schottkydiodes, with their low voltage drop, can also be used for this, and theyare available in higher current ratings. You can also make your shuntresistance a bit higher and put a resistor in series with your meter tolimit that current so you don’t have to get high current diodes.  Becareful if you’re using pulses to drive your meter: The diodes may limitthe pulses if the duty cycle is low and they are greater than the usualFSD voltage of the meter. If you are using 100% duty cycle = FSD thenyou don’t need to worry.

There are those who say that you shouldput a backwards diode across your meter especially if it is being drivenwith pulses to limit back EMF, as you do with relays. I’ve never foundthis necessary. There is a resistance in series with the meter usually,and the inductance of a meter and the current in it are usually muchsmaller than those of a relay coil. But it would not do any harm.

Lastly,if you ever have to ship a meter, especially a moving-coil one, or takeit places where it may get jarred or jolted, you should short out themeter terminals — and by this I mean the meter coil connections, notafter any resistors that may be in circuit. Doing this shorts out anyback EMF generated when the coil moves, and this tends to oppose anymotion and keep the coil from moving too much if it is jolted. If youdesign a meter into any equipment, consider putting a shorting switch onthe connections to the movement in the “off” position for this reason.Most analog multimeters have an “off” position which does just this —and you should use it when the meter is not in use.

Size does matter
Meterscome in all shapes and sizes. In the old days they used to be round. Ifind they look very dated these days, but Max with his steampunktendencies loves them. Here are some he bought at the Huntsville Hamfest recently (along with some other stuff…. and my wife says I’m a magpie… 🙂

Thesedays, meters tend to be square or rectangular, though there is a lot ofvariation. The usual needle deflection is 90º, though types with 110ºdeflection are common and you get them up to 270º. Here’s one which isprobably about 240º.

Thisis a moving coil meter with a rectifier. The non-linearity of the scaleabove 300 is probably due to the moving coil getting outside themagnetic pole-pieces. Note also the 300/5A mark in the bottom left —this indicates that this meter is used with a current transformer, andits actual FSD would be 5 Amps. I see a lot of meters like this inelectrical substation instrumentation.

How did I know this was a moving coil meter? From the symbols at the bottom below the 300/5A marking. From left, these mean

  • AC only (An AC/DC meter will have the squiggle with a line above or below it)
  • Moving coil, with rectifier (see below)
  • 1.5% accuracy
  • For use in the vertical position (the upside down T shows this — if it is to be used horizontally it will have a symbol like a table)
  • The star indicates the class or high voltage specification — a 2 inside the star indicates it can take 2 KV for 1 minute, a 3 indicates 3 KV (between the meter terminals and the grounded meter case).

Thesymbols are according to certain German DIN standards, and they arevery difficult to find information on. The below is the best I could getfor the meter types. Have a look at the moving iron 5A meter face (in part 1) andthat on the center-zero meter (in part 2) and compare the symbols on those.

Some of these symbols seem to be universal. Here is a face from a Simpson meter (probably made in the USA) from one of Max’s recent articles.

Notethe use of the moving coil symbol as above. Note also the writing“ES=10 DCV” at right bottom. As this meter was scaled 0-65 Kilovolts AC,obviously the multiplier, rectifier and/or other scaling circuitry wasexternal. I’d say this indicates that the meter FSD was 10 V DC.

Youcan also get edge reading meters, where the movement sits on its sideand the needle is bent through 90º to read on a curved scale like this:

Theysave a bit of space. There is an old microwave comms site I have beeninto with tons of these meters. I’d love to get my hands on them. Youget some very small meters like this but they are not good for much morethan tuning indicators or battery test meters.  

Most modernmeters are square or rectangular, and there are some standards. Themeter cases are usually rated by the diameter of the cylindricalprotrusion at the rear that houses the movement. So you get MU-38,MU-45, MU-52 and MU-65 types, though I’m not sure if these sizes areused in the US as well. MU-45 are probably the most common. Here are themain dimensions for each size:

 

Meter A B C D E F G
MU-38 38 45.3 50.5 31.8 31.8 19.5 4
MU-45 45 52.5 58.8 37.5 37.5 19.5 4
MU-52 52 60 64.5 48 48 25 4
MU-52E 52 66.2 80.2 63.7 48 28.4 4
MU-65 65 82.5 101 79.5 63.8 33.9 4

Alldimensions in mm, but there seems to be a bit of interpretation ofthese standards. For example the fixing bolt size (G) is given as 4mmbut I have definitely seen other sizes used. There is a chart of many sizes with dimensions here but I guarantee you will find others!

Mounting
Tofix your meter onto your case, you need to get the dimensions of thecase as above. For the main cylindrical part you can use a hole saw,chassis punch or a nibbling cutter (see my blog How to Make Holes in Things).You then need to make holes for the mounting bolts. You can put themeter in the hole and then roughly position the meter, rotating it backand forth slightly so the bolts leave marks on your case. Use a squareto mark the holes and then center-punch and drill them. Make sure youkeep any metal shavings, especially steel or other magnetic ones, awayfrom the meter — if they get inside your meter they can ruin it.

Meterterminals are usually substantial and quite able to hold a small PCB,so you can often dispense with other PCB mounting arrangements if youdesign your PCB with the appropriate hole spacing. There is an exampleof this below.

click for larger version

Myhomebuild 40-year old capacitance meter, showing the board mounted onthe meter terminals (by the two screws in the middle of the board). Thiswas stripboard, not PCB, and in retrospect I’m not too proud of theconstruction, but it cost me next to nothing and has worked fine for 40years! But for a fairly simple circuit like this, mounting the board onthe meter saves a lot of other work.

Rescaling
Ifyou adapt a meter and the scale on the faceplate is not what you need,you have a few choices. The Dymo-type tape labellers produce a veryacceptable text quality and you can relabel the scale using numeralsproduced on these labellers, cut with a fine knife or pair of scissorsand carefully positioned. This does not look really professional butworks fine, especially if you are just changing mA to V, or changing0-1-2-3-4-5 to 0-5-10-15-20-25. But if you have a meter scaled 0-1-2-3and you want to change it to one of the above scales, or to (say) a dBscale, you have a problem — the major divisions won’t line up. You canthen print a new scale as a label on a laser printer, or have it made asa plastic sticker (preferable, as paper discolors quite quickly) orhave a new scale professionally printed — economic if you’re doing it inquantity. Max and a few commenters discussed these questions further inhis blogs Creating New Faceplates for Antique Analog Meters and Analog Meter Faceplate Solution for Vetinari Clock.

Driving your meter
Usuallyyour meter will be bought or adapted for a specific measurement and youwill not have to make special efforts to get it to read what you want.Or your schematic will specify the sensitivity of the meter required.When designing yourself, you may design a circuit around a meter thatyou have. Bear in mind here that meters will not usually indicatechanges at more than about 10 or 20 Hz, so you can use a pulsed signalto drive them. I did this with the capacitance meter shown above that Ibuilt when I was about 18. It was pretty simple — a 74123 and a 555which presented a pulsed DC to the meter, the pulse width depending onthe capacitance. I still have it today and though it looks very crude,it still works fine and has not bad accuracy. A good rule to follow whendesigning is to make sure that the maximum current through the meter —even under fault conditions — is not more than around 1.5 times the FSD —any more may damage the meter. The lower the better.

So if youare using an MCU to drive your meter you have a couple of options: Ifyou have a DAC port you can use that, or you can use a PWM port. Makesure it has enough drive capacity or buffer it to be sure. Arrange it sothat when your port is putting out full voltage, or 100% modulation,the meter reads full scale. You should then be able to drive it to 1%resolution or better, and you’ll be doing well if you can readdifferences less than that. Again, this is getting off topic, and Maxhas done more on this in his recent articles, so that is enough about this here.

Whatare your experiences with analog meters? Do you love ‘em, hate ‘em?Have you any stories about analog meters? Please comment below.

Be sure to read the other installments in this series:


Join over 2,000 technical professionals and embedded systems hardware, software, and firmware developers at ESC BostonMay 6-7, 2015, and learn about the latest techniques and tips forreducing time, cost, and complexity in the development process.

Passes for the ESC Boston 2015 Technical Conferenceare available at the conference's official site, with discountedadvance pricing until May 1, 2015. Make sure to follow updates about ESCBoston's other talks, programs, and announcements via the Destination ESC blog on Embedded.com and social media accounts Twitter, Facebook, LinkedIn, and Google+.

The Embedded Systems Conference, EE Times, and Embedded.com are owned by UBM Canon.

4 thoughts on “Analog meters – Getting the most out of your meter

  1. “Davidnn “So germanium diodes (which start conducting at around 200 mV) are ideal for protecting sensitive meters but make sure the FSD voltage is less than 200mV! Put a couple of diodes back to back across the meter, and it will not get more than about

    Log in to Reply
  2. “Davidnn”I'd guess someone would still make them in one form or another. “nnI started poking around. I discovered Digikey doesn't uniquely recognize the word “germanium” and returns Schottky diodes. I did discover at least one- the 1N34A http://w

    Log in to Reply
  3. “@Antedeluvian… “Can you still get germanium diodes…?” I really don't know! I have tons of them around, removed from old computer boards many years ago. Maybe I should sell them on EBay as “meter protection diodes”… nnI'd guess someone would

    Log in to Reply
  4. “Nice work Aubrey! 50mA average forward current ain't bad. There is some (fairly skimpy) data on some of the OA types herennhttp://pdf1.alldatasheet.com/datasheet-pdf/view/166123/ETC1/OA91.htmlnnBut really, with meters, you are interested in l

    Log in to Reply

Leave a Reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.