Fullscreen HTML5Med Res (??MB)HTML4

By Nicholas Vinen Using our Voltage/Resistance/ Current Reference for . . . Checking & calibrating multimeters Do you trust your favourite multimeter? Does it always tell the truth? Maybe it doesn’t! We tend to assume that the readings are accurate but are they? They can drift off calibration quite markedly as the years pass by. So how old is your multimeter and has it ever been checked? Now you can use our Voltage/Resistance/ Current Reference to check your multimeter and if necessary, precisely calibrate it on DC voltage, current and resistance. D EPENDING ON how much you paid for you multimeter, it may have been very accurate when new. But if it’s a few years old, its readings might now be far from accurate. How would you know? You need to check it regularly against a reference to ensure it still meets its specifications. Ideally, you need several references; at least one voltage reference, one siliconchip.com.au resistance reference and one current reference. Luckily, our Voltage/Resistance/Current Reference described elsewhere in this issue gives you all three, without you having to spend much money. Of course, professional multimeter calibration operations need to have more complex equipment than this; they will have very accurate refer- ences that are themselves periodically checked against even more accurate devices and which can be adjusted to give voltages/currents/resistances/etc that allow each range of a multimeter to be calibrated accurately. While our project only supplies a fraction of that, it’s still a lot better than nothing. So assuming you’ve built our Reference module, how do you go August 2015  45 If you don’t have a service manual for your multimeter, you will have to figure out which pot does what by a process of trial and error. about checking and calibrating your multimeter? Checking voltage accuracy First, if you have a very accurate multimeter that itself has been recently calibrated in a NATA-approved facility, you can use it to test your reference device and write down the measurements. If you don’t have such a meter, you will simply have to assume that the Reference is perfectly accurate. If using our reference device, this is not a bad assumption – we measured the nominal 2.5V output error (on a 7.5-digit multimeter) as just -0.0005% (see photo overleaf)! Similarly, the nominal 1kΩ resistor was just 0.0134% low (compared to a 0.1% specification). Now set the multimeter to be calibrated into DC voltage measurement mode and set the range to the lowest range that will read the test voltage (if it’s auto-ranging, it will select this automatically). Connect the probes to the OUT+ and OUT– terminals on the reference, switch it on and check the reading. If it is as close to the expected value as the meter can read, you 46  Silicon Chip know it’s properly calibrated. You can reverse the probes and check that the negative reading is equally accurate. You can also check that the reading is correct on higher settings, although the number of digits shown will of course be reduced so this will be a less accurate test. Still, it’s worth doing. Note that a typical DMM typically only has a single adjustment for its DC voltage mode so if it is out in some ranges and not others, you probably won’t be able to improve the situation without actually replacing some of its on-board multiplier resistors. For checking lower voltage ranges, where the output of the reference will give an over-range error, you could connect a resistive divider or potentiometer across the reference outputs, measure the resulting voltage on the higher range and then check that the lower range gives a similar reading. Resistance mode Checking the resistance reading is a similar process. Set the DMM on its lowest mode that can read 1kΩ (this will usually be the 2kΩ, 4kΩ or 5kΩ range) and check that the reading is as close as possible to the actual value. To check higher ranges, you could use the same resistor however it’s better to pick a “random” resistor which is just below the maximum you can read on the current range, note its value, then switch to the next higher range and verify that the reading is very close. You can then pick a resistor with 10x the value as the last and repeat the process up through the ranges. Ranges below 1kΩ can be checked using the same procedure, ie, pick a resistor with a value that’s towards the upper limit of the lower range (eg, 180Ω), read this value on the same range as you used to check the 1kΩ calibration resistor, then switch into the lower range and verify that the reading is correct. Then using a smaller value again, proceed down through the lower ranges. Of course, the ideal situation would be to have a precision resistance box or a series of individual precision resistors but in practice, this cheaper method should do the job. Ammeter checking Testing an ammeter with the current source on our reference board is a little more involved because it has a high output impedance of 1kΩ. That means that, depending on the multimeter’s range setting, its shunt resistance (and by implication, burden voltage) will affect the reading. However, you can easily compensate for this. The simplest method is to use a second multimeter to measure the shunt resistance of the meter being tested. The current is nominally 2.5mA for the 2.5V unit (5mA for the 5V unit, etc) so it should be suitable for testing both milliamp and microamp ranges (if present). To measure the shunt resistance, set the DMM on the range being tested, then connect the second meter in resistance mode between its current measurement terminals. On our example meter, we got a reading of 101.28Ω on the microamps range, 2.2Ω on the milliamps range and 0.077Ω on the amps range. You can then calculate what the meter should read in each range by adding the calibration resistor value to the measured shunt resistor values and dividing into the reference voltage. In our case, our calibration resistor measured 999.866Ω (an error of just -0.013%!) and our reference voltage siliconchip.com.au While the nominal accuracy of the Maxim voltage reference is ±0.04%, typically it will be much better, as demonstrated by this readout on the Keysight 34470A bench multimeter reviewed in this issue. 2.499987V. Thus the expected readings for this meter are: (a) 2.499987V ÷ (999.866Ω + 101.28Ω) = 2.270mA in µA mode (b) 2.499987V ÷ (999.866Ω + 2.2Ω) = 2.4948mA in mA mode (c) 2.499987V ÷ (999.866Ω + 0.077Ω) = 2.5mA in A mode We didn’t calibrate the example DMM but we did check its readings against these and got 2.270mA, 2.495mA and 2.5mA respectively. So it seems it doesn’t need any adjustments for now. Performing calibration If any of your checks give results with a noticeable deviation from the expected values (ie, more than ±1), you will probably want to trim the meter to make it more accurate. Unfortunately, the procedure for doing this will be different for each meter but there are some common steps. First, you need to gain access to the trimpots on the PCB(s). This usually involves removing the back of the meter. If it is in a rubber holder, remove that first, then look for screws on the back. There are usually 2-4 screws holding the back on. You may also need to remove the battery cover first. Usually, having undone the screws, the back will pull off quite easily. Modern DMMs are usually built on a single board but some may comprise two PCBs joined with headers or some other form of connector. Inspect the board(s) and locate any trimpots. We’ve seen as few as one and as many as 12! If you’re lucky, a service manual will be available on the internet for the model of multimeter you are calibrating which details the location and function of each trimpot. For example, we had a look for the manual for our venerable Fluke 77 and found it at the Fluke website. It confirms that the single trimpot is used to adjust the DC voltage reading. They suggest using a test voltage of 3V, which our reference board can provide with a suitable reference IC, however 2.5V should work fine too. There should be a manufacturerprovided service manual available for just about every modern, brand-name DMM on the market. If you have a re-badged DMM, you may have some luck if you do a web search to find out the original manufacturer’s model number for that product, then look up the service manual for that product. If you can’t find a manual for your meter but there’s only a single pot, chances are that, like the Fluke 77, it adjusts the reading in the DC voltage mode. In that case, it’s just a matter of hooking the reference up and tweaking it until the reading is correct. It may or may not also affect the current and resistance readings. If there are multiple trimpots though, it’s unlikely they will be labelled with anything other than a code. If you can’t find a service manual for your DMM, you’ll have to figure out what they do the hard way. First, take a photo of the trimpots so you can see which position each one is in, in case you can’t easily re-calibrate it later. Then, switch the meter into each mode in turn and adjust each trimpot. You’ll probably have to hook something up to the input terminals in each mode to make changes appar- ent. Once you figure out what a given trimpot adjusts, write it down and move onto the next mode. Hopefully, by the end of this process you have a full list of what each trimpot does. You’ll also likely have a meter that’s way out of calibration! So calibrate the voltage, resistance and current pots using the previous explanations for how to check the operation of each mode. All you have to do is adjust the appropriate trimpot until each reading is correct (or as close as you can get it). If there are any pots that you can’t calibrate, refer to the photo you took earlier to set them back into their original positions. Note that in some cases, the pots themselves may not be directly accessible without removing the PCB or unplugging a sub-module, however you may find that you can adjust them from the back through holes in the board. Generally it’s impossible to calibrate a multimeter without being able to observe the display while making adjustments so there’s usually a way to do it with the board still in the case. By the way, do not be tempted to use the 230VAC mains or other highvoltage sources to calibrate a DMM. It isn’t safe to connect a DMM to the mains with the case open. You could get a lethal shock if you do. Digital calibration Some modern DMMs use digital calibration. There’s no need to open the unit up; calibration is performed by manipulating the buttons on the front panel. For example, our Agilent U1252A and U1253B multimeters use this procedure. In this case, you’ll need the service manual for instructions on how to enter adjustment mode and perform the calibration. It’s usually a similar process to adjusting trimpots, except that the up/down adjustments are made using pushbuttons instead. You’ll still need the reference board to make the SC adjustments. Issues Getting Dog-Eared? Are your SILICON CHIP copies getting damaged or dog-eared? Keep them safe, secure and always available with these handy binders REAL VALUE AT $16.95 * PLUS P & P Order now from www.siliconchip.com.au/Shop/4 or call (02) 9939 3295 and quote your credit card number. *See website for overseas prices. siliconchip.com.au August 2015  47