3
\$\begingroup\$

I have a cheap multimeter, which gives me different readings between the two inputs. I would like to set 100 mA using a potentiometer in a circuit, which runs from two AA batteries. If I measure the circuit using the standard slot which works up to 200 mA, it shows about 80-90 mA. In the same circuit, if I plug the cable into the 10 A slot, and switch the multimeter to the 10 A mode, it shows 0.11 A.

Could you tell me why is it happening and how should I know that:

  1. is my multimeter actually modifying the current, or just showing it?
  2. if not, then which measurement is the correct?

As a small question: if I buy a new multimeter, what should I look for if I want to buy a better one? Should I buy one from a known brand? Is there any way to identify the ones I should avoid?

Update: I think I should tell you what is my circuit: 2xAA batteries ---- 10 Ohm ----- IR LED (specified at 1.35V/100mA) --- multimeter ---- batteries

Thank you for all the detailed answers! Based on markrages' answer, I think in my case the multimeter is indeed modifying the current, because

  1. without the multimeter its 10 Ohms
  2. with the multimeter in 10 A mode its 10.1 Ohms in total.
  3. with the multimeter in 100 mA mode its 12.5 Ohms in total.

Which would prove the fact why am I seeing 80-90 mA vs. 110 mA in the two modes, as thats the same 25% difference as the difference in resistance between 2. and 3.

Even more extreme was when I tried to run the 1.35V LED from a single AA battery. In that case I used a 1 Ohm resistor, and the two readings in the multimeter were 35 mA vs. 100 mA. Now I understand why, because the multimeter + 1 Ohm acted as a 3.5 Ohm resistor.

\$\endgroup\$
  • \$\begingroup\$ You are probably seeing effects of the AA cell's internal resistance. Compare open-circuit voltage with loaded voltage. The one-ohm resistor is not a great circuit as it will be very sensitive to battery aging. \$\endgroup\$ – markrages Apr 21 '11 at 16:48
6
\$\begingroup\$

Measuring the cheap meter on my desk, the 10A range uses a 0.1 ohm shunt resistor and the 400 mA range uses a 2.5 ohm shunt resistor and the 4mA range uses a 100 ohm shunt resistor.

100 mA through 0.1 ohms is 10 mV drop and 100 mA though 2.5 ohms is 250 mV drop. So depending on the impedance of your circuit, the current could be lower in the lower range just because of the increased series resistance.

You can measure the voltage drop across the ammeter with another voltmeter.

When you can't tolerate voltage drop, try a feedback ammeter: http://www.keithley.com/data?asset=6169

\$\endgroup\$
  • \$\begingroup\$ He's pushing about 100 mA from two AA batteries... Which works out to about 30 ohms. A 2.5 ohm shunt would contribute close to a 10% error right there! \$\endgroup\$ – Toybuilder Apr 21 '11 at 8:48
  • \$\begingroup\$ Are you sure it is 0.1 ohms? It is usually 0.01 ohms in almost every cheap meter I've ever seen. \$\endgroup\$ – Thomas O Apr 21 '11 at 13:20
  • \$\begingroup\$ I think this is indeed what's happening in my case! Thank you for the answer! I have never measured this multimeter, because I have never thougth that is has so much resistance in the 200 mA range. I have updated my question with findings. \$\endgroup\$ – hyperknot Apr 21 '11 at 14:31
  • 1
    \$\begingroup\$ @Thomas, depends on the cheapness of the meter I suppose. \$\endgroup\$ – markrages Apr 21 '11 at 16:23
  • 5
    \$\begingroup\$ @zsero You can eliminate this kind of surprise by using your own shunt resistor as part of the circuit, then measure its voltage and use Ohm's law to get current. \$\endgroup\$ – markrages Apr 21 '11 at 16:24
5
\$\begingroup\$

Screen Error

You need to keep in mind that a larger range will result in less resolution. Auto-ranging multimeters help you get around this issue by automatically switching to the highest resolution range that it can use for your signal. But without auto-ranging, you are just stuck with rounding error that has to be in place in order to display the value on the screen.

Lets say you have 4 positions for numbers to go, on a 0 to 200mA scale this means you can display at most a resolution of .1 mA. Ex, 199.9mA is 4 digits, and unless you change the screen, theres no way to add more digits. Now if you switch to the 10A mode, you have the same 4 digits that now have to represent 0 to 10,000 mA. This means you can display 9.999A or 9.998A but you can't show 9.9985. This makes your step size 10 mA where as before your step size was 0.1mA. That is a factor of 100 difference in resolution!

So purely by design of the LCD screen you will be able to get more accurate data at a lower range.

Note: I have seen cheap multimeter that have the first digit be either a '1' or turned off. This might be the case here. Even though it says it can go up to 200 mA, it could mean that it can go up to 199.9mA

Measurement Error

First of all, 0.1mA resolution can be pretty hard to measure, especially on cheaper meters. Some might show you something that might make you think it is reading that exactly, but I wouldn't trust it.

Second, 80-90mA is a pretty significant swing in readings at that range. This makes me think you either have a very crappy meter or there is something else going on. I have found that when you are at a range like that, how well your connections are between multimeter and circuit start to show up as variability. For example, I have seen people try to stick the pointy end of a DMM probe into a breadboard and just barely make contact internally. In this situation, as you move the probes around making more and less contact, the resistance between the probe and breadboard changes which then of course causes the current to change.

Ideally you should be able to set up your measurement such that you can take your hands off of the project and just look at the measurement. Hopefully if you set yourself up to be able to do this you will find a reading that is pretty steady.

Range Differences

On DMMs, as you change your range, it is changing what components are being used to take the measurement. Every component is going to have some chance or error associated with it. There is a chance that by luck of the draw your meter ended up with higher accuracy components for the 10A range while it had much worse in the 200mA range. But at the same time it is just as likely to happen in any configuration you can imagine.

It is also a possibility that the manufacture was willing to spend a bit extra to get nicer components for the 200mA range (like 1% or .5% tolerance) while allowed for 5%-20% tolerance to be allowed in the 10A range in order to save money. But even in this situation it is possible that a 20% tolerance component ended up being exactly correct while a .5% could be its full .5% off.

Now nicer meters will generally use components that have higher accuracy and thus will yield more accurate results.

Moral to the story

I think David's answer pretty well sums it up.

\$\endgroup\$
3
\$\begingroup\$

On the 10A scale, you are asking it to read a current value at about 1% of full scale. To get a difference at that level of only about 20% is pretty good if you think about. Having a perfectly linear circuit is always fairly difficult to do. Most components will only have at best 0.5% to 1% tolerance and will also vary with temperature (which is usually compensated for). A better-brand will probably give a closer result but I would still expect a 5% - 10% error when you are working at the bottom of the scale.

\$\endgroup\$
2
\$\begingroup\$

The different measurement ranges will have different accuracies. Generally, the cheaper the multimeter the more the two ranges will give different measurement values. But, to more directly address your questions:

  1. You're multimeter is not modifying the current, what it's showing is off.

  2. Generally, the lower range (200 mA in this case) will probably give you the more accurate reading. This isn't 100% true, but maybe 95% true.

  3. It's hard to go wrong with a Tek or Fluke. Other major brands are probably fine. I'd stay clear of the no-name brands, but I'm sure someone will disagree with me on this one. My experience is that you get what you pay for on this one.

\$\endgroup\$
  • 1
    \$\begingroup\$ Hmm, you can't really be sure of #1, can you? They do add shunt resistances and cause a voltage drop, and we all know that resistors act as current limiters for this reason. I'm also surprised no answer has mentioned the least-significant-digit accuracy spec (i.e. 0.1 % +/- 3 least significant digits), as the 10 A range usually has a rather low accuracy, in part for this reason (the LSD has a high value). \$\endgroup\$ – exscape Mar 17 '12 at 8:48
  • \$\begingroup\$ @exscape I guess it could, but ideally the change in current would be smaller than the accuracy of the meter. That's easy enough to test using a second multimeter set to measure the resistance through the first. I should also mention that not all multimeters use a shunt resistor. Some use hall effect sensors, which wouldn't have any appreciable resistance. \$\endgroup\$ – user3624 Mar 17 '12 at 16:13

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.