Why does multimeter show different readings for same current?

[geratheg]

I have a manual multimeter and I'm measuring current through an LED. Using DCA, when I used  the 20m setting, I adjust the current to 19.51 which I think is 19.51 mA (Is this correct?).

Using the same current, when I put the multimeter to the 200m setting, the current shows 8.3 for some reason.

Why are these numbers different for the same current?
If I went over 20m and had to use the 200m setting, it won't be accurate?

[Rerouter]

It is called burden voltage, as the current is measured across a resistor shunt it drops some voltage, and thus your measured current, different multimeters and different ranges on the same multimeter use different resistor values, some using larger values to use cheaper measuring techniques, and some very low such as daves uCurrent, where reducing this effect was the designed purpose,

[IanB]

Your multimeter is broken or inaccurate.

A higher current range has a lower resistance, which means that current on the 200 mA range should read higher than on the 20 mA range. If the current goes down on the 200 mA range you are not reading the true current.

[geratheg]

I checked again with 6 V and 2k resistor. Plugged the multimeter into the circuit for current reading.
I predicted a current of 3 mA.

The 200m setting shows 2.9 and the 20m setting shows 7.29.

The 200m setting is the accurate one. The 20m setting is the inaccurate one.
Well that's weird considering 2.9 is within range of the 20m setting.

[wiss]

What is the voltage over your multimeter while you measure?

[Matje]

I checked again with 6 V and 2k resistor. Plugged the multimeter into the circuit for current reading.
I predicted a current of 3 mA.

Actually should be less - unless you have a special LED that doesn't exhibit at least 1.5 V of forward voltage drop.

The 200m setting shows 2.9 and the 20m setting shows 7.29.

The 200m setting is the accurate one. The 20m setting is the inaccurate one.
Well that's weird considering 2.9 is within range of the 20m setting.

Hmm, seems weird indeed, although there is no actual guarantee for higher current ranges having lower resistance/equivalent burden voltage to lower ranges at full range (they often do not have that...), but I have never seen a DMM with specs to explain what you see. A crazy person could build one though ;-)

What type of multimeter are you using? Do you have (access to) another one to check or to measure voltage drop over the DMM in the two ranges?

[onlooker]

Why it still needs to be explained after the 6V/2kR test? As had already been pointed out, the meter is the problem.

Assuming the OP is reporting correct measurements,

  Range    6V/2kR     LED         ratio
  20mA    7.29mA   19.51mA   2.68
200mA    2.9  mA     8.3  mA   2.86

The best trust-worth measurement above is 2.9mA;  The ratio difference can be explained by the burden voltage difference (or the instability of the bad range). One can conclude: the meter's 200mA range is good and the 20mA range is bad.