2512 SMD resistors rated at 20 mΩ

As I was developing my new electronic load, I needed some low-resistance resistors. There was no rush, so I decided to order them from AliExpress.
Find out how that turned out in this mini-review.


The seller offers several resistance values: 10, 20, 50, 100, 200, 220, 330, 470, and 500 mΩ, and 1 Ω. I needed 20 mΩ. The lot contained 50 pieces, and the total price including shipping came to $2.61. Even so, it was cheaper than buying them locally, so I decided to give them a try.

They came in a plain little bag with the label handwritten.


I planned to use them in the power stage of the electronic load to measure current in the feedback loop of the current-sharing circuit between the power transistors. In other words, the circuit has a control voltage at its input, and the load is then distributed across a large number of transistors through multiple channels. Each resistor is connected in series with the transistors and acts as a current sensor, helping distribute the load evenly.


I've already built a power stage that works on this principle, and I'm completely happy with it. The highlighted parts in the photo are these resistors, made up of a large number of small, precision resistors.

Based on my estimates, each channel was to use 4 transistors, with a channel current of up to 25 A, or about 6 A per transistor. With a resistance of 20 mΩ, the voltage drop across each resistor would be about 125 mV, and it would dissipate up to 0.8 W. The resistors are rated for 1 W, and with active cooling they should be perfectly fine.


They are size 2512 and come on standard tape.


There's not much point measuring them; their dimensions match the datasheet.


SMD resistor markings depend on their resistance value. Standard resistors are usually marked as shown below.


Very small resistors are sometimes marked with an alphanumeric code, but in my opinion it's much harder to read without a cheat sheet.


Low-resistance resistors use a slightly different marking: first comes the letter R, followed by the resistance value in milliohms.


And of course, what would a review be without some tests? It was very awkward to hold and photograph them, so I used a battery holder to mount a few samples for testing.


The seller claimed the resistors had a 1% tolerance, but my measurements showed a much larger deviation. For starters, the actual resistance is not 20 mΩ but 18–19 mΩ. Also, 1% is 0.2 mΩ, while 1 mΩ is already 5% in this case.


But I was much more interested in their temperature stability—the temperature coefficient of resistance (TCR).
Without removing the last resistor from the holder, I started heating it with a soldering iron.
1. The soldering iron tip is about 3–4 mm from the resistor.
2. Here I briefly touched the resistor with the tip on the side opposite the resistive layer.

The result is very disappointing. I didn't care much that the resistance differed from the stated value—currents would just vary a little between channels—but a change in resistance due to heating is a bigger problem.


To rule out measurement error, I soldered a pair of wire leads to one of the resistors and used a different meter.
1. Resistance before heating.
2. Heated with a soldering iron. It might look in the photo as if the tip is resting on the resistor, but it isn't; there's a gap of about 1.5–2 mm between the resistor and the tip.


I can certainly still use them. The feedback loop will cover the entire power module, but their high TCR and poor initial accuracy are disappointing. They won't work as measurement resistors.

The worse part is that I've already confirmed receipt and left a positive review. Now I'm trying to figure out how to update it to mention the shortcomings.

That's all from me. I hope you found this information useful.