View source for On the Topic of Resistor Precision

On August 26th, 2016, I was making a 
[https://en.wikipedia.orgwiki/Resistor_ladder resistor ladder DAC] 
for an arduino. I needed a lot of resistors with nearly identical values. The only resistors which I had more than enough of were 1/4 watt, 470 ohm, 5% precision. I had a bag of 200 of these. To get a smaller number (about 40) of nearly identical resistance, I measured the resistance of every single one.

I measured with a $20 [https://amzn.com/B000JQ4O2U Mastech MS8268], which is not a particularly precise multimeter, but it produces reasonably repeatable values.

I am writing this to record some useful observations of the precision of these "5%" resistors. Namely, that out of 200 resistors, all were within 2.5%, about half were within 1%, and all but one were below the rated 470 ohms.

* The minimum value resistor was 459 ohms.
* A roughly equal number of resistors were 5-10 ohms low, 4 ohms low, and 3 ohms low.
* Roughly half as many resistors were 2 ohms low as were 3 ohms low.
* Roughly half as many resistors were 1-0 ohms low as were 2 ohms low.
* The maximum value resistor was 472 ohms. It was the only resistor over 470 ohms.

I was expecting a normal distribution of resistances, centered on 470 ohms, with a small number of resistors close to being off by as much as 5% (23.5 ohms too high or low). However, the actual distribution is much better than that. For many hobbyist projects, it may be worthwhile to not purchase 1% resistors, and to instead purchase 5% resistors and assume that they are 2.5% resistors, or even that they are 1% precision resistors with an average value 99% of the rated resistance.