One Percent Resistors for One Cent?

I love using these 1/8 watt values for small projects, and was really looking forward to having some of my missing 1% values filled in, but first, I wanted to verify these resistors really were 1%. So, I sampled some of the values. The result was that some of the values (Like 3.0K, and 680K) fit within the 1% window for accuracy, but some other values (like 1.0K, 10.0K) are far out of spec, and in fact are 5% parts labeled (brown stripe) as 1% parts. Even as 5% parts, they are well worth the penny apiece price, but it’s wrong to sell them on Amazon as all being “1%” when many of them are obviously far from it. I would have rated this product as 4 stars if the seller had honestly listed them as “5%”. In the video I show more details on the analysis. If you buy this box, you will have to decide, on a value by value basis, which to mix with your other resistors, which to keep separate, and which to discard.

Video Errata:
(1) The snips I use are not $50, though the best ones can cost that much. You can get good ones for $23.
(2) I called out the 10K resistor bands backwards while I was pointing at them. Oops!

Here are some of the charts shown in the video. They are a bit easier to read here. The resistors in the assortment were 20 resistors per value. I used a resistance fixture to quickly check each resistor’s value while dictating the number to a voice-to-text app, and then pasted the list into Excel. The chart shows three red lines, to mark the nominal value and the high and low 1% limits. The photograph was made with my cell phone through my microscope. I randomly chose a value to begin with: 330 ohms.

The 330 ohm value was mostly within the 1% limits.
The 56K resistors were mostly outside the allowed limits.
The 3.0K resistors were nearly all OK. Notice the visible corrosion on the leads, possibly indicating age.
Now on to a more commonly used value, 10K. Which might be the most common value of all. The results are terrible! These are clearly not 1%, despite their markings, and must be surplus/trashed parts thrown onto the aftermarket for salvage. The color bands are not well painted.
So that made me think, what about another common part? Like 1K? Amazingly, this too was TERRIBLE. I ended up throwing these away.
So I thought, what if my meter has gone bad suddenly? I should check a known good resistor. So I checked these full price resistors bought from Mouser. ALL OF THEM fit well within the 1% tolerance. Oddly, they all turned out low, as if there was some kind of selection happening.

I guess the thing that most dismays me is the lying. I think there is a certain type of competition in Amazon and similar stores that amounts to — whoever lies best, wins. So we the buyers have to deal with it, and we have to assume that dishonesty is the norm, not honesty. Sadly, the lying seems to be the most prevalent in the China listings. The assumption that is becoming widespread is that Chinese listings are lying or cheating or fake in some way. On Ebay this is doubly true, and even Ebay’s geographical selection feature allows you to view only sellers in North America, if you wish.

Anyway, it was interesting to try to be a detective and figure out what this seller had done to put this set together. Clearly, they had lucked onto a large supply of manufacturer discards and then added a few more values to make the assortment. Then they felt they needed to lie about it and exaggerate its quality. But labeling this as a 5% set, while more honest, would have been problematic, since the 5th band, brown, shows it was intended to be a 1% part. So, it was always going to be the case that the resistors were visibly tainted. As far as my collection goes, I decided that I would only add one of the values of the resistors to my 1/8W 1% set if (1) I actually needed that value, and (2) I had checked each resistor. I ended up adding 15 of the values to my collection, but some of the values had many of the resistors failing the 1% test. The worst was the 2M resistor – only 1 of 20 was within 1%. The best was 270 ohms, with all 20 resistors passing. The other 13 values I checked had a range of compliance from terrible to very good. They were all over the place.