Capacitor values

[Charlie]

Curious... does anyone here know why capacitors generally have odd values? 22mfd, 33mfd, 47mfd, 82... and so on. Years ago, it seems they were mostly values such as 20, 30, 40, 50, 80, 100mfd. Why the change?

[bob91343]

The standards association, EIA or whatever it was called at the time, created the list of 'standard EIA values' that were to apply to components. This happened, I think, in the 1950s or thereabouts.

The idea was to have succeeding values about the same ratio so that any component would fit into a bin and only one bin, depending on the tolerance. So a 20% spacing made for a set of values that could have a 10% tolerance, and so on.

For that reason, you won't see a 51 Ohm resistor with a 10% tolerance. All 51 Ohm resistors would then be 5% tolerance or better. A 51 Ohm resistor would also fit into a 47 Ohm 10% bin.

It's not a perfect system, but it works most of the time.

[old_tv_nut]

Also, electrolytics had/have very wide tolerances, so it was more economical in mass production of radios and TVs to specify the minimum value needed and let the cap maker cut the materials down to an amount that would guarantee that minimum. So, the stamped values were determined by a manufacturer's mass order, not by pulling from standard-value bins. Especially true for multi-section caps of course.

[Charlie]

I understand.... somewhat. It just seems like it would have been easier to stick with nice round numbers stamped on the component such as 20 instead of 22, or 30 instead of 33.

Often, I do notice that multi-section cans give readings that are a good bit higher than the value on the can, but still within the 20% tolerence. Individual electrolytics, I've noticed, are typically spot-on what the stamped value is.

The example of the 51 ohm NOT having a 10% tolerance makes a lot of sense. I would think if it circuit calls for a 51 ohm resistor, it needs to be 51 ohms... or really freakin close.

I recently went thru a lot of resistors I got a while back. They were all nicely cataloged in small mailing envelopes with value, wattage, and tolerance written on the envelope. There were at least 75 different values... most of which were somewhat odd... 91, 110, 18, 13, 34... and so on. They were all 5%. Just so happens... as I am looking at them, there's a 51 ohm 5% envelope (as in your example) that's staring at me.

[old_tv_nut]

Another thing to note about resistors - most modern resistors are made to the tolerance stated. The carbon composition resistors, however, could not really be made to 5% or tighter tolerance, and therefore 5% resistors were picked out by testing at the end of the process line. This means that the 10% tolerance resistors of the same nominal value would be the leftovers, and would have a bimodal distribution, at least plus or minus 5% off from the nominal value.