Resistor Tolerances

The following information was originally posted by Terry on Vrat 1.0 in July 2012 see here but buried away and hard to find. Due to its usefulness, its now included in Vrat 2.0 as a service article for the  Service | General Tips section.


By Member: Terrykc.

The subject of tolerance crops up here frequently and is obviously a much misunderstood topic. A typical post will say something like “I changed R47 because it had dropped in value from 220k to 201k.” This is the equivalent of saying “I changed all the tyres on my car because the tread depth had dropped from 8mm to 7mm …” But you wouldn’t do that, would you?

Manufacturing accurate resistors has been possible for many years but it wasn’t always so. Most electronic designs, particularly the vintage kit we are usually talking about here, is pretty tolerant – if you’ll pardon the pun – of component value variation and a tolerance of ±20% was considered acceptable very early on. Attempting to make resistors of, say, 10, 11, 12, 13, 14, 15 ohms and so on, was obviously a waste of time, given such wide tolerances and an exponential series of values was evolved to make keep the available range within sensible limits. Take a 10Ω resistor, for example. 10Ω plus 20% = 12Ω. Similarly, a 15Ω resistor minus 20% is also 12Ω – so there is no point making anything between 10V and 15Ω . With a ±20% tolerance, each decade can be covered using only six values – the E6 series – because the values progress exponentially thus: 10, 15, 22, 33, 47 and 68. Hands up if you thought all those strange resistor values were chosen at random …! If you check them, you’ll find that the tolerance of each pair of adjacent values either just meet or slightly overlap each other.

No doubt as manufacturers found improved methods, accuracy improved, leading to the familiar E12 series – 10, 12, 15, 18, 22, 27, 33, 39, 47, 56, 68 and 82. Note that although, in the main, these were classed as ±20% components, the values actually overlap at the 10% point. My experience of resistors in the early 60s was of ±20% Erie carbon composition resistors, then RS ±10% resistors – still carbon composition but half the size – and then to ±5% carbon film resistors which, with very rare exceptions, I used throughout my working life. The 5% tolerance leads on to an E24 series of values – 10, 11, 12, 13, 15, 16, 18, 20, 22, 24, 27, 30, 33, 36, 39, 43, 47, 51, 56, 62, 68, 75, 82 and 91 but, for virtually all practical purposes, the additional values can be ignored. The only E24 value I can ever recall in a (vintage) radio or TV was a 43k resistor used by Bush in the line stage of the TV125 series (and probably the earlier TV105 and TV115 series before. Anybody involved with RF work will, of course, find the 75Ω value useful!

The picture below shows the three series with the upper and lower limits for each value at the appropriate tolerance. Hopefully it will help to clarify the issue and stop much non-essential resistor changing in the future! Spend your time (and money) finding the real fault! Resistor values can vary between hot and cold – but it is rare (and few resistors actually get warm, anyway!)

Resistor Tolerances 1