Why Do Resistors Have A Color Code?

One of the first things you learn in electronics is how to identify a resistor’s value. Through-hole resistors have color codes, and that’s generally where beginners begin. But why are they marked like this? Like red stop signs and yellow lines down the middle of the road, it just seems like it has always been that way when, in fact, it hasn’t.

Before the 1920s, components were marked any old way the manufacturer felt like marking them. Then in 1924, 50 radio manufacturers in Chicago formed a trade group. The idea was to share patents among the members. Almost immediately the name changed from “Associated Radio Manufacturers” to the “Radio Manufacturer’s Association” or RMA.  There would be several more name changes over the years until finally, it became the EIA or the Electronic Industries Alliance. The EIA doesn’t actually exist anymore. It exploded into several specific divisions, but that’s another story.

This is the tale of how color bands made their way onto every through-hole resistor from every manufacturer in the world.

Dots Then Bands

Ésistances anciennes annees 50 by François Collard, CC-BY-SA 4.0

By the late 1920s, the RMA was setting standards and one of them was the RMA standard for color-coding. The problem was that marking small components is difficult, especially back in the 1920s.

The solution was color bands, but not quite as we know them today. The standard for colors was the same, but the body of the resistor acted as the first band. Then there would be two or three other bands to show the rest of the value. In some cases, the third band was actually a dot. So the bulk of the resistor would be the first band color. The “tip” of the resistor would be the 2nd band and a dot would be the multiplier. Radios using this scheme started to appear in 1930. Here’s the color code chart from the 1941 Radio Today yearbook:

Ads in that magazine promoting resistors were careful to note that they were RMA color-coded. The code soon extended to capacitors (condensers, in the contemporary parlance).

The dot, as with printed piece of text on the cylindrical, might be hidden from view depending on the position of the resistor. So eventually, everyone switched to bands.

The colors are meant to follow the visible spectrum (remember ROY G BIV?). However, the RMA omitted indigo because apparently many people don’t distinguish blue, indigo, and violet as three different colors; indigo is really a tertiary color, anyway and Newton included it because of his interest in the occult, apparently. That leaves four slots, so dark colors represent the low end (black and brown) and bright colors the high end (gray and white).

Of course, none of this was funny if you were color blind. Reading a resistor with a meter or a bridge out of the circuit was certainly an answer. Reading one in a circuit, though, was another matter.

The Origin of E-Series Values

In 1952 the International Electrotechnical Commission (IEC, another standards group) defined the E-series which dictates what values resistors come in so that you get equal spacing on a log scale for resistors. If that sounds confusing, consider an example.

The E12 series is for 10% resistors and the values on it give you 12 values per decade. The base values are

1, 1.2, 1.5, 1.8, 2.2, 2.7, 3.3., 3.9, 4.7, 5.6, 6.8, 8.2

That’s why you can get, say a 4.7 K or 47 K resistor but not a 40K resistor.

However, consider the tolerance. A 10% 39 K resistor could be off by 3.9 K. If the error pushed the resistance up that would be 42.9K, making a 40 K resistor unnecessary. That is, a 39 K resistor might well be a 40 K resistor, anyway. A low 47K resistor, on the other hand, could be 42.3 K, which is less than a high-value 39 K unit.

As you might expect, the number of values goes up as the tolerance goes down. At 2%, for example, you’ll use E48 which has 48 values per decade (if you’d guess E96 — the standard used for 1% has 96 values, you’d be right). Using E48, the values near 40 K are 38.3 K and 40.2 K. That’s 39.06 on the high side and 39.2 on the low side.

Next Time

Next time you pick up a resistor and read the code from it, you can recall the history behind it all. The legacy of color bands carries over into the surface mount realm, not as color but as three digits representing the first two numbers and multiplier for the resistor’s value. These days many electronics like wireless modules and lithium batteries include a datamatrix (something like a QR code) on them. Honestly, I’m surprised that all components — through hole and surface mount — don’t have some form of micro data matrix on it that lets you point your phone at them and see their complete datasheet. Maybe one day.

219 thoughts on “Why Do Resistors Have A Color Code?

  1. Wow the body colour + band + dot is genuine improvement over current beige plus colours which could be anything scheme.

    It always seemed ludicrous that we marked them so given 10% of the male population is frankly colour blind (red-green) and many more have subtle shifts or simple poor colour discrimination. Aren’t electronics the territory of men?

    Then I realised actually this kind of work rapidly became female dominated (and continues to be as I understand in the far east’s sweat shops). – I might check those stats actually. And woman are far better than men in this kind of subtle detail orientated skill as well as genetically less likely to suffer from colour blindness and be more careful.

    I’ll put my hand up here. I use a cheap DMM to test resistance then write it on the resistor strips. I know the colours, how the marking is supposed to work but frankly in practice rarely can be confident if that’s a black, brown or green (especially with small cheap resistors I end up buying).

    1. I suffer colour blindness too. Biggest trouble is between brown and red. I must check with a meter every time. I have found, however, that if I take a picture with my digital camera and zoom in, I can discern the colours. Perhaps my camera (unintentionally) shifts colours or perhaps it’s from having a larger sample?

      1. The brown and red are difficult are difficult on aged components for everyone I think. Also the white can go yellow, the yellow pale enough you wonder if it’s dingy white, and the blue and violet can fade so much you wonder if they’re grey.

    2. THIS. There is no consistency among manufacturers and I often find it easier to just DMM it if there is a question of value which makes the color code pointless. About time it gets an update. One man’s brown may be another man’s orange and so forth. Hopefully next will be legible transistor labeling that does not require an electron microscope to decipher lol.

    3. I think even more common than colour blindness, is bad colour rendering from a lot of cheap high-efficiency lightbulbs. Often orange and red will look brown, and I’ve even had trouble differentiating purple from blue or black in a few cases.

  2. Whoops sorry forgot to include “many people don’t distinguish blue, indigo, and violet”.

    The shocking answer to this is this defect apparently is in large part psychological (there are subtle colour shift defects in people especially due to lens transparency).

    But just as painters or those you spend all day around colour swatches seem to have a colour discrimination super power so do those where the language makes explicit differences in colours by name.

    Native Russian speakers are faster than English speakers at discriminating light and dark blue (siniy or goluboy in Russian)
    [ references https://www.newscientist.com/article/dn11759-russian-speakers-get-the-blues/ and https://www.pnas.org/content/104/19/7780 ] precisely because the words are so different in Russian (tested by showing those who learn Russian and speak it are better than those who never learn Russian). Wow.

        1. Well, cyan (“blue”) is between green and blue (“indigo”). Cyan’s not a word I ever heard til I got into computers, it’s “turquoise” over here, and Americans call it “teal”. So no wonder it’s not in popular use. And yeah it’s greenish but I suppose if you take your sample further towards blue it looks less green and more light blue. It’s a language thing anyway, apparently Eskimos have trouble distinguishing, visually, some colour, because they don’t have a word for it in their language. Which isn’t surprising really, I bet they’ve got lots of words for “white”.

          The language / vision connection is real. If there isn’t a word, you often don’t discern the colour well. According to Stephen Fry on QI, the ancient Greeks didn’t have a word for “blue”, to them the sky was “bronze”. But being a QI factoid that’s about 50:50 whether it’s true or not. Same thing applies to most of what technology expert Stephen Fry says in general.

          1. In printing too, probably most well-known. Including consumer printers. But still most people aren’t printers or professional photographers. Apart from maybe manufacturing your own colour film, do photographers need to use those colours much? Photography uses the RGB colour model, and that’s what modern camera sensors output. Even if internally they use more innovative colours in their colour filters.

            A chap called John Savard / Quadibloc has the most interesting personal website of mental bric-a-brac, and he’s a very smart guy. He has a page on the filters used by cameras, it’s not just Bayer, and his own suggestions. You can use other colours that let more light through, then figure out the colour of a pixel mathematically from each of the 3 sensors, with a bit of adding and subtracting. Letting more light in nearly always means more quality and accuracy. And of course allows shorter exposures with less noise.

          2. Teal is blue-green, cyan is greenish-blue.

            Only color-blind Americans call cyan teal, because regular people just say blue-green or “blue” or “green” as a category, they don’t worry about being that specific between similar colors.

            Teal is an uncommon word in America outside of artists, but any American who can see color knows what cyan and teal are. But they still call it all “blue” or “green” most of the time.

  3. It is just me, or do others have problems discerning color codes where the resistor body is some weird brownish or greenish color? But have also noticed that female techs and engineers have less trouble distinguishing colors than their male colleagues.

    With SMT stuff, this is becoming moot.

    1. I think this is partly related to color blindness which is more prevalent in the male population. And more prevalent in the Red and Green part of the spectrum (brown has a lot o red).

      My experience it seems also that women see more colors than man, because I have been tested for color blindness and am normal, but it is very difficult for me to distinguish the shades of pink. Colors typically referred as salmon, peach, flesh…

      1. That second X chromosome has a lot of good stuff for color vision on it. I’m colorblind and I absolutely despise old color-coded leggy resistors. The worst. Can’t we stop doing that and just print tiny text on them like we do with every other type of component now?

        I’ve rigged up a little gizmo consisting of a throw-away multimeter and two Y-shaped cradles made of wire that I can just drop a resistor into and quickly read out its value. Of course it doesn’t work for reading values of resistors that are already part of a circuit. I mean one in ten males is colorblind, and even more have difficulties with color even if they don’t have a full-on disorder—how did this labeling scheme ever get a pass? Or at least why does it still get a pass? This isn’t the seventies anymore, we can print fine detail on these things.

        Also they chose the absolute worst colors for both the beige backdrop and the most commonly used bands. They really wallowed around in browns and greens and reds and oranges and yellows and other trash colors that are the ones most struggled with among deuteranopic eyes.

        1. They can print fine detail on things, but could you read it? SMD resistors are flat rectangles, and often bigger than through-hole resistors. Stripes are surely better suited for tiny cylinders. There’d also be the cost of the printing, which has to be very cheap for the profit resistors make. Tiny fine printers might be more than they could manage. The printing on SMDs is wobbly enough.

      2. One of my eyes sees things with a bit of a blue cast while with the other eye I see more of a red hue. So it’s likely one eye is a bit red deficient while the other is lacking some in blue. With both eyes everything has (I presume) the proper colors.

        1. Maybe you’re developing cataracts! Or else your lenses are yellowing a bit with age. There’s some famous artist, I forget whom, who suffered something like that as he got older. Experts have looked at his paintings, which have a very different palette, from when he was older, compared with the ones he did in his youth. They analysed the colours, along with experts, and it exactly matched some age-related colour degeneration. Paintings from inbetween show a progression between the states, even along the right sort of timescale.

      3. Pretty sure there’s little to no difference between male and females normal colour vision.
        Some women have tetrachromacy, which allows better colour vision, but that’s certainly not the norm.

        There can be other forms of reduced colour vision than the common red/green, though.

        1. So 10% of men are colour blind( mostly dichromacy like 0.1% of women), theoretically 15% have tetrachromacy (0% of men).

          Pretty big difference there imo

          But no as far as I know men and women have always had vision differences in any study I’ve seen from colour shifts to outright better discrimination in women. Some of which is probably psychology like I described above.

          1. “Another study suggests that as many as 50% of women and 8% of men may have four photopigments and corresponding increased chromatic discrimination compared to trichromats”

    2. Yeah, I discovered I had this problem years ago. The solution was to make sure my entire workshop was outfitted with daylight fluorescent lights. Incandescent lighting only made the problem worse..

      1. Due to the toxic nature of marking dyes, many modern SMD resistors are shipped with no markings at all. Switching to a different type of bulb is not going to help you any more, you’ll have to get organized.

        1. The vast majority of SMD components are applied by machines, which don’t bother trying to read them. Through-hole is still what hobbyists and prototypers use, since they work with breadboards.

        2. Did nobody tell you, that you are not supposed to eat electronic components? :-) But even if you do, two of the most common white pigments, titanium dioxide or zinc oxide are non toxic. I think the are even listed as food additives. Only lead based pigments should be avoided.
          I think it is just a factor of cost to print on the small 0402 or even smaller components. resistors in 0603 and up are still marked.

          1. Here in Australia I see s lot of foreign tourists who ask how to tell which snakes are poisonous.

            I just tell them that I don’t know because I don’t eat snakes.

      2. That’s might just be some luck on your part. Often fluorescent lighting is what causes the problem, since the peaky spectrum can have dips that align with the colour bands, making an orange or red look brown, etc.

        High wattage incandescent lighting (not dim, yellowish 40w bulbs, etc) will have “perfect” colour rendering, so is actually the best for differentiating colours.

        Good quality fluorescent or LED can be almost as good, so if your fluorescents either have a good even spectrum, or have the peaks in the right places not to cause any interference, you’re good, but if you’re picking lighting for your workbench and read a lot of colour codes, it’s best to try whatever lights you plan to use before you commit to them.

        1. I would say, perfect colors you have only with sunlight (6000K). Incandescent bulbs are unfortunately limited by the melting point of tungsten. So I prefer 6000K LED for some rooms, although some of them can have a really awful color rendering, you have to choose good ones with high CRI.

          1. LEDs adjust the color temperature mainly by varying the amount of blue light in a very narrow frequency band – so it’s the same low CRI light with more blue. It doesn’t help you see colors, because the eye iris responds mainly to blue light, so by having a strong peak there causes your eye to restrict the amount of light going to the retina, which causes the loss of color contrast at the other end of the spectrum due to the Purkinje effect (red colors become black).

      3. That’s ironic, cos incandescents, by their nature, have a full spectrum. Though they can be lacking power towards blue, since the light is often yellowish. Might be throwing more photons at the problem helped, fluorescents are brighter, although yes the cool white ones have more blue too.

        1. Halogen bulbs have slightly higher color temperature.

          When the light levels go down, the eye sensitivity to red wavelengths drops and blue increases. We have an evolved response to seeing colors “correctly” under sunset conditions when the natural color temperature starts to shift towards red.

    3. I don’t think it’s necessarily color blindness (@zé) or lighting (@Medix). Some of those cheap resistors with the almost-lime-green bodies are really terrible. I think the paint they use for the stripes is a little too transparent or maybe it’s applied while the body paint is still wet so it mixes. Either way the colors mix and become something that only a person with a lot of experience mixing paint palettes might be able to read.

      1. Lime green!? I’ve seen light blue but most resistors are officially “salmon”. I suppose having to find an eleventh colour, that doesn’t confuse too much with the other ten, and they’d already used gold and silver, must’ve been a challenge! Would have been easier to genetically engineer a few more cone cells into engineers.

        1. Well, not exactly lime green, a little darker but that’s the closest shade that I have a word for. I’ve received them when I bought really cheap stuff from Chineses Ebay sellers. I’ve also seen them soldered onto PCBs in consumer electronics. They are really hard to read, I often have to resort to just using a meter. I hate doing that though because then I don’t know if this is the actual intended resistance or if it’s gone off spec. If the latter will it be stable? It’s better to just buy new. It’s getting hard to do that now though if you don’t want to wait a month or more for it.

      2. I have above average color discrimination when doing art, but I often squint at a resistor under bright light, and it just isn’t the same color as the other manufacturer used, and it is somewhere in between the two colors. Their bad colors make me start to question my vision.

        I find myself very hesitant to say, “Well, it is 5% closer to this color, so it must be this one!”

        I try to remember before squinting too hard that the DMM is within reach and faster anyways.

    4. I’m not color blind, but I agree that the modern resistor color codes are much harder to read than the resistors from the ’60s that had a dark brown body and opaque color stripes.

      1. Cataracts and retinal dystrophy due to age also cause loss of color vision. Especially in the first case, everyone’s lenses get a little foggy as they age, and this causes a loss of contrast and saturation with color vision – the brain gets used to it, so you don’t notice until it gets really bad.

        1. I believe Bill doesn’t speak about his 50+ years old memories of resistors.
          You can compare a vintage resistor from a vintage device with a modern one to see a difference Bill speaks about.
          I’m dealing with lots of vintage tech and I can confirm his words.

          1. Thanks for confirming it is true. I dislike when people get doubted with random guesses like above, that suggest serious health problems as cause (or other human “defects”). Luke’s answer is (maybe unintentionally) quite a hostile reaction.

            It’s a common reaction indirectly suggesting the “observer”/human is always at fault (for being imperfect), defending problematic technology/solution/approaches.

            A mindset that is too common among technicians of all kind.

          2. I don’t think a person can be unintenionally hostile. Maybe insensitive. Though on the other hand they can also be paranoid, prickly, and over-sensitive.

            If you were looking at 50 year old resistors, it’s not unreasonable to assume that was 50 years ago. There’s a lot of codgers on here! Might not be the case for you but as a guess based on a forum post it’s an acceptable guess. You’re free to refute it, as you did.

            Deteriorating eyesight happens to all of us. Especially people who spend a lot of time looking at screens and the rest squinting at tiny coloured stripes. So that’s not an unreasonable assumption either.

            Again, you can refute it. That’s fine. No need to feel bad, the guy doesn’t know you. It was just a suggestion based on partial information from a forum post. For plenty of people who can’t see the colours on modern resistors, it IS their age! Happens to the best of us. Wait til your eyes start going and you start finding grey hairs in places all over your head and body!

            Bizarrely, my eyes have sort-of lost flexibility as I become middle-aged. Optician says it’s normal. But it’s in such a way that I can now read better, including the computer, without my glasses. I’m cured! Except if something’s far away I still need them. So I tend to peer over them like some wise wizard when I’m messing with my phone or reading something, then push them back up the nose when I’m actually seeing. In front of the computer, I just take them off. Put them on the desk. There they are!

          3. It’s never good to make assumptions about health.

            My vision faded to the point that I could barely see even with glasses.

            My lung function dropped to 52% and I was diagnosed with COPD and was looking forward to a short life in a wheel chair as my lung function faded to the point that I could no longer support my vital organs.

            Now my lung function is 98% and a recent eye test shows that I have better than average vision for my age and that is without glasses that I don’t ware anymore.

            I discovered my condition was the result of toxin exposure and had nothing to do with age. I left the environment where the toxins were.

            Never assume that a health condition is from age or you could well be writing yourself off and consigning yourself to a debilitating condition that you don’t need to suffer.

    5. I now do have problems discerning some of the colors, but wasn’t always this way.
      Stupid old aging old eyes ;}
      I didn’t think about how the resistor body color may change things, I’ll have to pay attention next time.

      But even without that my red/orange blends, and my brown/purple too.
      Just the other night I had a 5-band resistor where I swear the gold band was solidly yellow, and yet again cursed my eyes, until I noticed the 4th band was purple…
      Brown black black purple ‘gold’ would be 1 billion ohms which (thankfully) makes no sense.
      Was an odd feeling realizing that for once I got the color right and was reading the silly thing backwards. It was a 1% 470.

      1. 1 Gig-Ohm Resistors definitely exist. Normally they are not color-coded but I am pretty sure that I have seen at least one with normal ‘striping’.
        I do a lot of work in audio-equipment-electronics and in the preamps for condenser microphones, very high value resistors are pretty common. Sometimes up to ten Gigohms.

          1. I do have some zero ohm resistors. Of course the singular black band lacks tolerance info.
            But your comment really makes me want to paint on my own silver or gold band at the end, and somehow try to get them back out into the wild.
            Could you imagine the look on the face of whoever runs across those a decade from now? >:}

          2. Once upon a time I ordered a reel of 1206 zero ohm resistors, and Digikey sent me a reel of 1206 fuses. I realized that I couldn’t say they sent me the wrong part, just not the part I ordered.

          3. The joke is that when you look at the datasheet of a zero ohm resistor, many times they DO specify a tolerance.

            https://uk.farnell.com/yageo/rc0603fr-070rl/res-0r0-1-0-1w-0603-thick-film/dp/2693558

            The product code for that resistor is RC0603FR-070RL where RC0603 is the size and F is ±1%, R is reel type, 07 is reel size, 0R is zero Ohms, and the final L means customized label. When ordering these parts, you can technically order a zero Ohm resistor with a 5% tolerance if you wish. What you will get is anyone’s guess.

          4. Ima buy a cheap Volt or Leaf with a blown battery, and a container full of 5% tolerance zero ohm resistance, test them all for the ones that are up to 5% less than an ohm, run the car off them and sell the rest back.

          5. @RW: 5% of zero is still zero.

            Also please don’t forget, some of them have a really high TK, up to 4380ppm/K .
            But for most “zero ohm jumpers” they don’t specify a tolerance but a max. value of e.g. 50 mOhm.
            Also odd was the “max. working voltage” specification in case of a 0R jumper from one manufacturer of 200V – that would be a tremendous amount of power.

        1. Oh I know they exist, it’s just I have never purchased or used any myself, so wouldn’t make sense to be in my parts bin.
          Of course the parts bin can be a lot like the junk drawer. It would only mildly surprise me to find one in there somehow.

    1. I would guess it’s because the SMD parts have an orientation, there’s an ‘up’ that is recognisable from any angle with the parts that bother to have a code, and no room on the parts too small for a code.

      1. Through-hole resistors don’t have a “left” or “right”. Or an “up” come to think of it. You work out which is the start end from the gold / silver / none stripe at one end. Same thing would work as well on a rectangle as it does a cylinder.

    2. Don’t fucking tempt them! They have actual numbers which is way better than those awful, useless colors! They should be adapting those numbers the other way around, putting them on resistors.

      1. Useless. Until you’re working with low-TC parts that ARE printed with the value but the machine that loaded them put all the values against the PCB. The schematic doesn’t help either because the designators are under the part. Dig through the documentation until you find the board layout, then cross-reference the schematic, and 5 minutes later you know the value and can continue troubleshooting–if you can remember why you needed the value.

        1. I don’t think SMD resistors should be installed inverted either. The resistive element is usually on top, so the maximum dissipation is likely worse if it’s mounted against the board.

  4. Feh!
    I learned how to read the basic resistor color code when I was considerably younger, and never looked back. Of course I still need to look something up when I need a strange value that’s somewhere in my parts kit, but, it is one that wasn’t used before. Al I do confess that I first noticed that oddity in a booklet that Radio Shack had published years ago on standards.

  5. Interesting article – thanks! But I’m still confused about the E series and the idea of tolerance. I’ve heard two interpretations:
    1. Most common is that a 10% resistor is random and can have up to a 10% error, e.g., a 100 ohm resistor could be anywhere between 90 to 110 ohms due to manufacturing processes I suppose. However, I’ve never been able to turn up information about underlying statistics. Does this mean a resistor with a mean of 100 and a standard deviation of 10 and a Gaussian distribution? Alternatively, the distribution could be uniform between 90 and 110 ohms? If anyone has a good reference I’d appreciate it.
    2. The “non-random” interpretation of E series is that if a design calls for a resistance R, there is always a nominal value within 10% of the desired R value, and using 10% resistors the designer will never have more than a 10% error between desired and nominal resistor values. Does this mean that no resistor randomness is assumed? (i.e., actual R values are very close to nominal values otherwise there are two sources of error, the difference between actual and nominal, and the randomness of the actual R value). Perhaps a 10% 100 ohm resistor actually has a very small standard deviation, maybe an ohm or less?

    Any clarification would be great! Not being an electronics guy my apologies in advance if this falls into the “dumb question” category!

    1. There’s a third version:

      3. Manufacturers measure and bin the resistors out of the same production lots. First they remove the 1% resistors, then 2% resistors, then 5%, then 10%, and discard the rest or sell them to off-brand fly-by-night retailers in China.

      This means a 10% resistor is almost never less than 5% off the actual value, except when the factory had an order to fill and shifted some of the 5% or better bins to the 10% bin. In other words, you can expect to never get close to the actual value unless you bought the 1% resistors, or the 0.1% resistors.

      1. Good article, still don’t really know why the bands are needed in the first place. New resistors say their value on the bag, and about the only time you replace one is if it burned up and you can’t read the bands. If you are trying to clone a circuit, you remove the parts so you can see the traces. I guess could be useful for recycling but I can’t imagine resistor manufacturers would want to help with that. Maybe they were more useful when a dmm cost more than the batteries that come with it?

        I remember a class I once had where we identified resistors for a lab. They used some little 1/8w resistors that must have been 20+ years old. Couldn’t tell brown from yellow from orange from gold, even with a microscope. Instructor didn’t even know what they were so he just gave everyone 100%. Very useful stuff!

        1. Hm, well the bands work far better than a number would for cylindrical components.
          But all of your arguments aren’t really about the color bands, they are all just arguing against any labeling of any form really.
          I guess if you never had to reference a parts labeling once it was in circuit, more power to you! (pun totally intended) but I for one am glad they are there.
          If you think about older chips with engraved markings that are a pain to see, that’s more akin to a problem with the colors specifically. But now imagine that problem, plus even the poor engraving only shows up at 1 of 360 degrees of rotation of your head!

          1. Hmm. I work a lot with old German and Russian gear and I have no problems with resistors sporting their value in ‘plain text’. But I have to admit that I also get by with the color codes.

          2. In a way that is correct; it’s a part that probably doesn’t need markings. We live in a world where half the chips we get on a board don’t have datasheets available and even a lot of discreet components lack any kind of identification. I know those stipes can’t cost much to put on, but I also know resistors are very cheap so it could be a significant percentage saved? If nothing else, think how many hours schools waste on this…get rid of the markings and those classes turn into something useful!

            For my part, I generally buy the cheapest resistors that fit my requirements. Sometimes they have blue bodies that make the lines impossible for me to read even new. Don’t care at all, actually like those from an visual appeal standpoint. Cut $0.01 per 100 units off the lowest price and that’s the part I’ll buy, even if it has no markings.

            Or here’s a bit of an oddball idea that manufacturing might like… offer every resistor value in a few different colors. That way if your product has maybe 3 thru hole resistors you can buy them in red, green, and blue bodies and it’s both harder to use the wrong one and easier to check that the right one was used! Industrial machine vision cameras are expensive and the price difference between a camera that can check the correct body color and a camera that can read the bands is massive. Not like we see hundreds of thru hole resistors on modern boards… usually they are just for a few high current items if they are used at all.

        2. Ever encounter a 30 ohm resistor that apparently decided to have its own little rebellion and resisted even more until it read up 330 ohms? It was the cause of a VCR not working correctly because it was in the circuit to the end of tape detector IR LED.

        3. Yeah but once you take them out of the bag, they’d be effectively useless! Sure, you’d use a multimeter, but with the stripes they’re independent of needing any sort of measuring.

          As far as their Gaussian distribution, I think you get “random” as in “random”. Unknown. That’s the point. So you don’t know the exact value of this one, or any of the other hundred. It means what it says and nothing more. Even if there were some sort of known random distribution in manufacturing, how would you guarantee that one they were bagged up? Random is random, there is no more information!

      2. I once had a cheap DMM blow a resistor.
        According to the schematic, it needed to be something like 111.1 ohms.
        (I had the schematic, because I built it from a Vellemann(?) kit.)
        I went through my resistors with a trusty DMM and found one at that value and soldered it in!

      3. That’s an interesting idea. But I wonder if by making various series it turns out not to be such a big deal?

        One mans 10% could be another’s 1% if the target value is shifted only slightly.

        I’d guess the change from the scenario you describe to the better one I do is down to scale more resistors produced and sold in a variety of series the easier it is to make low tolerance sets

      4. The distribution of values depends heavily on the manufacturer. Back in the bad old days of carbon composition resistors (a category Allen-Bradley came to dominate) a good manufacturer of carbon comps could get almost all of a whole batch within 5% or better as they left the factory, even if they were labelled 10% and sold as such. If the truck paused for a day beside a lake they might pick up enough moisture to fall in value below tolerance limits, and they’d have to be baked to get back in spec.

        Carbon film resistors were better and could be abraded to good accuracy before coating; the same is true of metal film and cermet and probably whatever technology is common today. Much modern trimming is done with lasers instead of abrasion, and there’s no excuse for a batch of 10% resistors having none within 5% if the manufacturer is not garbage.

        There are other problems. Resistors have temperature coefficients and values drift with age. If the resistor has to handle power near its rated power, its going to age faster. If you really need a 1% resistor, you don’t want to buy a resistor using the same technology that produces cheaper 10% resistors and was sorted to meet spec.

    2. Neither of those options is totally correct.

      Resistance is set by material properties and dimensions, both of which need to be controlled during manufacture. The chemistry is relatively easy to get consistent – volumetric and mass measurements have been well-understood for centuries, and a few labor-expensive measurements are used to make a large batch of components. The forming of the physical resistor is more difficult to do consistently and requires either measurement and adjustment of every part (very expensive) or the expectation that parts are allowed to miss the target by some allowed tolerance. That would infer a gaussian distribution.

      In time, the costs associated with that process have plummeted. Early resistors could have a tolerance of 20%, but the automation of manufacturing has allowed for tighter tolerances to be achieved at ever-decreasing cost. However, when you’re manufacturing 100,000 pcbs each with 100 resistors, a one cent price difference per resistor is $100,000, so looser tolerance parts are still preferable in most cases.

      The series were chosen such for a given tolerance, a measured resistance is almost certain to be closer to its specified resistance than the next value in the range (in either direction) could be. That is all. If I select a 47K 10% resistor, it could be 42.3K to 51.7K. The next value down in E12 (39K) could be 35.1K to 42.9K, and the next value up (56K) could be 51.4K to 61.6K. Even at the extreme edge of a 10% tolerance, if I want a 47K resistance, I am most likely to get it by selecting 47K. Now consider if the tolerance was 20%, but still selecting from E12. The 47K could be anything from 37.6K to 56.6K. The overlaps of 39K (31.2K to 46.8K) and 56K (44.8K to 67.2K) are significant – if I want a 47K resistance any of the 3 values might provide it. It is possible that a resistor marked 39K has a measured resistance higher than one marked 47K, and a resistor marked 56K a measured resistance lower than one marked 47K. To put that another way – it make no sense to select a 20% tolerance resistor using the E12 range because it brings you no more certainty of hitting your desired value than selecting from the E6 range would. Similarly, for a 10% tolerance, it rarely makes sense to select from the E24 range because it gives you no more certainty than the E12 range does. And so on…

      Even as process control has improved and the variance has reduced, the tolerance value remains only a guarantee. Although a 47K 10% resistor made today is much more likely to measure 47K than one of 50 years ago, an outlier could still be 42.3K. In mass production, guarantees matter. For most circuit design, resistor values beyond that guarantee are a moot point. Good circuit design does not require a component to meet a tight value unless absolutely necessary, e.g. high end analogue electronics such as pro audio, or test and measurement equipment.

      There are tricks that can be employed that rely on the above. Suppose I have a widget that I want to sell in three versions, a normal spec, a value spec, and a super high-end spec, and the design relies on a somewhat critical 50K resistance. I could achieve all 3 specifications whilst using 10% components from the E12 range. First, I design the circuit such that the 50K resistance is split across multiple components. The normal spec gets 47K and 330 Ohms in series. I rely on the modern tight variance to create a specified 50.3K resistance. During QA, that resistance and the resulting product performance are verified to specification. The budget version has looser product specifications that can be met by 99.9% of components, dealing with the egregious outliers through warranty programs. The super specification model has the 300 Ohm resistor replaced with a 470 Ohm trimming potentiometer allowing for factory and aftermarket calibration.

      1. I’m pretty sure your plan to use multiple components wouldn’t help, actually making things worse. If you’re unlucky enough to buy the widget where all the resistors were at the low end of their range, you end up with an even less accurate value. On average I don’t think it’d make any difference. But overall it will cause extra problems and solve none.

        Also you meant 3.3K but we all knew that anyway.

        1. Yeah, I meant 3k3. Brain-fart at the end of a long post.

          Two resistors vs. one creates no extra problems beyond the costs associated with its sourcing and installation. The distribution of two added Gaussian distributions is itself Gaussian. The means are summed, as are the variances.

          However, there is no 50k resistor in E12. Nor in E24. Or even E48. So when selecting a “50k” resistor, would you choose a 47k mean distribution or a 50.3k mean distribution with a doubled variance?

          This is easy to visualize in R:

          x=seq(40,56,length=500)
          plot(x,dnorm(x,mean=47,sd=sqrt(2)),type=”l”,lwd=2,col=”blue”,main=’Normal Distribution’,xlim=c(40,60),ylim=c(0,0.5),xlab=’R’,ylab=’φμ, σ²(X)’)
          curve(dnorm(x,mean=50.3,sd=2), add=TRUE,type=”l”,lwd=2,col=”red”)

          E24 series (typically 5%) would get you to 51k. Even there – two 5% resistors totalling 50.3k is a better choice:

          x=seq(40,56,length=500)
          plot(x,dnorm(x,mean=51,sd=1),type=”l”,lwd=2,col=”blue”,main=’Normal Distribution’,xlim=c(40,60),ylim=c(0,0.5),xlab=’R’,ylab=’φμ, σ²(X)’)
          curve(dnorm(x,mean=50.3,sd=1.41), add=TRUE,type=”l”,lwd=2,col=”red”)

          You’d have to go to E96 to get 49.9k, and that still isn’t tight enough to meet a super-spec for which you want the ability to calibrate.

          I did a cursory digikey search on 1/8W through hole resistors (because this topic was about color codes, after all). It’s just about impossible to buy 10% resistors today. So… a Stackpole 1/8W 5% resistor is $0.00589, a similar Vishay Dale 1% resistor is $0.0675.

          If you built 100k widgets it would cost $5572 to use a single 1% resistor vs. two 5% resistors. It would be a perfectly respectable design choice if that met the market needs of the widget.

          Realistically, my example was contrived – no-one in their right minds would select a 10% tolerance for a “somewhat critical” component, let alone a critical one, given the vastly better choices out there today. Horowitz and Hill would be aghast. It’s not just tolerance that matters – other properties such as temperature coefficients and moisture resistance are factors. In my defense, I didn’t say that I would, only that I could. The choice of the values in each resistor series harkens back to the days when you might be using 10% tolerance. Even still, the trick is still valid but today you’d be selecting two 1% resistors to get the mean closer to a nominal value only available in E192.. Commentary on eevblog suggests that variance in resistors is generally much better than could be expected from their tolerance specifications.

      2. Also “The series were chosen such for a given tolerance, a measured resistance is almost certain to be closer to its specified resistance than the next value in the range (in either direction) could be.”

        I think that’s for the manufacturers’ sake. It means they can bin all their resistors, whatever their value, and nearly every resistor they make will be within 10% of one or the other values. As others have said, the ones within 1% get sold as 1% but with a 10% tolerance there’s always a place for just about every resistor.

        The ones with 20% tolerance will, logically, have been the ones almost exactly in between the other values. Assuming everything that falls within 10% is removed and sold in the better tolerances.

        1. Chip resistors are batch sampled, not individually binned. A normal distribution still exists within each batch, due to NiCr thickness variations and the precision of the optical and motion control systems aiming the lasers used to trim them. Those in turn will be dependant on machine age. Newer machines have better initial specifications than older machines did, and machines incur wear that reduce their performance over time. My point is that it is generally more advantageous to replace a process line with newer equipment, and introduce a new product line with better specifications than it is to try and cherry pick batches from older lines. Older tolerances eventually become obsolete as a result.

          An example – Vishay TNPU chip resistors are available with a +/- 0.02% tolerance and 5ppm/C temperature coefficient. No amount of binning an older 1% tolerance product line could achieve that. It’s notable that higher precision chip resistors are typically thin-film NiCr, not thick-film, I’d guess that film thickness is the controlled variable most responsible for the increased tolerance.

          Precision wire-wound resistors are typically individually measured. The manufacturing process for them hasn’t changed much over the decades. The process variables there are core (diameter and cylindricity), wire (diameter, cylindricity and tension), and the mechanical precision of the spot welding and winding processes. Most of these are easy to control allowing for batches to be made with high precision. However, a spool of wire can only wind one resistor at a time. The resulting high cost of production makes the cost of individual measurement less significant.

          In all of this, it’s important to remember that good circuit design does not require critical tolerances unless necessary, and even those circumstances are much less frequent today. Few people would now build an A to D converter out of discrete components, it is generally more effective to buy an IC. Wheatstone bridge? IC. Audio preamp? IC. Analog system have been replaced with digital systems. Bipolar transistors with MOSFETs. Discrete resistors have mostly been relegated to supporting roles, such as pull-ups and protection, where critical tolerance is not required. I wouldn’t be shocked if there were more 102 and 103 valued 5% tolerance chip resistors in use than the total of every other value in any tolerance.

    3. I’m not sure about resistors, but I’ve spent an embarrassing number of hours hand-selecting 1% surface mount capacitors for a 0.05% tolerance application. (They were going to be used at low temperature, and it wasn’t possible to get more precise components.) Their distribution was definitely not Gaussian. In a batch, you’d get around half clustered tightly around some very specific value and a bunch of outliers. The next batch would have a very different central value. We didn’t bother to record the ones that didn’t meet spec, but my qualitative impression is that it was a huge spike at some value different from the labeled value by a fixed amount, with very wide tails in both directions.

      1. Again, they pick the better toleranced parts out of the lot first.

        It’s basically a gaussian distribution that is skewed to one side by some amount because the process isn’t entirely accurate, and then the bell curve is split in two at the nominal value where they took all the highest precision parts out to a different bin.

        If the cluster you measure is above the nominal value, then it should have a tail to the right, and if below the nominal value then to the left.

  6. What I don’t get is that nearly every beginning electronics book explains the resistor color code near the beginning. But hardly any of them mention a thing about how to interpret capacitor markings! Sure, electrolytics are pretty obvious most of the time but not so much disc capacitors.

    As a kid, growing up before parts were cheap and ordered online this and the lack of any markings at all on most inductors was a strong limiting factor in my budding electronics hobby. I dreamed of re-using parts from junk devices but most of them I couldn’t identify the values!

    And no.. L/C meters were not an option on a kid’s budget back then. Everything is so much less expensive now!

    1. Ceramic capacitors use the same system as SMD resistors… but give the value in picofarads.

      104 ⇒ 10 0000pF ⇒ 100nF
      473 ⇒ 47 000pF => 47nF
      331 ⇒ 33 0 pF ⇒ 330pF

      1. Sure, and it’s not too hard to find that these days given the internet. But why isn’t that information included in more beginning electronics texts when resistor color codes are practically everywhere?

      1. You couldn’t find the resistor colour code? Really every beginning electronics text gives it. Or there’s that rhyme which we’re not going to mention here… I’m really surprised at that, “how to find a resistor value” or something on a web search must surely come up with it pretty quickly! Or you’d only need to ask a single electronics geek. I think you were looking more off than on!

        Stuart, I was thinking there was something wrong, since the 1940s guide above mentions microfarads. But then I noticed, it’s spread over 2 lines and it’s not a typo… it says “micro-micro farads”. Didn’t they have ancient Greeks in the 1940s? I suppose they just didn’t have ISO standards!

        1. No. That’s not what is being talked about here at all! How did you read this far into the comments and not know what’s in the article?

          What isn’t in those books is how to read the old, antique style resistors where the color code is not stripes. Even if you know the colors what order do you read them in when they are body color, end color and dot? It’s not the usual left to right stripes.

          When was the last time you saw a beginning electronics text that included that information? I might give you the benefit of the doubt, thinking you have only read really really old electronics texts but then I’d have to wonder how micro-micro wasn’t old news to you.

          I’m thinking that somebody’s mind is operating a few kilo-cycles slow today huh?

        2. “Didn’t they have ancient Greeks in the 1940s?”

          I’ve never met a single person who learned Ancient Greek, or consulted an ancient Greek, before learning about electronics.

          And I even read Anabasis.

          1. You know mega, micro, pico. That’s the sort of Greek I’m talking about! “micro-micro farads”, like they hadn’t invented the other scale words, even though they had the capacitors that needed them. Wasn’t entirely serious.

            Yeah as far as the point about reading OLD resistors, mea culpa. Who knows what my brain was up to?

          2. “I’ve never met a single person who learned Ancient Greek, or consulted an ancient Greek, before learning about electronics.”

            Have you met married persons that did?
            B^)

    2. Some analog meters had capacitance ranges. Connect the capacitor in series with the meter and a source of 115 VAC 60 cycles.

      It would have been worth your trouble to build your own nulling bridge for capacitors, resistors, and inductors. With a few known good components for reference you could even have a calibrated measuring device.

  7. OMG Thanks Al!

    I have a couple cigar boxes full of things that look like the “Résistances anciennes annees 50.jpg” picture. I wasn’t sure if they were resistors or inductors and I had no idea how to read their color codes. All I ever knew was bands!

    1. I would guess than inductors that age would have visible wire windings. Although so might a wire-wound resistor! I really wouldn’t use those museum pieces in actual circuits though. I suppose if you must, measure their actual resistance with a meter. I suppose if they’ve drifted over the years they’re not gonna drift any further, but really, you can buy cigar boxes full of modern resistors for not very much. Maybe you could sell the old ones a few at a time to people restoring antique radios. Although again, anything that connected to the mains I wouldn’t want to rely on components that old. Even if I had an antique radio that still worked, I wouldn’t leave it plugged in unsupervised.

      They’re not valuable or rare enough to be useful to a museum, although some small technical museum might want them. They’re good as an example, you could show them, and modern resistors, and SMDs, to show how they’ve shrunk over the years. That we still use the same colour code is interesting in itself. You could keep them and hope they become valuable before your grandchildren are dead. Or after some future apocalypse where nobody can get components any more, and you build a 2-way radio and are hailed as a hero in your community. Apart from that though… I wouldn’t use them cos they’re too rare but just as much cos they’re surely useless and unreliable.

      1. “Maybe you could sell the old ones a few at a time to people restoring antique radios.”

        What do you think I had in mind the day I bought them?

        Everything I have read says that resistors in antique radios are USUALLY ok. It’s the capacitors that are best off replaced. So.. if I did have a radio with a burnt up resistor, perhaps the result of a bad cap failing closed then there is a good chance that one of these resistors would be a good, period correct replacement.

        I wouldn’t use it if it has drifted far though. I’d be afraid to trust it. Now that I know how to read their markings I can finally test them! If they are usable or useless.. I’ll finally know for sure.

        Still, though I appreciate and like to look at antique electronics I spend far more time with modern stuff. I only really own one antique radio, a family heirloom that I would like to restore one day. Even if repairing antique radios became my new main hobby I likely have a 3-lifetime supply of these resistors.

        They might be great for circuit sculptures. They are far more decorative than anything modern. Or maybe for making little QRP rigs that don’t get enclosures. The Michigan Mighty Mite comes to mind here.

        1. Yes but perhaps some parts of the radio can get hot enough to be dangerous within the power supplied through the fuse. Some parts were supposed to get at least warm. They’d sometimes use just resistors to drop mains voltage down. They used cloth insulation and wax back then, and safety standards weren’t what they are now. I’d be suspect of anything that old, really. Not saying you can’t enjoy using it, but maybe don’t put it on the shelf next to a bottle of methylated spirit, and stay in the room, listening to it, while it’s plugged in.

          I’d also perhaps want to rig up a little short-range AM transmitter and play some 1940s / 50s music through it, listening to the garbage on the radio now through a virtual antique would annoy me!

  8. A (possibly) interesting side note: one of the early radio manufacturers (Philco?) installed new lighting, probably something along the lines of mercury vapor lights. The increased light level was welcome, but it made certain colors hard to distinguish.
    Their solution was to have the engineers change some of the resistor values to less problematic colors!

          1. 40W halogen is not that much of light output, I prefer 40-50W of fluorescent or LED light.
            Although a few years ago I needed a short term solution for a night of SMD assembly. The 300W or 500W halogen lamp served this purpose quite well, except for it’s high heat output – it was already early summer begin and dripping sweat does not serve as flux :-)

      1. I think it depends on the phosphors. A single yellow phosphor, plus the blue LED they’re based on, would give “white” but I wouldn’t want to light my house with it. Especially not have to work under it identifying colours. Although, again, I’d just use a multimeter where that’s possible.

        If they add extra phosphors you get better light. The ones from proper manufacturers with English or European-sounding names, at least names in the Latin alphabet (!), designated for lighting, are likely to be better for CRI. Put in some red, green, etc, phosphors to get as wide a band as possible. LEDs are fairy monochromatic. Cheap noname Ebay LEDs, like every Ebay vendor that sells commodities, do it purely on price, which you can search by. Price – quality = profit so quality is the enemy there!

        It might be a problem a lot of people notice but don’t know exactly why. One of those things where public ignorance, and disinterest, allow in shysters. Like politics!

        1. The “better” ones you can find at any supermarket without special order are CRI >80. They basically just add a red phosphor to the yellow one. It’s difficult to find CRI > 90 bulbs anywhere because 80 is kinda the price cut-off point where these things are cheap enough to manufacture and most consumers don’t notice or don’t care.

          But that’s still pretty terrible. Even the old compact fluorescent tubes got CRI > 87 and to have actually good color rendering you want greater than 92.

    1. I had to do something similar once when we found out that our assembly workers were treating 270K (red violet yellow) and 4K7 (yellow violet red) resistors interchangeably. The 270K resistor was in an arbitrary RC time constant so we switched to an adjacent value (220K or 330K). Near enough!

        1. It does feel like the vast majority of old resistors on the planet began brown-black, red-red, or yellow-purple. If I had to guess, the next most common values would start blue-grey.

          Using just those E3 values in a gain circuit R/r, you can approximate gains important in log10 math – 1, 2, and 5. Using E3 series from just 3 decades, ie 9 total values, allows for gains approximating 0.01, 0.02, 0.05, 0.1, 0.2, 0.5, 1, 2, 5, 10, 20, 50, 100, 200, 500.

          A combination of two E3 in series ( or one with a 0 ohm jumper) can generate: 1, 2.2, 3.2, 4.4, 4.7, 5.4, 5.7, 6.9 and 9.4 which is a healthy start in a linear range. Adding that 6.8 value allows 7.8 and 9.0 to be generated too.

  9. I like the simple three-letter system for SMD parts, but the ”new” EIA-96 is a PITA really. No way I’d remember that stuff, always have to look it up. Measuring SMDs in-circuit does not always work, and removing them without damage when they are glued down and in confined spaces isn’t fun either.

  10. This is all going away soon anyway. Resistor manuf. have started to stop printing values on SMT resistors.
    For mainly a few

    #1) they can save $0.0000000002 by skipping the print process and save the cost for the ink.
    #2) caps don’t have values on them, and no one complains so why are we still doing it to resistors.
    #3) A machine installs them mostly, it does not need to read the values.

    In the coming years, you will pay a premium for resistors with markings.

    1. Removing marking is not in the plans for any of the numerous resistor manufacturers I deal with on a daily basis (at least 9 I can think of due to working at an electronics contract manufacturer)

      1) Cost is irrelevant for the value added by having the marking for inspection and other post processing. Resistor material is ideal for the added steps for marking, so that is why it is a standard process for resistors.

      2) Caps don’t have marking because of the material composition. Marking capacitors with standard methods does not lead to very legible results. You can get marked caps by the manufacturer using laser based processes, but that adds noticeable cost so it is not standard.

      3) Machines are only as perfect as the humans loading them and the traceability tools being used by a manufacturer. Flying probe or ICT style tests to validate assembly are not always used post SMT and functional tests can’t always capture when a 10k resistor was loaded in place of a 20k, but can affect the end user. AOI, AXI, and visual inspection utilized is a much more cost effective solution in managing quality.

      4) Yes, there are limitations to legibility on 01005 as well as most 0201s.

  11. The mnemonic – Better Be Right Or Your Great Big Plan Goes Wrong – (Black ,Brown, Red etc -Purple for Violet) has helped many learners. Of course a crusty variant also exists but this “clean ” version suits mixed company!

        1. Al, I’m glad you were able to step up to the plate for brother!
          B^)

          I just came here to say I found the file where I had stored it decades ago…

          Zij Zwart (black,0)
          Bracht Bruin (brown,1)
          Rozen Rood (red,2)
          Op Oranje (orange,3)
          Gerrits Geel (yellow,4)
          Graf Groen (green,5)
          Bij Blauw (blue,6)
          Vies Violet (violet,7)
          Grijs Grijs (grey,8)
          Weer Wit (white,9)

          Something like: She brought roses to the Grave of Gerrit with dark grey
          whether.

          1. Hm, you’ve got BRacht for BRuin, GErrit for GEel, and Grijs for Griijs of course! So the ambiguous ones, that start with the same letter, are well enough distinguished! Bij and Blauw don’t give much of a clue, but it’s the only B that isn’t bruin.

            That’s pretty good! If I spoke Dutch I’d learn that! I wonder how close the phrase is to working in German? I wonder what the French and everyone else uses as little mnemonics for technical stuff? Particularly the resistor code, which is international even if the little poems aren’t. Interesting post, really it’s no surprise that other nations would have the same thing, but it’s interesting to see it.

    1. The one that is stuck at the back of my head from when I first learned about the bands is totally unrepeatable. But I’ve never forgotten it once in 40+ years, because it is totally and utterly inappropriate.

      1. My old electronics lecturer always swore he would give up lecturing when ladies joined the class because he could not teach the resistor code, and I am told he changed subjects.

      2. Yeah, I learned that one too, and I never forgot it. Guess that was the point.
        I now use “Better Be Right…” with our new hires. The “classic” one is in Wikipedia, I made sure of that.

        A female co-worker once confided that she could never remember which was the male connector and which was the female. I paused for a minute and tried to think of an answer which would meet with HR approval (i.e.: not get me fired), and was unable to. I think I answered something like “Male is the pins and female is the sockets”, which satisfied her. But it was a close thing.

        1. It’s not always that easy. Some families of military connectors can have either pins or sockets in the male connectors and vice versa. The determination of sex had to do with the shape of the connector’s shell. Not fun.

      1. The evil, filthy, and offensive version is actually better. Because BAd BOys coincides with BlAck and BrOwn. BUt goes with BlUe. So the 3 “B” colours are encoded in the rhyme. It’s a very useful mnemonic! And of course “Violet” is violet!

        One where the first two are both “bye” doesn’t have that quality. And the evil, Nazi rhyme that kills babies is of course very memorable! It’s not like anybody was actually Red in the making of the rhyme. I’m pretty sure Violet isn’t even a real person. It doesn’t have anything to do with the actual crime so I dunno who’s actually supposed to be so bothered about it. I think originally it was anti “bad” language campaigners cos it’s a bit smutty, but frankly fuck them, we’re grown-ups. I don’t think any actual victims of Red would genuinely feel trivialised or mocked by an old electronics geeks’ aide memoire.

  12. > yellow lines down the middle of the road

    That just means it’s an alpine road. Only place I’ve seen such markings is around the Snowy Mountains, notably Charlotte’s Pass, Threadbo, Mt Selwin… Worth noting that a lot of these places are a bit on the over-done side having just gotten a roasting in the recent past.

      1. Where yellow means “traffic coming the other way on the other side of this line” and white means “two lanes going the same way”. Standard says it must be a double yellow.

        1. In some other countries, yellow lines means “do not overtake”. White line means “Intersection, don’t cross the line anymore”, and dashed white line means “Drive wherever you want.”

  13. Some additional comments:

    Early resistors were generally 20% tolerance (or worse). Sometimes a 4th band was used for tolerance, but it was not standardized. “Standard” values varied between manufacturers. For example, Centralab used a 100 200 300 400 500 600 750 800 900 1K sequence (so you could get a 400 ohm 20% resistor from one manufacturer; but not another).

    When EIA coding began, a 4th band was added to designate tolerance. Silver = 10%, gold = 5%, brown=1%, red=2% orange=3%, yellow=4%. Military standard MIL-R-11E added a 5th band designating the failure rate.

    A wide first band indicates that it is a wire-wound resistor.

    Capacitor coding was never as standardized as resistors. My pre-WW2 manuals only need a half page to document resistor color codes, but 5 pages for capacitors!

    It’s fairly easy to measure inductor and capacitor values if you are willing to fiddle a bit. Build a bridge, with a potentiometer in one leg, and two capacitors (or two inductors) in the other; one a known value, the other unknown. Apply any source of AC; audio for larger values, RF for small ones. Adjust the resistors to find the null between the two legs of the bridge. The ratio of the resistors matches the ratio of the known and unknown capacitor (or inductor) values. Headphones or an audio amp will find the null for audio; or an AM radio for RF. One of my earliest projects was such an RLC bridge, built from plans in a magazine and an old AM radio.

  14. The color code was developed so electronics hackers could have the neatest almost naughty password ever that they thought would never ever ever be guessed by anyone. BBROYGBVGW

  15. “….Newton included it because of his interest in the occult, apparently. ”

    What a slanderous accusation! Newton must have had “interest in the occult” because he liked the number 7? Have you never read a Christian or Jewish Bible? The number 7 is everywhere! Newton was a Christian! It is historically ignorant to say he had “interest in the occult!” But of course, modern irreligious people have “interest” in historical revisionism, even in a seemingly-innocuous (and otherwise) excellent article on the history of component color coding.

        1. Right! It’s original meaning is “hidden, obscured, blocked from view”. Now it mostly refers to magic and heebie-jeebies etc, but it comes from the idea of science / religion (both accepted as truth back then) having a hidden obscured side that you have to look for and research using all sorts of weird, but seemingly appropriate, methods.

          If you’re already superstitious enough to completely believe a supernatural, religious explanation for how and why the world is, it’s not a great stretch to go a bit further and look for the magic powers that have been mumbled about and written down in arcane grimoires, for years. By someone who was dafter and madder than you were! Plenty of ancient books are full of nonsense on how to magically achieve the impossible. After all, how else could one possibly bend a spoon?

          They didn’t have the Scientific Method, and most of the scientific community was involved in the same kind of daft old rubbish. It’s during the Renaissance and after, that things got more systematic, and “scientists” started to join together to try make the whole thing more rational, and based on demonstrable and tested principles. Newton got a lot further figuring out gravity than he did summoning angels to do his bidding.

          1. They did have -a- scientific method, which is the application of logic and evidence to deal with matters of rational interest. That’s what’s always differentiated the occult and the magical – the fact that you don’t question it.

            THE scientific method we have today is actually pretty young, dating to around 1950’s with Karl Popper. The standards and particulars have been changing all along, and it’s not an universally held standard like the faces on TV would like to tell you.

          2. Also, Newton didn’t really “figure out” gravity – he developed the mathematical methods to describe it accurately and efficiently. Other people had already described the properties, like the fact that falling objects accelerate, or that objects have inertia. Newton formalized these disparate observations into a mathematical framework that explained the motions of celestial objects etc.

            It’s a kind of a myth that Newton alone came up with the whole thing when an apple fell on his head. The apple part obviously, but more importantly the part where Newton single-handedly “invented” gravity and the laws of motion. He too was standing on the shoulders of giants.

    1. Everybody was a nominal Christian back then, if you didn’t want taking off and hanging somewhere. Not long before Newton you had to be the correct subtype of Christian or else off to the torture chamber you would go. Similarly Galileo got so much shit for pointing out that it’s obvious when you actually measure it, that the Earth isn’t the centre of the Universe.

      Besides which you can’t really be interested in the occult without some sort of religious or other irrational belief, it’s the same sort of idea. Particularly the sort of Kabbalistic / summoning angels and demons stuff Newton was definitely into, along with many of his fellows. For someone who lives in a religious world, where religion is accepted as truth, it would be no different from any other sort of science. Nor would the occult, simply the less popular, “hidden” side of the supernatural. Trying to figure out Enochian, and alchemy, was done alongside all the other science. I’m sure plenty of people thought gunpowder must have ground-up demons dissolved into it.

        1. That said… thanks for the info! I’ll look that up, because absolutely everybody “knows” Galileo was persecuted for his heliocentric theory. If that’s not the case I’d like to know what is, and also how the incorrect version of the story got out so far.

      1. He was clearly deeply devout, and his views were actually pretty mainstream at the time.

        The word “occult” would only occur in a modern analysis. It was merely “bible study,” and he was doing it the same way as most others. Many people wouldn’t consider it occult at all unless the views were held in a later period when they were no longer standard.

        He was born Anglican, but his personal views were more mainstream at the time than Anglican teachings.

      2. >”Similarly Galileo got so much shit for pointing out that it’s obvious when you actually measure it, that the Earth isn’t the centre of the Universe.”

        Actually, no. Galileo tried to support the heliocentric view with a false theory about the tides, claiming that the tides are caused by the centrifugal force of the earth around the sun, but this would only explain one of the two tides, where the other one is explained by the gravity of the sun which was unknown at the time.

        So Galileo’s theory and the observations about the tides were actually mismatched, and Galileo couldn’t explain why – but he kept on pushing the point and criticizing the church anyways. It’s only in the modern narrative that Galileo is presented as a martyr of science, when in reality his behavior would give him the nomination of a crank even today. It’s one thing to be half-right, and another thing to prove as much, because the proof is the pudding.

        1. No, because “the proof of the pudding is in the eating”! Makes perfect sense, the definitive method of determining if a pudding is good or not, is eating it. I do hate when half-forgotten sayings end up spreading into popular use. Americans are terrible for it. Either the original version has been forgotten or somebody picked up half of the phrase to start with, and didn’t bother checking before they started using it.

          So you end up with “I could care less” instead of “I couldn’t care less”, where the incorrect one doesn’t actually make sense for the point it’s intended to make. We have access to endless realms of knowledge in our pockets now, it takes 2 minutes to look something up! People, make the bloody effort!

      1. There’s those “psychic” “churches” that keep trying to bother the dead. I bet it’s like when the Jehovah’s Witnesses come knocking, the dead all hide behind the couch.

      2. All churches did numerology until recently, and many still do. But not during the standard service. In the consultation rooms, on days with no casuals around.

        Alchemy is a solved problem now. Churches don’t have any reason to do it, because why would gold made by man be more desirable than gold made by God? They don’t have a use case. Newton wanted to understand the physics involved in the transmutation, he had a solid use case.

        1. Man made gold is just too damn expensive compared to natural one. Yes, you can use mercury as a precursor, but it has to be a special rare isotope. Otherwise your gold is radioactive and decays. Also not everybody has a nuclear reactor at hand.

        2. Why would gold made by man be…? Because lead is common and cheap, and gold is rare and extremely expensive! So if you could make it from lead, using whatever chemistry you had, you’d be infinitely rich! It’d be like the time Musa Mansa went to Mecca, and spent that much gold he depressed the value of it for ages after.

          What country wouldn’t want infinite gold? Why do you think they spent all that effort looking for Eldorado? Most of Europe was at war back then. A system to produce unlimited money, it’d be like cheating in some RTS game!

          If it’s indistinguishable from the divinely manufactured stuff (a real case of “fiat” money!) then of course you’d want it! Of course there’s a use for mountains of gold! Alchemists aimed for making gold entirely for personal gain. Same way they were searching for immortality and a panacea to cure all ills. Getting angels and demons involved would seem like a good idea, cos they could work outside the normal rules of reality.

          1. They weren’t stupid. Of course if you could manufacture gold, it would lose all value because you’d be up to your ears in it. That’s the whole point of why alchemy was an occult business – whoever discovered it wouldn’t have wanted to reveal that they got it.

          2. Of course. It wouldn’t lose all value unless you made mountains of it. Yeah there’s supply and demand, but gold is special because it was used as currency. Back then it had no practical use, beyond making decorative items like jewellery. Even now that’s mostly the case, the amount plated onto socket contacts isn’t much in the scheme of things.

            So yeah you’d keep it secret yourself so your “friends” didn’t come round, stab you, and try work out how your equipment worked. But if you approached the King you’d be welcome to make tons of the stuff, and your government would just lie to the world and claim they had a big gold mine. Overall the relative value of gold might be depressed but your country would still have more than everyone else. Still a massive strategic advantage. You’d live in luxury as the King’s best mate with your own money supply. Your job would be to do your special thing to the big piles of lead you’d be given each day.

            The alchemists probably did think they were doing some sort of semi-divine work, they were religious men like everyone was. And open to trying any daft method.

        3. Just as a footnote, Alchemy had 3 main aims…

          1 – Chrysopoeia, the transmutation of base metals into gold.
          2 – Creation of an elixir of immortality, and as a corollary a panacea that can cure all ills, so you don’t spend eternity with the flu.
          3 – Create an alkahest, a solvent that dissolves anything. The corollary to that would surely be finding something you could store it in. It’s bad enough dealing with hydrogen fluoride! They usually tend to mix that up on demand rather than storing too much of it. I suppose you could do the same with your alkahest, keep it as 2 separate parts, til you needed something to gradually sink a hole through the Earth’s crust, and end up with your own volcano.

          Mixing up every chemical they could find, and mixing in religion with science and chemistry was common. One guy pissed on some sand, since humans are the most noble creature and gold is the most noble metal. Cos why not? He baked it and ended up discovering phosphorus, so it wasn’t entirely wasted as an effort.

          Apparently also when Britain went to war in the 17th Century, every hole-in-the-ground toilet was dug up for it’s contents, so the nitrates could be used for gunpowder. It would be interesting to find out exactly how that was done, would surely be fun for some gun fan to produce his own black powder. You could fire it with some ancient blunderbuss or matchlock or something.

          1. During the Civil War (USA v. CSA)
            The Confederate Army had people working in a cave (at one time a part of the Union Army was encamped a half mile away) putting guano (bat crap) onto “mangers” filled with straw to extract the saltpetre(?) needed for their gunpowder.

          2. Yeah I think you pour lots of water over the shit and piss, and collect it at the bottom. Presumably the straw is to filter out the solids. I dunno how well-rotted it needs to be. Apparently composting toilets use lots of straw, or sawdust. I think to allow lots of air in for aerobic bacteria. There’s (apparently) no unpleasant smell wafting up. So maybe the straw was mixed in and left for a while for the bacteria to break everything down into simple nitrates.

            Then where the potassium came from, to add to the nitrate for saltpetre, I dunno. I suppose literal pot-ash would do for that, dissolved in water.

            You can find saltpetre as a mineral but it’s rare. So back in the day it was synthesised from organic sources. Sulphur I dunno, can you dig it up? Charcoal for the carbon is the easy one!

            As a kid I made black powder with a friend who was a big chemistry geek, reading a book on inorganic chemistry that was meant for much older students. Lots of fun! And a few other things, amazing what the garden centre will sell you! Ammonium nitrate for ANFO is usually not available! I think farmers have to be licenced to keep tight custody of it. This is because the IRA used to either nick it from farms, or else they, or their friends, were actually farmers and could order large quantities of it. Stick it in the back of a van, made into ANFO, and bang goes part of London!

            And now I’m definitely on a few lists for using those keywords! Dammit! If I told you the way to make ANFO, the recipe is in the name for fuck’s sake, I’d probably end up in prison for encouraging t-rr-r-sts!

          3. The process of making saltpeter from manure is about bacteria breaking down urine and feces, producing ammonium gas. The ammonium reacts with calcium and potassium carbonate from wood ash, bits of limestone, etc. to produce calcium and potassium nitrates, which are then washed down with water to extract the nitrates. The water is boiled off and the somewhat mixed saltpeter is collected. Potassium nitrate is the preferred one, while calcium nitrate would work (but it’s a bit sketchy for gunpowder).

            With guano and other bird droppings, the feces is already rich in ammonium, calcium and potassium because the birds excrete these in a concentrated form. With further breakdown of the uric acid, they are a potent source of ammonium for making pure potassium nitrate. Within the guano mines, you can also find the stuff on the walls of the cave in abundance.

    2. As it also says behind the link, the real reason he used 7 colors was to make a “color octave.”

      The stuff about the occult is just slander based on a misunderstanding of his study of alchemy, which was actually a wild goose chase trying to discover quantum mechanics, if you peruse his writings on the subject and apply “theory of mind.”

      1. How would he be trying to “discover quantum mechanics” which is a 20th century concept in the first place?

        They weren’t even aware of atoms yet (it was postulated but not in the same sense), so the questions of quantum mechanics wouldn’t have made any sense.

  16. I’m a colorblind technician and it’s no fun. Over the years I’ve memorized a number of rules for dealing with resistors, such as: If the first band is Yellow (4) the second is always Violet (7), never Blue (6). Blue (6) is followed by Red (2) or Grey (8). Brown (1) – Red (2) is a valid combination, but Red (2) – Brown (1) isn’t. White (9) is always followed by a Brown (1) and a Green (5) band is always followed by either a Brown (1) or a Blue (6), never a Red (2) or Violet (7). Of course, this only works up to the E24 series (which is what I see most often) and it’s not a perfect solution since Brown-Brown and Red-Red are both possible combinations.

    The worst time I ever had was terminating multiple 25 pair telephone cables, often mixing up blue with violet, green with yellow and brown with red. I ended up building myself a test box with a 50 pin connector and 50 terminals to which I could attach a sounder and “Ring-out” the cables from across the room.

    I get especially annoyed when a manufacturer uses a single multi-colored LED to indicate status. Is it green, is it red, is it yellow? I often haven’t a clue and have to depend on coworkers or a phone app, which sometimes makes mistakes if the LED is flashing or particularly bright.

    I’ve often thought colorblindness should be considered a disability, since there are so many occupations a colorblind person can’t do. No medical work (is the patient jaundiced or flushed?), no piloting of aircraft or ships (are those marker lights green or red?), no police work (what color was the car?), no biology, geology or chemistry (what color was the sample/solution?), painting (can you match this color?), video work and computer graphics (do those faces look a little green to you?), and so many more. Many occupations depend on good color vision and there really is no way to properly compensate for colorblindness.

    And yes, I’ve tried those expensive colorblindness glasses. They alter what you see so you can tell that two colors are different and will allow you to pass an Ishihara test, but they won’t tell you which color is which.

    1. I agree, colorblindness is no less debilitating than partial sightedness or partial deafness, and could be much better accommodated by society. It’s ridiculous that so many systems rely on the ability to distinguish two colors that 8% of males of Northern European descent do not have.

    2. I’m color blind to so use other techniques.

      Early resistors were a challenge to read by shade. Using a consistent high color temperature light helped and you could get used to the variations between different manufactures.

      Then for many decades the shade was consistent between manufactures under a 5700 k light.

      Then when China started making them things were impossible.

      I also worked in communications.

      High density cables like 1200 pair had other reading methods.

      The color bands had different lengths and every second color band had jagged edges.

      The plastic bundle wrap identifier had a different width.

    3. You sound severely colour blind, much more than the ordinary kind. I know red-green is the most common but I think that’s more of an axis, surely most colour blind people can tell one from the other. My brother is colour blind, and he has trouble distinguishing some colours but he can tell the basic colours of the rainbow.

      Perhaps the glasses could help you if you practised, once they’re on, teach yourself some red and green things, look at lots of examples. They say it’s hard for people to distinguish colours they don’t have words for in their language. In your case you’ve missed out on a lot of practise, since you were little I suppose and learning your colours. So you don’t have the practice remembering colours cos you can’t see them!

      The glasses are interesting, optical-wise they’re notch filters. They cut out colours in such a way as to remove ambiguity, colours are separated more clearly. But certainly part of colour perception is psychological.

      The phone cable thing sounds a nightmare! Now you can get testers for network cables that’ll tell you which one is which out of 8. A gadget goes on one end and the handheld tester on the other. It sends out some kind of coded pulses. You could do the same thing with an Arduino and a few shift registers. Or even the 8-wire one might help if you split them into groups of 8. All the other wires just reading as not connected. That’s if you find yourself in the same spot, or a time machine. Actually these days telephony can all be done over a network cable, so 8000-pair cables are happily obsolete, I’m sure that’s a source of joy, both to the people who had to work with them, and the maufacturer who had to figure out how to print a rainbow of dozens of colours onto each one!

      It might be debilitating but it’s not really a disability, though it excludes you from some jobs. And again, most colour blind people aren’t as severe as you sound. Mostly though you could surely just ask someone to tell you what colour a thing is. You can’t say the same for borrowing someone’s legs, or their brain functioning, or their skeleton, or lots of other things. As things go it’s a mild disability at best, bordering on inconvenience compared to being blind or having limited mobility. You can live a pretty full life and if you can’t do a particular job there’s surely something similar you could do. You managed in electronics despite having a pretty bad case of it.

    4. Actually… you’ve worked out the rules that this-colour is never followed by that-colour… what you’ve done is encode the E12 or E24 series into a set of rules.

      Why not just get a list of the values for each series, and write down “470K yellow violet yellow” as words, to the right of each value? Make a chart. For E12 and E24, depending on which range you’re working with. Then you’d be able to reference your efforts to check the colours with a list of what they could possibly be.

      Perhaps split it into two, one part being the value from the first two stripes, the next being the multiplier. So you’d need 12 or 24 on the left, and 10 values on the right.

      1. Wow!

        If you’re number blind then what number is 2?

        If your colorblind you don’t bother with colors at all. If the colors fit into a grey scale then why bother converting the shade to a color and then the color to a number. You just see the shades as numbers and you know the valid combinations. The shades don’t have names like colors do.

        If the shades don’t fit into a grey scale then grab the DMM.

        1. Colorblindness does not work like that – things aren’t shades of gray. They still exist on a spectrum of hue and brightness, but instead of three reference points you typically have two.

          In other words, a color that is equally saturated red or green appears to be the same color. This is because an earlier evolution version of color vision relied only on the difference between yellow and blue.

          The difficulty is that “equal saturation” depends on the ambient lighting spectrum. You do not see a consistent “shade”.

          1. Wow, Seriously?

            Let’s get some basic facts right.

            Color blind people know more about color than others just as a blind person with a walking stick will know more about the pavement than others.

            Color blind people don’t have any lack of perception.

            Color doesn’t actually exist, the sky is not blue and the grass is not green. These are only perceptions created by your personal biology. Other people have different personal biology and this creates a different perception.

            The distribution of perceptions falls under a bell curve where a much higher percentage of people are clustered towards the center.

            This has caused all conventions to do with color like the frequencies or bands to be chosen to have names or become more common or the frequencies or bands to be emitted by lighting or rgb LCDs or LEDs and absolutely everything else to be designed to suite those people at or close to the center of bell curve of perception.

            I am in the less than one percent region of the extremities of the bell curve.

            If all of the above things were designed to suite my biology rather than the center of the bell curve then 99% of the population would be colorblind.

            I have techniques that better enable me to fit in with a world that is designed for people that have far different perception than I. I use these techniques simply because they work for me.

            You as a normal color perceptive person are telling me as a color blind person that my techniques are not valid. That they are incorrect.

            Like I said color blind people have a far greater understanding of color than normal color perceptive people.

            We get stupid questions like … If your color blind then what color is red.

            Then we have to explain to normal color perceptive people in terms they can understand within the framework of their colour perception that color is actually a lot more complex than their understanding of it and then explains to them how their own color perception actually works so that they can understand the difference.

  17. If anyone can invent a better system of marking resistors, I suggest they patent it. I can inspect a resistor, and instantly know it’s value…Don’t need no damn sentences to know. Does the ohmmeter say less of more ohms? Parallel or resisdule voltage throwing off reading? Just like eliminating gasoline powered vehicles, the color codes will be here forever

  18. I recommend have a multi-meter handy at all time. I know the color codes for some standard values such as 1k, 10k e.t.c. but always check the resistance of a resistor with a multi-meter. I’m not color blind but I’ve seen many resistors with color bands that could be interpreted as more than one color e.g. either red or orange for instance or red or brown.

    1. I’m not colourblind and I never bothered learning the colour bands. I know THE RHYME but that involves counting on your fingers as you recite it, and then you get to Violet and start wondering what her phone number is, and then you forget what you’re doing. Also of course it’s a giant faff having to squint so much. Fortunately multimeters are cheap. You could buy ten and scatter them about the house. Or else maybe we could crowdfund getting them near-cost if we buy thousands of them, and just plop them around the streets of cities. Then nobody would have to bother learning it.

      1. The RGB colour theory is based on our receptors, yep. If we had an extra one it might be RYGB or the like. Maybe infra-red would give us more colours. And yep, when you’re looking at an RGB monitor or TV, you’re actually seeing separate R, G, and B. The colours don’t actually “mix” into yellow or purple, that’s all done in the brain. Which, as ever, can be tricked.

        But EM radiation is as solid as anything else in science. It’s definitely a real thing, and a rainbow is a rainbow and will always be the same, even if it is perceived differently. That’s our problem, not light’s. Rainbows came first, we evolved after.

        1. A rainbow is a continuous color fade.

          The color bands you see are from the peek frequency sensitivity of the three different frequency cones in your eyes.

          The peak sensitivity frequency points are different from person to person. When this variation is a lot different to average then they are called colorblind.

          There is color addition such as a TV screen and color subtraction such as paints.

          Paints don’t emit RGB. They absorb all colors except the specific color they are. If you shine a red light an a red wall the paint will look very bright. If shine a red light in a green wall the paint will look very dull.

          If you are color blind and use shade or contrast then narrow spectrum light like RGB lights will give you false contrasts as the color sensitivity peaks any you eyes don’t match the peak points if the light source.

          I always use a high color temperature broard spectrum light so that I can see by shade.

          1. >The color bands you see are from the peek frequency sensitivity of the three different frequency cones in your eyes.

            Peak and through. Some of the colors that are not present in the rainbow are caused by binomial distributions of the spectrum. Blue and red, but not green, produces pink and purple.

          2. >Paints don’t emit RGB. They absorb all colors except the specific color they are.

            They emit lots of colors – not just one wavelength corresponding to the specific color that they are.

            It’s more appropriate to say that they modify the spectrum of light that strikes on them, adding some peaks and subtracting other wavelengths by absorption, but never completely. Otherwise all objects would look neon-colored like in the movie Tron.

          3. Fair enough.

            Paints present a very narrow band of frequencies rather than a specific frequency. This still however completely different to emitting frequencies as far apart as red, green and blue.

Leave a Reply to preamp Cancel reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.