38

I always wondered why color coding is still used on resistors in 2014.

Here is Wikipedia's word on the original reason why :

Colorbands were commonly used (especially on resistors) because they were easily printed on tiny components, decreasing construction costs. However, there were drawbacks, especially for color blind people. Overheating of a component, or dirt accumulation, may make it impossible to distinguish brown from red from orange. Advances in printing technology have made printed numbers practical for small components, which are often found in modern electronics.

However, like pointed out in this quote, printing small numbers on electronic is now quite an easy thing (or so it seems) and it would, in my opinion, be much more convenient, especially for colour-blind people.

Is there reason why we still use colour coding on resistors in 2014?

The Photon
  • 126,425
  • 3
  • 159
  • 304
DylanM
  • 543
  • 1
  • 5
  • 12
  • 21
    The only reason I can think is that I can read color codes without a magnifying glass. I've seen lettered markings on 1/8 watt and 1/4 watt resistors, and they are very small. Colored bands on the same size part are much larger and easier to read. (Actually, I don't need a magnifying glass. I'm so near sighted that I just need to take off my glasses and hold the parts up close to my eyes. Still, I don't have to do that to read color codes.) I can read markings 0603 size parts. I think they stop marking parts below 0603. – JRE Aug 21 '14 at 08:42
  • I agree with JRE - on leaded resistors, colour bands are far easier for me to read. – Andy aka Aug 21 '14 at 09:19
  • 3
    @JRE I would also not be surprised if the color coding is not only more easily seen but requires less thought to distinguish parts (i.e., friendlier to pattern recognition than a series of numbers) at least when the diversity of parts of similar size and shape is small. This *might* be useful when prototyping (when the convenience of using fewer part types might be more important than the cost, durability, or efficiency of the result). (Just a guess; I have never designed or developed electronics.) –  Aug 21 '14 at 13:07
  • 3
    The equipment used for making axial-lead resistors was probably made thirty years ago! – Spehro Pefhany Aug 21 '14 at 14:20
  • 2
    Have to disagree here - I'm colour blind and the red/orange/brown distinction especially is a nightmare (nevermind after 'overheating' per above). – OJFord Aug 21 '14 at 19:12
  • 5
    That I can answer in one word - TRADITION! – Bob Jarvis - Слава Україні Aug 22 '14 at 11:28
  • Despite the Wikipedia author's opinion, color banding worked well enough and was popular until this century. There weren't all that many possible combinations when parts were 10% tolerance, so you could guess the value anyway. (270 or 370? Only 270 was a standard value). With tolerances of 5% or better, there are a lot more standard values, so now they just print the numbers. SMT packages also had a part in this. – gbarry Aug 24 '14 at 06:11
  • @ Spehro Pefhany only 30 yrs? but use of such resistors with colorbands seems more older (from the time of valve radio).https://en.wikipedia.org/wiki/File:Old_Radio_Resistors.jpg#/media/File:Old_Radio_Resistors.jpg – Always Confused Jul 02 '16 at 04:51
  • @ Spehro Pefhany Wikipedia tells its around 1960s. https://en.wikipedia.org/wiki/Resistor#Carbon_composition – Always Confused Jul 02 '16 at 04:57
  • Some of this doesn't hold water. Tiny numbering is far more likely to become illegible over time, heat, dirt, etc. than quite thick colour bands wrapped right around the component. – user207421 Jul 02 '16 at 05:36

6 Answers6

46

The reason for using color bands on through-hole (axial) resistors is simple -- when they are inserted into the PCB, you can't guarantee their orientation -- there is no top or bottom. So you need a way to mark the value so it can be seen no matter how the part is oriented with the board. Color bands are perfect for this.

For this reason, I don't expect to see color bands disappearing from through-hole resistors.

With surface mount parts, there is a top and bottom, so the parts can have the value stamped on the top.

tcrosley
  • 47,708
  • 5
  • 97
  • 161
  • 5
    +1 - good point! Not so obvious, yet still obviously true... –  Aug 21 '14 at 12:06
  • 16
    just print the number three times... – nbubis Aug 21 '14 at 20:46
  • 5
    Axial capacitors still have text printed on them, or radial capacitors with legs bent 90 degrees. They just print the value multiple times. I think resistors just stick with tradition, nothing more. – user Aug 22 '14 at 10:21
  • @MoJo Just about all axial electrolytics, even small values like 1 or 2 µF are much larger than a 1/8 watt resistor (4.5mm dia x 10mm L for a 1 µF cap, 1.7mm dia x 3.3mm L for the resistor). – tcrosley Aug 23 '14 at 11:44
  • or just put a black mark on one side of resistor to indicate the -/+ and print number in between of the resistor – Muhammad Umer Aug 23 '14 at 18:15
  • 1
    @MuhammadUmer -/+ of a resistor? Do you mean capacitor? – tcrosley Aug 23 '14 at 21:34
  • i meant it to show orientation, even though it is pointless. – Muhammad Umer Aug 24 '14 at 15:47
  • FWIW, Vishay (Dale, IIRC) do sell axial resistors marked with text rather than color bands. If users found an advantage to this marking scheme, they'd either capture the whole market or their competitors would start marking with text. – The Photon Jul 15 '17 at 04:37
16

My opinion.

Through hole parts are often cylindrical. Hence applying stripes is simple in production. Resistors may have five stripes, tolerance, 3 significant digits, and order of magnitude. That is a lot of information to write, "22811" 22.8ohm 1%? In san-serif font what is 88818? It would be possible to read 1, 2 or 5, 6 or 9, and 8 rotated, i.e. upside down, and be confused.

As pointed out by JRE, a letter in the text is used to help ensure the text is the correct way round, e.g. 22K8.

However, while colour blindness is a problem for stripes, dyslexia is for text. I know of people who would write 'k' backwards without noticing.

Edit: There are many written languages which do not use the Arabic digits (in fact, Arabic digits is a misnomer, because the Arabic digits are different shape or value to commonly recognised digits in 'the west'). So printing digits isn't automatically better than a level of indirection via colour which is language independent.

One pretty thing about 7 of the resistor colour codes is it is in the same sequence as the colours of the rainbow (in some cultures), so many people learn that part of the sequence as children.

The through hole cylindrical part could have its wires bent in any direction, and coloured stripes are still legible from any direction.

If written text is applied in only one location, the part value may be obscured or invisible when it is in circuit. That would be a disaster for repair and inspection.

Hence, written text will have to be applied on all sides in order to be legible, which may well be more difficult and hence expensive to produce, and still awkward to read. Stripes don't suffer from this orientation problem. Stripes are simple to use.

Manufacturers have already invested in machinery to product components with colour stripes.

Where is the competitive advantage? There needs to be 'new' money to fund the change over to printed numbers, or it won't happen. I don't see anyone financial benefit.

I haven't seen any new robots doing pick & place with through hole parts. It is hard to imagine such a robot would have enough financial advantage over SMT to make it worthwhile.

The assembly staff who use the components will have to be retrained, and what benefit does the assembly company get?

Through-hole with printed numbers would have to show big benefits to replace coloured rings. AFAIK, SMT has displaced through-hole in the majority of products; robots are cheaper than labour costs humans for mass production.

"Show me the money" - It seems a very tenuous benefit to printing numbers vs lots of cost, especially when the investment has been made already, in a shrinking sector of industry.

gbulmer
  • 10,040
  • 20
  • 29
  • 1
    The through hole parts I've seen that had printed values were only printed in one place. Depending on how it landed, some of the markings were underneath the part and couldn't be read - I had to unsolder one end to read the value. Score another point for color bands. – JRE Aug 21 '14 at 10:33
  • @JRE - "some of the markings were underneath the part and couldn't be read" that is part of the point about writing on multiple sides. I've edited the answer, hopefully making it clearer. Thanks for providing an instance where it wasn't done!-) – gbulmer Aug 21 '14 at 10:39
  • The point about being hard to tell which way is up was also solved - the values all had a letter ar part of the value, and the letters show you which way is up. 2K2 or 22K or 1K1, that kind of thing. When the letter is right side up so is the number. – JRE Aug 21 '14 at 10:45
  • @JRE - I understand. Part of my point is, for resistors containing 5 pieces of information, text is getting quite long. – gbulmer Aug 21 '14 at 10:59
  • 7
    I don't know why you say "Robots don't do pick & place with through hole parts" -- pick and place was *invented* long before SMD became common. – Dave Tweed Aug 21 '14 at 11:50
  • Through-hole electrolytics are printed though, even the small axial ones which are more likely to lie down. Admittedly they're still usually larger than a 1/4W resistor, and the value is printed at least twice. Of course you don't need to squeeze as much data on a capacitor. Resistors of 1W and up are usually printed IME, even if cylindrical, but of course there's more area for the text. – Chris H Aug 21 '14 at 13:27
  • @DaveTweed - I agree my answer isn't clear. I'll fix that. However, I would very much like to see references to through-hole pick & place electronic-assembly robots. I have occasionally visited electronics industries shows, and I have never seen one. It seems like a more complex, higher-precsion, and slow process than SMT. – gbulmer Aug 21 '14 at 15:04
  • 2
    @ChrisH: Electrolytics are usually insulated with a heat-shrink wrapper, which is printed on *before* being applied to the part. – Dave Tweed Aug 21 '14 at 15:42
  • @DaveTweed good point (should have gone with my other example - some diodes). But still printing onto heatshrink that then must be shrunk puts limits on the text size that would be tight on a component the size of a 1/4W resistor. – Chris H Aug 21 '14 at 16:19
  • 1
    Axial through-hole parts are generally supplied on reels or in "ammo packs", with a tape at each end connecting all the parts together (so the parts form the "rungs" of a "ladder"). As such, machines don't "pick and place" them, but instead use a die to cut and bend the leads and force them into the holes on the board (for each lead, the machine would have an open-half-code opening which would be placed against the hole, so once the lead was forced into the cone, its elasticity would hold it against the cone as it was coerced downward until it went through the hole). – supercat Aug 21 '14 at 16:36
  • they should put qr code, and i should have virtual reality augmented contacts lenses with 60 mp camera that reads the tiny qr code that i can barely see and highlight the value on resistor for me in my eyes. – Muhammad Umer Aug 23 '14 at 18:17
  • @supercat - thank you for that explanation. Have you any references to companies or machines that do that. I think it is pretty clear that SMT involves less mechanical complexity, and SMT circuits have some better properties than using through-hole parts. – gbulmer Aug 23 '14 at 18:24
  • @gbulmer: I don't think there would be much reason for many companies to invest in new equipment for inserting axial through-hole parts; equipment for automated surface-mount assembly would be simpler and thus cheaper to produce. On the other hand, since equipment to auto-insert through-hole parts already has been built, it exists, and it works pretty well, it often makes more economic sense for a company to keep using an existing machine than to invest in a new one. Further, single-sided boards can be cheaper to produce than double-sided boards, and through-hole parts can go over traces... – supercat Aug 23 '14 at 19:02
  • 1
    ...in a way that surface-mount parts cannot. – supercat Aug 23 '14 at 19:03
  • @supercat - *"since equipment to auto-insert through-hole parts already has been built, it exists, and it works pretty well, it often makes more economic sense for a company to keep using an existing machine than to invest in a new one"* exactly the point my answer is making, "Show me the money". Unless their is economic benefit to changing to printing numbers instead of stripes, then "it ain't gonna happen". Even if there is, it has to outweigh cost of current investment and provide realistic incentives to change. – gbulmer Aug 23 '14 at 19:36
  • One point to the `88818` thing: well, while it may be a problem with certain fonts, it would be adviseable to use a useful one, and one which even distinguishes between 6 and mirrored 9, etc. Or just write `88818.` to mark the orientation. – glglgl Aug 24 '14 at 10:04
  • "what benefit does the assembly company get?" They can employ even color blind people. – glglgl Aug 24 '14 at 10:05
  • @gbulmer Videos of through-hole assembly: [Pick'n'place for large radials](https://www.youtube.com/watch?v=SNGIBCdh6rQ) | [Slo-mo of bending/insertion radial](https://www.youtube.com/watch?v=w2B0XpaxgII) | [Radial LEDs and various axials](https://www.youtube.com/watch?v=2sm_qdeconI) | [Various axial components](https://www.youtube.com/watch?v=LJ7OLJBTGQE) | [High-speed radial resistors](https://www.youtube.com/watch?v=oR8hPWa-ASM) | [Different style machine](https://www.youtube.com/watch?v=F1d9Y7SixD8) | [Scary fast panasonic](https://www.youtube.com/watch?v=Os2WEhqpj_k) – marcelm Jul 15 '17 at 12:02
13

My guess (and it is just that) - colour bands are easy to apply with a basic machine (rollers dipped in paint, component rolls past) and if the machine works, why change it?

The alphanumeric markings on things like SMD devices require laser-etching, printing or similar. I'd further guess that either the laser etching / printing is already an inherent part of the production of SMD components, or that it's easier than trying to colour band or print or whatever onto a flat/square component, or that when they developed the production system they decided it was no great extra effort to move from colour bands to alphanumerics.

TBH most components used, by bulk sales, will be fed in reels into a pick & place machine, as long as the label on the reel or package is correct the machine is not reading colour codes or part numbers.

John U
  • 7,041
  • 2
  • 21
  • 34
0

I would say that mass production also plays a part in why there are still Color bands as well.

There is probably a lot of these still left from some company mass producing them a long time ago, and if they are still good and usable why throw them away or attempt to print numbers on them, there is no need to change something that people are already acquainted with, that still serves the purpose.

Otherwise, I would say it is because they are round and have no top or bottom like @tcrosley said.

Malachi
  • 103
  • 5
0

TRADITION. Nothing else.

Maybe the banding technology is older and cheaper than alphanumeric etching technology (and in some alphanumeric-coded capacitors and resistors where the value is printed with a blackish-ink, and NOT etched), the if the ink of 1 character got erased some-how, the small parts becomes almost unusable.

But if once the modern etching methods get popularized, it will no longer remain so-costly.

Color-codes are quite useful in identifying Qualitative informations (Live-wire, Neutral-Wire etc), but they're indeed not good to encode a quantitative value.

If the both methods carry some or some advantages and disadvantages (Such as in alphanumeric method, there are few disadvantages,such as; some characters may be eroded away with time, or could not seen from all directions, and there is also the scope of confusion in similar-shaped letter such as among numbers: 6,9; 2,5; 1,7 etc; and among letters (for-say if letters required in here or any-other alphanumeric code) such as b,d,p,q; C(Capital),c(small); and among Letters and numbers (O (capital-O for Owl),0(zero) ; l (Small-eL fore Logic) , I (Capital-aye) for iodine, and 1 (one) etc. ) ; the solution should be done by a new convention or meet, in few very simple way, like 1. Using both-the method on same time (so-that, from colored bands, the value could be seen from all-direction), as well alphanumeric method will be the main-basis to determine values; and 2. Using some distinction-mark (strike-outs, dots etc). to distinguish similar-looking characters (as they done sometimes) with well-publishing and popularizing the meaning of such notations.

Always Confused
  • 1,591
  • 6
  • 25
  • 49
0

Why change as we had a good system for over 50 years and most of us had a rhyme to remember the order of the colour codes. With the old system all I had to do was right down the coloured bands look up the colour codes and within a couple of minutes you had the value of the resistors.. Unlike today with the smd's you have to scroll through multiple columns of code to find what you are after. You have even got an app on your phone for the colour codes..