We’ve all been there. A big bag of resistors all mixed up. Maybe you bought them cheap. Maybe your neatly organized drawers spilled. Of course, you can excruciatingly read the color codes one by one. Or use a meter. But either way, it is a tedious job. [Ishann’s] solution was to build an automatic sorter that directly measures the value using a voltage divider, rather than rely on machine vision as is often the case in these projects. That means it could be modified to do matching for precise circuits (e.g., sort out resistors all marked 1K that are more than a half-percent away from one nominal value).
There is a funnel that admits one resistor at a time into a test area where it is measured. A plate at the bottom rotates depending on the measured value. In the current implementation, the resistor either falls to the left or the right. It wouldn’t be hard to make a rotating tray with compartments for different values of resistance. It looks like you have to feed the machine one resistor at a time, and automating that sounds like a trick considering how jumbled loose axial components can be. Still, its a fun project that you probably have all the parts to make.
An Arduino powers the thing. An LCD screen and display control the action. If you want some practice handling material robotically, this is a great use of servos and gravity and it does serve a practical purpose.
We have seen many variations on this, including ones that read the color code. If you ever wanted to know where the color code for resistors came from, we took a trip to the past to find out earlier this year.
22 thoughts on “Resistors Sorter Measures Values”
Just throw them in the bin
May you be cursed with remembering the exact value of discarded components, the moment you need them for a time critical project.
Sure wish they would finally ditch the color code that only exists for weird EE cred now and use printed numbers like every other component. At least a lot of them use blue instead of beige for the background now.
-a colorblind person
I did a test once at work and let 4 different people read the resistor color code, just the colors, and I got 4 different answers. +1 for printing the values.
What kind of work do you do exactly?
Assuming that at least one person got it right, this means that 75% is color blind? Although it is statistically more likely that the subjects of your “test-group” aren’t as technically skilled as the component itself requires. Or perhaps you biased the “test” by tricking them in using the most difficult to read resistor you could find.
Printed values aren’t perfect either and the color band system worked perfectly fine in the electronics industry for many decades. If you want numbers, switch to large SMD somponents. Reading color bands is a skill (assuming you aren’t color blind) and every skill needs practice to produce reliable results. Keep practicing…
I work for a electronics manufacturer.
3 of the 4 coworkers were skilled technicians.
The problem was that the manufacturer of the resistors (I think it was Vishay or Yageo) used a colors that were hard to distinguish from each other. In this case red and brown…
Yeah I really don’t understand why they can’t just have both. 3 character code in the centre with colour bands too. Not why this isn’t a thing
Asking for color code to be killed is like asking for the metric system to be killed; it is a globally recognized marking system. As for blue bodies, they have been around for many years, but more so with high precision or mil-spec resistors. In the past few years, they are becoming more common for standard values as you say though.
As for your comment about being colorblind, I completely understand your frustration. Resistor color code can be a major pain as opposed to an aid for everyday usage. I have worked with colorblind people that routinely used TH resistors, but only one had come up with a solution. She made a setup with a some sort of vision system connected to a desktop that showed the part magnified on screen as well as identified the value based upon color recognition. The resistor had to be manually located to the center of a holder, but it worked well. I have no clue what the system was, but it was roughly 10 years ago.
Nowadays, I would think you could do something similar very easily with an RPi.
They moved on and changed more than 20+ years ago, but you are still using through holes packages. SMD resistors have printed values (until they got too small).
I’ve got no problems with smd parts and use them every day at work. But a lot of our customers have both smd and trough hole parts incorporated into their designs.
I don’t know their exact reasons for this.
You don’t have to be colorblind. The numeric codes on SMD parts are soooo much more pleasant.
A million times this. I’m not even colourblind and I struggle to see the colours on them most of the time. Plus, enough time passes that I can never remember how to read the code so I have to look it up every time. It’s a mess.
You might not think you are color blind, but could actually be. I only found out when my optometrist had me do a “paint pots” test, where you have about two dozen round blocks, each with a small patch of color on top. You are supposed to line them up so that the closest colors are together, which results in a circle. I ended up making a figure-eight. Basically, while I have no problem with good pure reds and greens, I have trouble telling brownish shades apart, especially in small patches that are not connected, worse with a bad background color, which is what you get with the modern beige resistors.
I didn’t have as much problem with the classic dark brown cylinder resistors, and those usually had wider stripes too. Yesterday I was looking at some 1-watt resistors salvaged from a power supply or something, and the stripes were hairlines 1mm apart, with a vaguely bluish background. I used a DVM to check their values.
I feel your pain, oh so so much. I’ve got a bin of thousands of these through hole resistors that I save for emergencies where I aim to enlist my daughter or anyone else nearby that isn’t colorblind, in identification. Admittedly, I haven’t yet gone to the bin, but I know that day will come eventually. Knock on wood, it will hopefully be years from now since I’ve been stocked up for the long haul.
I find it is kind of ‘zen’ to sort resistors :) But then again I can recognize the value on sight with the E12 series. E192 or 1% resistors I just measure. I usually sort them by multiplier first, then put them in the tray according to value.
When I make a pcb, I usually use SMD resistors unless it’s a special value or type (like in low noise applications) or needs higher currents.
Hmm nice idea but for color marked devices it’s best to use the color codes.
1 The color codes aren’t JUST the resistance value, they are also which e series is being used see
2 The color code can be used to see if the resistor is
A) defective (IE the value is wrong or the resistor is flawed)
B) usable (after testing the value it has)
The best way to think of this as empirical testing versus supplied information.
It is best to use both.
Marking is a serious issue, first it is likely the colors are designed to be in a specific light, it is unlikely they they choose random colors for marking (and also just plain ignorant). It would make sense if they used the printing color gamut and a calibrated light source to be certain to match the colors. At least to me. Some people throw out sensible things because “that’s the way we use to do it” attitude. Well OK crash your head into a wall and repeat history I guess. Is my response :D
The background color of the resistor can be part of the problem however, the eye has to distinguish between the colors. It’s easiest to distinguish colors with a white background or light background. Blue is a bit on the harsh side for the eyes,
I really don’t get the use of this, other than maybe a student project demo, or to get views.
I mean, if you have to handle the resistors one by one anyway, then just sort them yourself — much quicker and easier. If for whatever reason you can’t read the color code, or desperately need a go-nogo tester, then rig the same measurement system up and light a led over the appropriate output sorting bin. No need to lift the resistor up and drop it in the chute and wait for the measurement cycle and servo to move.
Im color blind, never had an issue using proper RETMA colors on a brown body. This new junk however……
Howabout well aged ones where red, orange, brown and maybe yellow are a sort of tan color, that’s half a muddy shade off what might be dirty white, grey, or was it blue or violet? Green and black are still usually greenish and blackish though.
Not that faded or flaked off numerals on ceramic caps of the same age are much easier sometimes… hurrr, last number it’s a curved top and a smudge, was that a 2, 3 or 8 even???
It seems like lifetimes ago, I built a thing for the company I was working for. This was for the girls on the hand stuffing lines. You took a resistor out of a bin, and it would tell you if another one was “the same”. I wound up building another one and made it so you could crank up the gain on the differential amp and you could sort matched pairs down to a very small fraction of a percent pretty quickly.
Well, I am Ishaan the one who built this project.
But how is it posted here without my acknowledgement.
Ok I understood
Please be kind and respectful to help make the comments section excellent. (Comment Policy)