WEBVTT 00:00:00.000 --> 00:00:02.940 Welcome back to The Deep Dive. Today, we've got 00:00:02.940 --> 00:00:05.360 a mission tailored specifically for you. Yeah, 00:00:05.360 --> 00:00:06.919 especially if you're prepping for a hardware 00:00:06.919 --> 00:00:10.179 design meeting or honestly just insanely curious 00:00:10.179 --> 00:00:11.939 about the hidden infrastructure of the modern 00:00:11.939 --> 00:00:15.099 world. Exactly. It's going to be right up your 00:00:15.099 --> 00:00:18.660 alley. We are going to decode a hidden language. 00:00:18.960 --> 00:00:21.640 And the wildest part about this is that this 00:00:21.640 --> 00:00:25.120 language is hiding in plain sight inside almost 00:00:25.120 --> 00:00:27.800 every single electronic device on planet Earth. 00:00:27.879 --> 00:00:29.780 Right, whether it's the smartphone in your pocket, 00:00:30.000 --> 00:00:32.579 the laptop on your desk, or the microwave in 00:00:32.579 --> 00:00:35.380 your kitchen. This secret code is etched right 00:00:35.380 --> 00:00:37.850 into its core. To do this, we're pulling from 00:00:37.850 --> 00:00:40.429 a really comprehensive breakdown of something 00:00:40.429 --> 00:00:43.310 called the RKM code directly from our Wikipedia 00:00:43.310 --> 00:00:45.549 source material. And it's a topic that seems 00:00:45.549 --> 00:00:48.289 incredibly niche at first glance. But once you 00:00:48.289 --> 00:00:50.350 begin to unravel it, you start to realize how 00:00:50.350 --> 00:00:53.670 the smallest, seemingly insignificant logistical 00:00:53.670 --> 00:00:57.250 decisions can fundamentally shape global manufacturing 00:00:57.250 --> 00:01:00.170 for decades. It really connects physical engineering 00:01:00.170 --> 00:01:03.539 with pure data management. It does. It's a fantastic 00:01:03.539 --> 00:01:06.540 subject to dig into. It really is. So let me 00:01:06.540 --> 00:01:09.560 ask you this. Have you ever cracked open a broken 00:01:09.560 --> 00:01:12.439 radio? Oh, yeah. Or maybe taken apart an old 00:01:12.439 --> 00:01:14.540 remote control when you were a kid. Or perhaps 00:01:14.540 --> 00:01:16.760 you just look really closely at a computer motherboard. 00:01:17.040 --> 00:01:19.239 If you have, you've definitely noticed those 00:01:19.239 --> 00:01:21.540 tiny little rectangular components. Right. The 00:01:21.540 --> 00:01:24.340 resistors, capacitors, inductors, all soldered 00:01:24.340 --> 00:01:27.250 onto the green fiberglass board. And if you squint 00:01:27.250 --> 00:01:29.250 it hard enough, you might have seen some very 00:01:29.250 --> 00:01:32.170 cryptic alphanumeric codes printed on them. Something 00:01:32.170 --> 00:01:37.730 like 4R7 or 47K. Yeah. And at first glance, it 00:01:37.730 --> 00:01:39.989 almost looks like a factory typo. You might wonder, 00:01:40.189 --> 00:01:42.370 why don't engineers just print standard numbers 00:01:42.370 --> 00:01:46.689 in units? Why not just write 4 .7 ohms? OK, let's 00:01:46.689 --> 00:01:48.829 unpack this. Why are we dealing with cryptic 00:01:48.829 --> 00:01:51.430 letters instead of plain numbers? What's fascinating 00:01:51.430 --> 00:01:53.810 here is that this notation isn't just some random 00:01:53.810 --> 00:01:57.250 manufacturer's quirk or a misprint on a particular 00:01:57.250 --> 00:02:00.750 assembly line. It's a rigorously defined, globally 00:02:00.750 --> 00:02:03.370 recognized shorthand. The standard. Exactly. 00:02:03.549 --> 00:02:05.989 Standardized by the International Electrotechnical 00:02:05.989 --> 00:02:08.530 Commission, specifically in a standard known 00:02:08.530 --> 00:02:14.090 as IEC 60062, all the way back in 1952. Wow, 00:02:14.189 --> 00:02:17.840 1952. Yeah, so we aren't looking at some new 00:02:17.840 --> 00:02:20.659 internet age tech speak here. This is a system 00:02:20.659 --> 00:02:22.780 that has been governing electronics since the 00:02:22.780 --> 00:02:25.280 early days of the Cold War. That really puts 00:02:25.280 --> 00:02:27.780 things in perspective. And according to the source 00:02:27.780 --> 00:02:31.340 material, this entire global standard stems from 00:02:31.340 --> 00:02:34.580 a very real, very physical problem. The dot. 00:02:34.699 --> 00:02:36.340 The problem with the dot, the standard decimal 00:02:36.340 --> 00:02:39.419 point. Imagine you're manufacturing a microscopic 00:02:39.419 --> 00:02:42.069 resistor. and you need to print its resistance 00:02:42.069 --> 00:02:45.530 value on the side. Say the value is 4 .7 ohms. 00:02:45.669 --> 00:02:48.550 Right. To do that, you have to print a tiny number 00:02:48.550 --> 00:02:53.490 4, a microscopic dot, and a tiny number 7. In 00:02:53.490 --> 00:02:55.930 the physical world, on a component that might 00:02:55.930 --> 00:02:58.349 be the size of a grain of rice, or even smaller, 00:02:59.449 --> 00:03:01.689 that decimal point is your absolute worst enemy. 00:03:01.789 --> 00:03:04.349 Because a single dot of ink is incredibly fragile. 00:03:04.669 --> 00:03:07.289 On a factory floor, these components are tumbling 00:03:07.289 --> 00:03:09.409 through solding machines. They're being grabbed 00:03:09.409 --> 00:03:12.110 by robotic arms. Or handled by technicians with 00:03:12.110 --> 00:03:14.710 grease on their gloves. A tiny scratch. A bit 00:03:14.710 --> 00:03:17.169 of wear and tear. Or just a slightly misaligned 00:03:17.169 --> 00:03:19.710 print job. And that microscopic decimal point 00:03:19.710 --> 00:03:22.610 easily rubs right off the component. Which completely 00:03:22.610 --> 00:03:24.189 changes the value of the part if you're trying 00:03:24.189 --> 00:03:27.169 to read it later. But the source notes... It 00:03:27.169 --> 00:03:29.050 isn't just about the physical components, right? 00:03:29.430 --> 00:03:32.150 The documentation is where this gets truly disastrous. 00:03:32.449 --> 00:03:35.069 Precisely. The vulnerability of the decimal point 00:03:35.069 --> 00:03:37.349 was a major problem in the engineering office 00:03:37.349 --> 00:03:40.780 itself. Think about how information moved in 00:03:40.780 --> 00:03:44.439 the 1950s, 60s, and 70s. Circuit diagrams were 00:03:44.439 --> 00:03:46.599 drawn by hand. Right, and then they were constantly 00:03:46.599 --> 00:03:49.199 being photocopied, duplicated, and faxed across 00:03:49.199 --> 00:03:52.060 the world. Decimal points have a notorious habit 00:03:52.060 --> 00:03:54.879 of completely disappearing when schematics are 00:03:54.879 --> 00:03:57.080 photocopied multiple times. A speck of dust on 00:03:57.080 --> 00:04:00.219 the copier glass, low resolution toner, a coffee 00:04:00.219 --> 00:04:03.020 stain, and poof. The dot is gone. And losing 00:04:03.020 --> 00:04:05.300 a decimal point on a schematic isn't just a minor 00:04:05.300 --> 00:04:07.460 typo that you can brush off. Not at all. If you 00:04:07.460 --> 00:04:09.560 lose the decimal point on a diagram calling for 00:04:09.560 --> 00:04:12.620 a 4 .7 ohm resistor, the next person down the 00:04:12.620 --> 00:04:15.580 line reads it as a 47 ohm resistor. That is 10 00:04:15.580 --> 00:04:18.100 times the intended electrical resistance. 10 00:04:18.100 --> 00:04:21.019 times. If a technician builds a circuit based 00:04:21.019 --> 00:04:24.920 on that faulty, degraded photocopy, the entire 00:04:24.920 --> 00:04:27.860 circuit could malfunction. Depending on the application, 00:04:28.100 --> 00:04:31.220 it could overheat, blow a fuse, or literally 00:04:31.220 --> 00:04:34.079 catch fire. The stakes were surprisingly high 00:04:34.079 --> 00:04:36.779 for a tiny, vulnerable dot of ink. They really 00:04:36.779 --> 00:04:39.860 were. So the engineering world needed a bulletproof 00:04:39.860 --> 00:04:42.540 way to write these numbers that could survive 00:04:42.540 --> 00:04:46.100 terrible photocopiers and greasy fingers. And 00:04:46.100 --> 00:04:48.060 the elegant solution they came up with in the 00:04:48.060 --> 00:04:51.899 RKM code is brilliant in its simplicity. It is. 00:04:51.939 --> 00:04:54.519 You take that fragile, vulnerable decimal point 00:04:54.519 --> 00:04:57.279 and you completely eliminate it. Instead, you 00:04:57.279 --> 00:04:59.480 replace it with a sturdy, highly visible letter 00:04:59.480 --> 00:05:01.560 from the alphabet. The letter acts as both the 00:05:01.560 --> 00:05:03.720 multiplier prefix and the decimal separator all 00:05:03.720 --> 00:05:05.680 at once. Let me walk you through how this actually 00:05:05.680 --> 00:05:08.079 works, because visualizing it makes it click 00:05:08.079 --> 00:05:10.699 instantly. It's a very clever typographical trick. 00:05:10.800 --> 00:05:13.959 It really is. So let's take that same 4 .7 ohm 00:05:13.959 --> 00:05:16.920 resistor. Instead of writing 4 .7, the RKM code 00:05:16.920 --> 00:05:19.939 dictates you write 4R7. The R drops right into 00:05:19.939 --> 00:05:22.279 the space where the dot used to be. You literally 00:05:22.279 --> 00:05:25.740 cannot miss a giant capital R on a bad photocopy. 00:05:25.879 --> 00:05:28.000 And it scales up depending on the value of the 00:05:28.000 --> 00:05:29.860 component. Right. So if it's a larger value, 00:05:29.939 --> 00:05:33.379 say 4 .7 kilombs, the k becomes the decimal point. 00:05:33.420 --> 00:05:35.540 You write it as 4k7. Makes sense. And if you 00:05:35.540 --> 00:05:37.300 have a solid number where there is no decimal 00:05:37.300 --> 00:05:40.279 point needed, like 470 ohms, you just stick the 00:05:40.279 --> 00:05:42.600 letter at the end to confirm the value, making 00:05:42.600 --> 00:05:45.959 it 470R. And for components requiring high precision, 00:05:46.160 --> 00:05:49.120 you might have trailing zeros. A value like 15 00:05:49.120 --> 00:05:52.420 .0 megohms uses the M for mega as the decimal 00:05:52.420 --> 00:05:55.899 marker, making it 15M0. The system guarantees 00:05:55.899 --> 00:05:58.180 that an alphanumeric character is always anchoring 00:05:58.180 --> 00:06:01.100 the value, providing a visual failsafe. Exactly. 00:06:01.259 --> 00:06:03.139 That makes a lot of sense. But looking at these 00:06:03.139 --> 00:06:05.680 examples, something is conspicuously missing. 00:06:06.000 --> 00:06:10.000 In 4R7 or 15M0, there is no mention of the word 00:06:10.000 --> 00:06:12.579 ohm for resistance or farad for capacitance. 00:06:13.220 --> 00:06:15.439 The actual unit of measurement is just gone. 00:06:15.779 --> 00:06:18.519 That is entirely by design. For the sake of brevity 00:06:18.519 --> 00:06:21.040 on these tiny surfaces, the code usually drops 00:06:21.040 --> 00:06:23.660 the actual unit names entirely. You're just expected 00:06:23.660 --> 00:06:25.399 to know what unit you're dealing with. Right. 00:06:25.740 --> 00:06:27.660 Based on the physical appearance of the part 00:06:27.660 --> 00:06:29.839 you are holding and the specific letters being 00:06:29.839 --> 00:06:32.459 used. Which brings us to the very deliberate 00:06:32.459 --> 00:06:34.899 choices made regarding the alphabet. Let's look 00:06:34.899 --> 00:06:36.759 at resistors first, since they make up the bulk 00:06:36.759 --> 00:06:40.680 of this. Resistors typically use uppercase letters. 00:06:40.939 --> 00:06:44.379 So as we discussed, you see K for kilo, M for 00:06:44.379 --> 00:06:47.819 mega. I imagine G is for giga, and T is for tera. 00:06:47.959 --> 00:06:50.220 That is correct. They follow the standard SI 00:06:50.220 --> 00:06:52.720 prefixes for the most part. But wait, what about 00:06:52.720 --> 00:06:56.120 that R we used for basic ohms? The standard scientific 00:06:56.120 --> 00:06:59.689 symbol for ons is the Greek letter omega. Everyone 00:06:59.689 --> 00:07:01.709 remembers that horseshoe shape from high school 00:07:01.709 --> 00:07:04.009 physics. Oh sure. Why didn't they just use 4 00:07:04.009 --> 00:07:06.730 omega 7? This is a perfect example of how the 00:07:06.730 --> 00:07:09.410 technological limitations of an era dictate the 00:07:09.410 --> 00:07:11.949 standards we still use today. You have to remember 00:07:11.949 --> 00:07:14.810 the context of early computing, early typewriters, 00:07:15.029 --> 00:07:16.850 and early drafting systems. Right, they didn't 00:07:16.850 --> 00:07:18.970 have special character menus. The Greek letter 00:07:18.970 --> 00:07:21.410 omega was simply missing from most character 00:07:21.410 --> 00:07:24.509 encodings. It wasn't in early ASCII, and it certainly 00:07:24.509 --> 00:07:26.730 wasn't available in early CAD -CAM environments 00:07:26.730 --> 00:07:28.970 that's computer -aided design and manufacturing. 00:07:29.290 --> 00:07:31.069 You just couldn't type an omega on a standard 00:07:31.069 --> 00:07:34.370 keyboard. No, or render it reliably on early 00:07:34.370 --> 00:07:37.689 cathode ray tube screens and dot matrix plotters. 00:07:37.949 --> 00:07:40.389 So they were forced to find a stand -in from 00:07:40.389 --> 00:07:42.870 the standard Latin alphabet. Exactly. The letter 00:07:42.870 --> 00:07:45.329 R was chosen for two very practical reasons. 00:07:46.069 --> 00:07:48.949 First... Visually, a capital R looks a little 00:07:48.949 --> 00:07:51.370 bit like the omega glyph if you squint and use 00:07:51.370 --> 00:07:53.350 your imagination. I can kind of see that. But 00:07:53.350 --> 00:07:55.930 much more importantly, it serves as a fantastic 00:07:55.930 --> 00:07:58.490 mnemonic for the word resistance across several 00:07:58.490 --> 00:08:01.470 major languages. It was a pragmatic compromise 00:08:01.470 --> 00:08:04.589 when the ideal symbol was technologically unavailable. 00:08:04.829 --> 00:08:07.149 A workaround for mid -century keyboards that 00:08:07.149 --> 00:08:09.509 we were still stamping onto modern microchips. 00:08:10.079 --> 00:08:12.160 That is amazing. Now I was looking through the 00:08:12.160 --> 00:08:14.480 notes on the resistor alphabet and I found this 00:08:14.480 --> 00:08:17.019 exception that threw me for a loop. A L exception. 00:08:17.139 --> 00:08:19.519 The L exception. Say you want to write a tiny 00:08:19.519 --> 00:08:21.740 resistance, something measured in milliohms. 00:08:22.040 --> 00:08:25.000 Millie means 10 to the power of negative 3. Normally 00:08:25.000 --> 00:08:27.459 you'd use a lowercase m as the standard prefix. 00:08:27.720 --> 00:08:30.620 But the RKM code insists on maintaining an uppercase 00:08:30.620 --> 00:08:33.500 only rule for resistors to avoid visual clutter. 00:08:33.899 --> 00:08:35.639 Right. And you obviously can't use a capital 00:08:35.639 --> 00:08:38.379 M for milli, because M is already permanently 00:08:38.379 --> 00:08:41.240 assigned to mega, which is a million ohms. Mixing 00:08:41.240 --> 00:08:43.879 up a millio and a megaohm would be catastrophic 00:08:43.879 --> 00:08:46.320 in any circuit design. Total opposite ends of 00:08:46.320 --> 00:08:48.620 the spectrum. One is essentially a dead short, 00:08:48.720 --> 00:08:50.919 and the other is a massive roadblock to the current. 00:08:51.480 --> 00:08:54.960 So to solve this clash, the code completely throws 00:08:54.960 --> 00:08:58.440 out the M and uses a capital L for milli instead. 00:08:59.039 --> 00:09:01.970 YL. It seems it just fits the uppercase requirement 00:09:01.970 --> 00:09:04.710 without clashing with the M for mega. It shows 00:09:04.710 --> 00:09:06.970 how they were willing to abandon standard prefixes 00:09:06.970 --> 00:09:09.110 to protect the integrity of the code. And that 00:09:09.110 --> 00:09:12.330 willingness to adapt led to some other less ideal 00:09:12.330 --> 00:09:15.389 variations out in the wild. You mean the E anomaly. 00:09:15.730 --> 00:09:18.669 I saw this rogue non -standard variation in the 00:09:18.669 --> 00:09:21.129 source material. Apparently sometimes you will 00:09:21.129 --> 00:09:23.830 see a capital E used instead of an R for basic 00:09:23.830 --> 00:09:27.009 ohms. Yes. So a component might read 4E7 instead 00:09:27.009 --> 00:09:29.889 of 4R7. Where did that come from? It is a fascinating 00:09:29.889 --> 00:09:32.509 historical holdover. The reason E was sometimes 00:09:32.509 --> 00:09:35.409 used is purely visual. A capital E looks quite 00:09:35.409 --> 00:09:37.809 a bit like a sideways lowercase Greek omega. 00:09:38.080 --> 00:09:41.679 And historically, well before World War II, engineers 00:09:41.679 --> 00:09:45.240 actually used the lowercase omega to denote ohms 00:09:45.240 --> 00:09:48.320 before the uppercase omega became the universal 00:09:48.320 --> 00:09:50.460 standard. Oh, I see. So it's an old school visual 00:09:50.460 --> 00:09:52.399 pun. They just rotated the letter in their minds. 00:09:52.539 --> 00:09:55.080 It is. However, I must stress that this variation 00:09:55.080 --> 00:09:57.379 is strongly discouraged today. Because it breaks 00:09:57.379 --> 00:10:00.240 things. It causes massive headaches in modern 00:10:00.240 --> 00:10:04.120 computing. The problem is scientific e -notation, 00:10:04.159 --> 00:10:06.759 which is standard on calculators and in programming 00:10:06.759 --> 00:10:09.809 languages. Ah, right. In that context, E stands 00:10:09.809 --> 00:10:13.269 for times 10 to the power of. So if a modern 00:10:13.269 --> 00:10:16.649 computer reads the string 4E7, it interprets 00:10:16.649 --> 00:10:20.870 that as 40 million, not 4 .7. Yikes. It's a perfect 00:10:20.870 --> 00:10:23.730 storm of conflating notations that can completely 00:10:23.730 --> 00:10:26.789 break automated design software. Note to self, 00:10:27.110 --> 00:10:29.750 do not use E unless you want to completely confuse 00:10:29.750 --> 00:10:32.259 the software. Okay, so we've established the 00:10:32.259 --> 00:10:34.820 uppercase world of resistors. Let's look at the 00:10:34.820 --> 00:10:37.220 other major component crowding these circuit 00:10:37.220 --> 00:10:39.179 boards capacitors. They seem to flip the script 00:10:39.179 --> 00:10:41.360 entirely. They do. To differentiate themselves 00:10:41.360 --> 00:10:43.919 at a glance, capacitors generally rely on lowercase 00:10:43.919 --> 00:10:47.600 letters. You will see P for picoforads, N for 00:10:47.600 --> 00:10:51.600 nanoforads, and M for milliforads. The only major 00:10:51.600 --> 00:10:54.200 exception is F for forog, the base unit, which 00:10:54.200 --> 00:10:56.620 remains capitalized. But even within this lowercase 00:10:56.620 --> 00:10:58.899 system, they ran into another type of graphical 00:10:58.899 --> 00:11:01.919 wall. You're referring to the U workaround. This 00:11:01.919 --> 00:11:04.159 one is my favorite because it feels so scrappy. 00:11:04.700 --> 00:11:07.820 The standard prefix for micro, as in microfarad, 00:11:07.899 --> 00:11:10.379 which is an incredibly common capacitor value, 00:11:10.820 --> 00:11:13.559 is the lowercase Greek letter mu. Yes. It looks 00:11:13.559 --> 00:11:15.840 like a U with a long tail on the front left side. 00:11:16.399 --> 00:11:19.299 But again, just like omega, mu was a nightmare. 00:11:19.580 --> 00:11:22.659 to print, type, or render on early computer systems. 00:11:22.980 --> 00:11:25.139 So the standard officially sanctioned a very 00:11:25.139 --> 00:11:27.879 pragmatic substitution. If a system cannot handle 00:11:27.879 --> 00:11:29.980 the Greek letter Mu, you are allowed to substitute 00:11:29.980 --> 00:11:32.360 it with a standard Latin U, either lowercase 00:11:32.360 --> 00:11:34.799 or uppercase. Wow. And this wasn't just an isolated 00:11:34.799 --> 00:11:37.259 decision by the IEC. It lined with other early 00:11:37.259 --> 00:11:40.580 data processing standards like ISO 2955 back 00:11:40.580 --> 00:11:43.840 in the 1970s. They also authorized up as a stand 00:11:43.840 --> 00:11:46.299 -in for Mu when dealing with basic Latin -only 00:11:46.299 --> 00:11:48.779 text terminals. It's amazing how much of this 00:11:48.940 --> 00:11:52.899 highly technical standard is basically just Generations 00:11:52.899 --> 00:11:55.159 of engineers saying look our typewriter doesn't 00:11:55.159 --> 00:11:57.259 have that key Let's just use a letter that looks 00:11:57.259 --> 00:11:59.529 close enough so we can go home It is the very 00:11:59.529 --> 00:12:02.990 essence of engineering, finding a reliable, robust 00:12:02.990 --> 00:12:05.509 solution within the strict constraints of your 00:12:05.509 --> 00:12:07.769 available tools. Absolutely. But if we connect 00:12:07.769 --> 00:12:10.710 this to the bigger picture, the RKM code isn't 00:12:10.710 --> 00:12:13.309 just about printing legible text on tiny parts 00:12:13.309 --> 00:12:16.730 or avoiding bad photocopies. Right. As the industry 00:12:16.730 --> 00:12:19.350 digitized, this alphanumeric shorthand revealed 00:12:19.350 --> 00:12:22.690 a secondary, almost secret benefit. It functions 00:12:22.690 --> 00:12:25.750 as a logistical superpower. A logistical superpower. 00:12:26.090 --> 00:12:28.559 That's a strong claim. How does swapping a dot 00:12:28.559 --> 00:12:31.240 for a letter give you a superpower? Because it 00:12:31.240 --> 00:12:33.799 completely transforms how parts are handled in 00:12:33.799 --> 00:12:36.740 bulk databases. Imagine a modern manufacturing 00:12:36.740 --> 00:12:38.480 plant. They are building complex electronics 00:12:38.480 --> 00:12:40.860 with tens of thousands of different tiny components. 00:12:41.120 --> 00:12:44.460 They organize all of this inventory using a massive 00:12:44.460 --> 00:12:46.840 database called a bill of materials, commonly 00:12:46.840 --> 00:12:49.700 known as a boom boom. Sorting a bill of materials 00:12:49.700 --> 00:12:52.200 is where the RKM notation truly shines as a data 00:12:52.200 --> 00:12:54.100 management tool. Wait, let me see if I can follow 00:12:54.100 --> 00:12:56.259 the logic here based on the source text. Let's 00:12:56.259 --> 00:12:59.100 say you're an engineer and you have a raw list 00:12:59.100 --> 00:13:01.600 of standard resistance values for a new device. 00:13:02.419 --> 00:13:05.759 You've got a 3 .3 ohm resistor, a 3 .3 kilom 00:13:05.759 --> 00:13:09.820 resistor, a 3 .6 kilom resistor, and a 4 .7 kilom 00:13:09.820 --> 00:13:13.279 resistor. In a standard dumb spreadsheet that 00:13:13.279 --> 00:13:15.620 doesn't natively understand engineering units, 00:13:16.019 --> 00:13:18.600 if you just sort purely by the numbers, things 00:13:18.600 --> 00:13:21.110 are going to get incredibly messy. Exactly. Because 00:13:21.110 --> 00:13:23.769 a standard text sort will look at 3 .3 and group 00:13:23.769 --> 00:13:26.629 it with 3 .3 kilo, it ignores the attached word 00:13:26.629 --> 00:13:29.470 kilo. So your 3 .3 ohm resistor gets sorted right 00:13:29.470 --> 00:13:31.809 next to a resistor that is a thousand times more 00:13:31.809 --> 00:13:34.210 powerful just because the prefix numbers match. 00:13:34.330 --> 00:13:36.049 It separates components that should be grouped 00:13:36.049 --> 00:13:38.409 and it makes scanning the database a chaotic 00:13:38.409 --> 00:13:40.309 experience. But look at what happens when you 00:13:40.309 --> 00:13:42.230 convert those values into the RKM code before 00:13:42.230 --> 00:13:44.590 you hit the sort button. The 3 .3 ohm becomes 00:13:44.590 --> 00:13:48.600 3R3. Yep. The 3 .3 kilom becomes 3K3. The 3 .6 00:13:48.600 --> 00:13:52.340 kilom is 3K6, and the 4 .7 kilom is 4K7. Now 00:13:52.340 --> 00:13:55.220 when your computer database sorts this list alphanumerically, 00:13:55.720 --> 00:13:57.940 it's beautiful. The R values group together. 00:13:58.200 --> 00:14:01.399 And all the kilom values, 3K3, 3K6, 4K7, are 00:14:01.399 --> 00:14:03.820 locked together by the K. perfectly arranged 00:14:03.820 --> 00:14:06.299 in ascending order. It forces the database to 00:14:06.299 --> 00:14:09.139 behave intuitively. And why should you care about 00:14:09.139 --> 00:14:11.899 this? If you are a supply chain manager handling 00:14:11.899 --> 00:14:14.860 millions of dollars of inventory, this sorting 00:14:14.860 --> 00:14:17.779 behavior is invaluable. It allows you to scan 00:14:17.779 --> 00:14:20.539 a massive spreadsheet and instantly spot redundant 00:14:20.539 --> 00:14:23.350 parts. You might look at your cleanly sorted 00:14:23.350 --> 00:14:26.289 list and say, wait, why are we buying 10 ,000 00:14:26.289 --> 00:14:29.629 3K3 resistors and 10 ,000 3K6 resistors? Can 00:14:29.629 --> 00:14:32.029 we just adjust the electrical design slightly 00:14:32.029 --> 00:14:36.070 to use the 3K3 for both? Oh, I see. By optimizing 00:14:36.070 --> 00:14:38.129 your inventory like that, you reduce the number 00:14:38.129 --> 00:14:40.509 of unique parts you have to source, buy stock 00:14:40.509 --> 00:14:43.210 on shelves, and load into the automated assembly 00:14:43.210 --> 00:14:45.529 machines. Exactly. That saves a tremendous amount 00:14:45.529 --> 00:14:47.929 of money in overhead. It's an accidental database 00:14:47.929 --> 00:14:50.750 optimization tool that was invented decades before. 00:14:50.759 --> 00:14:53.500 supply chain databases were even a thing. Precisely. 00:14:53.700 --> 00:14:56.179 And because it was so undeniably useful, the 00:14:56.179 --> 00:14:58.440 notation couldn't be contained just to resistors 00:14:58.440 --> 00:15:00.779 and capacitors. It naturally bled into all sorts 00:15:00.779 --> 00:15:02.539 of other areas of electrical engineering. Right, 00:15:02.659 --> 00:15:05.580 like inductors. Yes, inductors, which are coils 00:15:05.580 --> 00:15:08.620 of wire used in circuits, adopted it. You might 00:15:08.620 --> 00:15:11.649 see an inductor labeled 4R7. which would mean 00:15:11.649 --> 00:15:15.450 4 .7 microhenries. And even more surprisingly, 00:15:15.769 --> 00:15:18.070 the system was co -opted for expressing voltages. 00:15:18.750 --> 00:15:20.789 Right. The source mentions this. When you are 00:15:20.789 --> 00:15:23.110 writing a schematic on a computer, or naming 00:15:23.110 --> 00:15:25.649 a digital file for a circuit design, or even 00:15:25.649 --> 00:15:27.889 assigning a label to a tiny pin on a microchip, 00:15:28.289 --> 00:15:30.830 a standard dot can completely break the computer 00:15:30.830 --> 00:15:33.720 code. In software, a dot usually indicates a 00:15:33.720 --> 00:15:36.279 file extension. Or a command in an object -oriented 00:15:36.279 --> 00:15:39.059 programming language. So you can't name a software 00:15:39.059 --> 00:15:41.779 variable 3 .3 volts. The compiler would crash 00:15:41.779 --> 00:15:44.179 trying to read it, so instead, engineers use 00:15:44.179 --> 00:15:47.539 the RKM style and simply write 3v3. No dots, 00:15:47.679 --> 00:15:50.620 no spaces, just 3v3. And there's another variation 00:15:50.620 --> 00:15:52.279 for voltages in the text that I found really 00:15:52.279 --> 00:15:54.600 interesting. Sometimes, instead of a V, you will 00:15:54.600 --> 00:15:57.100 see a P used as the decimal separator. So they 00:15:57.100 --> 00:16:00.019 might write 1P8 to represent 1 .8 volts. The 00:16:00.019 --> 00:16:02.639 P in that context typically designates a positive 00:16:02.639 --> 00:16:05.019 voltage or indicates a specific power supply 00:16:05.019 --> 00:16:07.379 rail on the board. It serves the exact same function. 00:16:07.610 --> 00:16:10.250 Eliminating the fragile, software -breaking decimal 00:16:10.250 --> 00:16:13.029 point while conveying essential technical information 00:16:13.029 --> 00:16:16.110 in a clean, machine -readable format. OK, so 00:16:16.110 --> 00:16:18.149 we've got the base values, we've got the multipliers, 00:16:18.210 --> 00:16:20.090 we've got the voltages. But if you look at a 00:16:20.090 --> 00:16:22.909 full component code, there is often more printed 00:16:22.909 --> 00:16:25.929 on these tiny parts than just their core value. 00:16:26.629 --> 00:16:29.210 I want to dig into tolerances because this brings 00:16:29.210 --> 00:16:31.470 in a whole different layer of history. Tolerance 00:16:31.470 --> 00:16:33.789 is a crucial concept in manufacturing. It essentially 00:16:33.789 --> 00:16:36.009 tells you how precise the component actually 00:16:36.009 --> 00:16:39.529 is. When a factory makes 100 ohm resistor, it 00:16:39.529 --> 00:16:43.710 is rarely going to be exactly 100 .00 ohms. Right. 00:16:43.789 --> 00:16:46.350 The manufacturing process always has slight variations. 00:16:46.570 --> 00:16:50.090 Exactly. It might be 101 ohms or 99 ohms. So 00:16:50.090 --> 00:16:52.529 the RKM standard tax on additional letters at 00:16:52.529 --> 00:16:54.870 the very end of the code to indicate this tolerance 00:16:54.870 --> 00:16:57.039 percentage. and they assign specific letters 00:16:57.039 --> 00:16:59.179 of the alphabet to represent different percentages. 00:16:59.440 --> 00:17:01.659 For example, if you see the letter F at the end 00:17:01.659 --> 00:17:04.099 of a code, it means the part is highly accurate, 00:17:04.299 --> 00:17:07.019 guaranteed to be within plus or minus 1 .0 % 00:17:07.019 --> 00:17:09.539 of its stated value. But if you see a J, it means 00:17:09.539 --> 00:17:12.559 it's a bit looser, within plus or minus 5 .0%. 00:17:12.640 --> 00:17:15.279 What's truly fascinating here is the embedded 00:17:15.279 --> 00:17:17.660 history of those specific tolerance letters. 00:17:18.339 --> 00:17:20.079 The standard committee didn't just pull them 00:17:20.079 --> 00:17:22.579 out of thin air in 1952. Where did they come 00:17:22.579 --> 00:17:24.940 from? Many of these symmetrical tolerance letters, 00:17:25.160 --> 00:17:28.220 specifically G, J, K, and M, actually stem from 00:17:28.220 --> 00:17:31.700 mid -1940s military specifications. Oh, the source 00:17:31.700 --> 00:17:33.859 mentions the American War Standard and the Joint 00:17:33.859 --> 00:17:36.980 Army Navy specification. Yes, the AWS and JN 00:17:36.980 --> 00:17:39.480 specs. These were rigorously developed during 00:17:39.480 --> 00:17:41.759 World War II to ensure that military equipment, 00:17:42.000 --> 00:17:45.660 from radar systems to field radios, could be 00:17:45.660 --> 00:17:48.420 mass produced with interchangeable standardized 00:17:48.420 --> 00:17:51.799 parts across different allied factories. Wow. 00:17:52.200 --> 00:17:54.680 So when you look at a modern circuit board inside 00:17:54.680 --> 00:17:57.640 your smartwatch today and you see a J indicating 00:17:57.640 --> 00:18:00.680 a 5 % tolerance, you are actually looking at 00:18:00.680 --> 00:18:03.440 a direct linguistic descendant of World War II 00:18:03.440 --> 00:18:06.380 military logistics. That is wild. It's like finding 00:18:06.380 --> 00:18:08.740 an industrial fossil sitting inside your smartwatch. 00:18:08.859 --> 00:18:11.259 But the code still doesn't stop there. The source 00:18:11.259 --> 00:18:13.319 notes that letters can also denote things like 00:18:13.319 --> 00:18:15.539 temperature coefficients. Now, to make sure we 00:18:15.539 --> 00:18:18.359 are all on the same page, a temperature coefficient 00:18:18.359 --> 00:18:21.259 basically measures how much the part's electrical 00:18:21.259 --> 00:18:23.819 value shifts as it physically heats up or cools 00:18:23.819 --> 00:18:27.480 down. It's measured in parts per million per 00:18:27.480 --> 00:18:30.700 Kelvin. And this is incredibly important. Imagine 00:18:30.700 --> 00:18:33.380 you are building a precision medical sensor that 00:18:33.380 --> 00:18:36.619 monitors a patient's vital signs. As the device 00:18:36.619 --> 00:18:39.339 runs, the electronics naturally heat up. And 00:18:39.339 --> 00:18:41.839 if your resistors change their value wildly as 00:18:41.839 --> 00:18:44.339 they get warm. The entire calibration of the 00:18:44.339 --> 00:18:46.859 device shifts, and the machine might give a doctor 00:18:46.859 --> 00:18:49.039 a false reading. So knowing how stable the part 00:18:49.039 --> 00:18:51.359 is under heat is critical. And again, they use 00:18:51.359 --> 00:18:53.119 a letter code for that. But the part of the code 00:18:53.119 --> 00:18:55.339 that I found the most amusing deals with production 00:18:55.339 --> 00:18:57.079 dates. Here's where it gets really interesting. 00:18:57.599 --> 00:18:59.960 How do you encode the exact date a microscopic 00:18:59.960 --> 00:19:02.619 part was made using just a couple of characters? 00:19:02.859 --> 00:19:06.319 The date codes are a brilliant, if slightly chaotic, 00:19:06.700 --> 00:19:09.539 study in human factors engineering. To track 00:19:09.539 --> 00:19:11.920 when a component was manufactured, the standard 00:19:11.920 --> 00:19:14.660 uses a repeating 20 -year cycle represented by 00:19:14.660 --> 00:19:17.140 letters and numbers. But they didn't just use 00:19:17.140 --> 00:19:19.960 the alphabet straight through from A to Z, because 00:19:19.960 --> 00:19:22.099 humans are the ones reading these microscopic 00:19:22.099 --> 00:19:25.599 codes, often in poor warehouse lighting. Or staring 00:19:25.599 --> 00:19:27.519 at components that have been running in a hot 00:19:27.519 --> 00:19:30.259 machine for five years. Exactly. The architects 00:19:30.259 --> 00:19:32.700 of the standard deliberately banned certain letters 00:19:32.700 --> 00:19:36.140 to avoid visual confusion. They removed the letter 00:19:36.140 --> 00:19:39.650 G because On a tiny component, it looks entirely 00:19:39.650 --> 00:19:42.049 too much like the number 6. They removed i because 00:19:42.049 --> 00:19:44.329 it looks like a 1. And q and z were banned because 00:19:44.329 --> 00:19:46.869 they look like 0 and 2. Which is perfectly logical. 00:19:46.950 --> 00:19:49.490 You want zero ambiguity on the factory floor. 00:19:49.750 --> 00:19:53.250 But then the source text reveals this absolutely 00:19:53.250 --> 00:19:56.609 hilarious contradiction about how they encode 00:19:56.609 --> 00:19:59.250 the specific months of the year. So in the date 00:19:59.250 --> 00:20:01.410 code, the first character is the year and the 00:20:01.410 --> 00:20:03.630 second character is the month. For the months, 00:20:04.029 --> 00:20:06.390 digits 1 through 9 obviously represent January 00:20:06.390 --> 00:20:08.890 through September. But what do you do for October, 00:20:09.250 --> 00:20:11.549 the 10th month? You can't use the number 10 because 00:20:11.549 --> 00:20:14.329 that requires two characters and space is the 00:20:14.329 --> 00:20:16.329 most valuable commodity on these components. 00:20:16.809 --> 00:20:19.430 Right. So they decided to switch to letters for 00:20:19.430 --> 00:20:22.269 the final three months. And for October, they 00:20:22.269 --> 00:20:24.769 chose the letter O. Which is incredibly counterintuitive 00:20:24.769 --> 00:20:26.829 given the rules we've literally just discussed. 00:20:27.069 --> 00:20:29.930 Exactly. The letter O was famously on the banned 00:20:29.930 --> 00:20:31.890 list for the years because it looks exactly like 00:20:31.890 --> 00:20:34.049 a zero. So why on earth would they use it for 00:20:34.049 --> 00:20:36.109 the month? That's a great question. Well, the 00:20:36.109 --> 00:20:38.170 source text explains they did this specifically 00:20:38.170 --> 00:20:41.369 because October is the 10th month. The number 00:20:41.369 --> 00:20:45.480 10 contains a zero. The logic was... If they 00:20:45.480 --> 00:20:47.319 used a different letter that happened to look 00:20:47.319 --> 00:20:50.079 like a one, a technician might confuse it with 00:20:50.079 --> 00:20:53.960 January, the first month. So by using O? By using 00:20:53.960 --> 00:20:57.039 O, even if the technician's brain misreads it 00:20:57.039 --> 00:20:59.539 as a zero, they still associate it with the tenth 00:20:59.539 --> 00:21:01.960 month, effectively preventing it from being confused 00:21:01.960 --> 00:21:04.420 with the start of the year. It is this bizarre 00:21:04.420 --> 00:21:07.019 twisted psychological logic that somehow works 00:21:07.019 --> 00:21:09.339 perfectly in practice. Following that, November 00:21:09.339 --> 00:21:12.160 naturally becomes N and December becomes D. It 00:21:12.160 --> 00:21:15.140 is a wonderfully pragmatic contradiction. They 00:21:15.140 --> 00:21:17.839 willingly broke their own cardinal rule of typography 00:21:17.839 --> 00:21:20.740 to solve a very specific human error problem 00:21:20.740 --> 00:21:23.059 regarding the calendar. I just love the mental 00:21:23.059 --> 00:21:25.819 gymnastics behind that. Me too. Now before we 00:21:25.819 --> 00:21:28.019 wrap up, we have to touch on what happens when 00:21:28.019 --> 00:21:31.079 things get truly microscopic. Because technology 00:21:31.079 --> 00:21:34.220 hasn't stood still since 1952. We are talking 00:21:34.220 --> 00:21:37.619 about components today like surface mount resistors 00:21:37.619 --> 00:21:39.900 that are barely larger than a grain of sand. 00:21:40.279 --> 00:21:43.079 For these, even the standard RKM shorthand is 00:21:43.079 --> 00:21:46.299 sometimes too long. Yes, when you simply do not 00:21:46.299 --> 00:21:48.339 have the physical real estate for three or four 00:21:48.339 --> 00:21:51.480 characters, the industry relies on standard preferred 00:21:51.480 --> 00:21:54.480 values, commonly known as the E -Series. Let's 00:21:54.480 --> 00:21:56.680 break that down for a second. The E -Series is 00:21:56.680 --> 00:21:58.859 basically the industry's master list of standard 00:21:58.859 --> 00:22:01.180 values, right? Right. It exists so manufacturers 00:22:01.180 --> 00:22:04.400 don't have to endlessly produce a 100 ohm resistor, 00:22:04.920 --> 00:22:07.619 100 .1 ohm resistor, 100 .2 ohm resistor, and 00:22:07.619 --> 00:22:10.460 so on. They just produce a set spacing of values 00:22:10.460 --> 00:22:12.940 that cover all the necessary bases. Precisely. 00:22:13.019 --> 00:22:15.619 It is a logarithmic scale of preferred values. 00:22:16.019 --> 00:22:17.859 And because engineers know these exact values 00:22:17.859 --> 00:22:20.240 are standardized across the globe, modern updates 00:22:20.240 --> 00:22:25.240 to the standard, like IEC 600, 62 .20, and 16, 00:22:25.640 --> 00:22:27.460 allow them to compress the code even further 00:22:27.460 --> 00:22:30.509 by using a specialized lookup table. Exactly. 00:22:30.569 --> 00:22:33.269 They drop down to a highly compressed three -character 00:22:33.269 --> 00:22:35.829 code for tiny resistors, which usually involves 00:22:35.829 --> 00:22:37.990 two digits representing the base value from the 00:22:37.990 --> 00:22:40.529 table and one letter for the multiplier. And 00:22:40.529 --> 00:22:42.609 for microscopic capacitors. They compress it 00:22:42.609 --> 00:22:45.450 down to an almost impossibly dense two -character 00:22:45.450 --> 00:22:48.730 code. They squeeze out every single drop of unnecessary 00:22:48.730 --> 00:22:52.390 ink. So what does this all mean? We started today 00:22:52.390 --> 00:22:56.009 by looking at a tiny, easily ignored string of 00:22:56.009 --> 00:22:58.970 characters on a circuit board. And we ended up 00:22:58.970 --> 00:23:01.450 uncovering a master class in pragmatic engineering. 00:23:01.750 --> 00:23:04.789 It really is. The RKM code isn't just an arbitrary 00:23:04.789 --> 00:23:07.509 set of rules forced on manufacturers. It is a 00:23:07.509 --> 00:23:10.450 deeply evolved global language. It was invented 00:23:10.450 --> 00:23:13.109 simply because physical decimal points are too 00:23:13.109 --> 00:23:16.250 fragile for reality, and mid -century photocopiers 00:23:16.250 --> 00:23:18.900 were unreliable. It's a language built to survive 00:23:18.900 --> 00:23:21.240 the harsh, gritty environment of the factory 00:23:21.240 --> 00:23:23.980 floor. To tame the chaos of supply chain databases. 00:23:24.240 --> 00:23:26.440 And to work around the limitations of early computing 00:23:26.440 --> 00:23:29.359 keyboards. It really is a testament to how practical 00:23:29.359 --> 00:23:32.759 problem solving creates lasting, invisible infrastructure. 00:23:32.900 --> 00:23:35.220 It's incredible to think about. And it leaves 00:23:35.220 --> 00:23:37.480 you with something fascinating to consider long 00:23:37.480 --> 00:23:40.460 after we wrap up today. This entire globally 00:23:40.460 --> 00:23:42.900 adopted shorthand spanning from World War II 00:23:42.900 --> 00:23:46.019 logistics to the motherboards and our most advanced 00:23:46.019 --> 00:23:48.759 AI super computers today is entirely defined 00:23:48.759 --> 00:23:51.400 by the physical limitations of mid -20th century 00:23:51.400 --> 00:23:54.099 tools. Yeah. It makes you wonder, as we move 00:23:54.099 --> 00:23:56.900 into the future of computing, will entirely new 00:23:56.900 --> 00:24:00.119 paradigms like quantum computers still be forced 00:24:00.119 --> 00:24:04.059 to inherit and use these exact same 1950s typewriter 00:24:04.059 --> 00:24:06.839 shorthands simply to remain backward compatible 00:24:06.839 --> 00:24:10.160 with our massive, deeply entrenched global supply 00:24:10.160 --> 00:24:12.420 chains? Are we locked into the letter R forever? 00:24:12.680 --> 00:24:15.509 No. That is a wild thought to end on. Imagine 00:24:15.509 --> 00:24:18.430 a futuristic quantum array still using an R because 00:24:18.430 --> 00:24:21.009 someone's typewriter, 1952, didn't have an omega 00:24:21.009 --> 00:24:22.849 key. Thank you so much for joining us on this 00:24:22.849 --> 00:24:24.529 deep dive. We hope you never look at a broken 00:24:24.529 --> 00:24:26.609 radio or a piece of modern electronics the same 00:24:26.609 --> 00:24:27.890 way again. We'll catch you next time.