WEBVTT 00:00:00.000 --> 00:00:01.919 Welcome back to another custom -tailored deep 00:00:01.919 --> 00:00:05.400 dive into the source material. We are thrilled 00:00:05.400 --> 00:00:06.980 to have you with us again. Yeah, it's really 00:00:06.980 --> 00:00:09.380 great to be back. And you, our resident learner, 00:00:09.539 --> 00:00:11.099 you know, someone who always wants the inside 00:00:11.099 --> 00:00:13.419 track. You want the quick but thorough download 00:00:13.419 --> 00:00:16.079 on the invisible systems running the world without 00:00:16.079 --> 00:00:18.239 getting bogged down in some overwhelming info 00:00:18.239 --> 00:00:21.320 dump. Right, and today we are targeting a topic 00:00:21.320 --> 00:00:24.260 that is quite literally hiding in plain sight. 00:00:24.519 --> 00:00:27.980 It is highly likely within arm's reach of you 00:00:27.980 --> 00:00:30.660 right now. Exactly. Our source material today 00:00:30.660 --> 00:00:33.520 is this incredibly comprehensive Wikipedia article 00:00:33.520 --> 00:00:36.859 covering something called the RKM code. Which, 00:00:36.920 --> 00:00:39.280 I mean, it is entirely probable that there are 00:00:39.280 --> 00:00:41.859 hundreds, if not thousands, of instances of this 00:00:41.859 --> 00:00:44.159 sitting inside the room with you right now stamped 00:00:44.159 --> 00:00:46.500 onto the internal components of every electronic 00:00:46.500 --> 00:00:49.240 device you own. So our mission for this deep 00:00:49.240 --> 00:00:52.340 dive is to decode that secret, highly optimized 00:00:52.340 --> 00:00:55.119 shorthand. The language printed on those tiny 00:00:55.119 --> 00:00:57.299 components populating the circuit boards inside, 00:00:57.420 --> 00:00:59.539 you know, your laptop, your phone, your audio 00:00:59.539 --> 00:01:02.340 gear. This is just a phenomenal piece of engineering 00:01:02.340 --> 00:01:05.659 history. Most technical professionals look at 00:01:05.659 --> 00:01:08.319 these codes every single day, right? They rely 00:01:08.319 --> 00:01:11.340 on them implicitly without ever realizing the 00:01:11.340 --> 00:01:14.439 sheer amount of historical baggage, a technological 00:01:14.439 --> 00:01:17.950 compromise. Honestly, the accidental brilliance 00:01:17.950 --> 00:01:20.370 baked into those little alphanumeric characters. 00:01:20.709 --> 00:01:22.909 To really set the stage for you, we need to think 00:01:22.909 --> 00:01:25.209 about the physical reality of sharing technical 00:01:25.209 --> 00:01:29.189 information before the digital age. Anyone who 00:01:29.189 --> 00:01:31.689 has dealt with physical documentation has encountered 00:01:31.689 --> 00:01:34.349 the nightmare of the degraded copy. Oh, absolutely. 00:01:34.590 --> 00:01:36.709 You take a master document, you run it through 00:01:36.709 --> 00:01:40.810 a duplicating machine, and a tiny physical imperfection, 00:01:40.829 --> 00:01:43.950 a speck of dust on the glass, a smudge of ink. 00:01:44.510 --> 00:01:46.909 it completely alters the text. Yeah, I mean, 00:01:46.930 --> 00:01:49.730 a dropped comma can change the entire legal definition 00:01:49.730 --> 00:01:52.349 of a contract. Right. But in everyday text, you 00:01:52.349 --> 00:01:54.989 can usually figure it out from context. In electrical 00:01:54.989 --> 00:01:57.170 engineering, though, a disappearing punctuation 00:01:57.170 --> 00:02:00.340 mark is catastrophic. It really is. If a schematic 00:02:00.340 --> 00:02:03.379 or a bill of materials is degraded during duplication 00:02:03.379 --> 00:02:07.359 and a value is misread, it can lead to an overloaded 00:02:07.359 --> 00:02:10.259 circuit, a thermal event, or just the ruination 00:02:10.259 --> 00:02:13.099 of a massive manufacturing run. Okay, let's unpack 00:02:13.099 --> 00:02:15.939 this. Imagine the workflow of an electrical engineer 00:02:15.939 --> 00:02:19.219 back in the 1950s. They're designing an analog 00:02:19.219 --> 00:02:21.639 circuit, maybe for a television set. Or a radio 00:02:21.639 --> 00:02:23.620 receiver, yeah. Right, and they calculate that 00:02:23.620 --> 00:02:26.479 a specific trace needs a resistor with a value 00:02:26.479 --> 00:02:30.409 of 4 .7 ohms. That value gets written onto a 00:02:30.409 --> 00:02:32.409 paper schematic. Which then has to be duplicated 00:02:32.409 --> 00:02:36.349 probably via mimeograph or early, you know, primitive 00:02:36.349 --> 00:02:38.830 photocopying machines. And it gets passed down 00:02:38.830 --> 00:02:40.990 this long chain of command to the factory floor. 00:02:41.770 --> 00:02:44.150 Ultimately, it dictates what gets printed onto 00:02:44.150 --> 00:02:46.629 a physical component that might be no larger 00:02:46.629 --> 00:02:49.210 than a grain of rice. And the vulnerability in 00:02:49.210 --> 00:02:51.569 that entire chain of custody is the decimal point. 00:02:51.650 --> 00:02:54.169 The dot. Just that single tiny dot separating 00:02:54.169 --> 00:02:57.610 the 4 and the 7. It is incredibly fragile. It's 00:02:57.610 --> 00:03:00.169 notoriously difficult to print reliably onto 00:03:00.169 --> 00:03:02.710 the curved, imperfect surface of a carbon composition 00:03:02.710 --> 00:03:05.930 resistor. And when those paper technical documents 00:03:05.930 --> 00:03:08.310 are copied and recopied for different departments, 00:03:08.629 --> 00:03:11.669 that dot is almost guaranteed. to vanish at some 00:03:11.669 --> 00:03:14.430 point exactly and the moment that dot vanishes 00:03:14.430 --> 00:03:16.810 the factory floor is no longer building a circuit 00:03:16.810 --> 00:03:20.449 with a 4 .7 ohm resistor they are suddenly installing 00:03:20.449 --> 00:03:23.689 a 47 ohm resistor right you have just introduced 00:03:23.689 --> 00:03:27.090 a 10 -fold increase in resistance into a carefully 00:03:27.090 --> 00:03:30.210 tuned circuit which means the device fails qa 00:03:30.210 --> 00:03:33.000 testing or worse It fails in the hands of the 00:03:33.000 --> 00:03:35.319 consumer. The International Electrotechnical 00:03:35.319 --> 00:03:38.060 Commission, the IEC, saw this happening across 00:03:38.060 --> 00:03:40.419 the industry and realized they had a massive 00:03:40.419 --> 00:03:43.659 systemic vulnerability. So in 1952, they moved 00:03:43.659 --> 00:03:46.080 to solve it. They introduced a standard that 00:03:46.080 --> 00:03:50.460 would eventually be codified as IEC 60062. And 00:03:50.460 --> 00:03:52.680 the core innovation of this standard, which we 00:03:52.680 --> 00:03:55.319 now call the RKM code, took a rather drastic 00:03:55.319 --> 00:03:57.280 approach to the problem. Instead of trying to 00:03:57.280 --> 00:03:59.219 find better ways to print the dot, they just 00:03:59.219 --> 00:04:01.580 eradicated the decimal point entirely. They just 00:04:01.580 --> 00:04:03.960 banned it. Yep. Engineers were instructed to 00:04:03.960 --> 00:04:05.960 replace the decimal separator with the letter 00:04:05.960 --> 00:04:08.159 of the multiplier itself. The source material 00:04:08.159 --> 00:04:10.300 provides some excellent examples to help visualize 00:04:10.300 --> 00:04:13.219 this. If an engineer wanted to specify that risky 00:04:13.219 --> 00:04:16.540 4 .7 ohm value, they wouldn't use the dot anymore. 00:04:16.740 --> 00:04:18.519 I'll let you explain how the code transforms 00:04:18.519 --> 00:04:21.879 that. Sure. Under the RKM standard, 4 .7 ohms 00:04:21.879 --> 00:04:26.019 becomes 4R7. 4R7. Right. The letter R acts as 00:04:26.019 --> 00:04:28.420 the decimal separator, anchoring the value while 00:04:28.420 --> 00:04:30.779 simultaneously indicating that the base unit 00:04:30.779 --> 00:04:32.779 is standard ohms. And it scales up beautifully, 00:04:32.939 --> 00:04:36.160 too. If the circuit requires 4700 ohms, which 00:04:36.160 --> 00:04:39.360 is 4 .7 kilohms, the engineer drops the dot, 00:04:39.500 --> 00:04:42.579 grabs the K for the kilo multiplier, and writes 00:04:42.579 --> 00:04:45.079 it out as 4K7. It even handles fractional values 00:04:45.079 --> 00:04:47.939 strictly under a single ohm. Say you have 0 .47 00:04:47.939 --> 00:04:49.879 ohms. You just place the letter at the very front 00:04:49.879 --> 00:04:51.920 of the sequence. So it is written as R47. That 00:04:51.920 --> 00:04:54.199 is so simple, but so effective. What's fascinating 00:04:54.199 --> 00:04:56.720 here is the sheer density and elegance of this 00:04:56.720 --> 00:04:59.560 notation. Within a footprint of just three characters, 00:04:59.879 --> 00:05:02.560 the system is defining the base unit of measurement, 00:05:02.800 --> 00:05:05.600 providing the mathematical multiplier, and baking 00:05:05.600 --> 00:05:08.399 an absolutely unerasable decimal point right 00:05:08.399 --> 00:05:10.360 into the center of the string. A duplicating 00:05:10.360 --> 00:05:13.240 machine might drop a microscopic dot, but it 00:05:13.240 --> 00:05:15.279 is not going to perfectly erase a capital K. 00:05:15.819 --> 00:05:18.000 without leaving obvious signs of document damage. 00:05:18.100 --> 00:05:20.939 Exactly. It bulletproofs the schematic. The source 00:05:20.939 --> 00:05:23.639 also points out a brilliant expansion of this 00:05:23.639 --> 00:05:27.160 logic for indicating precision. If a manufacturer 00:05:27.160 --> 00:05:29.879 is making a highly precise component, they can 00:05:29.879 --> 00:05:32.500 append extra zeros to the end of the RKM string. 00:05:33.019 --> 00:05:36.699 Right, so a code written as 15M0 translates to 00:05:36.699 --> 00:05:40.259 15 .U megaohms. That trailing zero isn't just 00:05:40.259 --> 00:05:43.279 placeholder text. It is an active signal to the 00:05:43.279 --> 00:05:45.420 designer that this component is manufactured 00:05:45.420 --> 00:05:48.100 to a tighter tolerance. It communicates the intent 00:05:48.100 --> 00:05:50.139 of the original designer without requiring any 00:05:50.139 --> 00:05:52.519 verbose explanations on the schematic. But as 00:05:52.519 --> 00:05:54.680 you analyze the specific letters chosen for this 00:05:54.680 --> 00:05:57.519 shorthand, you start to uncover the severe technological 00:05:57.519 --> 00:05:59.899 limitations of the era. That brings us to one 00:05:59.899 --> 00:06:01.579 of the most interesting quirks in the entire 00:06:01.579 --> 00:06:03.730 Wikipedia article. The missing Greek letters. 00:06:03.970 --> 00:06:07.529 Yes. The multipliers make perfect intuitive sense 00:06:07.529 --> 00:06:10.470 based on the metric prefixes. K is for kilo, 00:06:10.709 --> 00:06:13.829 M is for mega, G is for giga. But the base unit 00:06:13.829 --> 00:06:17.189 for resistance presents a strange anomaly. Why 00:06:17.189 --> 00:06:20.649 on earth did the IEC choose the letter R for 00:06:20.649 --> 00:06:23.350 standard ohms? It seems counterintuitive at first. 00:06:23.670 --> 00:06:25.990 Right, because the universally recognized scientific 00:06:25.990 --> 00:06:28.769 symbol for electrical resistance is the Greek 00:06:28.769 --> 00:06:31.870 letter omega. An outsider would assume the code 00:06:31.870 --> 00:06:34.639 should be 4 omega -7. The absence of the omega 00:06:34.639 --> 00:06:37.439 symbol is a direct result of the encoding limitations 00:06:37.439 --> 00:06:41.139 of early CAD and CAM environments. When engineers 00:06:41.139 --> 00:06:43.339 were transitioning from paper drafts to early 00:06:43.339 --> 00:06:45.839 computer systems in the mid -20th century and 00:06:45.839 --> 00:06:48.519 even into the 70s and 80s, computer memory was 00:06:48.519 --> 00:06:51.060 incredibly expensive. System architectures used 00:06:51.060 --> 00:06:53.180 highly restricted character sets, right? Often 00:06:53.180 --> 00:06:56.259 limited to just basic ASCII. Exactly. They possessed 00:06:56.259 --> 00:06:58.920 the standard Latin alphabet, digits, and a handful 00:06:58.920 --> 00:07:01.399 of punctuation marks. They simply did not have 00:07:01.399 --> 00:07:03.480 the capacity to render Greek letters. We take 00:07:03.480 --> 00:07:05.100 for granted today that Unicode is everywhere. 00:07:05.300 --> 00:07:07.759 You can just drop an emoji, a Cyrillic character, 00:07:07.879 --> 00:07:10.100 or a Greek omega into any text field on your 00:07:10.100 --> 00:07:12.500 phone and it renders perfectly. But these legacy 00:07:12.500 --> 00:07:14.899 industrial systems required a stand -in from 00:07:14.899 --> 00:07:17.759 the standard Latin keyboard. So the committee 00:07:17.759 --> 00:07:20.199 settled on the capital letter R for two overlapping 00:07:20.199 --> 00:07:23.769 reasons. Primarily... It serves as a highly effective 00:07:23.769 --> 00:07:26.350 mnemoning for the word resistance across several 00:07:26.350 --> 00:07:28.709 languages. And secondarily, if you look at a 00:07:28.709 --> 00:07:31.810 capital R rendered on a low resolution, blocky, 00:07:31.829 --> 00:07:34.589 vintage terminal screen, the squared off top 00:07:34.589 --> 00:07:37.490 and the two descending legs actually bear a crude 00:07:37.490 --> 00:07:40.550 visual resemblance to the omega glyph. That is 00:07:40.550 --> 00:07:42.829 essentially an institutionalized visual pun. 00:07:43.389 --> 00:07:46.290 It's a functional, pragmatic compromise. And 00:07:46.290 --> 00:07:48.110 the standard doesn't just dictate which letters 00:07:48.110 --> 00:07:51.069 to use. It enforces a strict rule of cases to 00:07:51.069 --> 00:07:53.459 provide another layer of error -proofing. Yeah, 00:07:53.519 --> 00:07:56.160 the documentation explicitly mandates that uppercase 00:07:56.160 --> 00:08:00.139 letters K, M, G, T are exclusively reserved for 00:08:00.139 --> 00:08:02.800 labeling resistors. While lowercase letters P 00:08:02.800 --> 00:08:06.040 for pico, N for nano, M for milli are strictly 00:08:06.040 --> 00:08:08.480 segregated for labeling capacitors. That strict 00:08:08.480 --> 00:08:10.860 segregation creates an instant visual heuristic 00:08:10.860 --> 00:08:13.279 for the technician. If you are scanning a dense 00:08:13.279 --> 00:08:15.339 bill of materials or a crowded circuit board 00:08:15.339 --> 00:08:18.180 and you see an uppercase code, your brain immediately 00:08:18.180 --> 00:08:20.220 registers resistor without needing to see the 00:08:20.220 --> 00:08:23.060 word ohm. And if your eye catches... a lowercase 00:08:23.060 --> 00:08:26.060 multiplier, you immediately know you're dealing 00:08:26.060 --> 00:08:29.060 with a capacitor. It streamlines the whole cognitive 00:08:29.060 --> 00:08:32.179 load of reading a schematic. But capacitors introduce 00:08:32.179 --> 00:08:35.320 their own missing Greek letter dilemma. Right, 00:08:35.399 --> 00:08:37.360 because capacitors are frequently measured in 00:08:37.360 --> 00:08:40.179 microfarads, and the universal scientific prefix 00:08:40.179 --> 00:08:43.139 for micro is the Greek letter mu. And the standard 00:08:43.139 --> 00:08:46.519 ran into the exact same CAD -CAM wall. The drafting 00:08:46.519 --> 00:08:49.080 systems could not render a mu any more than they 00:08:49.080 --> 00:08:52.220 could render an omega. To solve this, the IEC 00:08:52.220 --> 00:08:54.960 officially authorized a workaround that is still 00:08:54.960 --> 00:08:57.980 in heavy use today. If the system lacks the mu 00:08:57.980 --> 00:09:00.379 character, engineers are instructed to substitute 00:09:00.379 --> 00:09:03.779 the Latin letter U. Either lowercase or uppercase. 00:09:04.419 --> 00:09:07.360 Visually, a lowercase u is essentially a mu that 00:09:07.360 --> 00:09:09.899 has had its leading descending tail chopped off. 00:09:10.100 --> 00:09:12.019 It is close enough to convey the meaning without 00:09:12.019 --> 00:09:14.320 breaking the encoding. And this specific character 00:09:14.320 --> 00:09:17.360 substitution is not isolated to electrical engineering. 00:09:17.960 --> 00:09:20.100 The standards adoption of U for micromirrors 00:09:20.100 --> 00:09:22.379 brought her ISO standards developed in the 70s 00:09:22.379 --> 00:09:24.419 and 80s for early information processing systems. 00:09:24.659 --> 00:09:27.740 You will even find this exact same U substitution 00:09:27.740 --> 00:09:31.519 actively utilized today within the HL7 healthcare 00:09:31.519 --> 00:09:34.200 standard for transmitting medical data and lab 00:09:34.200 --> 00:09:37.159 results. It became the universally accepted patch 00:09:37.159 --> 00:09:39.340 for systems that couldn't speak Greek. Here's 00:09:39.340 --> 00:09:41.840 where it gets really interesting. Because the 00:09:41.840 --> 00:09:45.539 RKM code was originally conceived solely to solve 00:09:45.539 --> 00:09:47.820 a localized printing and duplication problem, 00:09:48.100 --> 00:09:51.379 right? Smudged ink. Yeah, just to save the decimal 00:09:51.379 --> 00:09:54.539 point. But in doing so, it inadvertently unlocked 00:09:54.539 --> 00:09:57.179 a massive systemic advantage for supply chain 00:09:57.179 --> 00:10:00.100 logistics and manufacturing efficiency. It transformed 00:10:00.100 --> 00:10:03.000 from a simple visual safeguard into a powerful 00:10:03.000 --> 00:10:05.419 data management tool. The formatting of the code 00:10:05.419 --> 00:10:07.820 fundamentally changes how component data interacts 00:10:07.820 --> 00:10:10.740 with sorting algorithms. To visualize this, I 00:10:10.740 --> 00:10:12.440 want you to consider the friction of managing 00:10:12.440 --> 00:10:15.100 a massive spreadsheet. Say you are a procurement 00:10:15.100 --> 00:10:17.340 engineer and you need to sort a raw list of component 00:10:17.340 --> 00:10:19.799 values for a new product run. And your column 00:10:19.799 --> 00:10:25.399 contains purely numeric entries. If you run a 00:10:25.399 --> 00:10:27.580 standard sort command on those raw numbers, the 00:10:27.580 --> 00:10:29.299 database struggles with the shifting orders of 00:10:29.299 --> 00:10:31.899 magnitude. The units ohms versus kilohms are 00:10:31.899 --> 00:10:34.710 not aligned. You might end up with a 3 .3 ohm 00:10:34.710 --> 00:10:37.870 part sitting uncomfortably close to a 3300 ohm 00:10:37.870 --> 00:10:40.629 part on the screen, creating a disjointed, chaotic 00:10:40.629 --> 00:10:43.789 view of the actual physical requirements. Standard 00:10:43.789 --> 00:10:45.950 database logic doesn't inherently understand 00:10:45.950 --> 00:10:48.090 the physical relationship between those raw numbers. 00:10:48.330 --> 00:10:49.870 But watch what happens to the data structure 00:10:49.870 --> 00:10:52.429 when those identical values are translated into 00:10:52.429 --> 00:10:54.330 RKM notation before hitting the spreadsheet. 00:10:54.690 --> 00:10:58.750 1 .0 ohms is entered as 1R0. 3 .3 ohms becomes 00:10:58.750 --> 00:11:03.549 3R3. 3300 ohms transforms into 3k3. 3600 becomes 00:11:03.549 --> 00:11:08.230 3k6. 4700 becomes 4k7. Now how does a standard 00:11:08.230 --> 00:11:11.110 alphanumeric sort handle that column? An alphanumeric 00:11:11.110 --> 00:11:13.450 sort processes that string left to right, character 00:11:13.450 --> 00:11:16.309 by character. When it runs from A to Z, the translated 00:11:16.309 --> 00:11:18.710 values line up in perfect sequential harmony. 00:11:19.070 --> 00:11:22.149 1R0 is followed seamlessly by 3R3, which transitions 00:11:22.149 --> 00:11:25.649 to 3K3, then 3K6, and finally 4K7. The standard 00:11:25.649 --> 00:11:27.970 naturally forces the components to clump together 00:11:27.970 --> 00:11:30.230 into distinct, perfectly ordered groups based 00:11:30.230 --> 00:11:32.649 on their actual real -world electrical magnitude. 00:11:32.889 --> 00:11:34.730 It naturally groups them. If we connect this 00:11:34.730 --> 00:11:37.309 to the bigger picture, the seemingly minor formatting 00:11:37.309 --> 00:11:39.970 trick is a superpower for inventory. management. 00:11:40.250 --> 00:11:42.870 Consider a manufacturing facility preparing to 00:11:42.870 --> 00:11:46.169 build 100 ,000 units of a new consumer device. 00:11:46.610 --> 00:11:49.610 The engineers generate a bill of materials or 00:11:49.610 --> 00:11:53.019 BIM. When that BUM is sorted alphanumerically 00:11:53.019 --> 00:11:56.100 using RKM notation, a designer scanning the list 00:11:56.100 --> 00:11:58.960 will immediately spot a 4K7 resistor sitting 00:11:58.960 --> 00:12:02.120 directly adjacent to a 5K1 resistor. And this 00:12:02.120 --> 00:12:04.080 raises an important question for the design team 00:12:04.080 --> 00:12:06.620 during the optimization phase. Is the circuit 00:12:06.620 --> 00:12:09.299 truly sensitive enough to require two distinct, 00:12:09.519 --> 00:12:12.759 highly specific resistance values? In many pull 00:12:12.759 --> 00:12:14.639 -up or pull -down applications, the answer is 00:12:14.639 --> 00:12:17.620 no. The engineers will realize they can standardize 00:12:17.620 --> 00:12:20.659 the design by utilizing the 4 .7 kiloohm for 00:12:20.659 --> 00:12:23.200 both locations on the board. The cascading financial 00:12:23.200 --> 00:12:25.399 effects of that single decision are massive. 00:12:25.600 --> 00:12:28.039 The procurement team only has to source one component 00:12:28.039 --> 00:12:30.080 instead of two, allowing them to leverage bulk 00:12:30.080 --> 00:12:32.740 pricing. The factory floor saves crucial setup 00:12:32.740 --> 00:12:34.919 time because they only need to load one reel 00:12:34.919 --> 00:12:36.980 of parts into the automated pick -and -place 00:12:36.980 --> 00:12:39.360 machines. The entire assembly line is simplified. 00:12:39.940 --> 00:12:43.360 By forcing the data to sort cleanly, the RKM 00:12:43.360 --> 00:12:46.860 standard naturally guides engineers toward rationalizing 00:12:46.860 --> 00:12:49.480 their inventory, saving companies millions of 00:12:49.480 --> 00:12:52.129 dollars in logistical overhead. It is wild to 00:12:52.129 --> 00:12:54.309 think that a notation format designed to prevent 00:12:54.309 --> 00:12:57.309 smudged ink on mimeograph paper accidentally 00:12:57.309 --> 00:13:00.049 handed the manufacturing world the perfect database 00:13:00.049 --> 00:13:03.009 sorting algorithm. It really is. But as with 00:13:03.009 --> 00:13:05.210 any dense engineering document that has evolved 00:13:05.210 --> 00:13:08.230 over decades, the deeper you read into the standard, 00:13:08.429 --> 00:13:10.710 the more complex and contradictory the rules 00:13:10.710 --> 00:13:12.750 become. Oh, definitely. Let's look at the rules 00:13:12.750 --> 00:13:15.269 for defining a component's tolerance. Physical 00:13:15.269 --> 00:13:18.350 manufacturing is never flawless. There is always 00:13:18.350 --> 00:13:20.389 a margin of deviation from the stated value. 00:13:20.570 --> 00:13:22.870 And the IEC standard integrates this physical 00:13:22.870 --> 00:13:25.629 reality by appending a specific letter to the 00:13:25.629 --> 00:13:28.830 very end of the RKM string to denote the tolerance 00:13:28.830 --> 00:13:31.330 percentage. A technician might look at a component 00:13:31.330 --> 00:13:34.389 and see a code ending in the letter J, indicating 00:13:34.389 --> 00:13:36.450 the part has a symmetrical tolerance of plus 00:13:36.450 --> 00:13:39.090 or minus 5%. Alternatively, the code might end 00:13:39.090 --> 00:13:41.950 in K, denoting a plus or minus 10 % tolerance. 00:13:42.600 --> 00:13:44.960 The lineage of those specific tolerance letters 00:13:44.960 --> 00:13:47.899 is a fascinating piece of industrial archaeology 00:13:47.899 --> 00:13:50.600 in the source material. It really is. It traces 00:13:50.600 --> 00:13:53.179 the selection of J &K all the way back to the 00:13:53.179 --> 00:13:56.159 mid -1940s. They are direct descendants of the 00:13:56.159 --> 00:13:59.679 American War Standard, the AWS, and the Joint 00:13:59.679 --> 00:14:02.279 Army -Navy specifications. During World War II, 00:14:02.500 --> 00:14:05.679 the sheer scale of producing ruggedized radios, 00:14:05.740 --> 00:14:07.860 radar installations, and electronic equipment 00:14:07.860 --> 00:14:11.860 required a unified code for component interoperability. 00:14:11.919 --> 00:14:14.700 across different defense contractors. And those 00:14:14.700 --> 00:14:17.620 wartime military specifications were so deeply 00:14:17.620 --> 00:14:19.960 entrenched in the supply chain that they were 00:14:19.960 --> 00:14:22.480 seamlessly absorbed into the civilian standard. 00:14:22.820 --> 00:14:25.720 When you read a tolerance code on a modern motherboard 00:14:25.720 --> 00:14:28.240 today, you are interacting with a logistical 00:14:28.240 --> 00:14:30.500 framework forged during the Second World War. 00:14:30.879 --> 00:14:32.860 The military lineage is incredibly neat, but 00:14:32.860 --> 00:14:35.100 the section on production date codes reads like 00:14:35.100 --> 00:14:37.340 a deliberate logic puzzle. It really does. To 00:14:37.340 --> 00:14:39.580 track when a component was manufactured, the 00:14:39.580 --> 00:14:42.399 standard defines a rotating 20 -year cycle utilizing 00:14:42.399 --> 00:14:44.669 a single letter. The letter A, for instance, 00:14:44.850 --> 00:14:48.009 is assigned to the year 1970. However, because 00:14:48.009 --> 00:14:51.230 it is a fixed 20 -year loop, that exact same 00:14:51.230 --> 00:14:54.789 letter A is recycled to represent 1990 and then 00:14:54.789 --> 00:14:57.909 again for 2010 and 2030. The ambiguity built 00:14:57.909 --> 00:15:00.330 into that cycle is notoriously frustrating for 00:15:00.330 --> 00:15:02.889 repair technicians dealing with vintage electronics. 00:15:03.210 --> 00:15:05.889 An isolated component stamped with an A provides 00:15:05.889 --> 00:15:08.570 zero definitive chronological information on 00:15:08.570 --> 00:15:11.059 its own. The technician is forced to play detective, 00:15:11.279 --> 00:15:13.820 using context clues from the surrounding circuitry, 00:15:13.860 --> 00:15:16.299 the manufacturing style of the board, or the 00:15:16.299 --> 00:15:18.879 chassis designed to deduce which 20 -year epoch 00:15:18.879 --> 00:15:21.639 the component actually belongs to. And the logic 00:15:21.639 --> 00:15:23.460 behind which letters they are actually allowed 00:15:23.460 --> 00:15:25.600 to use for these year codes contains the most 00:15:25.600 --> 00:15:27.620 hilarious contradiction in the entire Wikipedia 00:15:27.620 --> 00:15:30.259 article. It is amazing. The standard is obsessed 00:15:30.259 --> 00:15:32.559 with preventing visual reading errors. Therefore, 00:15:32.740 --> 00:15:35.000 it explicitly bans the use of the letters I, 00:15:35.179 --> 00:15:38.539 O, Q, G, and Z for the year code. And the rationale 00:15:38.539 --> 00:15:40.879 for the band is perfectly sound. An uppercase 00:15:40.879 --> 00:15:43.659 I is easily mistaken for a numeric one. An O 00:15:43.659 --> 00:15:46.460 or a Q can quickly be misread as a zero. A G 00:15:46.460 --> 00:15:49.519 closely resembles a six. And a Z is dangerously 00:15:49.519 --> 00:15:52.399 close to a two. In a system designed to eliminate 00:15:52.399 --> 00:15:55.059 ambiguity, striking those characters from the 00:15:55.059 --> 00:15:57.980 ear pool is a rigorous, logical decision. It 00:15:57.980 --> 00:16:00.620 is an airtight, flawless rationale. That is, 00:16:00.700 --> 00:16:03.419 until you turn the page and see how the standard 00:16:03.419 --> 00:16:05.620 handles the code for the month of production. 00:16:05.779 --> 00:16:07.779 Oh, this is the best part. For the months of 00:16:07.779 --> 00:16:10.779 January through September, the standard predictably 00:16:10.779 --> 00:16:13.840 uses the single digits 1 through 9. But the system 00:16:13.840 --> 00:16:16.440 requires a single character code. When they hit 00:16:16.440 --> 00:16:20.379 October, the 10th month, a two -digit 10 is invalid. 00:16:20.980 --> 00:16:22.720 I'll let you explain how the committee solved 00:16:22.720 --> 00:16:25.629 the October problem. They immediately abandoned 00:16:25.629 --> 00:16:28.370 their own visual safety protocols. To represent 00:16:28.370 --> 00:16:30.870 October, the standard officially dictates the 00:16:30.870 --> 00:16:33.309 use of the uppercase letter O. They took the 00:16:33.309 --> 00:16:35.409 exact letter they had just aggressively banned 00:16:35.409 --> 00:16:37.590 from the year codes because it posed a severe 00:16:37.590 --> 00:16:40.409 risk of being misread as a zero and made it the 00:16:40.409 --> 00:16:42.970 mandatory code for the month of October. It is 00:16:42.970 --> 00:16:45.190 the ultimate hallmark of design by committee. 00:16:45.370 --> 00:16:47.629 One working group likely locked down the safety 00:16:47.629 --> 00:16:50.289 rules for the year codes. A separate group, or 00:16:50.289 --> 00:16:52.169 maybe the same group working on a tight deadline, 00:16:52.429 --> 00:16:54.870 realized they had run out of single Arabic numerals 00:16:54.870 --> 00:16:57.009 for the months. So they likely grabbed the letter 00:16:57.009 --> 00:17:00.070 O, simply because it provided a convenient phonetic 00:17:00.070 --> 00:17:03.029 match for October, completely disregarding the 00:17:03.029 --> 00:17:05.130 strict visual heuristic they had established 00:17:05.130 --> 00:17:07.990 just paragraphs earlier. Exactly. So what does 00:17:07.990 --> 00:17:10.380 this all mean for us today? We are no longer 00:17:10.380 --> 00:17:13.299 constrained by the memory limits of 1970s mainframes 00:17:13.299 --> 00:17:17.400 or the smudgy ink of 1950s mimeographs. We design 00:17:17.400 --> 00:17:20.599 circuits on 4K monitors. We have laser engravers 00:17:20.599 --> 00:17:22.960 that can print microscopic high -resolution text 00:17:22.960 --> 00:17:25.880 onto any surface. We can easily print decimal 00:17:25.880 --> 00:17:28.599 points, omega symbols, and mu prefixes. Has the 00:17:28.599 --> 00:17:31.869 RKM code become obsolete? Far from it. The RKM 00:17:31.869 --> 00:17:34.529 notation proved to be so robust, so elegant, 00:17:34.630 --> 00:17:37.109 and so deeply embedded in the cognitive workflow 00:17:37.109 --> 00:17:39.630 of electrical engineering that it has entirely 00:17:39.630 --> 00:17:41.960 broken containment. It's everywhere now. The 00:17:41.960 --> 00:17:43.779 community routinely uses it for applications 00:17:43.779 --> 00:17:46.420 the IEC never officially sanctioned. For instance, 00:17:46.559 --> 00:17:48.980 the standard does not strictly dictate this format 00:17:48.980 --> 00:17:51.339 for inductors, yet you will constantly see inductors 00:17:51.339 --> 00:17:54.240 labeled as 4R7. Any engineer reading that board 00:17:54.240 --> 00:17:56.559 immediately translates the dialect, knowing it 00:17:56.559 --> 00:17:59.160 specifies 4 .7 microenrays. It has essentially 00:17:59.160 --> 00:18:02.099 become open source slang. And the most mind -blowing 00:18:02.099 --> 00:18:04.839 detail from the source material is how this hardware 00:18:04.839 --> 00:18:07.559 labeling trick jumped the gap into modern software 00:18:07.559 --> 00:18:10.470 engineering. Yes. When programmers are writing 00:18:10.470 --> 00:18:12.710 code for embedded systems or hardware description 00:18:12.710 --> 00:18:15.630 languages, they have to define variables to represent 00:18:15.630 --> 00:18:18.809 physical voltages. But a compiler will throw 00:18:18.809 --> 00:18:21.529 a fatal error if you try to put a decimal point 00:18:21.529 --> 00:18:24.509 in the middle of a variable name. The software 00:18:24.509 --> 00:18:27.609 architecture strictly forbids it. To the compiler, 00:18:27.849 --> 00:18:30.529 a dot often represents a structural command or 00:18:30.529 --> 00:18:34.099 method call. It cannot process 3 .3 as part of 00:18:34.099 --> 00:18:36.200 a text string identifier. So when a software 00:18:36.200 --> 00:18:38.000 engineer needs to name a variable representing 00:18:38.000 --> 00:18:41.519 a 3 .3 volt power rail, they reach back to 1952. 00:18:41.680 --> 00:18:44.900 They utilize the RKM hack. They name the variable 00:18:44.900 --> 00:18:48.259 3V3. The letter V steps in to serve as both the 00:18:48.259 --> 00:18:51.410 decimal separator and the unit identifier. perfectly 00:18:51.410 --> 00:18:53.670 satisfying the compiler's syntax rules while 00:18:53.670 --> 00:18:55.750 remaining instantly readable to the human programmer. 00:18:55.930 --> 00:18:59.190 You also frequently see 3P3, where the P denotes 00:18:59.190 --> 00:19:01.410 a positive voltage line. They are deploying a 00:19:01.410 --> 00:19:03.329 solution designed for physical hardware labeling 00:19:03.329 --> 00:19:06.630 to bypass a 21st century software syntax restriction. 00:19:06.970 --> 00:19:09.009 It is the ultimate testament to a standard's 00:19:09.009 --> 00:19:12.549 utility. A truly robust protocol doesn't merely 00:19:12.549 --> 00:19:14.950 solve the immediate crisis it was drafted for. 00:19:15.089 --> 00:19:18.069 It establishes a logical framework that subsequent 00:19:18.069 --> 00:19:21.009 generations can adapt to solve entirely new categories 00:19:21.009 --> 00:19:24.109 of problems. It really is time to step back and 00:19:24.109 --> 00:19:26.349 survey the massive amount of ground we have covered 00:19:26.349 --> 00:19:30.000 today. What began as a desperate, pragmatic hack 00:19:30.000 --> 00:19:32.819 to protect the fragile decimal point from the 00:19:32.819 --> 00:19:35.980 ravages of 1950s duplicating machines evolved 00:19:35.980 --> 00:19:39.180 into a fiercely efficient global shorthand. We 00:19:39.180 --> 00:19:41.259 have seen how it bypassed the severe character 00:19:41.259 --> 00:19:44.039 limitations of early CAD systems, substituting 00:19:44.039 --> 00:19:46.779 ours and us for missing Greek letters. We explored 00:19:46.779 --> 00:19:49.079 how its alphanumeric structure inadvertently 00:19:49.079 --> 00:19:51.500 revolutionized supply chain logistics, allowing 00:19:51.500 --> 00:19:53.920 spreadsheets to naturally group components and 00:19:53.920 --> 00:19:55.859 rationalize million -dollar bills of materials. 00:19:56.250 --> 00:19:58.609 We traced its tolerance codes back to the ruggedized 00:19:58.609 --> 00:20:00.970 demands of World War II manufacturing. And we 00:20:00.970 --> 00:20:03.150 watched it survive to become a syntactic workaround 00:20:03.150 --> 00:20:05.789 for modern software compilers. The evolutionary 00:20:05.789 --> 00:20:08.069 arc of this standard is phenomenal. And as we 00:20:08.069 --> 00:20:10.569 wrap up, I want to leave you with one final lingering 00:20:10.569 --> 00:20:13.730 thought to mull over. Consider how deeply foundational 00:20:13.730 --> 00:20:16.589 the RKM code is to the global infrastructure 00:20:16.589 --> 00:20:19.930 of hardware design and procurement. Now, run 00:20:19.930 --> 00:20:23.089 a thought experiment. If early computer architectures 00:20:23.089 --> 00:20:25.670 had possessed the memory to easily render the 00:20:25.670 --> 00:20:29.130 Greek omega and mu symbols, or if those mid -century 00:20:29.130 --> 00:20:31.329 duplicating machines had simply been a fraction 00:20:31.329 --> 00:20:34.009 more reliable at preserving tiny decimal points, 00:20:34.190 --> 00:20:37.410 how much of this highly optimized, globally spoken 00:20:37.410 --> 00:20:39.970 engineering language would even exist today? 00:20:40.170 --> 00:20:42.630 That is a fascinating question. It raises an 00:20:42.630 --> 00:20:44.890 important question that extends far beyond circuit 00:20:44.890 --> 00:20:47.630 boards. It forces us to wonder how many of the 00:20:47.630 --> 00:20:50.630 unbreakable rules, optimized workflows, and universal 00:20:50.630 --> 00:20:52.930 standards that govern our modern technical world 00:20:52.930 --> 00:20:55.769 are not actually the result of pure forward -thinking 00:20:55.769 --> 00:20:59.150 design, but are instead permanent invisible monuments 00:20:59.150 --> 00:21:01.769 built to patch a temporary technological failure 00:21:01.769 --> 00:21:04.430 from 70 years ago. We are truly operating inside 00:21:04.430 --> 00:21:06.789 the fossilized workarounds of the past. Thank 00:21:06.789 --> 00:21:08.529 you so much for joining us on this exploration 00:21:08.529 --> 00:21:10.650 of the source material. We look forward to having 00:21:10.650 --> 00:21:12.089 you with us on the next deep dive.