Oh my gosh, I recognized that project box, and in the shots showing the catalog number brought up so many memories of helping my Dad do inventory at the Radio Shack store he mananged in the late 70's, early 80's. I saw the 276 prefix and new exactly where in the store on Florence Blvd, Florence, AL I would go to grab that project board.
I just watched a video by TubeTime where he delved into this and found that it wasn't a bug but rather was intentionally just completely left off the chip to make it as simple as possible. So more of an oversight of how important that instruction could be. He was able to find one with the die exposed and confirm that the transistors for ROR just plain didn't even exist in the early version.
I watched the same video. A sad thing about the Internet is that once incorrect information is disseminated and accepted by the public, subsequent attempts to correct it don't get as much traction as the original misinformed story.
Need to search for that video! The fact that was considered a bug could be because the chip did have a ROL instruction so is odd they missed the ROR but the fact that they came with a revised chip so soon after launch adding it does lead me to believe that was omitted on purpose to get to market early, in the 70’s you needed the product ready to present it in the key conferences to be able to market it properly
I love when Adrian is at a loss for words. I'm so thankful that he takes the time to make these videos. It's more than simple nostalgia. Many of us grew up loving these 8-bit systems too and the memories are reminders of moments of joy and the learning that came from experimenting. He's not just preserving history but he's sharing that love and joy with the world. I can't think of a better way to spend your time.
Well said, moments like these are why this is my favorite tech channel. Adrian isn’t some noisy RUclips personality churning out low effort videos, he genuinely loves this stuff and it comes through in his work. It’s so nice of him to share it with us. Adrian if you read comments (so many so probably not) I have a bugged Pentium 60 in a bin somewhere if you want more buggy silicon
I have a white 1976 cpu in the basement along with the Jameco receipt for $75 for it and the 6522 pia I bought in high school my sophomore year. I lived in PA not far from MOS fab. I called and an engineer sent me programming manuals and data sheets. Wire wrapped it up and had a tic tac toe program working on it using an old point of sale terminal. Later ran tiny basic on it in college. Fond memories.. I will probably dig it out when I retire in 3 years. But your video made me remember the gogo days of home computers
I work in E-Waste ... I see things like this all the time and have to take them apart to sort the parts, that board would have been thrown right in the low grade box. I try and keep an eye out for special things, but I am sure I don't always recognize that it's special. I am able to save a few things but I can't save it all. I rationalize a lot of it by saying to myself if we didn't recycle it nothing would be rare. :)
It'd be awesome to see what you've collected... even if it's just a pan over a bunch of stuff on the floor! Hope you don't mind my subscribing in hopes one day you do.
I've been thinking lately that at this point, a lot of e-waste places probably purposely employ people who are at least aware of some of the obviously-valuable stuff so they can turn it over to the secondhand market for a lot more than they make just processing garbage.
@@stevethepocket It's not about individual pieces, it's about volume. The place I work is an intake facility. We buy product from the public and some commercial and basically sort and package it. We sell it to other recyclers who break it down into useful metals who in turn sell it to foundries. Trying to sell individual items on the second hand market is way to slow. So if I don't save it or someone down the line doesn't save it, it will be destroyed. I do know that the place we sell PCBs has a machine that photographs them and then points out any IC's that can be sold intact. They will recover those IC's and then the rest of the PCB and components get ground up, sent through several different washes, then melted down to recover the metals, specifically gold and platinum. ... Disclaimer : Other recycle companies may do things differently :)
Your reaction was the nicest thing to watch. I really enjoy your channel. Keep up with the good work. Rest assured that it is really greatly appreciated.
Truly Adrian when Franks program (iz8dwf Hi Frank) popped up declaring this was in fact a bugged MOS 6502…Man I burst into tears…I’m surprised you contained your exposure. Simply wonderful…
Adrian, just found your channel and haven't looked at your other videos yet but I just had to comment. You have taken me back to 1975 where at that time we were programing industrial controllers with paper tape readers. When the 6502 came out a whole new digital age was launched in basements and garages, it was a magical time. Thank you for sharing this story for this old man.
Thanks for the mention! Wow. Now that is a very lucky find. I have one very early 6502, but it's soldered into my OSI 300 trainer. I don't know if I could devise a way to test for the ROR bug as I'm reluctant to remove it. Very cool, Adrian! There may not be real buried pirate treasure, but there is still lots of digital treasure buried away in devices like this waiting to be found!
Look on the top for a Date code: Do you have access to the assembly/monitor? Its only like 5 instructions to perform the test. The message is 3x the size of the test. Is your ROM soldered in?
You could use the switches on the trainer to enter a simple program: LDA #$22, ROR A, STA $8 (assembly is A9 22 6A 85 08; use Windows Calculator to translate to binary). You'd then read from $8 the result, which will be $11 (or maybe $91 if carry was set) if ROR works correctly, or it will be $44 if ROR behaves like ASL (the bug).
I'm glad that you didn't throw away because imagine the regrets afterwards and like Indiana Jones said "it belongs into Museum". Very impressive since I never seen this very early version of MOS 6502 before since only the later version of the Chip especially for the one of the most if not the most popular CPU of all time for 8 bit Computers. What a great find my friend.
like he said wouldn't have known to regret it. the amount of lost stuff in this world is kinda insane though, have to stop myself from thinking about it too much. entire civilisations have grown, thrived and then disappeared without a trace remaining. especially wilful destruction of information is infuriating but it no doubt pales in comparison to the amount of ignorance, incompetence and accidents. we will never know the richness of our lost history.
@@snooks5607 How would you now that, when: everybody in fact "wouldn't have known to regret it"? Sorry for that smarty-pants comment about the hole in your arguments logic:) Of course I agree. It is probably a lot of technology that is unwittingly wasted. But there is definitely more that is deliberately thrown away. To preserve everything simply was and is not been possible with our actual consumer society since the 1950s. Otherwise we would just suffocate in trash. It is at least good if something is reused or recycled[1]. The majority of tech/electronic-waste and old equipment end up in countries like Africa ... so that they suffocate, in dimensions as large as entire states in garbage dumps. Unfortunately, this is the reality:/ [1] Like from us retro-freaks or repair enthusiasts/technicians. But by all love, this is just a drop in the sea ...
Thank you to you and Frank IZ8DWF for the video and code. I had two white ceramic 6502s that I wasn't sure if they had the ROR bug. One did and one didn't and without that code (or writing something in machine language) I would not have known if they were early enough. Your reaction is quite fitting as they are truly historic and a joy to be the caretakers of them.
AY-5-1013 is a UART from General Instruments, which eventually became Microchip, famed for their PIC line on microcontrollers. I used these UARTs extensively in the 70s, they were ubiquitous in those days.
you beat me to it, i vaguely recognised the number and googled it, i may have one or 2 lurking around, definitely havent got any ceramic 6502s, i did find a ceramic 8008 on a board....
That chip is also a piece of history itself The Taiwan silkscreen shows that the chip might be the first ever batch of chips that Taiwan manufactured in history, cause the semiconductor division of ITRI(which later dissociate from government and became UMC/Mediatek until now) finished their first batch of chips around the end of 1977
I loved this @adrian! I was studying electronics in 1974-1977, and then did a PhD in computer design. My one quibble with the video is that the 6502 is NOT a simplified 6800. Yes most of the original design team came from motorola, and yes the 6501 was pin compatible to the 6800. However it's architecture is very different to the 6800 and it has a fraction of the instructions. For example it has 2 index registers, and far fewer instructions. Chuck wanted to build a better chip for control, because he found that people just balked at the $300+ price of the 6800! Chuck Peddle's vision was for a cheaper chip to do simple interface and washing machine, etc. control. In other words using the microprocessor as replacement logic. I think he would have loved this application of the chip! And I love your sheer joy and excitement of having it work! It made my day!😊
I agree. Sometimes there was an old way of doing something w/o a microprocessor, but now always. I used mainly a Z-80 in that time period but I remember looking at the 6502 instruction set, etc., and I remember how relatively bare it was. Sometime later than that my most hated micro was a PIC of some kind where instead of conditionally jumping to someplace you had to write the jump instruction and then branch over it if you Didn't want to jump. Arrrg!
This is so cool. I appreciate how you were able to convey the importance of the discovery to people less well steeped in 6502 lore. That is a really awesome discovery, and even more awesome that it runs (aside from the bug) correctly. Thanks so much for sharing.
I recognized the 1702 EPROM right away, I was in that electronics hobbyist era in the early to mid '70s, went to tech school, and it became my career. Working at different jobs, seeing so many esoteric parts that are no longer made and have no equivalent. So glad you were able to save that little bit of history.
A first revision MOS 6502 from the first year of production. You're not kidding, binning that would've been a tragedy, that thing belongs in a museum display!
I remember the times when you could go down to Radio Shack and get raw boards and etching to make your own circuit boards. Those were the days, really Miss the old Shack :(
@BumbleBee Maplin was already pretty useless by the late 00s sadly. A few connectors and plugs but not much else to my recollection. Or maybe I just had a rubbish tiny one near me.
I remember when you could go to Radio Shack store and buy the parts to make a radio. I knew the end was coming the day I went to Radio Shack and they were more concerned about who my cell phone carrier was than whether or not they had a germanium diode for my kid's project.
I spent hours in Rat Shack as a kid. They had all the common stuff in stock and could order in just about anything else. The employees were quite knowledgeable. There is no more customer service in the US.
I have a Pentium (1) with the FDIV bug! It was removed from an IBM 9595 server while I worked at IBM in the 90s. If you want it I can send it to you Adrian although it will take some time since I live in Buenos Aires. Keep up the great work!
Do you know if it also has the meltdown bug. When meltdown was published, the CPU designers were awfully silent about the status of older chips. The official Meltdown test programs tend to use non-buggy unrelated features of newer chips, but it's basically a question if a specific documented security check is done before hitting cache and external bus or not.
@@johndododoe1411 Meltdown would date back to the Pentium Pro and the P6 architecture. So the Pentiums would be unaffected. Incidentally, I also have a FDIV bugged Pentium 90.
@@steeviebops Are you sure the Pentium doesn't let the Meltdown affected operation hit the cache, thereby leaving a side effect of the supposedly hidden value?
Adrian, I'm so glad that I stumbled into your channel so many years ago, I think from an LGR recommendation, when I really just started on modern PC gaming channels and slowly migrated to this time period. At first I thought it was a bit too technical, despite being an electrical and computer engineer, I never went so far as to have the practical and working knowledge you have and went into the business side. Now every week I'm eagerly waiting for your next video to drop, whether it's this channel or your 2nd channel. It's so hard to describe how over time, you became my all time favorite RUclipsr, but I think the reaction in this video encapsulates it. We all live vicariously through you. We are in awe about how you self taught your way to being one of the most knowledgeable early computer repair people on the planet, and yet still be humble enough to reach out for help, while always letting your true passion and excitement come through in your videos. There are some days I see when you drop a 45 or 55 minute video, and I get giddy thinking I need to save this for a long relaxing viewing session before I go off to sleep. I guess what I'm trying to say is just... thank you. Thank you for reigniting my love of this hardware which originally pushed me to get the degrees I did all those years ago, and remember why I did. Everytime I learn something new from you every episode, and it's got me into tinkering again. I feel like I'm rambling, so I'll stop. Never change, because whatever it is you are doing... it freaking works.
So well said. Adrian is one of a short list of people who represent to me the best of the internet - passionate people who unapologetically and humbly share that passion with the world. It's a great little community we have here in retrocomp land. Aside: sigh, anyone here watched Aussie50? RIP Ed. He was the one who showed me it's OK to be passionate about tech. I think he was pivotal in really kicking off the amateur tech community. I consider channels such as these to be an offshoot of that.
Actually back in the day the ability to write simple machine language programs and design simple circuits is considered basic skills, like the ability to write HTML these days.
The workmanship was awesome. I was just grinning ear to ear !! The code and project just for an interface. This guy was on the top of his game. Reminds me of the "City on the Edge of Forever": ...stone tools and bear skins. Radio Shack really deserves more credit than they are getting. We took them for granted. They even took themselves for granted.
@@18000rpm Or, they could've simply been an electrical engineer or computer engineer who worked in the industry and had access to prototypes and other parts like this that were quickly discarded as updated versions came out.
So glad for you Adrian. Your happiness is contagious and we all feel like we have touched this magnificent piece of history. Hope you located it in a privileged place on your bench, so we can admire it in future videos.
The 6502 was also in the British Broadcasting Corporation home computer - the BBC Micro 32K, a machine that most people totally ignore. There was an American version, the ABC which was virtually identical and sold in the States, but as I said, it virtually unknown today. The BBC was proceeded by the Acorn Atom, the Acorn Electron. The BBC Micro came in a 16K version - the BBC Model A, the Model B which was either 16K, 32K, or 64K and some rare modified ones had 128K. Later it was surpassed by the BBC Master 128K and the BBC Master Compact. These machines could include a 512K add-on board that ran a very early version of DOS, otherwise the machines all had tape and floppy disk access, as well as the first Winchester hard drive, and Laserdisc system. The BBC Micro also supported a local network, TELETEXT, PRESTEL, telephone modem access, a 6502 co-processor and many other adaptions, including a MIDI system. I had a friend who worked for a software company, Acorn Computers Limited in Britain, and she would send me sales brochures. Eventually this machine was supplanted by the Archimedes, the world's first RISC system, which was in effect an early PC clone. I believe there is an emulator for it that runs on PC called BeebEM (Beeb being the affectionate name given to the BBC machine by its users). Of note is the fact that the worlds very first 3D space combat and trading game, the legendary ELITE by Ian Bell and David Braben was developed on a BBC Micro in the eighties that was ported to virtually every other computer platform of the time and which still remains popular - in fact, the modern remake of the game is called ELITE Dangerous.
The early 6502 cpus did NOT have the silicon circuitry to support a ROR instruction. so it was NOT a bug but not even implemented. The ROR instruction was added later in the next revision of silicon. (It DID have a shift right but not the rotate right). the reason that the code almost works is that the motorola style processors such as 6502 and 68xx were random logic decoded so as to operate much closer to the clock frequency. the Intel style (8080 and Z80) used microcode which take a lot more clock cycles per operation (often 10 or 12x instead of 2x or so). If you take a 6502 or 6800 processor and list all of the operations in a table you will see patterns and missing opcodes for stuff like a STA immediate (which doesnt make much sense in real code)
There have been a lot of recent comments about the missing ROR being "not a bug." I feel this is semantics, pure and simple. The feature is not fully implemented on the first revision of the chip, as pointed out by TubeTimeUS in his recent video, but there is plenty of evidence it was the intention of the designers to have it there from the start. As pointed out by TubeTimeUS, there was just no space for the transistors to make it work properly. It's like they ran out of time/space or made a mistake implementing it, not realizing they would need some more room for the extra logic. So when they decided to release the CPU without ROR, they just removed it from the August 1975 datasheet. (That last part entirely speculation.) The issue is, there is just a lot of evidence it was supposed to have been there from the start: * Robin has a great tweet showing the decode table for the 6502 with ROR. Notice how it fits in so perfectly? twitter.com/8BitShowAndTell/status/1629754457172111361?s=20 * The Jan 1976 hardware manual for the 6502 lists ROR and says it will be working after June 1976. That is so soon after the chip's release to already be saying it will be implemented. * There is some wiring for the instruction that doesn't go anywhere, but there was some work done to implement something. Why wouldn't this have been ROR but they couldn't fit it, so they just gave up on ROR for that first revision? * It seems unlikely to implement ROL but not ROR - -especially since the 6800 (the "father" of the 6502) had both. The real discussion to have is this: Is an unimplemented feature that was supposed to be there a bug or not? There is an argument to be had either way and both are valid points. For me personally, I look at it as a bug.
I love that this was saved and it works and looks perfect. Not only did you save an extremely rare chip, but you know what it is. This one couldn't have went to a better person. You'll treat it like the work of art it is.
I think very early on MOS Tech was selling a boxed set of CPU manuals and 2 of those 6502s at shows. I had one as a hand me down and passed it along to a 3rd party years later. Yep, two ceramic 6502s from that year.
This is why I love watching this channel, your enthusiasm is always sure to bring a smile to my face. What a great story of the buggy chip. A survivor indeed!
That is a fantastic find. Keep preserving our digital history. You rock! In my years of experience hand building boards, I came to loath stamped pin sockets and wire wrap. I did build hundreds of boards with that method and sockets. Virtually all of the builds had all manner of weird problems traceable to faulty connections. Machined pin sockets are much pricier than the stamped pin variety. BUT, they pay for themselves by eliminating many hours of wasted time spent troubleshooting an unnecessary problem. The one pin you found replaced on the socket was, most likely, because it was intermittent or outright open. ALL stamped pin sockets are a GUARANTEED built in failure point. Wire wrap also corrodes severely in a short time and CAN become intermittent as well. Depends on the alloy of the pins and wire combinations. Galvanic corrosion results with mismatched materials. FYI, wire wrap was developed for computers in missiles and rockets due to lacking a stress point where the wire exits the solder on a normal connection. This also saved weight by eliminating all that lead solder in the rocket. Missiles and rockets usually had very high QC to bypass the initial bad connections and corrosion was not an issue as these things didn't sit around for very long before being used. Once.
I'm so glad you did take the time to look in that box. Not only because of the chip -- which IS super awesome -- but because it obviously has so much meaning for you. Well done on saving such an interesting piece of history!
23:36: If you were to test the Apple II + 1975 6502 more extensively, maybe on your second channel, I'd be happy to watch that. I'm now wondering if Apple accounted for the ROR bug and wrote their Apple II firmware so as to test for the bug and work around it if present. Of course, it's a safe bet most Apple II software did not do the same, but I'd love to know if Apple's early Apple IIs were "ROR bug-tolerant".
@@SidneyCritic ...Even though by that time, the faulty chips would likely have been out of stock and harder to get hold of than the good ones. Still wouldn't put it past him though.
@@SidneyCritic most probably there'd have been a version of the ROM code specifically for use with that chip if so. A fairly simple solution would be to provide a ROM routine to perform ROR by executing ROL seven times instead but I don't know how that would affect general compatability and performance. You couldn't really intercept the opcode though. You'd probably have to prepatch everything.
Amazing story. Your videos take me back to high school in 1980. Kids today have no idea how primitive and wonderful things were way back in the late 1970s.
I've really enjoyed this video Adrian! Thanks for sharing this fabulous CPU! Now you need to buy all old project boxes from the '70s and I'll pay for them :)
I read many years ago that the low cost of the 6502 was one of the main reasons Jobs and Wozniak used it in their early Apples instead of the 8080. And thus the great divide was born... Apples vs PCs. I still have my original Poly-88 micro based on a 2MHz NEC 8080 from 1977. The Poly was one of the few early micros that didn't require manual entry of a bootstrap program, as it had a 1k ROM for boot and I/O functions. I have a soft spot in my heart for these old machines, IMSAI, Altair, SWTPC. I still have the advertisement cut sheets for them, as well as a number of late 70s micro periodicals: Byte, Dr. Dobbs, etc. Great video!
I always liked the look of the old ceramic packages of early chips. This is definitely a super cool find. As a teenager I used to experiment with TTL chips on breadboards, but higher end chips like this were definitely out of my budget in the old days. I really want to get back into that again. Great video, and I appreciate the work and research done by you and others to preserve these pioneering pieces of history.
That's an absolutely amazing find! Most people don't really know the value of old equipment (especially when it still works) and it's always a bit sad knowing how much working legacy hardware lies buried in landfills.
I really enjoyed watching tis video, because you transport your excitement about the 6502 to your viewers. The 6502 in white color and golden contacts has really something magical and i’m sure that this chip has found its rightful owner now!
What a beauty of a chip!! CpuGalaxy would be really jealous of you for that white ceramic first cut 6502! Considering also that it is the first CPU for the masses it takes a very very special place in computer history and my heart also as a computer enthusiast with extremely low budget. The other wow thing about that chip is that it is fully functional also with its cute bug... My suggestion is to put it in a working pretty rare system that can use it and make a transparent case for the whole computer unit with this little white beauty displaying its body and also running the system in exhibitions or videos for people to see! Greetings from Greece, you are a happier man after that and keep on doing your stuff for us to learn and enjoy!
Oh damn Adrian, what an outstanding find!! Now I am jealous, I am still missing this CPU in my collection 😮. Great that you didn’t throw the box away. And your emotions definitely swapped over to me here in Austria. We share the same passion. Cheers, Peter
Ohhhh! The wire wrapped back of the board!! Get a deep shadow box and maybe put the board in tilted forward with a mirror behind it and an angled mirror next to it. So you can see the back as well as the front. That would look so cool! Great find, and good presentation on the history. I never knew about that early bugged version.
Makes sense. Thanks for clarifying. For those who don't understand, he is talking about where ROR instruction is used in Apple ROMs, and "monitor" is memory-editing software, not a display device.
It's interesting that the ROM boots, its likely that whatever operation uses the ROR function doesn't have its result verified, so the ROM is 'running' but at some point you're going to try to use something in memory that's been run through the buggy ROR routine and you'll receive unexpected results. Still very neat to see it preserved and the short history lesson! Thanks for sharing!
I always thought the gray ceramic DIP ICs were like little socketed pieces of jewelry, but the white one looks especially pretty. It's always a pleasure to open a machine and find a ceramic chip inside. #TotallyJealous
I bought my first Apple ][ in 1979 for $2250. It started me on a lifelong career as a programmer and Network Engineer. I have an original Gold Intel Pentium as a collectable.
Excellent idea. Seconding this! Could incorporate the board's edge connector into the side of the shadow box to hold the board a little off of the mirror to make it easier to see the reflection.
I've just started watching the video and looking at the box that Adrian nearly threw away. At first glance it looks like the box is "The Internet" as seen on the British Channel 4 comedy programme The IT Crowd 🌐 😂 Cool to see an early 6502
Beautiful video. I have known the assembler of the 6502 since it entered the business, and then moved on to the older brother: the 6510, of the Commodore 64. I was amazed that the clock is generated by a 555. Great job man !
This is one of the most amazing videos I've seen for a long time, I'm afraid your title didn't really convey how special and I procrastinated over watching it (I'd assumed that the rare and special processor would be something later or less interesting to me.) I loved your excitement and emotion and was 100% with you on that. The 6502 is special to me too, for the same reasons as yourself, and I'm very glad to have seen this specific early example, learned a little about its history and seen it running.
Great video Adrian. In 1975 I had a small electronics startup and had ordered some 8080s for some programmable boards we were doing for Lockheed. They were not yet available and the price was $360 each. Before delivery, Intel dropped the price to around $220. The chips were white and gold, just like your 6502. Thanks for bringing back some great memories.
@@samuelcolvin4994 Yes, a chain of shift registers (up to 2 K bytes!) with a continuously scrolling set of data. Other chips sit on the same bus and watch for particular values in this sequence (command codes), then take the data and put the results back.
You are lucky it worked as we were taught that technician handling the chip must ground his body by wearing a conductive wrist strap which is connected to a conductive mat which at ground potential. Thank you for all your videos including bad and great decisions that we all learn from!
This whole video is amazing. I love hobby projects like this, I totally admire the old-school skills of drawing up a schematic and wire-wrapping a proto board (btw can high-res pics of the top/bottom be preserved as well?), the historical aspect is fascinating, the chip itself is beautiful, the custom ROM is genius, and, on top of all of that, "it freaking works!"
Amazing find! So glad you didn't throw that beautiful 6502 out! :D I just came across an early 80s PET printer interface and it looks very similar to the self-made board you have there except for using "proper" RAM. And there is actually a (quite a bit newer) 6502 on there, too!
Hi Adrian, you tugged on my heart strings as well my friend. Not only is this my 1st micro (well second, if you count that the 1st was not completed and working yet too much), but my 1st running and working computer was the Kim-1. I worked for the National Science foundation here at a college in Massachusetts in 1975, was part of a small group that created an 8008 based computer to help teach for students on a project called CaLM. Computer and Laboratory Math. The 8008 was our design and needed so much more work, then we found out about the Kim-1 by mostek and put our old 8008 design on the shelf. We used the Kim1 for many presentations about computer programming for NSF and some classes until not to long after the project was canceled. The late 70s was filled with home computers in a way back then, at least for us in that project. I got into the 6502 so much myself in that it directed my future in programming as well as hardware with that chip. I was involved with the 6502 with my career in a Game company, a computer publishing company, my own development business and so on. Your right buddy. That chip tugged on my heart I believe in a way as powerful as yours. It was the foundation that changed the world. Don't let it go. It's almost #1. Thanks for sharing my friend. Btw,, awesome find about ror. I've heard about it once, but never saw that issue. The Kim's we had was abit later than that. I think we had 02s in mid 76. The 8008 was in 75.
That would go well in my RevA Kim-1! It has the grey slate, but I've always wished it had the white ceramic. I also have a 6530 in white ceramic for it!
Wow, not only do you have a very early version of the 6502, but you have a rare one with the bug. This is an incredible piece of history -- congratulations!
@@_B_K_ Sorry that you have problems with reading what you wrote. "not only do you have a very early version of the 6502, but you have a rare one with the bug" You clearly say that only the very early ones with that bug are rare while all the very early ones had that bug.
Hey Adrian, that's an amazing find! The ceramic MCS6502s are highly sought after by Apple-1 replica builders for their original appearance. They are virtually unobtainium these days and sell for upward of $3000 on eBay.
@@Okurka. That's probably the ultimate question. There is one that did sell on October 26th for $1500 bearing a 3975 date code though, but that may be a unicorn.
@@rjolly87 Sold mine direct to a collector for $1500 (don't like dealing with ebay) ..I had bought it in '75 for $20, new for a project that never happened..
4:46 That . . . is a ceramic MOS 6502 if I'm not mistaken. While I am interested in older computers I'm not that well versed, but if I recall correctly just the fact that it's ceramic rather than plastic makes it 1.) rare, 2.) old. 5:16 1975? Wasn't the 6502 introduced in '73 or something? Hot damn. 5:46 Ah, it was '75, not '73. And yeah, the 6502 WAS the home computer revolution. It dropped the price from, IIRC, something like $300 for a processor to $30. That's not adjusted for inflation, so that was still expensive but affordable for a home user. 7:00 LowSpecGamer has a pretty good video on the history of the 6502 too. Ben Eater's series on building a 6502 computer is great if you want to dig into the nitty gritty of how it works from the ground up. 13:05 I looked up the datasheet for the top left chip and it's a UART, so yeah, I/O chip. 13:58 Not to shill, but Ben Eater has an EEPROM programmer (TL866 II Plus) for sale on his website for like $70. Fantastic video, thanks for making it!
Wow, what a great find of this piece of history! And it couldn't have gone to a more deserving person. You have done so much (and continue doing so) for the retro community as well as myself. Congrats and thanks for sharing this moment with us.
How cool is that, Adrian?! You have a piece of history in your hands! What would be nice would be to get some good closeups of the board and the wiring and maybe some of us can try and reverse engineer the schematic for the board and the box so the rest of the story isn’t lost! Excellent!!!
Amazing find! I got goosebumps when you popped that cover off. The 6502 and it's variants powered almost all of my childhood and early teen years. So cool this one ended up in very deserving hands! You'll have to show us your shadowbox when you have it done!
I'm just now getting into the hardware side of things after a decades long programming career and really geeked out about this. It's sad that this ended up in e-waste and awesome that you caught it! Looking forward to your next discovery!
What other cheaper microcontrollers were available (in small quantites) in the mid-late 70s? I don't think the 8031 came along until the 80s - 8048 perhaps ?
Intel 4004 was available. I lusted after one in our electronics lab but was informed by the guys there if I so much as touched it I would lose my thumbs. Things worked out for me though: Bought an 1802 Super Elf and was happy with it.
The first PICs were 1976, but they were mask programmed, except for expensive development and field test versions which had ROMless microcontrollers or ICEs allowing the use of the developer's own PROMs. Dev cycle was ICE then Field test with ROMless PIC plus PROM, then get GI to run an initial mask version for validation, then production. EPROM-based microcontrollers, let alone flash, were still some way off still. I certainly don't miss these aspects of the development cycle! One bonus though, no firmware updates!
Other than the Zilog Z80... which started rolling out to market around the same time... not much else in a reasonable price range. Definitely a collector item... the early 6502 bug made early adoption difficult versus the easier to use and stable Z80. The combination of the MOS 6502 and the Zilog Z80 hitting the market at around the same time was the cost catalyst for hobbyist to take their babies to commercial viability. I myself started with the Z80... but I knew about the 6502. This is the first one I have actually seen covered in a video... admittedly I never looked before. Thank you for sharing your find with us Adrian.
Great story, enjoyed the video. Interestingly, Jan Beta recently posted a video where he was refurbing an Epson RX-80 that had installed an aftermarket interface card which also had a 6502. My guess this box might have been a project from a magazine like Popular Electronics in the 70's. It would be an interesting find if true. What made this video so great is it shows your genuine love and passion for these computers, and experiencing it along with you.
@@DaftdogUK Wow thanks, awesome great find, I'll page through it. These type magazine project box solutions were the essence of computing in the 70's, that gave birth to computer companies like Apple, where enthusiast transformed to full blown companies with the right idea and funding. Modern day Raspberry Pi and Arduinos projects are the great grand-children of this stuff and great to see flourishing today.
Those 1702 Eproms are "fun" to program.. you need a programming voltage of -48V. On ALL pins, so, the logic levels are -48V, too. Except the Vgg pin, that needs -35V. You then have to apply +12V to the Vbb pin, first set the adress pins to the inverse of the address you want to program, then invert them back to the "real" adress, then give this programming pulse for 1ms and go the next adress. You then have to program the 256 bytes 32 times. Programming 32ms at once destroys the memory cells. Still beats mask rom though ..
I had a piece of equipment with sixteen 2708s (and a ceramic 8080A from 1977) that I finally got to read out this week. The 2708 "only" needs +12 and -5 volts to read it. But I had already fried a "test" 2708 (with known data) trying to read it with a bodge board on an EPROM programmer, so I was definitely worried. I built my own reader from scratch with a "blue pill" board to run it. In the meantime I learned that if the -5 isn't there, the chip could quickly fry itself, so I put switches for the +5 and +12. I had also heard that too high of a +12 could be bad, so I added a diode drop to make sure that the +12 was less than 12 volts. There's a very good reason why you rarely see EPROM programmers that support anything smaller than the 2716, even ones from the 90s. There are just too many weird voltages to deal with.
I was in second grade in 75. You know it is astounding what happened in such a short time. To think we would have an affordable home computer in the early 80s that we could rally around as youngsters. Thank you. That was fun!
Thank you for letting us visit your basement Adrian and letting us experience this very special occasion with you. Honestly, I can't tell you how much it means to us that we get to feel your emotion with you.
I love wire wrapped circuits too. They look like mini telephonic centrals. I knew about the "secret" instructions that 6502 have depending of the manufacturer (probably more a side effect of the way instructions were decoded). But I never heard about that. Very cool.
great video Adrian, nostalgic for me as well. the 6502 was the first uP I learned to program while working as a college intern. I recognized the AY-5-1013 straightaway as a UART as you were talking, so I had a bit of a smile about that. it resembles a project I worked on very similar to the serial-to-parallel converter you described, it was for a selectric-style terminal (our school was an IBM shop) with a proprietary parallel interface. I wonder if that's what this box was for, some non-Centronics hardcopy device? thanks for the walk down memory lane!
Yes, I remember the AY-5-1013 (and the AY-3-1015) as well! It was a UART that you could use in circuits without a processor. All settings were done by jumpering pins, and one could easily clock a byte in or out of the chip.
OMG, THIS IS SO COOL! I had to watch it again. Way to go Frank! I was getting a little teary eyed. Oh the memories that this video unearthed in my mind. Thanks for sharing this Adrian, you're a good man.
Thought for sure you were going to burst in to tears man!. I can hardly understand most of what you were talking about but I was excited right along with you. It's good to find and appreciate rare things. Others may not see the value that you do. That's okay. Enjoy telling this story many times.
I took a digital design fundamentals class here in college and im so glad i did. I have such an appreciation for these computers and components and what they can do. I’m 29 and I live in a time that we still have these jewels around and we can test them and appreciate what they are. I really want to take more classes in computer science.
Glad to see that mobo being useful! Be careful who you show that 6502 to when visitors are in your home - that chip is worth a small fortune on ebay. I had the pleasure, if that's the right word, of unwirewrapping a Singer computer (the owner had scavenged all the RAM off it for his own S100 machine) back in the early 1980s. While I appreciate the historical significance of wirewrapping, think Apollo Guidance Computer, it's frankly a PITA to deal with. What a wonderful find!
This is an unbelivably cool find. SO glad you took the time to check inside the box, who would suspect such a treasure would be inside? The 6502 is precious to me for the exact same reasons as you (though my first serious love was with the 6510 variant in the C64) and I almost nerd-gasmed when I saw what was inside that box and spotted the 75 year code. 🤤 That makes this CPU only 3 years younger then myself, and that's kind of terrifying and amazing at the same time. GREAT score, and take good care of that old ancestor of our childhood dreams!
The 1st revision of the 6502 doesn't have a bug, your tittle is wrong; as Chuck said "it worked the first time" the ROR function was never implemented by design. As you said ROR was never written in the instruction set..
Oh my gosh, I recognized that project box, and in the shots showing the catalog number brought up so many memories of helping my Dad do inventory at the Radio Shack store he mananged in the late 70's, early 80's. I saw the 276 prefix and new exactly where in the store on Florence Blvd, Florence, AL I would go to grab that project board.
I just watched a video by TubeTime where he delved into this and found that it wasn't a bug but rather was intentionally just completely left off the chip to make it as simple as possible. So more of an oversight of how important that instruction could be. He was able to find one with the die exposed and confirm that the transistors for ROR just plain didn't even exist in the early version.
I watched the same video. A sad thing about the Internet is that once incorrect information is disseminated and accepted by the public, subsequent attempts to correct it don't get as much traction as the original misinformed story.
It may be due to other reasons (e.g. they had to release it by a certain date)
I saw the same video. The one where they even looked at the transistors of the chip, right?
Need to search for that video! The fact that was considered a bug could be because the chip did have a ROL instruction so is odd they missed the ROR but the fact that they came with a revised chip so soon after launch adding it does lead me to believe that was omitted on purpose to get to market early, in the 70’s you needed the product ready to present it in the key conferences to be able to market it properly
And yet, compilers accepted the instruction and the CPU does something with it, even though it is incorrect.
I love when Adrian is at a loss for words. I'm so thankful that he takes the time to make these videos. It's more than simple nostalgia. Many of us grew up loving these 8-bit systems too and the memories are reminders of moments of joy and the learning that came from experimenting.
He's not just preserving history but he's sharing that love and joy with the world.
I can't think of a better way to spend your time.
Bruh that's still nostalgia, but nonetheless it's always magical X3
It Freaking Works is fantastic !!!!! I hope Adrian knows how much joy he brings to us !!!!!!! Thank you Adrian !!!!!!!
Well said, moments like these are why this is my favorite tech channel. Adrian isn’t some noisy RUclips personality churning out low effort videos, he genuinely loves this stuff and it comes through in his work. It’s so nice of him to share it with us. Adrian if you read comments (so many so probably not) I have a bugged Pentium 60 in a bin somewhere if you want more buggy silicon
I have a white 1976 cpu in the basement along with the Jameco receipt for $75 for it and the 6522 pia I bought in high school my sophomore year. I lived in PA not far from MOS fab. I called and an engineer sent me programming manuals and data sheets. Wire wrapped it up and had a tic tac toe program working on it using an old point of sale terminal. Later ran tiny basic on it in college. Fond memories.. I will probably dig it out when I retire in 3 years. But your video made me remember the gogo days of home computers
I work in E-Waste ... I see things like this all the time and have to take them apart to sort the parts, that board would have been thrown right in the low grade box. I try and keep an eye out for special things, but I am sure I don't always recognize that it's special. I am able to save a few things but I can't save it all. I rationalize a lot of it by saying to myself if we didn't recycle it nothing would be rare. :)
It'd be awesome to see what you've collected... even if it's just a pan over a bunch of stuff on the floor! Hope you don't mind my subscribing in hopes one day you do.
@@ericwazhung ruclips.net/video/BGnNlVzUCZM/видео.html This is a video of some things I saved and sent to Adrian.
@@Neodra Nice finds! Thanks for sharing with him to share with us.
I've been thinking lately that at this point, a lot of e-waste places probably purposely employ people who are at least aware of some of the obviously-valuable stuff so they can turn it over to the secondhand market for a lot more than they make just processing garbage.
@@stevethepocket It's not about individual pieces, it's about volume. The place I work is an intake facility. We buy product from the public and some commercial and basically sort and package it. We sell it to other recyclers who break it down into useful metals who in turn sell it to foundries. Trying to sell individual items on the second hand market is way to slow. So if I don't save it or someone down the line doesn't save it, it will be destroyed. I do know that the place we sell PCBs has a machine that photographs them and then points out any IC's that can be sold intact. They will recover those IC's and then the rest of the PCB and components get ground up, sent through several different washes, then melted down to recover the metals, specifically gold and platinum. ... Disclaimer : Other recycle companies may do things differently :)
Adrian literally teared up over a novelty microchip - with a fatal flaw no less. I love it!
An old novelty....
Fatal?
He's literally made of soy, it's a fucking chip
I'd hardly call the 6502 a novelty
@@crappyatlife I’m sure you feel superior after making that comment.
Your reaction was the nicest thing to watch. I really enjoy your channel. Keep up with the good work. Rest assured that it is really greatly appreciated.
Truly Adrian when Franks program (iz8dwf Hi Frank) popped up declaring this was in fact a bugged MOS 6502…Man I burst into tears…I’m surprised you contained your exposure. Simply wonderful…
Adrian, just found your channel and haven't looked at your other videos yet but I just had to comment. You have taken me back to 1975 where at that time we were programing industrial controllers with paper tape readers. When the 6502 came out a whole new digital age was launched in basements and garages, it was a magical time. Thank you for sharing this story for this old man.
Thanks for the mention! Wow. Now that is a very lucky find. I have one very early 6502, but it's soldered into my OSI 300 trainer. I don't know if I could devise a way to test for the ROR bug as I'm reluctant to remove it. Very cool, Adrian! There may not be real buried pirate treasure, but there is still lots of digital treasure buried away in devices like this waiting to be found!
Look on the top for a Date code: Do you have access to the assembly/monitor? Its only like 5 instructions to perform the test. The message is 3x the size of the test. Is your ROM soldered in?
@@joeturner7959 My 6502 is date code 3875. It is soldered in place.
@@TechTimeTraveller
So processor and ROM are both soldered. :(
@@joeturner7959 There's no ROM on the OSI 300, sorry.. just CPU, a 6810 RAM and some support logic. But yes.. everything is soldered.
You could use the switches on the trainer to enter a simple program: LDA #$22, ROR A, STA $8 (assembly is A9 22 6A 85 08; use Windows Calculator to translate to binary). You'd then read from $8 the result, which will be $11 (or maybe $91 if carry was set) if ROR works correctly, or it will be $44 if ROR behaves like ASL (the bug).
I'm glad that you didn't throw away because imagine the regrets afterwards and like Indiana Jones said "it belongs into Museum". Very impressive since I never seen this very early version of MOS 6502 before since only the later version of the Chip especially for the one of the most if not the most popular CPU of all time for 8 bit Computers. What a great find my friend.
like he said wouldn't have known to regret it. the amount of lost stuff in this world is kinda insane though, have to stop myself from thinking about it too much. entire civilisations have grown, thrived and then disappeared without a trace remaining. especially wilful destruction of information is infuriating but it no doubt pales in comparison to the amount of ignorance, incompetence and accidents. we will never know the richness of our lost history.
man I loved it when Indiana Jones appeared and said "Its d.i.p.ing time"
@@snooks5607 How would you now that, when: everybody in fact "wouldn't have known to regret it"? Sorry for that smarty-pants comment about the hole in your arguments logic:)
Of course I agree. It is probably a lot of technology that is unwittingly wasted. But there is definitely more that is deliberately thrown away. To preserve everything simply was and is not been possible with our actual consumer society since the 1950s. Otherwise we would just suffocate in trash. It is at least good if something is reused or recycled[1]. The majority of tech/electronic-waste and old equipment end up in countries like Africa ... so that they suffocate, in dimensions as large as entire states in garbage dumps. Unfortunately, this is the reality:/
[1] Like from us retro-freaks or repair enthusiasts/technicians. But by all love, this is just a drop in the sea ...
@@snooks5607 flashbacks to when i was at the library of alexandria when it burnt down. It was so sad watching all that knowledge perish
Thank you to you and Frank IZ8DWF for the video and code. I had two white ceramic 6502s that I wasn't sure if they had the ROR bug. One did and one didn't and without that code (or writing something in machine language) I would not have known if they were early enough. Your reaction is quite fitting as they are truly historic and a joy to be the caretakers of them.
AY-5-1013 is a UART from General Instruments, which eventually became Microchip, famed for their PIC line on microcontrollers. I used these UARTs extensively in the 70s, they were ubiquitous in those days.
Funny, when I saw AY and GI, I thought sound chip... 😅
you beat me to it, i vaguely recognised the number and googled it, i may have one or 2 lurking around, definitely havent got any ceramic 6502s, i did find a ceramic 8008 on a board....
Came here to post the info on the UART. Good info, @Nezbrun
That chip is also a piece of history itself
The Taiwan silkscreen shows that the chip might be the first ever batch of chips that Taiwan manufactured in history, cause the semiconductor division of ITRI(which later dissociate from government and became UMC/Mediatek until now) finished their first batch of chips around the end of 1977
I pulled out my 1978 GI catalog to look it up.😊
I loved this @adrian! I was studying electronics in 1974-1977, and then did a PhD in computer design.
My one quibble with the video is that the 6502 is NOT a simplified 6800.
Yes most of the original design team came from motorola, and yes the 6501 was pin compatible to the 6800.
However it's architecture is very different to the 6800 and it has a fraction of the instructions. For example it has 2 index registers, and far fewer instructions.
Chuck wanted to build a better chip for control, because he found that people just balked at the $300+ price of the 6800!
Chuck Peddle's vision was for a cheaper chip to do simple interface and washing machine, etc. control. In other words using the microprocessor as replacement logic.
I think he would have loved this application of the chip!
And I love your sheer joy and excitement of having it work! It made my day!😊
I agree. Sometimes there was an old way of doing something w/o a microprocessor, but now always. I used mainly a Z-80 in that time period but I remember looking at the 6502 instruction set, etc., and I remember how relatively bare it was. Sometime later than that my most hated micro was a PIC of some kind where instead of conditionally jumping to someplace you had to write the jump instruction and then branch over it if you Didn't want to jump. Arrrg!
This is so cool. I appreciate how you were able to convey the importance of the discovery to people less well steeped in 6502 lore. That is a really awesome discovery, and even more awesome that it runs (aside from the bug) correctly. Thanks so much for sharing.
I recognized the 1702 EPROM right away, I was in that electronics hobbyist era in the early to mid '70s, went to tech school, and it became my career. Working at different jobs, seeing so many esoteric parts that are no longer made and have no equivalent. So glad you were able to save that little bit of history.
A first revision MOS 6502 from the first year of production. You're not kidding, binning that would've been a tragedy, that thing belongs in a museum display!
I remember the times when you could go down to Radio Shack and get raw boards and etching to make your own circuit boards. Those were the days, really Miss the old Shack :(
@BumbleBee Luckily we still have one called Jaycar all around Australia and a few other individual ones scattered about.
@BumbleBee Maplin was already pretty useless by the late 00s sadly. A few connectors and plugs but not much else to my recollection. Or maybe I just had a rubbish tiny one near me.
I remember when you could go to Radio Shack store and buy the parts to make a radio. I knew the end was coming the day I went to Radio Shack and they were more concerned about who my cell phone carrier was than whether or not they had a germanium diode for my kid's project.
I spent hours in Rat Shack as a kid. They had all the common stuff in stock and could order in just about anything else. The employees were quite knowledgeable. There is no more customer service in the US.
The 20th Century was a great time for electronic hobbyists. From shortwave radios to TVs and the first computers, everything was so amazing.
awesome for you Adrian. I could feel the nostalgic emotion. Love it
I have a Pentium (1) with the FDIV bug! It was removed from an IBM 9595 server while I worked at IBM in the 90s. If you want it I can send it to you Adrian although it will take some time since I live in Buenos Aires. Keep up the great work!
Do you know if it also has the meltdown bug. When meltdown was published, the CPU designers were awfully silent about the status of older chips.
The official Meltdown test programs tend to use non-buggy unrelated features of newer chips, but it's basically a question if a specific documented security check is done before hitting cache and external bus or not.
@@johndododoe1411 Meltdown would date back to the Pentium Pro and the P6 architecture. So the Pentiums would be unaffected. Incidentally, I also have a FDIV bugged Pentium 90.
@@steeviebops Are you sure the Pentium doesn't let the Meltdown affected operation hit the cache, thereby leaving a side effect of the supposedly hidden value?
I have a Pentium 60 with this same problem. My first ever computer. I kept the CPU
Who cares about that piece of junk?
Adrian, I'm so glad that I stumbled into your channel so many years ago, I think from an LGR recommendation, when I really just started on modern PC gaming channels and slowly migrated to this time period. At first I thought it was a bit too technical, despite being an electrical and computer engineer, I never went so far as to have the practical and working knowledge you have and went into the business side. Now every week I'm eagerly waiting for your next video to drop, whether it's this channel or your 2nd channel. It's so hard to describe how over time, you became my all time favorite RUclipsr, but I think the reaction in this video encapsulates it.
We all live vicariously through you. We are in awe about how you self taught your way to being one of the most knowledgeable early computer repair people on the planet, and yet still be humble enough to reach out for help, while always letting your true passion and excitement come through in your videos. There are some days I see when you drop a 45 or 55 minute video, and I get giddy thinking I need to save this for a long relaxing viewing session before I go off to sleep.
I guess what I'm trying to say is just... thank you. Thank you for reigniting my love of this hardware which originally pushed me to get the degrees I did all those years ago, and remember why I did. Everytime I learn something new from you every episode, and it's got me into tinkering again. I feel like I'm rambling, so I'll stop.
Never change, because whatever it is you are doing... it freaking works.
So well said. Adrian is one of a short list of people who represent to me the best of the internet - passionate people who unapologetically and humbly share that passion with the world. It's a great little community we have here in retrocomp land.
Aside: sigh, anyone here watched Aussie50? RIP Ed. He was the one who showed me it's OK to be passionate about tech. I think he was pivotal in really kicking off the amateur tech community. I consider channels such as these to be an offshoot of that.
You are awesome! Thanks for the memories and the new answer to the question of how many lefts make a right.
Damn dude, whoever built that box really knew their stuff.
Actually back in the day the ability to write simple machine language programs and design simple circuits is considered basic skills, like the ability to write HTML these days.
@@18000rpm As a 90s kid, I feel like I missed out on a golden age.
The workmanship was awesome.
I was just grinning ear to ear !!
The code and project just for an interface.
This guy was on the top of his game. Reminds me of the "City on the Edge of Forever": ...stone tools and bear skins.
Radio Shack really deserves more credit than they are getting. We took them for granted. They even took themselves for granted.
@@18000rpm Or, they could've simply been an electrical engineer or computer engineer who worked in the industry and had access to prototypes and other parts like this that were quickly discarded as updated versions came out.
So glad for you Adrian. Your happiness is contagious and we all feel like we have touched this magnificent piece of history. Hope you located it in a privileged place on your bench, so we can admire it in future videos.
I can hear the utter awe in your voice and I am absolutely living vicariously through this video. ❤
The 6502 was also in the British Broadcasting Corporation home computer - the BBC Micro 32K, a machine that most people totally ignore. There was an American version, the ABC which was virtually identical and sold in the States, but as I said, it virtually unknown today. The BBC was proceeded by the Acorn Atom, the Acorn Electron. The BBC Micro came in a 16K version - the BBC Model A, the Model B which was either 16K, 32K, or 64K and some rare modified ones had 128K. Later it was surpassed by the BBC Master 128K and the BBC Master Compact. These machines could include a 512K add-on board that ran a very early version of DOS, otherwise the machines all had tape and floppy disk access, as well as the first Winchester hard drive, and Laserdisc system. The BBC Micro also supported a local network, TELETEXT, PRESTEL, telephone modem access, a 6502 co-processor and many other adaptions, including a MIDI system.
I had a friend who worked for a software company, Acorn Computers Limited in Britain, and she would send me sales brochures. Eventually this machine was supplanted by the Archimedes, the world's first RISC system, which was in effect an early PC clone. I believe there is an emulator for it that runs on PC called BeebEM (Beeb being the affectionate name given to the BBC machine by its users). Of note is the fact that the worlds very first 3D space combat and trading game, the legendary ELITE by Ian Bell and David Braben was developed on a BBC Micro in the eighties that was ported to virtually every other computer platform of the time and which still remains popular - in fact, the modern remake of the game is called ELITE Dangerous.
Wow! Treasured history in your hands and you almost threw it away! Glad your Spidey senses were tingling and you opened the case to check.
The early 6502 cpus did NOT have the silicon circuitry to support a ROR instruction. so it was NOT a bug but not even implemented. The ROR instruction was added later in the next revision of silicon. (It DID have a shift right but not the rotate right). the reason that the code almost works is that the motorola style processors such as 6502 and 68xx were random logic decoded so as to operate much closer to the clock frequency. the Intel style (8080 and Z80) used microcode which take a lot more clock cycles per operation (often 10 or 12x instead of 2x or so). If you take a 6502 or 6800 processor and list all of the operations in a table you will see patterns and missing opcodes for stuff like a STA immediate (which doesnt make much sense in real code)
There have been a lot of recent comments about the missing ROR being "not a bug." I feel this is semantics, pure and simple. The feature is not fully implemented on the first revision of the chip, as pointed out by TubeTimeUS in his recent video, but there is plenty of evidence it was the intention of the designers to have it there from the start.
As pointed out by TubeTimeUS, there was just no space for the transistors to make it work properly. It's like they ran out of time/space or made a mistake implementing it, not realizing they would need some more room for the extra logic. So when they decided to release the CPU without ROR, they just removed it from the August 1975 datasheet. (That last part entirely speculation.)
The issue is, there is just a lot of evidence it was supposed to have been there from the start:
* Robin has a great tweet showing the decode table for the 6502 with ROR. Notice how it fits in so perfectly?
twitter.com/8BitShowAndTell/status/1629754457172111361?s=20
* The Jan 1976 hardware manual for the 6502 lists ROR and says it will be working after June 1976. That is so soon after the chip's release to already be saying it will be implemented.
* There is some wiring for the instruction that doesn't go anywhere, but there was some work done to implement something. Why wouldn't this have been ROR but they couldn't fit it, so they just gave up on ROR for that first revision?
* It seems unlikely to implement ROL but not ROR - -especially since the 6800 (the "father" of the 6502) had both.
The real discussion to have is this: Is an unimplemented feature that was supposed to be there a bug or not? There is an argument to be had either way and both are valid points. For me personally, I look at it as a bug.
Good point about the 6800 having both ROR and ROL!
I love that this was saved and it works and looks perfect. Not only did you save an extremely rare chip, but you know what it is. This one couldn't have went to a better person. You'll treat it like the work of art it is.
I think very early on MOS Tech was selling a boxed set of CPU manuals and 2 of those 6502s at shows. I had one as a hand me down and passed it along to a 3rd party years later. Yep, two ceramic 6502s from that year.
This is why I love watching this channel, your enthusiasm is always sure to bring a smile to my face. What a great story of the buggy chip. A survivor indeed!
"So without further ado, let's get right to it!"
*Plays opening titles delaying getting right to it, adding further ado*
Lol irony
That is a fantastic find. Keep preserving our digital history. You rock!
In my years of experience hand building boards, I came to loath stamped pin sockets and wire wrap. I did build hundreds of boards with that method and sockets. Virtually all of the builds had all manner of weird problems traceable to faulty connections. Machined pin sockets are much pricier than the stamped pin variety. BUT, they pay for themselves by eliminating many hours of wasted time spent troubleshooting an unnecessary problem.
The one pin you found replaced on the socket was, most likely, because it was intermittent or outright open. ALL stamped pin sockets are a GUARANTEED built in failure point. Wire wrap also corrodes severely in a short time and CAN become intermittent as well. Depends on the alloy of the pins and wire combinations. Galvanic corrosion results with mismatched materials. FYI, wire wrap was developed for computers in missiles and rockets due to lacking a stress point where the wire exits the solder on a normal connection. This also saved weight by eliminating all that lead solder in the rocket. Missiles and rockets usually had very high QC to bypass the initial bad connections and corrosion was not an issue as these things didn't sit around for very long before being used. Once.
I'm so glad you did take the time to look in that box. Not only because of the chip -- which IS super awesome -- but because it obviously has so much meaning for you. Well done on saving such an interesting piece of history!
23:36: If you were to test the Apple II + 1975 6502 more extensively, maybe on your second channel, I'd be happy to watch that. I'm now wondering if Apple accounted for the ROR bug and wrote their Apple II firmware so as to test for the bug and work around it if present. Of course, it's a safe bet most Apple II software did not do the same, but I'd love to know if Apple's early Apple IIs were "ROR bug-tolerant".
In order to test for ROR, you have to run the ROR, so it WILL show up in the code.
It wouldn't surprise me if WOZ programmed around the ROR so they could save a buck using faulty parts - lol -.
@@SidneyCritic exactly my thought!
@@SidneyCritic ...Even though by that time, the faulty chips would likely have been out of stock and harder to get hold of than the good ones. Still wouldn't put it past him though.
@@SidneyCritic most probably there'd have been a version of the ROM code specifically for use with that chip if so.
A fairly simple solution would be to provide a ROM routine to perform ROR by executing ROL seven times instead but I don't know how that would affect general compatability and performance.
You couldn't really intercept the opcode though. You'd probably have to prepatch everything.
This is my favorite video of yours so far, and you have so many great ones. So cool and Kudos to Frank as well.
Amazing story. Your videos take me back to high school in 1980. Kids today have no idea how primitive and wonderful things were way back in the late 1970s.
I've really enjoyed this video Adrian! Thanks for sharing this fabulous CPU! Now you need to buy all old project boxes from the '70s and I'll pay for them :)
Nice job, OM !! 73!
I read many years ago that the low cost of the 6502 was one of the main reasons Jobs and Wozniak used it in their early Apples instead of the 8080. And thus the great divide was born... Apples vs PCs.
I still have my original Poly-88 micro based on a 2MHz NEC 8080 from 1977. The Poly was one of the few early micros that didn't require manual entry of a bootstrap program, as it had a 1k ROM for boot and I/O functions.
I have a soft spot in my heart for these old machines, IMSAI, Altair, SWTPC. I still have the advertisement cut sheets for them, as well as a number of late 70s micro periodicals: Byte, Dr. Dobbs, etc.
Great video!
I always liked the look of the old ceramic packages of early chips. This is definitely a super cool find. As a teenager I used to experiment with TTL chips on breadboards, but higher end chips like this were definitely out of my budget in the old days. I really want to get back into that again. Great video, and I appreciate the work and research done by you and others to preserve these pioneering pieces of history.
Ceramic chips are awesome, from stuff like this 6502 to early Athlon chips.
That's an absolutely amazing find! Most people don't really know the value of old equipment (especially when it still works) and it's always a bit sad knowing how much working legacy hardware lies buried in landfills.
I really enjoyed watching tis video, because you transport your excitement about the 6502 to your viewers. The 6502 in white color and golden contacts has really something magical and i’m sure that this chip has found its rightful owner now!
What a beauty of a chip!! CpuGalaxy would be really jealous of you for that white ceramic first cut 6502! Considering also that it is the first CPU for the masses it takes a very very special place in computer history and my heart also as a computer enthusiast with extremely low budget.
The other wow thing about that chip is that it is fully functional also with its cute bug...
My suggestion is to put it in a working pretty rare system that can use it and make a transparent case for the whole computer unit with this little white beauty displaying its body and also running the system in exhibitions or videos for people to see!
Greetings from Greece, you are a happier man after that and keep on doing your stuff for us to learn and enjoy!
Oh damn Adrian, what an outstanding find!! Now I am jealous, I am still missing this CPU in my collection 😮. Great that you didn’t throw the box away. And your emotions definitely swapped over to me here in Austria. We share the same passion. Cheers, Peter
Ohhhh! The wire wrapped back of the board!! Get a deep shadow box and maybe put the board in tilted forward with a mirror behind it and an angled mirror next to it. So you can see the back as well as the front.
That would look so cool!
Great find, and good presentation on the history. I never knew about that early bugged version.
A glass bell jar with a rotating base would be better.
The original monitor, at least, uses it only once: in the disassembler. FPBASIC uses it for the floating point code, ONERR and hi-res graphics.
Makes sense. Thanks for clarifying.
For those who don't understand, he is talking about where ROR instruction is used in Apple ROMs, and "monitor" is memory-editing software, not a display device.
47 years old and still works. Same here, just! Very interesting video, thanks.
It's interesting that the ROM boots, its likely that whatever operation uses the ROR function doesn't have its result verified, so the ROM is 'running' but at some point you're going to try to use something in memory that's been run through the buggy ROR routine and you'll receive unexpected results. Still very neat to see it preserved and the short history lesson! Thanks for sharing!
I always thought the gray ceramic DIP ICs were like little socketed pieces of jewelry, but the white one looks especially pretty. It's always a pleasure to open a machine and find a ceramic chip inside. #TotallyJealous
I bought my first Apple ][ in 1979 for $2250. It started me on a lifelong career as a programmer and Network Engineer. I have an original Gold Intel Pentium as a collectable.
Your enthusiasm over this made me smile :)
When you get the shadow box, maybe put a mirror on the inside surface so you can see some of the wire wrap
Excellent idea. Seconding this! Could incorporate the board's edge connector into the side of the shadow box to hold the board a little off of the mirror to make it easier to see the reflection.
@@spagamoto That's a brilliant idea
My heart stopped when you approached the ceramic with that weird screwdriver to lift it out from one side! Take care!
I've just started watching the video and looking at the box that Adrian nearly threw away. At first glance it looks like the box is "The Internet" as seen on the British Channel 4 comedy programme The IT Crowd 🌐 😂 Cool to see an early 6502
Exactly what I thought...
Beautiful video. I have known the assembler of the 6502 since it entered the business, and then moved on to the older brother: the 6510, of the Commodore 64.
I was amazed that the clock is generated by a 555.
Great job man !
That's insane!! Also interesting to see the AMD logo on the EPROM, from back in the early days when they were a general-purpose semiconductor company!
I'm so happy for you Adrian!!! You're never lost for words, this means a LOT!!!! Excited much!
It’s hard to stress enough how many electrical gizmos/adapters/connectors had that Archer logo on them in the 70s.
Don't those guys still make tv antennas?
@@samuelcolvin4994 Archer was an in-house RadioShack brand. I don’t think it exists anymore.
This is one of the most amazing videos I've seen for a long time, I'm afraid your title didn't really convey how special and I procrastinated over watching it (I'd assumed that the rare and special processor would be something later or less interesting to me.) I loved your excitement and emotion and was 100% with you on that. The 6502 is special to me too, for the same reasons as yourself, and I'm very glad to have seen this specific early example, learned a little about its history and seen it running.
What a great video! Crazy great find and I love the idea of making that a display piece. Would love to see that some day. Great job Adrian!
Great video Adrian. In 1975 I had a small electronics startup and had ordered some 8080s for some programmable boards we were doing for Lockheed. They were not yet available and the price was $360 each. Before delivery, Intel dropped the price to around $220. The chips were white and gold, just like your 6502. Thanks for bringing back some great memories.
Soviet calculators often used chains of shift registers as RAM.
How does this work? Like an electronic version of a delay line?
@@samuelcolvin4994 more like FIFO ram I’d guess
@@samuelcolvin4994 I think with some shift registers you can both write out to them as well as read their state back in again
@@samuelcolvin4994 Yes, a chain of shift registers (up to 2 K bytes!) with a continuously scrolling set of data. Other chips sit on the same bus and watch for particular values in this sequence (command codes), then take the data and put the results back.
@@watchmakerful Japanese calculators did this for a while too! In the 70s mostly.
You are lucky it worked as we were taught that technician handling the chip must ground his body by wearing a conductive wrist strap which is connected to a conductive mat which at ground potential.
Thank you for all your videos including bad and great decisions that we all learn from!
This whole video is amazing. I love hobby projects like this, I totally admire the old-school skills of drawing up a schematic and wire-wrapping a proto board (btw can high-res pics of the top/bottom be preserved as well?), the historical aspect is fascinating, the chip itself is beautiful, the custom ROM is genius, and, on top of all of that, "it freaking works!"
It's extremely well laid out and built for a hobby project. I'd love to know who built that. Someone really took pride in their work.
Thanks for such a feel good video. This rarely happens on tech channels.
Amazing find! So glad you didn't throw that beautiful 6502 out! :D I just came across an early 80s PET printer interface and it looks very similar to the self-made board you have there except for using "proper" RAM. And there is actually a (quite a bit newer) 6502 on there, too!
Hi Adrian, you tugged on my heart strings as well my friend. Not only is this my 1st micro (well second, if you count that the 1st was not completed and working yet too much), but my 1st running and working computer was the Kim-1. I worked for the National Science foundation here at a college in Massachusetts in 1975, was part of a small group that created an 8008 based computer to help teach for students on a project called CaLM. Computer and Laboratory Math. The 8008 was our design and needed so much more work, then we found out about the Kim-1 by mostek and put our old 8008 design on the shelf. We used the Kim1 for many presentations about computer programming for NSF and some classes until not to long after the project was canceled. The late 70s was filled with home computers in a way back then, at least for us in that project.
I got into the 6502 so much myself in that it directed my future in programming as well as hardware with that chip. I was involved with the 6502 with my career in a Game company, a computer publishing company, my own development business and so on. Your right buddy. That chip tugged on my heart I believe in a way as powerful as yours. It was the foundation that changed the world. Don't let it go. It's almost #1. Thanks for sharing my friend. Btw,, awesome find about ror. I've heard about it once, but never saw that issue. The Kim's we had was abit later than that. I think we had 02s in mid 76. The 8008 was in 75.
That would go well in my RevA Kim-1! It has the grey slate, but I've always wished it had the white ceramic. I also have a 6530 in white ceramic for it!
The AY5-1013 is a UART. - Jim Ferguson
Wow, not only do you have a very early version of the 6502, but you have a rare one with the bug. This is an incredible piece of history -- congratulations!
Umm, all the 6502s from 1975 have that bug.
@@Okurka. And how many of them are around?
@@_B_K_ Who knows, it could be thousands.
You made it sound like only some of the very early ones had the bug.
@@Okurka. Clearly not what I stated. Reading is hard.
@@_B_K_ Sorry that you have problems with reading what you wrote.
"not only do you have a very early version of the 6502, but you have a rare one with the bug"
You clearly say that only the very early ones with that bug are rare while all the very early ones had that bug.
Fantastic video. Great to see this olde 6502 in action! Great work Adrian and Frank.
Hey Adrian, that's an amazing find! The ceramic MCS6502s are highly sought after by Apple-1 replica builders for their original appearance. They are virtually unobtainium these days and sell for upward of $3000 on eBay.
Do they sell for upward to $3000 on eBay or are they asking upward to $3000 on eBay?
@@Okurka. Excellent question!
@@Sloxx701 That's one of the earliest possible, you can't compare it with this one.
@@Okurka. That's probably the ultimate question. There is one that did sell on October 26th for $1500 bearing a 3975 date code though, but that may be a unicorn.
@@rjolly87 Sold mine direct to a collector for $1500 (don't like dealing with ebay) ..I had bought it in '75 for $20, new for a project that never happened..
That processor needs to be in a museum! I'm sure we're _all_ glad you didn't throw it out!
Checked ebay for similar 6502's, and found several. Found one that sold recently for $1500. There are some still listed for higher prices than that.
4:46 That . . . is a ceramic MOS 6502 if I'm not mistaken. While I am interested in older computers I'm not that well versed, but if I recall correctly just the fact that it's ceramic rather than plastic makes it 1.) rare, 2.) old.
5:16 1975? Wasn't the 6502 introduced in '73 or something? Hot damn.
5:46 Ah, it was '75, not '73. And yeah, the 6502 WAS the home computer revolution. It dropped the price from, IIRC, something like $300 for a processor to $30. That's not adjusted for inflation, so that was still expensive but affordable for a home user.
7:00 LowSpecGamer has a pretty good video on the history of the 6502 too. Ben Eater's series on building a 6502 computer is great if you want to dig into the nitty gritty of how it works from the ground up.
13:05 I looked up the datasheet for the top left chip and it's a UART, so yeah, I/O chip.
13:58 Not to shill, but Ben Eater has an EEPROM programmer (TL866 II Plus) for sale on his website for like $70.
Fantastic video, thanks for making it!
Wow, what a great find of this piece of history! And it couldn't have gone to a more deserving person. You have done so much (and continue doing so) for the retro community as well as myself. Congrats and thanks for sharing this moment with us.
absolutly fantastic
Thanks for taking the time to make the video and show this fantastic microchip......
How cool is that, Adrian?! You have a piece of history in your hands! What would be nice would be to get some good closeups of the board and the wiring and maybe some of us can try and reverse engineer the schematic for the board and the box so the rest of the story isn’t lost! Excellent!!!
Amazing find! I got goosebumps when you popped that cover off. The 6502 and it's variants powered almost all of my childhood and early teen years. So cool this one ended up in very deserving hands! You'll have to show us your shadowbox when you have it done!
Looking at ebay solds, a 3975 dated 6502 sold Buy it now for $1500. A 1976 6502A sold for $1450.
Sept 30 is my birthday. I'm definitely older than this chip. But seeing it run is a nice present. Thank you.
I'm just now getting into the hardware side of things after a decades long programming career and really geeked out about this. It's sad that this ended up in e-waste and awesome that you caught it! Looking forward to your next discovery!
It’s Logan, just seeing this haha.
I actually smiled during the reveal. I'm not usually that enthusiastic. I'm an old 6502 guy, greatly appreciate this vid.
What other cheaper microcontrollers were available (in small quantites) in the mid-late 70s? I don't think the 8031 came along until the 80s - 8048 perhaps ?
Intel 4004 was available. I lusted after one in our electronics lab but was informed by the guys there if I so much as touched it I would lose my thumbs. Things worked out for me though: Bought an 1802 Super Elf and was happy with it.
The first PICs were 1976, but they were mask programmed, except for expensive development and field test versions which had ROMless microcontrollers or ICEs allowing the use of the developer's own PROMs.
Dev cycle was ICE then Field test with ROMless PIC plus PROM, then get GI to run an initial mask version for validation, then production.
EPROM-based microcontrollers, let alone flash, were still some way off still.
I certainly don't miss these aspects of the development cycle! One bonus though, no firmware updates!
Other than the Zilog Z80... which started rolling out to market around the same time... not much else in a reasonable price range.
Definitely a collector item... the early 6502 bug made early adoption difficult versus the easier to use and stable Z80.
The combination of the MOS 6502 and the Zilog Z80 hitting the market at around the same time was the cost catalyst for hobbyist to take their babies to commercial viability.
I myself started with the Z80... but I knew about the 6502. This is the first one I have actually seen covered in a video... admittedly I never looked before.
Thank you for sharing your find with us Adrian.
Great story, enjoyed the video. Interestingly, Jan Beta recently posted a video where he was refurbing an Epson RX-80 that had installed an aftermarket interface card which also had a 6502. My guess this box might have been a project from a magazine like Popular Electronics in the 70's. It would be an interesting find if true. What made this video so great is it shows your genuine love and passion for these computers, and experiencing it along with you.
It's a PET to ASCII/Baudot interface. This interface is in Kilobaud_Microcomputing_1979_September.pdf (which you can Google), p100-102.
@@DaftdogUK Wow thanks, awesome great find, I'll page through it. These type magazine project box solutions were the essence of computing in the 70's, that gave birth to computer companies like Apple, where enthusiast transformed to full blown companies with the right idea and funding. Modern day Raspberry Pi and Arduinos projects are the great grand-children of this stuff and great to see flourishing today.
Those 1702 Eproms are "fun" to program.. you need a programming voltage of -48V. On ALL pins, so, the logic levels are -48V, too. Except the Vgg pin, that needs -35V.
You then have to apply +12V to the Vbb pin, first set the adress pins to the inverse of the address you want to program, then invert them back to the "real" adress, then give this programming pulse for 1ms and go the next adress. You then have to program the 256 bytes 32 times. Programming 32ms at once destroys the memory cells.
Still beats mask rom though ..
Cripes!
I had a piece of equipment with sixteen 2708s (and a ceramic 8080A from 1977) that I finally got to read out this week. The 2708 "only" needs +12 and -5 volts to read it. But I had already fried a "test" 2708 (with known data) trying to read it with a bodge board on an EPROM programmer, so I was definitely worried.
I built my own reader from scratch with a "blue pill" board to run it. In the meantime I learned that if the -5 isn't there, the chip could quickly fry itself, so I put switches for the +5 and +12. I had also heard that too high of a +12 could be bad, so I added a diode drop to make sure that the +12 was less than 12 volts.
There's a very good reason why you rarely see EPROM programmers that support anything smaller than the 2716, even ones from the 90s. There are just too many weird voltages to deal with.
Holy Moly! I'm not sure if this helps my theory it might've been programmed via toggle switches.
@@ericwazhung I mean, you probably could with the assistance of a timer circuit for the actual programming pulse. It would be pretty hellish though.
I was in second grade in 75. You know it is astounding what happened in such a short time. To think we would have an affordable home computer in the early 80s that we could rally around as youngsters. Thank you. That was fun!
Thank you for letting us visit your basement Adrian and letting us experience this very special occasion with you. Honestly, I can't tell you how much it means to us that we get to feel your emotion with you.
I love wire wrapped circuits too. They look like mini telephonic centrals. I knew about the "secret" instructions that 6502 have depending of the manufacturer (probably more a side effect of the way instructions were decoded). But I never heard about that. Very cool.
great video Adrian, nostalgic for me as well. the 6502 was the first uP I learned to program while working as a college intern. I recognized the AY-5-1013 straightaway as a UART as you were talking, so I had a bit of a smile about that. it resembles a project I worked on very similar to the serial-to-parallel converter you described, it was for a selectric-style terminal (our school was an IBM shop) with a proprietary parallel interface. I wonder if that's what this box was for, some non-Centronics hardcopy device?
thanks for the walk down memory lane!
Yes, I remember the AY-5-1013 (and the AY-3-1015) as well!
It was a UART that you could use in circuits without a processor. All settings were done by jumpering pins, and one could easily clock a byte in or out of the chip.
OMG, THIS IS SO COOL! I had to watch it again. Way to go Frank! I was getting a little teary eyed. Oh the memories that this video unearthed in my mind. Thanks for sharing this Adrian, you're a good man.
It's wonderful to witness your joy here Adrian!
Thought for sure you were going to burst in to tears man!. I can hardly understand most of what you were talking about but I was excited right along with you. It's good to find and appreciate rare things. Others may not see the value that you do. That's okay. Enjoy telling this story many times.
I like how you got all sentimental at the end, looked like you almost teared up :)
I took a digital design fundamentals class here in college and im so glad i did. I have such an appreciation for these computers and components and what they can do. I’m 29 and I live in a time that we still have these jewels around and we can test them and appreciate what they are. I really want to take more classes in computer science.
Glad to see that mobo being useful! Be careful who you show that 6502 to when visitors are in your home - that chip is worth a small fortune on ebay. I had the pleasure, if that's the right word, of unwirewrapping a Singer computer (the owner had scavenged all the RAM off it for his own S100 machine) back in the early 1980s. While I appreciate the historical significance of wirewrapping, think Apollo Guidance Computer, it's frankly a PITA to deal with. What a wonderful find!
Pure Gold! Its a shame 99.998% of the planet has no clue how amazing and rare your find is.
That was emotional for all of us!❤
This is an unbelivably cool find. SO glad you took the time to check inside the box, who would suspect such a treasure would be inside? The 6502 is precious to me for the exact same reasons as you (though my first serious love was with the 6510 variant in the C64) and I almost nerd-gasmed when I saw what was inside that box and spotted the 75 year code. 🤤 That makes this CPU only 3 years younger then myself, and that's kind of terrifying and amazing at the same time. GREAT score, and take good care of that old ancestor of our childhood dreams!
The 1st revision of the 6502 doesn't have a bug, your tittle is wrong; as Chuck said "it worked the first time" the ROR function was never implemented by design. As you said ROR was never written in the instruction set..