6502 Hacking Hack Chat

• Hack Chat Transcript, Page 1

  04/03/2024 at 20:44 • 0 comments

  Anders Nielsen  3:00 PM Thanks Dan! Hello everyone! Super excited to answer any questions. Really been itching to jump in ahead of time By day, I'm the MacAdmin at Aalborg University, Denmark, handling about 1000 Macs. In my free time I like to play with old and new IC's and to get as much as possible out of as little as possible - most fun per transistor or most fun pr $ are favorites of mine. I've played with electronics since 2014.. or since early childhood if disassembling anything with a battery counts.. I made my first 6502 computer in 2021, followed it up with an entry for the 2022 Hackaday Prize, and was super happy the to see that go to the semifinals. Since then I've made the 65uino, which is a lot easier to manufacture and put together, but has it's own limitations - like the 8K address space and 128 bytes of RAM. Like an Atari 2600. And I almost have a PCB ready for a super super cheap (E(E))PROM programmer.

  Anders Nielsen  3:00 PM Oh - and I have a small Youtube channel where I try to explain most of the things I make, so it's easier for other people to replicate.

  Anders Nielsen  3:01 PM But ask away - I'm not an expert on any of it though.. Except my day job.

  Dan Maloney  3:02 PM Was the 6502 a nostalgia play for you? IOW, did the chip feature in any of the machines of your younger days?

  Anders Nielsen  3:03 PM Absolutely. When I was a kid in the 90s I had a C64 but I didn't have any of the knowledge to actually use it for anything but games.

  wheagy  3:03 PM Do you ultimately expect this to be a drop-in replacement in machines like the Apple II?

  Anders Nielsen  3:04 PM The Apple II should accept mostly any old NMOS 6502 I believe?

  Andy Geppert  3:06 PM The 65uino is awesome and very inspiring. Thanks for imagining and making it available! I think 65uino Mega is a good name for a future version which adds more memory, a full 6502 maybe some ADCs… ;) and a cartridge slot…

  Anders Nielsen  3:06 PM The waste stream I refer to is slightly abstract. Millions of Atari 2600's have already been recycled and I have a feeling there's still piles of boards in a container somewhere. If we buy those from the usual markets, we encourage the scrappers to save the chips instead of just melting them down.

  Anders Nielsen  3:06 PM @Andy Geppert Someone on Twit.. I mean X.. Suggested a 68Kuino :D

  Andy Geppert  3:07 PM Oh, and thanks for making the bootloader and dev work flow to go with it!

  Andy Geppert  3:07 PM Oh yes, 68Kuino needs to be done too!

  Dan Maloney  3:08 PM One vote for the 1802uino too. Probably the only one :-(

  Eric Hertz  3:08 PM What language do you most-use for 6502 programming?

  jdanen  3:09 PM @Eric Hertz - cool site, but there's an ocean in the way, shipping might get expensive :p @Anders Nielsen - ah, so basically picking the chips out by the invisible hand of the market, not a half bad idea

  Anders Nielsen  3:09 PM I try to explain most pieces of tech and code I use in my videos - but lemme tell you.. Editing video of 6502 assembly code and making it interesting is tough.

  Anders Nielsen  3:10 PM @Eric Hertz I'm slightly interested in that "phone" too :D

  Anders Nielsen  3:10 PM I mostly just go with 6502 assembly - cc65 does a decent job with C i head, but I haven't played with it much.

  Jacob Christ  3:11 PM It so easy to assemble by hand, why do you even use an assembler?

  Eric Hertz  3:12 PM LOL use toggle switches!

  Mitsuru Yamada  3:12 PM In my case, it is my homebrew floating point interpreter (7kB) or hand assembly.

  Anders Nielsen  3:12 PM @Jacob - almost getting there :) Actually want to build one like @Mitsuru's - but RAM only :D I have something to show you @Dan Maloney ...

  Andy Geppert  3:13 PM @Jacob Christ LOL. Reminds me of graph paper and a mechanical pencil to organize the bytes.. Ah, memories. I’m anxious to get back into 6502 assembly.

  Jacob Christ  3:14 PM @Mitsuru Yamada Nice builds!

  Mitsuru Yamada  3:14 PM I think 6502 is suitable for easy to creating machine language programs because of its small number of instructions.

  Xark  3:14 PM llvm-mos is an interesting C/C++ (and even Rust) compiler for 6602 also, but 6502 is a bit "resistant" to higher level code. 🙂 https://github.com/llvm-mos/llvm-mos-sdk

  Anders Nielsen  3:15 PM

  Mitsuru Yamada  3:16 PM I used to run a C compiler on APPLE2, but I had to change floppies many times to compile it.

  Anders Nielsen  3:16 PM @Dan Maloney I'm not sure if I'll get to it before or after or in parallel with the 8051 - but I'm getting there

  Dan Maloney  3:16 PM :heart emoji:

  Dan Maloney  3:18 PM I'd honestly love to build a replica of the COSMAC VIP I had way back when. No idea where it went. Must have cost my parents a fortune

  DrG  3:18 PM Anders / Mitsura - do either of you have any stories or experiences with the 6501 [six five zero one]?

  Anders Nielsen  3:18 PM Oh how I wish I had an Apple II to be frustrated with floppies about

  Anders Nielsen  3:19 PM @DrG As far as I know it barely existed. MOS created it as a gimmick to show it could be done cheaper but everyone was supposed to want the 6502 for the builtin oscillator driver

  Anders Nielsen  3:20 PM You'll notice the programming manual is actually called the MSC6500 Family Programming manual

  Xark  3:20 PM I read it never actually shipped.

  Xark  3:21 PM 6507! 😅

  Anders Nielsen  3:21 PM From what I've head they did bring it to WESCON but after that, not so much.

  CJ Keithley  3:21 PM I quite enjoyed the Rockwell R6501Q… quip (quad in-line package)

  Anders Nielsen  3:22 PM I wish the 6504 was easier to get my hands on. It should also be compatible with the #65uino if you have one of those.

  CJ Keithley  3:23 PM Rockwell also made a 68000 quip… they must have sold hundreds of those! :-)

  Xark  3:23 PM 6501Q....cool, I have not heard of those before.

  Mitsuru Yamada  3:24 PM In my experiments, the built-in oscillator did not have a good duty cycle.

  Pete Willard  3:24 PM Quad Inline is such an odd bird... and kind of a nightmare

  CJ Keithley  3:24 PM The R6501Q had (in a single chip) a 6502 variant, (a little bit of) RAM, parallel IO, and a serial port.

  Pete Willard  3:24 PM MC6800 also has some weird clock requirements, for comparison.

  Jacob Christ  3:26 PM So when I was learning 6502 assembly on an Apple II there was program that let you visualize an assembly program running that ran on an Apple II. It would show data moving from memory into registers and the registers feeding the ALU and to the accumulator then back to memory again. It was a great visualization tool. I'm wondering if others have seen it, knows what its called and know if there is a web page based equivalent that exists?

  CJ Keithley  3:26 PM https://web.archive.org/web/20221102041722if_/http://archive.6502.org/datasheets/rockwell_r6501q.pdf

  Anders Nielsen  3:26 PM @Mitsuru Yamada Exactly. It's there but everyone is better off generating the clock elsewhere.

  Andy Geppert  3:27 PM @Anders Nielsen What tricks have stood out to you as you use 6502 assembly to do fast pulses/timing stuff? I know some of them are in your video about driving serially addressable LEDs. Curious where you bump into timing limitations between 6502 and the modern area, and what tricks you use to bridge to more modern ‘tronics.

  Xark  3:27 PM @Jacob Christ Perhaps https://www.youtube.com/watch?v=aZvss4XnceU ?

  Anders Nielsen  3:27 PM @Pete Willard The big difference was the 6800 required you to input both phase clocks yourself

  DrG  3:28 PM

  DrG  3:28 PM My first computer - still have it and it still works (it is a rev b)

  Pete Willard  3:28 PM Yup, it did... and therefore... excess glue logic outside the CPU

  Xark  3:28 PM Nice!

  Anders Nielsen  3:29 PM @Andy Geppert I lean a lot on cool people like George Foot, Garth Wilson and good folks over at forum.6502.org

  Andy Geppert  3:29 PM @DrG That is SOOO on my shopping list.

  Jacob Christ  3:29 PM @Xark That's exactly the program!

  Andy Geppert  3:30 PM Oh, I recognize the Garth Wilson name - he has shared so many great resources.

  Pete Willard  3:30 PM Cool. Mine was an MEK6800... Still have it.

  Anders Nielsen  3:30 PM Timing can usually be fixed with a 74'74 and a gate if it's that bad - usually I try to avoid the edge cases. Driving ws2812's was one of those XD

  Anders Nielsen  3:30 PM @DrG Wow. Hang on to that one!

  Mitsuru Yamada  3:32 PM As for the difficulty with the 6800 clock, I solved it with the 6802, which can generate a stable clock using only an external crystal.

  Andy Geppert  3:33 PM I recently learned of, and purchased, and EMMA II which is a 6502 microprocessor trainer. Very similar to the KIM, but not a KIM. Anyone else here know about the EMMA II or have experience with one? I’m a 6502 enthusiast and I was surprised to learn about it recently. Apparently more popular in Europe, and I’m in the US.

  Anders Nielsen  3:34 PM @Mitsuru Yamada I see it was released in '76.. Wasn't that the reason MOS counter sued Motorola? :D

  Pete Willard  3:35 PM While the 6802 got a lot of things right, it still could not touch a 6502 in terms of price...

  Xark  3:35 PM I never heard of EMMA II, but pic from 6502.org (looks pretty nice). http://forum.6502.org/download/file.php?id=288&sid=30b5fe24dca784d18184aa56102bde21&mode=view

  Anders Nielsen  3:36 PM @Andy Geppert Hans Otten to the rescue: http://retro.hansotten.nl/6502-sbc/emma-by-l-j-technical-systems/

  Andy Geppert  3:36 PM It was sold with several mechatronic accessories and training manuals, originally, intended for tech/trade school use. From my understanding.

  Anders Nielsen  3:36 PM @Andy Geppert So someone made a remake?

  Andy Geppert  3:37 PM No, it’s an original. I purchased at a VCF. Costed much less than a KIM!

  DrG  3:37 PM

  DrG  3:37 PM +1 on the importance of 6502 price.

  Andy Geppert  3:38 PM Thanks for the Hans link.

  Anders Nielsen  3:38 PM We have way too few VCF's in Denmark :D

  Andy Geppert  3:41 PM Well, VCF is growing in the US right now and there are some people that would happy to help enable growth in Denmark with their experience! VCF does an amazing job of filling the niche.

• Hack Chat Transcript, Page 2

  04/03/2024 at 20:44 • 0 comments

  Anders Nielsen  3:42 PM Pot stirring time. The 65C02. I might sound like I have strong opinions about it sometimes... But.. I just feel it came around at a point where it was just "another micro".

  Anders Nielsen  3:43 PM Am I right the Rockwell C02's are pin compatible but the WDC's are not?

  Mitsuru Yamada  3:45 PM The 65C02 surprised me at the time with its low power consumption.

  J.B. Langston  3:45 PM yeah, the wdc is not pin compatible but can be made so with a few simple mods https://kerosmacmods.fc2.net/blog-entry-12.html

  Anders Nielsen  3:45 PM For my "bigger" #ABN6502 SBC R1, I had to add jumpers for the WDC but I believe a lot of relabelled ones that turned out to be C02's actually works without the jumpering.

  J.B. Langston  3:46 PM i have put one in my apple IIe that came with an original 6502 and it worked, though i later purchased an enhancement kit from reactive micro which came with a rockwell 65c and the updated roms

  Xark  3:47 PM From my understanding PIN 1 is slightly different in WDC 65C02 (it is an input and output so needs a resistor or similar). 🤔

  Anders Nielsen  3:47 PM @Mitsuru Yamada I have a USB cable that changes color with current - I spot a lot of "fake" 6502's from the unexpected power usage. Could also be bad silicon of course. It's fun when a 65C02 is labelled MOS6502A in super clear paint.

  Pete Willard  3:48 PM I've not imagined that there would be a benefit to updating the 6502 in my IIC

  Anders Nielsen  3:48 PM @Xark Pin 1 is GND so a resistor won't do :)

  J.B. Langston  3:48 PM some later software only worked with enhanced IIe... i think the IIc already came with the cmos chip

  Xark  3:48 PM Okay, never mind me, but some info here: https://www.westerndesigncenter.com/wdc/AN-002_W65C02S_Replacements.php

  J.B. Langston  3:49 PM yeah you have to clip pin 1 off and tie bus enable high

  Anders Nielsen  3:49 PM

  Anders Nielsen  3:49 PM Good chart!

  Pete Willard  3:49 PM BTW JB, thanks for the TMS9918 board for the RC2014...

  J.B. Langston  3:50 PM oh, my reputation preceeds me ;)

  Mitsuru Yamada  3:50 PM Regarding the subtle differences between the 6502 and 65C02, in my experiments, when I replaced the 6502 and 65C02 on my own computer board, the number of clocks from reset to the first instruction execution seemed to be different.

  DrG  3:51 PM Did all those C varieties maintain the bugs in the 6502?

  Xark  3:51 PM Yes, clock timing is indeed different on 65c02. It has an extra cycle on some instructions, e.g.

  J.B. Langston  3:51 PM no the C got rid of all the undocumented instructions and replaced them with NOPs

  Pete Willard  3:51 PM Geez. I never knew that

  J.B. Langston  3:52 PM also added a few other instructions

  Anders Nielsen  3:52 PM Or "new and interesting" instructions. Like STX, STY,

  J.B. Langston  3:52 PM there's a load zero iirc now

  Anders Nielsen  3:52 PM I mean PHX

  Xark  3:52 PM Yes, fixed (most of) the bugs. In fact one of the extra cycles added was to correct JMP ($xxFF) bug.

  Xark  3:53 PM Several very welcome new opcodes and some semi-useful new addressing modes.

  Xark  3:54 PM Most useful IMHO are as mentioned PHY/PLY/PHX/PLX, but also INC and DEC of A is handy, along with BIT #$XX and the new ($XX) addressing mode can sometimes be handy to not need to load Y with 0.

  Anders Nielsen  3:56 PM I always find it interesting to check the ones I get. You'd be surprised.. Well you might not.. Some might be surprised to find how many 6502's "fail" the BRA test.

  Xark  3:56 PM Nice rundown of the improvements (and examples of use): http://www.6502.org/tutorials/65c02opcodes.html

  J.B. Langston  3:56 PM i admit ignorance on most of the new opcodes... most of my interest in the 6502 is in the c64 demoscene so there's no C in that 6502 code :)

  J.B. Langston  3:57 PM and also, i've since gone to the dark side and i'm a much better Z80 programmer than I ever was at 6502

  Mitsuru Yamada  3:57 PM What I think is wrong with the 6502 in terms of readability is the lack of BRA, but with the 65C02, it is there.

  J.B. Langston  3:58 PM yeah people would just do clc, bcc, or if they were very clever they'd rely on a flag already being set or cleared by the previous instruction and use the appropriate branch when they wanted relative jumps

  Xark  3:58 PM @Mitsuru Yamada Yes, BRA is super welcome too.

  DrG  3:59 PM I personally remember stepping through code and being stymied by "An indirect JMP (xxFF) will fail because the MSB will be fetched from address xx00 instead of page xx+1. " It made my brain hurt. Then some *real* programmers told me, "oh yeah, that's a 6502 bug". ugh...good times

  Xark  4:00 PM I remember using it in "copy protection" (along with other tricks) to try to obfuscate the trail (but it was fairly well known).

  DrG  4:00 PM lol sweet

  Anders Nielsen  4:01 PM If I tell people to go out and buy a 65C02 and it turns out they get the "wrong" 65C02, things get frustrating fast. I have to admit I sometimes CLC, BCC but usually I have a flag available.

  Xark  4:03 PM WDC65C02 is nice because it is still made and available from suppliers (and ~14MHz).

  J.B. Langston  4:04 PM supermon64 used a very clever method to get relocatable code with regular jumps: https://github.com/jblang/supermon64

  J.B. Langston  4:04 PM discussed in the paragraphs right above the usage instructions

  J.B. Langston  4:05 PM there is a quote from jim butterfield explaining how it worked

  Anders Nielsen  4:05 PM There are plenty of cases where you need a factory new with warranty chip so I certainly appreciate it exists.

  asciimo  4:05 PM What's a reliable source for real 6502s?

  Anders Nielsen  4:05 PM Truth: Aliexpress

  DrG  4:06 PM @J.B. Langston I remember going through KIM cassette format and thinking - wow, neat. Then I read Butterfield's hypertape and went - WoW

  Xark  4:06 PM Apparently not too difficult to find pulls and such. Beware remarked parts...(e.g., 2+MHz may not really be...).

  Anders Nielsen  4:07 PM The 6507's I sell with the 65uino kits, I painfully have to test each and every one of them. And it seems they are rebadged to say whatever I order - but otherwise come from the same pile. Especially the 6532's.

  J.B. Langston  4:07 PM i have had decent luck buying used chips on ebay, but it's been a few years

  J.B. Langston  4:08 PM all of the TMS9918, SN76489, and YM2149 chips I bought turned out to be genuine. have never tried to get a 6502

  Anders Nielsen  4:08 PM Usually I can tell the actual manufacturer from the markings on the bottom.

  Dan Maloney  4:09 PM OK, sorry to break up the party but our time is up. I want to thank Anders for his time today and all of you for the excellent discussion. And everyone needs to start collecting those chips!

  asciimo  4:09 PM Awesome chat. Thanks!

  DrG  4:09 PM Thanks Anders and thanks Dan for hosting

  Xark  4:10 PM https://linuxjedi.co.uk/2023/12/01/fake-chips-i-got-stung/ might be useful

  Anders Nielsen  4:10 PM A lot of people assume that relabelled == fake, but that's obviously not the case. Worst case it's a different chip, but usually it's just to make the labelling uniform :D Oh - and at the end: Read ALL Don Lancaster's books. Thanks for having me!

  Xark  4:10 PM Thanks! 🙏

  Anders Nielsen  4:10 PM @Dan Maloney Thanks for the invite!

  Dan Maloney  4:10 PM My pleasure! Transcript coming up. Thanks all!

  Mitsuru Yamada  4:11 PM Thanks! from Mitsuru Yamada in Japan.

  DrG  4:14 PM I confess that chats like these make me wonder where all the excitement of the times went....It's weird being the same age as old people :)