2021 C64 Repair-a-thon #2: "Hmm, the mystery deepens!"
ฝัง
- เผยแพร่เมื่อ 6 ก.ค. 2024
- Welcome to my 2021 edition C64 repair-a-thon! In this video, I work on and fix machine #2 (of 5.)
At first I thought this would be a run-of-the-mill fix, but there turned out to be some tricky issues.
Also, I revisit machine #1 which had an additional fault that needed fixing.
Part 1 / Machine #1: • 2021 C64 Repair-a-thon...
Part 2 / Machine #2: This video
Part 3 / Machine #3 and 4: • 2021 C64 Repair-a-thon...
Part 4: Coming soon
0:00 Intro
0:46 Revisiting Machine #1 (And fixing it, again)
4:33 Machine #2 (all the way from West Germany)
36:02 Summary and outro
-- Video Links
Support the channel on Patreon:
/ adriansdigitalbasement
My video on damage-free chip removal:
• How to remove chips wi...
Jonard Tools EX-2 Chip Extractor:
amzn.to/2VazxDS
www.jonard.com/Products/EX-2-...
Wiha Chip Lifter:
amzn.to/3a9ftWw
www.wihatools.com/precision-c...
Link to desoldering video
Sven Petersen's Github (for the Diagnostic test harness files)
github.com/svenpetersen1965
C64 diagnostic ROMs (including Dead Test and regular diags)
blog.worldofjani.com/?p=164
8-bit Dancy Party:
csdb.dk/release/?id=200763
- Tools
Deoxit D5:
amzn.to/2VvOKy1
store.caig.com/s.nl/it.A/id.16...
O-Ring Pick Set: (I use these to lift chips off boards)
amzn.to/3a9x54J
Elenco Electronics LP-560 Logic Probe:
amzn.to/2VrT5lW
Hakko FR301 Desoldering Iron:
amzn.to/2ye6xC0
Rigol DS1054Z Four Channel Oscilloscope:
www.rigolna.com/products/digi...
Head Worn Magnifying Goggles / Dual Lens Flip-In Head Magnifier:
amzn.to/3adRbuy
TL866II Plus Chip Tester and EPROM programmer: (The MiniPro)
amzn.to/2wG4tlP
www.aliexpress.com/item/33000...
TS100 Soldering Iron:
amzn.to/2K36dJ5
www.ebay.com/itm/TS100-65W-MI...
EEVBlog 121GW Multimeter:
www.eevblog.com/product/121gw/
DSLogic Basic Logic Analyzer:
amzn.to/2RDSDQw
www.ebay.com/itm/USB-Logic-DS...
Magnetic Screw Holder:
amzn.to/3b8LOhG
www.harborfreight.com/4-inch-...
Universal ZIP sockets: (clones, used on my ZIF-64 test machine)
www.ebay.com/itm/14-16-18-20-...
RetroTink 2X Upconverter: (to hook up something like a C64 to HDMI)
www.retrotink.com/
Plato (Clone) Side Cutters: (order five)
www.ebay.com/itm/1-2-5-10PCS-...
Heat Sinks:
www.aliexpress.com/item/32537...
Little squeezy bottles: (available elsewhere too)
amzn.to/3b8LOOI
--- Links
My GitHub repository:
github.com/misterblack1?tab=r...
Commodore Computer Club / Vancouver, WA - Portland, OR - PDX Commodore Users Group
www.commodorecomputerclub.com/
--- Instructional videos
My video on damage-free chip removal:
• How to remove chips wi...
--- Music
Intro music and other tracks by:
Nathan Divino
@itsnathandivino - วิทยาศาสตร์และเทคโนโลยี
![ความรักไม่ได้น่ากลัวขนาดนั้น (TRUST ME) - LYKN [ OFFICIAL MV ]](http://i.ytimg.com/vi/9np-lAtCRpQ/mqdefault.jpg)
![Apple ][ plus: 8304 replacement from AliExpress, more keyboard fixes, and a huge surprise!](/img/n.gif)
In the end, I had a change of heart for those working MT RAM chips and I reinstalled them in the motherboard. At least they are socketed so it would be trivial to change a failed one in the future.
😎interesting
thanks for that ... Happy Easter from Croatia.
how can you write this a week before the video comes out ???
@@WilliamAndrewPhilipBodie Patreons have video 1 week in advance ...
That's the spirit :)
I was working on a software problem for my job tonight, and I hit a wall and gave up and came down to my couch. I started watching your video, and how you kept powering through this troublesome motherboard's issues... and was inspired to go back up and solve my problem. I did. Thanks Adrian, you saved my company's deployment.
You need to start a One leg in the grave bin for those chips that still have a pulse.
Yes; it can be quite useful to differentiate between "marginal" chips from "completely dead" ones. The CG ROM from the first one produced glitchy artifacts onscreen but otherwise seemed usable. I'm not sure what you would do with a glitchy component if you have a fully working alternative, but maybe one day down the road it could be the only one available...
A Schrödinger parts bin, for chips that are both dead and alive?
It would be interesting to see how a C64 works if it contains mostly or only marginal chips... :)
@@mrnmrn1 I'd love to see that
@@mrnmrn1 The G64, Glitch 64. The state of the machine will surprize you every time you turn it on! :D
Key phrase, "falsely accused chip." We must fight against chip injustice!
It is possible to determine bad RAM when you see different characters on the screen by looking up the character screen codes. In this case you could see $ in some places there should be spaces. The screen code for space is hex 20 and the screen code for $ is hex 24. So this indicates that bit 2 of the RAM is stuck high for some addresses.
Failed 6510 CPUs are very rarely seen in NTSC machines, and yet I've seen a handful of them on the imported PAL machines I've repaired over the years. I have a theory which could explain this; as over in PAL land the use of C2N cassette drives was far more widespread which connect through the CPU. Just like the famous theory of the exposed joystick port/joystick swapping being the cause of some failed CIA chips, its possible that over the years that something similar happens to PAL CPUs when cassette drives are plugged in/unplugged. Of course a cassette drive wouldn't be plugged in as much as joysticks, but its the only reason I can think of why PAL CPUs seem to have lead a harder life than our NTSC CPUs over here when they are otherwise the same. I've actually seen the bad zero page fault on a CPU before, as when they do fail (rarely) they tend to keep mostly working but glitch out on one specific test/function. Its strange.... -aphexteknol
Adrian, just thought i'd mention the weirdest CPU fault we had, it took ages to resolve, but we had a C64 which would never increment the score when playing "Paper Boy" no other faults and Diagnostics passed everything... eventually tried a new CPU and all was fine. So yep i've seen a weird CPU fault as well! Keep up the good work, Dave H UK
This channel is one of the few things keeping me sane recently. Only recently subscribed, so still got a huge amount of content to look back through.... Thank you!
Not sure if the human malware situation is getting to you or something else, but hang in there.
Looks at your username: are you sure you're sane?
@@YoureUsingWordsIncorrectly The '040 @ 40MHz upgrade was the best part so far.
Because of its clock rate, it beats my '040 Amiga/Mac Plus-based hackintosh that I posted a video about for #MARCHintosh. :)
@@YoureUsingWordsIncorrectly That was a great series. He just put a 68040 board in it. Thing is insane for an old Mac.
You should set up an online store and start selling off those dead chips as genuine Adrian's Digital Basement merch, along with Rammy t-shirts. :D
Haha! Like encase them in resin to preserve them forever?
@@adriansdigitalbasement Could make keychains from them. :)
@@adriansdigitalbasement Ya know, that's a pretty cool idea. I both collect glass/resin paper weights and display outdated hardware all over the pad. This would be a great combo of the two.
@@adriansdigitalbasement Then you can still use the bad parts bin, as is... Sell mementos to make more space, and support the channel at the same time.
I would suggest removing the epoxy from the dead chips to reveal the semi conductor underneath
encase the whole lot in glass or clear plastic and turn it into a desk ornament
one day, when mobile phone cameras are ultra-ultra-ultra high resolution, the grandkids can zoom in on the chips and see where the circuit failed
LOLOL
“I don’t need another working NTSC board, I have enough of those.”
...
Must be nice.
You're repairing pieces of my childhood to release into the world. That's something even I have failed to do. Love the channel:)
So I just got rickrolled by a tiny cute monitor... You know what? I'll allow it.
Hahah!
I hate the little monitor thing, I got rick rolled so many times, I can't take any more. I wish Adrian would take that thing back to the Goodwill store.
I love your opening credits. It looks like you reposted an old VHS episode. You even have some color bleed.
Great job as usual. Keep up all the fine work you do.
It’s sort of like the intro to a 80s cable public access program. 📺📼
Congratulations with 100 000 subscribers Adrian. It confirms that you are on the right track.
You need a bigger broken chips bin, when you get a new you should call it "The X-box" due to all the red X's 😃
He could build a nice new big box out of hot glue and RAM modules :D
Bad chips box out of an ugly broken trashed yellow breadbin- that should last for the next 5 years maybe 🤔 - and call it the X- box still ;)
@@0toleranz With all those dead parts, he would transform a breadbin into a deadbin 64. 😂
"Dead Test Redemption" followed by "Dead Test Resurrection" as T-shirt slogans?
Keeping with an Easter theme at least 🙂
Perhaps good game titles? 😉
Congrats on hitting 100k subs!!
With working 50 HR weeks this is what helps me get thru it thank you so much for the great content!
i know nothing about computer hardware but can't stop watching
My Commodore 64 is a made in W. Germany model. It came with all of the main chips (CPU, ROMS, CIA, SID, VIC, PLA) in sockets. Very handy as I killed more than a few CIA chips playing around with the User Port.!
I am looking forward to "Adrian is building a brand new computer from the Dead Parts Bin" video.
Would be like Frankenstein :-)
@@adriansdigitalbasement I know this is another very long shot of mine... But would it be possible to use some chips that are "dead"/half dead with those functions/legs that are still working? Say, an address line isn't working, design the system that it doesn't need that line and so on.
@@katho8472 that is sort of cheating. but interesting
@@highpath4776 and once a chip has begun to fail, it’s only a matter of time before the whole thing falls flat on its face.
Your' patience and enthusiasm for olde C64s is inspiring . So much so , that I have now bought a few olde computers and am in there solderring and desolderring to my hearts' content .
I thought about a nice idea for a video: an excursus about all the viable replacement for c64 chips (with eprom or even modern replacements)!
The dead parts bin had quite the feast on this episode
Take a shot every time Adrian says "bad" 😂
Great troubleshooting, as usual. I always love your enthusiasm, even when things fail!
Great, I've been waiting for you to upload all day! Thank you for your content and Greetings from Germany!
seeing these machines working properly is a wonderful sight ... and then when it's a Commodore 64 there is empathy
Woo. 100k subs. Well done Adrian. Keep up the great work and 200k will be just around the corner
Great troubleshooting Mr. Adrian. You impressed me a lot during this video. Thanks for sharing your masterful repair skills with all of us sir! Fred
I'm wondering if "cooking" that CPU could solve that marginal issue... Nice video as always, see you Adrian! 👍🏻😊
that sneaky RAM. goes bad and tries to take others with it LOL
Wow big congratulations on hitting your 100k subscribers milestone. Many more to come. Great channel great content. A lot to be learned from your videos. These repair series videos are some of my favorite on your channel.
Subscribed because of the intro. Great vids too. Takes me back.
Adrian, I've been following your channel for about 2 years now... I love what you do and how you present it. Don't change a thing buddy...! Tony K., Melbourne, FL :o)
Love your videos - keep 'em coming!
IIRC all of the PAL machines only had a cardboard/tin foil RF shield because the RF emission standards in Europe were less strict than the FCC ones.
Yes! I can't wait to watch! (have to wait, but I can't).
Thanks Adrian for the awesome videos!
If i remember well, there never was a soldered RF shield on the c64's in NL/Europe. I repaired dozens at least, back in the day.
They didnt have a real metal shield to begin with; it was just flimsy cardboard with a metal layer on it with the thickness of alu wrapper for food and that connected to the expansion slots metal casing with a sort of sliding clip.
Best spech Adrian: (15:18) "Look at that that looks a lot more normal!" hehe Super I like it!
I get absolutely amazed every time there's a Commodore repair video. Feels like C64's grow on trees over there or something. Never seen a Commodore in front of me before, only in magazines back in the day and I always lusted for one. At the time we mostly had TRS, Sinclairs or Apple clones. When the human malware is gone I'll try to get my damn dirty paws on one of those c64 maxis to get close enough to fullfil my childhood dream. Love seeing those repair videos!
My dad picked up a few C64s from a Charity shop in surrey. He had no idea what to do with them. I sold them for about £20 each on the internet, about 15 years ago.
The thing about diagnostic software, is to get the most out of it, you need to know how it does the tests, and how it interprets the tests. If you do not know how the test is done, you cannot trust the output. I used to have a whole suite of testing software when i was an IBM PC repair tech back in the day. And you could run several different ram test, and some would pass bad memory, and other would not. In then end, because i didn't have access to the source code of the testing software, you just learned which tests were good for detecting specific problems. For example some were really good at finding "slow" memory chips. Just because it says 250 doesn't mean it actually works at that speed. While others were very good at spotting "flakey" or intermittent bit errors.
In short, that dead test may be small enough to reverse engineer by hand, if the source code isn't available. If it is, then use that to determine why the tests are seemingly giving you bad information.
I agree. I'd trust a working game more than I'd trust that tester when it comes to the zero page.
In this case maybe the marginal CPU have one or more bad op-codes that are used by the tests and not in the tested game. I had a flaky 6502 many years ago that could not alter the D flag with SED/CLD. Since D flag is undetermined on power-on sometimes things worked but mostly not. Took me a while to track down. Maybe something similar happened here.
Great work. 🔨🔧🔩
Thanks so much for sharing. 😎👌🏼
Gratz on getting to 100K :D
I think by now I would keep a locally saved copy of the Dead Test manual.
I have seen all sorts of so called repairs I even had a cpu with a minty smell coming from it when it was warmed up some expert had used a blob of tooth paste instead of heat sink paste it did not do the cpu any good another was silver foil wrapped around a blown fuse and half a biscuit fed into a floppy drive lol always look forward to your videos
I love the see happy ending, both C64 fixed.
Hey Adrian! Waiting for the 100k subscriber video! ;-) CONGRATS!
Greetings, Doctor64!
I was given an Atari 800XL with almost the same issue as what was shown at the beginning of the video, and all it needed was Deoxit!
I was going to say that I didn't see any Deoxit this episode. I wonder if any of the massive and confusing failures could have had a bad connection involved.
@@jaoswald I was thinking it could be as much connections as chips.
keep up the great work Adrian 🕹 ♥
good job Adrian!!!!
This is wholesome content!
Have you tried the character generator in your ZIF-C64 too? It would be possible that there were problems with contacts in the old socket otherwise. Maybe the other two chips just worked because they scraped away some dirt/corrosion in the socket. I expected some Deoxid before throwing the chip is the bad parts bin ;-)
I have one of those C64 board, made in ´84 rev B. Bought it as "Spares and Repairs", replaced the caps and reflow the pins of the modulator and worked. Just yesterday brought it out for a "test run" and "poof" fuse blowed on it, replaced the fuse and now black screen. Those MT RAM on the board are hot to the touch, not sure why the fuse blowed though.
In this episode, ZIF 64 pulls jury duty.
Strange that the diag cart led you a merry dance.
BTW, Congrats on 100K subs. :)
I'm hoping that one day, one of the solutions to one of these problems is just going to be "cosmic rays", and a shield will have to be installed.
My NCR (GMbH) 386 was made in West Germany. I miss that thing.
I think I remember seeing that you had a chip burner. One of the functions is also testing chips. For your pile of ram chips you might consider 'pre testing' them and then putting the good ones in to the black static foam and then in to a chip box. It would be a interesting set of data to collect on how the population of ram chips vs how many MT chips failed.
Very entertaining as always Adrian. I have to admit I was a little worried about that thermal compound that was on the VIC, it looks like it could be conductive and some of it appeared to be bridging some pins on the VIC. But after looking up which pins they are (colour and luma output) it probably wouldn't make a difference anyway as the colour from these old machines bleeds into the luma anyway 😄
Heh yeah, that white stuff is safe and non conductive. It's usually gooped all over everything and I've never found it to cause any issues at all, even for delicate circuits.
@@adriansdigitalbasement it's likely zinc oxide aka old timey sunscreen
0:55 Murphy's Law at its best: just when u put the pieces back togethe and screw them...fail
R.I.P. six working RAM chips.
Oooooh Nice! Something todo later on!
Great video! I enjoy watching your mind at work - I literally ask a question in my mind about "why not try this?" and you do it as the very next thing. But you left a cliffhanger for me - I wanted to hear an actual working SID in the repaired machine - just to prove out everything was fine in the audio cct. :-)
wondering.. how many hours of "testing and soldering" were condensed into this 39 minutes video?
I think that RF shields are a US thing to comply with the FCC regulations. Other countries didn't have such rules, so products made for sale there often didn't have RF shields even if everything else was more or less the same as in the US version.
Excellent video as always. Sad there was no working 8-bit Dance Party due to the bad sid.
I am pretty sure one is coming up in the next part :-) (Or was it in part 4?)
Congrats on 100k subscribers. Now you need to get a bigger bad parts bin.
Hi Adrian. Always watching your videos packed with interesting informations on Commodore machines :-D. I have a question for the VIC-20 character set used on the Commodore C64. Is this just the binary for the Vic20 chr rom or is that something special? I would like also to burn a 27C64+adapter to have the VIC-20 font.
That PAL machine was all over the place. Nice work!
At this rate, the C64 community will end up developing a full drop-in RAM replacement. You clearly can't trust 40-year old RAM chips, regardless of brand.
It's no problem and has already been done: You can use a modern SRAM chip and some glue logic to make it compatible with the DRAM interface. However, because availability of 4164 is really good and it is cheap, it is pointless, it is much more elegant to use original 4164 RAM.
If the reference ZIF-board would develop an intermittent issue for any reason, Adrians sanity would probably plummet when checking for faults on new C64:s. Lol. I need to make myself something similar.
Wow, that second board was tough. When you put the RAM on the scope, you had some that showed weird data waves. I wonder if those were the bad ones.
Not all 6502/6510 cpu's have the same instruction set. There are "undocumented" instructions which are not part of the standard instruction set, but which can still be used. However these instructions are not available on every 6502/6510 out there. Maybe the test cartridge uses some of these instructions, which could explain why the test game runs fine, but the tests fails.
seems like you going to have to try some chip baking again with all those dead parts, im curious to see if the sid would come back or even if that bad cpu could be revived lol
Adrian, I have a bunch of SIPP memory in my parts bin; I’ll check which ones they are. The pins look like the Greek letter mu from the side, ie they clip onto the pcb on both sides
MT RAM? More like "empty RAM!"
"Dead-test Redemption"
That's another T-Shirt slogan along with the others 🥴🤷♂️
So the lesson with regards to the dead test cart and the diagnostics is this.. the ram test in the dead test only tests the first 4k of ram, and not very thoroughly. It does not need ram to do this test, whereas diagnostics needs at least part of the zeropage and stack to work in order to start up.
Due to not being able to use ram to keep track of anything, and the 6502 cpu only having 3 registers, the ram test in the dead test cartridge isn't just limited to the first 4k of ram, but also is not very thorough in how it tests ram.
Ram chips can fail partially, and fairly often do, the dead test cart will often not catch those. The diagnostics use a series of test patterns aimed at exposing errors which won't show with a trivial test, so will catch many errors not seen by dead test.
So.. always run diagnostics, and use dead test only for what it is intended for, figuring out why diagnostics wont start. If dt says ok, it only means it should be good enough to try run diags, not that the component(s) are actually good.
I REALY LOVE MY C64 ❤❤❤❤ FOREVER ❤❤❤❤
If I want to stack up on C64 rams for future repairs, what should I look for? There are 2 types of motherboards, one with 2 ram chips (probably 2x32k) and 8 chips (8x8k). What is the main proterty to search for if I don't want to replace with the exact same model number?
At 13:20, those are called Pinned sockets.
Nice video Andrian! Do you mind say where did you buy the RAM, PLA, VIC and others sockets (not the zif)?
You have to get a bigger "bad parts bin" or maybe "empty the trash".
Yep, the German PAL Rev. C C64s had the ugly cardboard heatshields. I always toss them in the bin :-)
The "marginal" CPU has most definitely it's I/O port damaged by ESD (this IS the type of board, that doesn't have any sort of ESD protection), that's why it does work, but zero page (the I/O port is memory-mapped to $01) shows damaged. Try testing it with a datasette - it most definitely wouldn't work. Other than that, there is no CPU-related stuff in the zero page.
We have a saying here in Derry, Maine: don’t look into the DeadTest
Good work on diagnosing the problems on the PAL machine and bringing it back to life. f you have a TL866 why don't you use that to do preliminary tests on some of the ICs instead of always testing in a machine?
isn't there a RAM test on the TL866II Plus Programmer you could use to see if the DRAM chips were working before putting them on your chip-tray thingy?
same should be possible with SN72x logic ICs
I feel cheated after watching this episode and not getting to see the dancing hands of the 8bit dance party.
Whipping out a "thingy of ram" chips is possibly the most bad A move in the tech world.
greetings from the UK ... looks like you're going to need a bigger bad parts bin before too long after all those bad ones were added lolol
LOL that bad parts bin overflow :-)
I have a feeling that the problem with that CPU was probably pin 27 (called "P2" on the pinout), which is bit 2 of the built-in 6-bit I/O port used for bank switching. There's three banks of RAM that you normally control: there's the RAM under the BASIC ROM ($A000-$BFFF), there's the RAM under the Kernal ROM ($E000-$FFFF), and there's the RAM under the I/O registers ($D000-$DFFF).
The dead test cartridge doesn't require the ROM chips to be installed, and probably starts with them already disabled by setting P0 and P1 low (the same as typing POKE 1,52 from BASIC). It seemed to work just fine with the ROMs removed. My educated guess is that pin 27 (P2), which controls access to the RAM under the I/O registers at $D000-$DFFF, is stuck high.
The entire instruction set probably works just fine, which is why that game worked; I bet Dead Test crashes when it tries to disable the I/O registers and starts testing $D000-$DFFF. It probably doesn't realize that the I/O registers are still mapped into that space. So I'm also going to guess that when the dead test cartridge writes to one of the registers that controls interrupts (like $DC0D or $DD0D), it's inadvertently "arming" an interrupt without first setting up the interrupt vector. So the next time the C64 executes that interrupt, it's jumping to some random location in memory, causing it to crash or get stuck in a loop, depending on what's at that memory location.
Dead Test disables the ROMs by means of Ultimax mode. The cartridge statically forces Ultimax mode, but in Ultimax mode only 4KB of the C64s RAM is enabled. Therefore Dead Test can only test the first 4KB of the RAM. In Ultimax mode, there is no way to access the RAM below $D000..$DFFF. (See www.c64-wiki.com/wiki/Bank_Switching )
@@danielmantione Interesting information. Testing the first 4KB will detect any stuck data bits, but it really doesn't detect if address bits 12-15 are stuck (since that memory isn't even mapped in Ultimax mode), so I was assuming that a full RAM test was actually a _full_ RAM test. Evidently not. I still think my theory is a strong one. We'll never find out unless Adrian runs more tests on the CPU, and what would be the point in that? We know it's _obviously_ got something wrong with it, even if the full instruction set works.
@@SpearM3064 I'm not sure what you mean with "full test". There is no logic on the Dead Test cartridge to get out of Ultimax mode, this means that from a software point of view there is absolutely no way to access RAM outside the first 4KB of memory as long as the Dead Test cartridge is inserted. For a "full" RAM test, you need the Diagnostics Cartridge.
that flashing is death, for the viewer
1:08
No, *THAT* pound symbol
West German breadbins are the best ones in my experience, a little less common than the English machines but still super common.
26:10 - Yes i am watching!
I think I saw cracked ceramic bypass capacitors on the RAM chips, might explain some of the wonkyness.
Just a quick note. MT usually stands for micron technology or micron
we're gunna need a bigger dead parts bin !!