2/12 C64 black screen repair: The chip was too fast for to work!

[@Cyro_Ciolfi]

Hi Adrain, about the first vídeo numbers, do not let YT worry you about visualizations, what make your channel great is the passion that you show when talk about retro computing. Do not let YT numbers drive your passion. I'm from Brazil and even here, where Portuguese is the main language, you have a lot of viewers who admires your work. Thanks for all you content,

[@roadsiderebels3039]

"I dont have any switches because I always seem to fix them" The testimony to why you are a successful content creator. Love the repair videos.

[@dronejunkieuk7168]

Thank you so much for his video! My C64 had the exact same fault and the same MOS chips. All in all, a simple fix and I can now say 'it freaking works!' 😊

[@LynxCarpathica]

Please Adrian, Never change the intro. I just love it so much, and when I hear it, I always get so excited for the epizode.

[@adriansdigitalbasement]

Aww thank you!

[@robertoXCX]

I agree, the intro is banging and always gets me hyped for whatever tech we're getting into

[@iroll]

3rd for its a banger, right up there with the old LGR intro (RIP), in addition being really well edited. but, changing is part of living!

[@LynxCarpathica]

@@iroll oh yeah, don't even get me started on those.

[@aaronjamt]

Plus hasn't anyone else noticed that the C64 that he opens at the beginning is perfectly timed to make it look like it pushes his logo off the top of the screen?

[@The8BitGuy]

Great work!

[@m4rgin4l]

I freaking love repairathons.

[@raggededge82]

Same!

[@sq1rlsqu4d]

Same here. They can only be improved by the inclusion of cleaning montages.

[@sjjjvideo1]

Same!

[@relic374]

same!

[@SockyNoob]

Same

[@necro_ware]

Hi Adrian! The switches are really hard to find, but I could already repair those couple of times. They usually just get dirty inside and have to be disassembled and cleaned properly. You have to desolder the switch though, then you can remove the metal front part by bending the brace on the side outwards. Be careful, since there is a small spring inside, which you have to remove and clean everything with some IPA and a cotton bud. May be add some contact cleaner as well. After reassembling the switches work usually as good as new.

[@Dukefazon]

19:55 - put it in a socket and put the socket in the chip tester

[@JulioE1984]

Hey Adrian, Good Video. NKK Switches Model No. M2022TXW41-FA is a good replacement available at Digikey. You have to trim the front pins for the swtch to fit flush with the board but otherwise is fine..

[@bobblum5973]

I just posted to check Digi-Key and Mouser! Thanks for posting the info with part number!

[@adriansdigitalbasement]

Awesome thanks!

[@gieselats]

It is so lovely that you give the repaired c64s back to the community. Such a nice guy. Keep up the good work.

[@InfiniteLoop]

How do we go about getting our grubby lil paws on and Adrian C64 ?

[@bobblum5973]

@@InfiniteLoop Well, first you wash your hands so they aren't grubby, then you... 😁

[@InfiniteLoop]

Yeah I imagine he had a literal pile of them in some room like petsci robots guy, I just hope nothing happens to them like all the rare Tetris minuets that Elliot lost

[@dhpbear2]

19:20 - Tech tip, Adrian. Insert the chip into a round-hole (high quality) socket before placing it in the tester.

[@BurleyBoar]

Love the new split screening in this series. Thank you for sharing your hobby and joy you get from working on these old machines.

[@stefannissens282]

Well, it's time to unreveal myself :) I've been watching you for good few years now... I wouldn't say since day one, but I surely did watch all your videos and went through all computer-related stuff in pretty much realtime. In old computer community, your channel is my favourite source of information, both related to history and pure technic side of stuff. I'm a DIY enthusiast and I utilised your techniques on soldering, chip diagnostics etc. in many cases, even if these are not related to computers in itself (recently I diagnosed a failed IC on digital scales, for example... too bad it was completely rubbish and not worthy of an replacement, but nevertheless...). I do enjoy old 8-bit computers and have a special place in my heart for C64, it was my first computer (a late European shortboard version, not the breadbox one, we do call it C64C here, but it's not the 64C you mentioned several times). It used to have problems with the SID being too quiet - a common issue, but imagine me being 12 years old and reading the magazines where the solutions would be like "put a D/A converter under address $xxxx"... Now I know what to do with it, partially thanks to your channel, but Commodores are already going at insane prices and I have very little place at home, so you're giving me an endless source of enjoyment, even if I can't have one physically. I will not mention what I did to my C64, since I deserve a special place in hell for that, but I guess that if (theoretically) I coould sent you the pitiful remnants, you would have been able to fix it :) Just a theory though. Anyway, keep up with the good work - I am always looking forward to the next episode :)

[@kjrchannel1480]

The pitfalls of propagation delays are always fun. I found long ago when using mismatched 257's, created many timing problems when correct signal timing was required on the same databus. Atari XY games were designed with that in mind.

[@gwyllymsuter4551]

Heres a trick for cleaning gold edge connectors. Years ago when I fixed pinball machines and arcade machines we used a soft pencil eraser to remove gunk. it works really well. you know those white ones on the end of your pencil or as a block from stationary store. no abrasion to the gold fingers

[@gwyllymsuter4551]

@Mr Guru old erasers have 2 types of rubber. abrasive and non. just use the other end for erasing pen ink. never use sandpaper as it removes all plating and leaves exposed copper

[@steve6375]

When I removed chips and solder in sockets, I always used to put back the original chips in exactly the same positions and check that the symptoms are still identical to what they were before. This confirms that I have not made any mistakes fitting the sockets (and if it starts to work then the fault could have been due to a dry joint or perhaps a PCB/trace issue). Only then do I try the new chips in the sockets.

[@bonemar66]

There is a pictorial c64 web site with common problems listed, screen shots of what to expect and hints on where to start looking. Rather than start from zero, start with that site and add Adrian's finds to it.

[@espressomatic]

If the site were a wiki that would be pretty easy to do. If anyone knows the author it may be worth reaching out. If a picture is worth a 1000 words, a video is worth an encyclopedia.

[@tylerpferrari]

You have such a talent for walking viewers through these repairs. Your entire production is really second to none and your on-screen personality is so easy to watch. I've been watching these videos for a couple of years (even though I have no retro computing background) and I can't get enough. The audio quality, narration and overall production value are what keep me coming back. The content is great (even though a lot of it is over my head) -- and I can only imagine how much the serious retro computing community must appreciate your work, Adrian!

[@dparks256]

Whoa! The four video frames @6:45 is next level editing. Good job!

[@joefarr3304]

I have a ZIF socket that I've ground down the top plate to the minimum I could get away with (remove the two screws first) and that allows me to test these ICs with the short pins. Or sometimes you can force the ICs into a standard socket and then place the socket+IC into the ZIF.

[@generessler6282]

Interesting stuff. Note the 74F series inputs have a .6 ma low input loading vice .1ma for LS. The 6x could also explain why the 74F didn't work.

[@ericwazhung]

Was gonna throw that out, myself... Just ran into this in a recent project. Also, IIRC [later-note: *wrongly*], F and LS are [NOT] just about the same speed, too. F being older than LS, so "fast" in comparison to 74nn, not in comparison to other 74Xnn's.

[@godfreypoon5148]

@@ericwazhung F is way faster than LS.

[@ericwazhung]

@@godfreypoon5148 !!! Indeed, I stand corrected! Look at those other numbers I was presuming incorrectly, as well... e.g. LS has the lowest drive-strength of the four (74xx, 74S, 74F, and 74LS). Heh, I guess I just never ran into reason to commit them to memory; by the time speed became a concern in my projects, those four were nowhere near the speeds of other newer series.

[@sapixiv2174]

i love watching you do repairs. it's a learning experience.

[@Electronics-Rocks]

I tried a 74F series IC in a project and found out that 8 bit machines do not like them back in the eighties unless I was trying to complete multiple tasks in a CPU clk cycle. Also the 74F and 74HC/T do not supply as much current as 74LS so may not hit the trigger level.

[@parrottm76262]

A repairathon vid, yes, I'm there! No matter how many of these I watch, I learn something.

[@bobblum5973]

My experience back at that same time was that although the _function_ of the various parts like those '257s were equivalent, the timing specs of the different families (plain TTL, Schottky, Low-power Schottky, Advanced Low-power Schottky, Fast, CMOS, High-speed CMOS) can cause issues. I even hit a case where the Texas Instruments version of a 74LS123 would not work properly because the reset pin acted differently from other manufacturer's devices. I had to look at the actual diagram in the TI Data Book to see it and compare to others.

[@adriansdigitalbasement]

I seem to recall that might have been an issue on the Apple II? That one particular vendor's 74LS chip wasn't working well and it took them a while to figure out which chip was the problem. Maybe I mis-remembering.....

[@BrokeDad1]

I remember back when I repaired Central Office switching cards that you couldn't substitute just any of those in many boards either (especially using the 74HC chips). Some of the manufacturers even called for a specific one like a 74S only etc. I was going to make this point but read comments first and saw your post. Thanks for another interesting repair video Adrian.

[@bobblum5973]

@@adriansdigitalbasement I wasn't aware of the issue with the Apple ][, but it certainly could be the cause. I'll need to look into that, I'm curious! It was probably around 1981 or thereabouts that I found the difference in TI's chip; the manufacturer I was working at just added a restriction to not use that specific vendor's part. It was fun trying to find it; if replacing it you happened to get the TI part it wouldn't fix the problem! By the way: Keep up the good work, Adrian!

[@bobblum5973]

@@BrokeDad1 Yes, I hit that too, specific requirements for a chip family. Even the "fanout" rating, stating how many inputs of each family could be driven from a particular family's output pins. An "L" output could drive, say, 10 "L" inputs, but only 5 "LS" inputs. (Example only, I do NOT recall actual values from 40 years ago! 🤔). It blew my mind a bit to see two TTL gate outputs tied together to drive an input, but the circuit used that as a "wired OR gate" or "wired AND gate"; you'd need both outputs High to see a High on the input, or either output Low would be seen as a Low on the input. It saved them from having to add another IC and its required power and space on the circuit card, something you'd want to avoid for something going into an aircraft, for example.

[@bobblum5973]

@@BrokeDad1 I worked on some of the earliest CMOS logic circuits, using the 4000-series chips. It took getting used to, when looking for problems on a new board you could still get an open trace due to damage during component insertion. You'd see an open input on a TTL chip tend to float at about 1.5 volts, so pull-up resistors were used to force them to be a hard "one". But then the CMOS chips came in and an open input would show basically zero volts. I created a probe with a resistor and could clip it to +5V or ground to act as a pull-up or pull-down. The later 74Cxx and 74HCxx series tended to act more like their TTL equivalents.

[@IDPhotoMan]

Great video, you went through and explained just about everything that could go wrong. i'll save this video for when i go through a few of my black screen boards.

[@mk500]

YES! More C64 repairs! Love these videos :-)

[@rickperez8044]

NCP must be proud of their work. They also like to be compensated for their work. In a former life, I used to repair C-64s. We would have to mark the boards to prevent customers from swapping a repaired mother board for a bad one and then demand warranty work. It is sad to say that a local school district tried this on us.

[@DanPellegrino486]

Great video, and the production quality keeps getting better.

[@mattrossesq]

When you said you used IPA on the ports in was half expecting to see a beer in your hand! Thanks for another great video!

[@rabidbigdog]

Thanks for breaking out the chip tester. I really wondered how much it can help. This was a great example of not just using a single tool.

[@OldmanGamerYT]

OMG! Willoughby is right next door to me! That's so cool! Small world.

[@alanharkleroad4376]

It wouldn't be a c64 repair video without an 8bit dance party for victory.

[@JoshBattin]

6:04 That's a hell of a "problem" to have! 🤣

[@BrianOliver73]

I love the 2/3/4 way split screen views!!

[@BigDaddy_MRI]

Great video! I use my RCT nearly every day, and while doing a Z80 project, I found a TI 74HCT245N had a bad bonding wire, I think. Pressing on the chip with a probe, it would pass. Remove the chip and just re-insert it into the ZIF socket, it would fail. I tried heating it up - no change. Freezing it - no change. Pressing it in the center of the chip with a probe - passed. Very strange problem I've never seen. Thanks for the repair-a-thon!! I really enjoy these!

[@absalomdraconis]

For emergency reference, any time that a chip is _just_ too fast, you can usually use a simple 2 component RC circuit (maybe with an additional diode-resistor serial circuit across the resistor if you need a direction-based speed difference) on one or more of the critical inputs (for buffers I'd look at the enables) to slow it down. Not a great situation, and high-leakage inputs won't always work with this, but it can be useful.

[@johndododoe1411]

If the output line is ringing, you need the RC on the output to take off the edge.

[@tw11tube]

Commodore actually did exactly that. Not at the mutliplexers, but at /RAMCAS. They needed /CAS to the RAM chip to be late enough to not out-race the multiplexers, and thus they added an RC circuit to the gated CAS signal originating from the PLA to make sure /RAMCAS only goes active after the multiplexers switched..

[@TechTimeTraveller]

My poor breadbin got tossed when I was a kid because of likely similar issues - could have saved it but didn't have the skills. It's probably siting in a landfill somewhere crying "WHY??" :(

[@adriansdigitalbasement]

I can't believe how much you hate old computers!!!!!!!! ;-) Just kidding, of course!

[@DavePoo2]

The metal in the chip and board has likely rusted away to nothing by now. The plastic case however, that will probably survive for longer than the human race.

[@stevescutt2865]

Quality diagnosis & Quality repair !! . I flippin love your channel. Thanks again Adrian. x

[@ultrametric9317]

Good job! You are now the world's foremost 8-bit stud.

[@aaronbritcher7121]

Excellent work again. Greetings from Adelaide, South Australia.

[@elmariachi5133]

There's no need ti replace the power switch, unless it's broken. You can easily open it, clear the internal parts, bent them back a bit for better contact if needed and put it all together again.

[@TonyAtkinspdx]

Excellent work with the different video views in the same frame!

[@sheppardtribe]

Love watching the methodical approach

[@godfreypoon5148]

74F devices have sharp edges (electrically... fast rise and fall times). This can wreak havoc in a cost-engineered design (i.e. lacking good internal planes in the PCB, perhaps not as much decoupling as one might normally use), making signals ring like bells and coupling into other signals etc.

[@da1otta]

I am still fascinated by the C64's sound capabilities.

[@AndrewTSq]

I think it was designed by the same man who later made synthesizers for ensoniq, so maybe thats why :)

[@jaycee1980]

@@AndrewTSq Yep, Bob Yannes went on to found Ensoniq

[@senilyDeluxe]

These MOS Mux chips look good on the scope even if they're bad. Had the same problem, all the outputs looked normal on the oscilloscope. My In-Circuit Chiptester gave different failure results for both Muxes so I tried a working C64, there I got the same failure results for both chips and one on the broken C64, so I just replaced the MUX with the different failure message and that completely fixed the machine.

[@DavePoo2]

The 9V line is also required for a working tape drive. On the C64c if the 9V fuse is blown or missing, the computer will still actually boot

[@FloppyDeepDive]

I learn so much watching your videos! Thank you for your great videos! Keep it up!

[@KorAllRBare]

*TIP* To Test chips that have their pins shortened you may need to insert them into a socket "Socket may need modification" and then that socket with the chip into whatever tester 😉 APU I dropped you a 👉👍👈

[@granitepenguin]

I used this trick to reattach a leg on a VIC-II. It works really well.

[@snarfusmaximus]

I have a bunch of breadbin 64's to fix at some point and your videos are building up a mental database of things I should check. Chock full o' valuable information.

[@RS-ls7mm]

The newer (or faster rated) the chip the faster the output edge rates and the lower tolerance for input edge rates. Since the old boards never considered controlling the trace impedance this usually means the signals ring. On the latest chips I spent most of my time just getting the signal from point A to point B without distortion, sub-nanosecond edge rates with a 15 year design life span under radiation conditions is challenging.

[@FlyingSurprise]

I never get tired on watching C64:s coming back to live.

[@phxbofh]

Always enjoy your C64 repairathons!

[@jason50146]

Bil Herd has stated frequently the DRAM timing of the C64 is really bad. Probably not surprising the faster chip messes things up. BTW - pretty sure the "F" is for Fairchild, not fast. But, yes, the Farichild logic is quite fast.

[@DeepThought007]

Adrian, excellent job on the new production format! Nice morphing transitions too!

[@dottedrhino]

Timing has to match between different signals in the machine. Haha U R the wiz, A!

[@ceebee23]

a dancing Adrian is always such fun!!!

[@neilkurzman4907]

Sometimes you can get the faster chips to work if you stick a few picofarads of capacitance on the inputs. Sometimes the faster chips allow glitches to get through that would actually get filtered by the slower chips. But there’s other timing issues that can happen too. I’m not sure any of the modern 74series chips are slow enough.

[@KimSeiji]

nice work Adrian!

[@tappel0]

I was able to source those power switches locally a few years ago, so I think they're still available. Part number is KS-SP-222A

[@bobblum5973]

I've seen similar ones made by C&K, but don't have a C64 handy to look. May be able to find them through Digi-Key or Mouser.

[@billybob884]

could you put the mos chips in a loose socket, then put that socket in the tester to get the legs to reach?

[@Charleshawn66]

Another good video! TY for the Commodore content!

[@rastislavzima]

Maybe its because its very hot weather lately but every your video makes me sleep (and I have to watch it several times from the point where I slept last time) although they are all very interesting and surely not boring at all, but your calm and self confident voice sounds almost lime when parent reads some nice story. 😁

[@ancipital]

Cheers! Just got 1 C64 working again after being mothballed for years, got another which is sort of dead though I think it was killed by one of the rubbish C64 brick PSU's which went bad and fried all the chips - should really give it a go and try to fix!

[@cs121287]

The 8 bit dance party is the best part 😁

[@okhouri]

Thank you so much for these videos

[@_nc513_]

I still have a good size stash of spare power switches for the C64. All shiny and new. I tend to charge about 3.50 EUR a piece whenever someone happens to ask.

[@malanvogt]

Always love a good repairathon

[@tigheklory]

Great video! I bought an assortment of LS74 series chips off Amazon, but I have yet to use one of them in anything as they aren't ones that I have needed. Looking forward to you getting a Coleco Adam!

[@maclemz]

Great video. Where did you get the chip pin out diagrams?

[@chaggydawg]

Repairathons are the best, regardless of what the algorithm thinks! I hope it favors these videos for you.

[@boydlewis8747]

AB, you get into the 8-bit dance too much 😁

[@Mclaneinc]

Adrian, you were asking for a database of C64 issues with screen shots made from your previous repairs, so I added a link to a page which had it all already done. When I looked the day after it was gone, I thought I must have not hit send so reposted the link (that I have no connection to) as I thought it would be very helpful. Would it stop people coming to the channel, no. But that link has gone again. A little explanation of why it's been removed would have been nice, just so I knew, it was just someone trying to help re a request you had made.

[@donaldblakley6796]

Nice work. With the help from your videos.. I'm doing alright. Lol. Fairly easy fixed on the c64s so far. Yikes

[@mforrest85]

I'll take on that challenge and compile a list. I'll probably put it up on a website.

[@KarlHamilton]

Very cool. Subscribed.

[@ModelLights]

'"The chip is too fast to work right!"' LOL Was going to say even before seeing the video that it sounds stupid but is definitely a thing. And exactly that, especially on DRAM, faster versions wouldn't work correctly with the timing that worked perfectly fine on slower versions. Usually one of the signals isn't quite right per spec but was reliable with the slow, and the newer faster DRAM had tighter specs and needed the timing to be more correct. A LOT, and I do mean a lot, of designers misinterpreted specs and would often drop the data lines at the same time as the latch lines, instead of holding them for the correct times. Used to see that kind of thing all the time, like half the people using DRAM didn't understand the data sheet.. Way back around 1991 I actually built a 1 megabyte C-64, with the info from Transactor for 512K in the 64 and 512K in the REU. Worked great even though it wasn't exactly useful, still was a lot of fun playing with all the memory banking. Between that and the GEN2 adapter and having a 20 MB Miniscribe hard drive my 64 was decked out, unfortunately the CMD etc hard drives came out right then too. If I'd been about a year ahead I would have made it into a product.

[@szahm1968]

Check Mouser 633-M2022TXW41-FA for the power switch.

[@helgew9008]

The power switch can be repaired. Just open it, clean and lube, and reassemble.

[@adriansdigitalbasement]

It's funny every time I try that, it ends up breaking worse! LOL! I found a place on AliExpress with them -- so will try those out.

[@juanluisperealopez5816]

Hi Adrian thanks for your videos. I have a C64 with similar symthoms, I have removed all multiplexor chips and checked with T48 programmer (logic chips test tool) and seems to be ok. It's T48 reliable? My C64 is still with black screen.😢

[@tbirdapalooza]

Love the video but more than that: inspired by the way you say, "I don't have time for that." We should all be more willing to say that.

[@MRSTU1210]

would like to see a video on how you made your demo you ran on the C64 was cool

[@Daveyk021]

If you had two "F" versions of the 257 chip, I wonder if that would have worked? The timing of the two mux chips would have matched.

[@andygozzo72]

quite possibly, would be interesting if someone tries it!

[@adriansdigitalbasement]

Unfortunately I only had the one chip -- my 74xxxxx stash is a bit chaotic so it takes me a while to find any of them. LOL!

[@tw11tube]

As Daniël Mantione explained below, the propagation delay theory I detailed here like is *not* the cause. The design of the C64 (according to schematics) doesn't seem sensitive to fast propagation through that multiplexer. Slow propagation would be an issue, but LS already is at the slow end of 74 series families, so substituting the LS257 by something else (except possibly an 74L257) should not violate timing constraints. I expect it would still have failed. The 257 chips are used to switch whether the row or the column address is routed to the RAM chips. The RAM chip requires that the row address is routed to them for the "row address setup time" before the RAS line gets asserted and kept their for the "row address hold time" after the RAS line got asserted. The RAS line will be asserted for the whole cycle, but the row address only has to be valid for setup+hold time around the time where RAS gets asserted. Then the 257 chips get switched to route the column address to the RAM chips. Again, the column address needs to be present the "column address setup time" before the CAS line gets asserted and it needs to stay stable for the "column address hold time". When you replace the 74LS257 by a 74F257, the routing changes faster after the column/row selection input to the multiplexers change, so when the row address is switched to the column address, the 74F257 switches from row to column earlier than the 74LS257 does. I guess what's happening here is that using the fast chip, the column address appears too early, violating the row address hold time. Let's dive datasheets to validate that assumption: Our beloved MT4264-20 RAM specifies a "row address hold time" of at least 25ns, so the output of the 74LS257 needs to stay stable for 25ns after the /RAS line has been asserted (pulled low). The 74LS257 datasheet (by ON semi) specifies a "typical" reaction time to the select signal of 14ns. The 74F257 datasheet (by ON semi again) tells us about a "typical" reaction time of 5.0 to 6.5ns (depending on whether the output goes high or low). The F257 datasheet even guarantees a minimum reaction time (i.e. how long the old value is guaranteed to stay at the outputs when the selection input changes) of 2.5 to 4 nanoseconds. Too bad that LS257 datasheet does not, but it is likely to be around 7 to 10ns. So using the F257 kills 5ns of the row address hold time. As the hold time needs to be fulfilled on every address pin, I don't think violating it on all pins (using two F257 chips) would fix it to compared to violating it only on half of the pins (using only one F257 chip). If the F257 would not violate the row address hold time, the system would work even with a single fast chip. Faster RAM requires less hold time, so possibly using 100ns or 120ns RAM (requiring only 15ns hold time) would be compatible with the F257 chip in the C64 mainboard, whereas 200ns RAM is incompatible

[@andygozzo72]

@@adriansdigitalbasement if you think your stash is 'chaotic' , as the saying goes, 'you aint seen nothing yet' 😉

[@danielmantione]

@@tw11tube I can't fully explain it: The 74LS257 multiplexers are switched by the CAS signal, so the bus is only switched at the very end of the RAS phase. Common knowledge is that the CAS timing is critical: The multiplexers have to be faster than the PLA, but that the RAS timing is also super critical is not common C64 knowledge, I didn't know and don't believe that RAS is also on the edge... CAS being on the edge is worse enough. It may also be a matter of parasitic inductance and reflecting signals on the wires, according to Bil herd they had problems with this and the resistors in the address lines are there to attenuate reflections. The 74F257 may behave differently in this field.

[@terryraymond7984]

Adrian can you find a C-128 that has a failed IEC serial port. I have a Flat 128 that has this issue and might give me a few ideas of what is wrong in the serial circuit.

[@easy288]

I continue to be amazed that 40 year old computers can still work.

[@douro20]

I'm thinking of sending you a 487SX I have. I just pulled it out of an NCR laptop which unfortunately had a bad power board. The chip is actually older than the laptop was, but that could very well be down to the fact that it just wasn't particularly popular. It says "USA" on the back and was made in 1991.

[@stephenscibbe1116]

bought an Amiga machine [A2000] from North Coast Programming, peripherals [disk drives and memory modules] and related software / games. Circa 1990-1995

[@CoreyDeWalt]

That thing seems to have really clean video and sound!

[@tuworlds]

great video

[@MikeDancy]

this is why I like my C64 Maxi. No issues :)

[@bentboybbz]

A comment about cutting or breaking the shield. I can understand keeping a pristine completely original machine that's in Really REALLY good condition 100% intact but.... something that's already being repaired will never be that machine again especially if you plan to actually use the machine. So good judgement I think is the best way to go.

[@jonshouse1]

Logically (excuse the pun) the faster chip should work if it was only the speed that changed between IC types. A faster car will still drive at 30Mph after all. The main issue is that the logic input transition point moves between the different types (See wikipedia 7400 series page). As the Vih Vil transition point changes between the ICs types that pushes the output slightly left or right (with respect to the original IC type). 74LS is bipolar hence is its logic transition point is based mostly on current while the CMOS style input (74AC 74HC 74HCT - and possibly 74F, depending on make) is driven mostly by voltage, switching one IC type for another works in many cases - but only if the output of one IC matches the input characteristics of the second without moving the gate output timings out of spec. A shorter way of saying the same thing is TTL is low impedance logic input, CMOS is high impedance, changing between them moves the relative output timing of the gate.

[@chirpetstudios]

If you haven't already you can find plenty of replacement switches just by searching "c64 replacement power switch".

[@ncc74656m]

Adrian, would you mind scoping the outputs on the good chips for comparison?

[@melkiorwiseman5234]

I wonder what would have happened if you'd "tacked" a 0.1uF capacitor across the power leads of the 74F chip? If the problem really was that the rise/fall time of the signal is too fast, there's a chance that the capacitor would compensate for it.

[@timblake5844]

Great video again Adrian! Thank you! Love the repair videos. It’s been so long since Adrian used the game instead of the test cart as a sid test. I’ve forgotten, is it donkey Kong intro?

[@KindredIndust]

Question: i'm looking for amiga 2000 up to the 4000. would you happen to have one to spare?

[@kcj1993]

If you put those stubby leg chips in sockets will the work in the tester?

[@EarlofBaltimore]

Normal consumable length = 40 to 50 mins. Go for it!

[@johndododoe1411]

Nah, I skip even high interest videos above 15 minutes. Made an exception because it was short for Adrian and chip speed issues are interesting.

[@EarlofBaltimore]

@@johndododoe1411 I suppose we all have different attention spans. I enjoy a long Saturday video personally.