This is my favorite kind of video. I don't care what you're working on. I just enjoy the diagnostic journey.

@michaelblair5566

Жыл бұрын

It is great to watch an excellent Technician work! I've been a computer tech for 27 years and I learn something from these kind of videos!

@neilhill2056

Жыл бұрын

Yeah, great journey. Pretty sure re-seating the Agnus was what fixed it btw. :) Some Amiga's had a strap over the Agnus to prevent the green screen issue.

@Okurka.

Жыл бұрын

Concluding that all RAM chips must be bad and then desoldering them wasn't very diagnostic.

We have all been there. The fact that you always show the whole process is one of the things I appreciate most about your videos.

@pipschannel1222

Жыл бұрын

Indeed. Every once in a while we all enter the twilight zone ;-) Very well put. I think that goes for the most of us 👌

@joshuahorton-campbell3554

Жыл бұрын

Except he often (not in this video) avoids showing the whole soldering process as it causes too much drama in the comments between folks arguing about flux. :)

Adrian has become such a powerful retro tech wizard that his mere presence is enough to repair vintage gear. That's my theory anyway.

@joeschoebel1458

Жыл бұрын

The Chuck Norris of the Amiga

@Bruces-Eclectic-World

Жыл бұрын

He is strong with the "Force"... 😆 LLAP 🖖

@TheGreatAtario

Жыл бұрын

My ex had a saying when something would work right when she tried to show me a problem: "It's afraid of you!"

@MarianoLu

Жыл бұрын

Solid theory. I concur

@fubaralakbar6800

Жыл бұрын

@@TheGreatAtario That's exactly what happens with my mom's tech. She has a problem, I come in there, poke at the machine, the problem goes away. I like to say that computers don't want to mess with me, because they have seen how many keyboards I have thrown across the room :D

I think it's the jumper setting at Agnus. You cleaned all 3 pads with desolder braid, that might have cleaned up bridges that were there by accident.... i think... or you found the first self healing Amiga :D

@GODAXEN

Жыл бұрын

Probably the cause is a short in that jumper due to the formation of dendrites due to electromigration and/or rusting, a real pain to diagnose if they are not very apparent.

@widicamdotnet

Жыл бұрын

That's the only explanation I could think of as well - those pads initially looked like they might have been all connected erroneously.

@Okurka.

Жыл бұрын

A real technician would measure for a short before cleaning that section.

@pragmax

Жыл бұрын

This gets my vote. The first set of solder jumper pads looked very sloppy before Adrian cleaned them. Having all three bridged was probably the root problem.

@dog61

Жыл бұрын

This is what I think, too. Cleaning up those jumpers did the trick.

What I find so refreshing about your videos is your humility and your honesty. That is the true mark of an ethical scientist. I used to have an Amiga 500 and did most of the upgrades myself and some with a local Amiga users group/club's help. We're lucky to have you and your spirit of exploration and rediscovery. You are continually learning and wonderfully surprised that you are and that is very encouraging to a lot us.

@ankan666

Жыл бұрын

Well said! I mean, why are we into this stuff? Fixing old computers? You said it. Spirit of exploration.

"this is a long video of me fiddling around" - this is literally why i subscribe. i dont care if the computer gets fixed or not. your channel is awesome and i always love your videos.

We have all gone down the rabbit hole repairing stuff. Your humility and honesty is why your channel is so popular. The Amiga platform was always the most difficult to diagnose and repair in my many decades of repairing personal computers.

80's self healing technique was awesome. Can't even remember how many times joysticks stops working, but all you need to do is to open them, look them (because you had no idea what is what (kids were stupid even back then)), put it back together and it worked again.

@prowlingfrost5588

Жыл бұрын

well at least we kids tried to fix the things before throwing them to landfill. it can be today's machines are too hard to fix. too small components etc.

Hi Adrian - the 68010 differs from the 68000 primarily in what it does when it receives a Bus Error. When the 68010 takes a Bus Error it drives a bunch of information onto the Stack (52 words worth of internal state if my memory serves (it was 40 years ago...) ) This information is the internal state needed for the processor to restart the instruction. Why is that important - you can put a Memory Management unit between the Memory and the CPU and implement Virtual memory with this mechanism. When a program fetches an instruction or data that isn't in memory - the CPU receives the Bus Error from the MMU and hands execution over to the OS typically. The OS must then read the missing info of disk (bring the page in from the memory- paging...) and then it can restart the instruction by loading the 52 words pushed to stack back into the CPU. The 68010 was the first CPU from Motorola that could be virtualized!

@Torbjorn.Lindgren

Жыл бұрын

Despite this a number of early UNIX workstations do actually use the 68000 and paged memory, AFAIK the normal solution is to have TWO 68000 processor where the second one dedicated to servicing the page faults and an external MMU (either MC68451 or much superior custom MMUs). Obviously this was both costly and complicated so everyone switched to 68010 when it came out 3 years later.

@barrybogart5436

Жыл бұрын

Similar to the 6809 I believe. Stacks saved to process fast interrupts for memory swapping. Something like that. It's been as long time......

@knurf79

Жыл бұрын

It also had this 6-byte(!) instruction cache which could speed up tight loops, and there was a simple mod to double the clock frequency to around 14 MHz

@stevenwilson1690

Жыл бұрын

@@Torbjorn.Lindgren - you are exactly correct! I received one of the engineering sample 68010s and used it to implement Virtual Memory with paging through a custom MMU. This was a commercial endeavor, and had unix ported it to it. It never got onto the market.

OK, Mr. Black, this video was NOT a waste, it was very instructive and informative. Thank you. Also, you should not feel bad, we all learned, and the Amiga is working properly too.

I know it's a small thing, but I do appreciate you putting the direct link in your emails from Patreon.

@50shadesofbeige88

Жыл бұрын

I'll second that.

@billesposito3482

Жыл бұрын

Thirded!

@acidhelm

Жыл бұрын

+

@ForTheBirbs

Жыл бұрын

Same here!

@pallsmortion4750

Жыл бұрын

Tnetennba

My Amiga 500 had this problem. The issue was the socket for Fat Agnus. Usually reseating it would get things working again, but eventually the socket had to be replaced. I even had to put a piece of paper in the bottom of the socket to act as a filler to get it to make a better electrical connection at one point.

My theory: one of the two control lines to the data input latches from Gary wasn't making good contact in the socket. Probably through the course of manipulating the board for soldering that contact was improved. I'd probably still Deoxit the Gary socket to prevent a recurrence.

Amigas really were ahead of their time. Apparently now is the time they start healing themselves 🙂

@barrybogart5436

Жыл бұрын

Resurrection will be next. The Second Coming of Intuition.

@Macs

Жыл бұрын

--From the book of Gary

Hahaha yes! My Amiga 500 took me all over the place with these same issues before. It turned out the 1 MB mod I performed as a teenager in the 90s was kind of unreliable and JP2 was the issue. Cleaning it up on mine and soldering it properly fixed it.

If nothing else, witnessing your troubleshooting process is enlightening. You not only show WHAT you are doing, but explaining WHY, and as a novice I really appreciate the insight even if the particular lead you are chasing ultimately goes nowhere.

You have to disable the RAM expansion to use it as chip RAM. That's what the cut jumper did. You can also disable it on the card itself, if it has a jumper or switch for this but the 501 doesn't. So you either have to cut this jumper or, on older boards, disable it on Gary directly for the 501 to be used a chip RAM.

@buffalomerkis7603

Жыл бұрын

I loved the upgrade options of the Amiga. I had a friend with an A1200 with just about every upgrade going and a hard disk. It was a true joy to simply boot the disk to workbench and just double click an “installed” game. He had some nice picture editing software and was able to modify his character’s picture on a game called hired guns. I also remember the cheat for hired guns was to type Amiga on the opening menu.

@noth606

Жыл бұрын

@@buffalomerkis7603 Heh, some of the upgrade things were strange, there was a box you plugged in to the side of the A500 that had SCSI and a HDD in a sort of extendor box, hideously ugly. I had a A1200 with a Blizzard 1228(?) turbo card, basically a double speed 020 CPU and fast mem expansion, the most compatible turbo card there was. A friend had a A4000/040 but with the Cyberstorm 060 crazy fast turbo card thing but a bunch of stuff just wouldn't run on it...

@Smartphonekanalen

Жыл бұрын

​@@noth606 the expansion bus slot is great and stil used a lot in the Amiga community. It's a interface possible to be used in all kind of ways for ram, for accelerator, as zorro 2, action replay, to override the CPU etc. And it looks awesome! :)

Dude I learned SO MUCH! Please don’t feel bad about videos like this! They are still amazing. You are amazing! Thanks a bunch!

It looks to me at 31:36, the 3 jumper pads are all shorted. Perhaps when you cleaned up the pads and set it back to the original configuration, the short was removed as well which fixed the problem? I've played with TerribleFire cards and multiple A500(+) boards, and setting those jumpers was a headscratcher for me too.

@FrankConforti

Жыл бұрын

I looked closely at the jumper using the new-ish zoom feature in the iPadOS KZread app and I think you’re right. It looks like someone scratched between one of the pads but it doesn’t look like they successfully cleared out the solder whiskers.

@temmayB

6 ай бұрын

To understand the address spaces that are printed to screen and what the jumpers change, you essentially need the knowledge about how computers work. I am guessing that the jumper changes the address space to which the RAM is assigned to. For those that are beginners in this territory: I learned this in school when we used the hades framework made and hosted by univerity of Hamburg, Germany.

It seems to me the problem was the EXTick jumper by the expansion slot. As you recall, it was NOT grounded and it was also not set to Default when you initially tested it. I would wager that if you broke the bridge at that jumper again, you would have the same initial problem again.

My suspicion is that the 512/1M jumper may have been linking all three connections together due to the 512 not being cut properly until you checked it by looking for the cut being there, it could have been a sliver of trace left bridging it.

That’s quite a time travel back to the 1980s when I had my own Amiga 500. I did this exact hack with converting the external memory to additional chip memory. So I actually got suspicious about it rather early in your video. I love that Amiga is still alive.

The value of learning new things while trying to fix old conputers is HUGE. Here the phrase "It's about the journey not the destination" is in place. Thanks Adrian, keep it up, we're watching!

Start of the video amiga wasn't working. End of the video amiga was working. I would call that a win.

I just had a 500 which gave me 10 green blinks which should be memory or Agnus. It was Gary which was just not quite in his socket. Took me 2 hours to figure that out …

@Okurka.

Жыл бұрын

Captain Hindsight figured it out in a second.

I went through this exact same issue last year. In my case it was bad contacts on the pins in the ranger port. Try cleaning the pins.

Best video ever. Two reasons. 1 - you said ITS FREAKING WORKING many times. 2 - makes us feel better about being bamboozled by our Amigas. Thank You. Il take the black magic explanation. Loved the troubleshooting explanations. Two thumbs up!

The Amiga has always been "magic". It was still interesting.

I had a dryer once that had similar behavior. Didnt work anymore so we (me and my dad) opened it up to see if we can see anything. All we did was a bit of cleaning, checking connectors and stuff while looking around. Screwed it back together after not finding anything we considered a fix and it worked. I know its not real, but sometimes devices get emotional and fake being broken just to check if you care enough to "repair" it.

@ronald3836

Жыл бұрын

My Galaxy S2 at some had a problem with the charging port, I think after it got a bit wet, that made it think it was charging all the time. It was still useable though, and over the course of a few weeks the periods in which it incorrectly thought it was charging became shorter and shorter until the problem disappeared completely.

Do not feel bad. It could have simply been, removing and re-seating the Agnus and the ram chips, in your case, reflowing the ram chips by adding sockets. That could have been the fix. Similar things happen with C64 and C128 repairs also. Awesome video regardless, because now you know and are more familiar with the layout and those particular jumpers for selection and detection of the add-on card, and so do the viewers. Thank you Adrian! 😃👍

Don't worry, Adrian. A500 chip/slow mem is one funny jungle. I was going to say that in my years of hobbying, I've met every combination of it. But I don't think it's possible. There are the board revisions and chip versions. Ways to set the jumpers on the motherboard, extra bodge wire jumpers, possibly with extra logic chips... 512k, 1M, 2M Agnus... 512k on MB, 512k with space for extra 512k, 1M on MB... And is your Rev 8 board populated and wired as an A500 or an A500+? It wouldn't be fun if it was too simple! 😊

Glad that you still post videos like this, it gives me a bit of comfort that this sort of thing doesn't just happen to me. 😁

Don't fret! In WWII the RAF personnel said they had "gremlins" in their machines. It happens! It is very good of you to still post this film, it shows you are a fallible human, and it is still very educational.

Even though this one was a mystery, I was interested in the inner workings of the Amiga so found this useful still.

A little late to the party, but I'm going to echo what many of the commenters are saying but... with a twist. We saw the memory test work once or twice. If there was a short on those pads, before you cleaned them, those tests would never have passed. What I think happened, and I have seen this exactly once, is there was a tiny bridge that moved with the flex of the board. When you flexed the board to insert the memory card, and the tests passed (or mostly passed) that was a temporary clearing of that short. I doubt it's heat induced or the results would have been far more chaotic then they were. Regardless if the short existed only when flexed, or was always present, you cleaning those pads is what fixed that machine. I doubt ANYONE would have even glanced at those pads as a first port of call, so to speak. Keep doing what you are doing, Adrian. I would far rather watch fifty-minutes of a "self-healing" computer fix then watch a ten minute, super cut down video. I want to see the faults, and mishaps and even the head scratches.

I can say I've been here before. I paid an electronics repairman $50 to tell me a power supply for a vintage computer I acquired was fine. He was right. The reason it wasn't powering up turned out to be 1 or 2 invisibly blown capacitors on the motherboard, right next to the power supply connection. Live and learn.

It seems like things started working after you rebridged the EXTICK jumper which we didn't see on the video, so I'm wondering if that wasn't grounded before and when you made the change you connected the ground which allowed the Gary chip to detect the extra memory properly.

@rommix0

Жыл бұрын

Not just detect the extension card but also to power it. Without power the card will not work.

@robmc3338

Жыл бұрын

It was originally floating. Odd really

@theinvisibleman5751

Жыл бұрын

I totally agree. If non of the connection was done Gary had no ground

@falcon-ng6sd

Жыл бұрын

@@rommix0 My guess is that the _EXTICK/_EXRAM signal, when asserted (low) causes Gary to assert the _RAMEN signal to Agnus when the CPU addresses the expansion RAM address range ($C00000 - $D7FFFF). _RAMEN tells Agnus that the CPU wants to access chip or slow memory. Having all 3 pads of JP2 bridged would force the CPU A23 and A19 lines to the same level, which /should/ allow the ROM, chipset register and CIA register ranges to still be accessed, as well as at least the lower regions of chip memory. On the other hand, having no connection between any of the pads would leave the Agnus A19 line floating, which could cause some problems. The video doesn't show a clear enough view of JP2 before the intervention; we can see the pads at 31:37, however I can't tell whether there's a bridge between the top and centre pad (no depth perception with a static 2D image!) As some commenters have mentioned, this could be tested by removing the solder bridge from JP2 and seeing how the machine behaves.

I remember as a young kid, reading about the trick to change the memory to all chip memory by cutting one or two rails on the PCB. Well I did it with a small knife on the floor of my bedroom of course, turning computer on. Just grey (or was it black I dont remember) screen. My head was turning to red and my heart was starting to jump. OK, with a shaking hands I took the soldering iron about a first time in my life and managed to fix the problem by some miracle, somehow the machine turned on and was working fine the rest of it's life (and it is probably still working somewhere).

As a viewer of yours with epilepsy, I appreciate the warning in the beginning so so much 🙏 Most KZreadrs are not as thoughtful to do so thank you 😊

Your transparency is commendable...well done! Even tho you weren't satisfied with the results of your work...we still enjoy watching the process develop and get to follow along and learn thru your thought process

Very nice documented debugging journey! The journey is often more interesting than the destination.

Amazing work, as always. What strikes me is how would you know all these tricks back then, as a teen and maybe no internet......

I wasn't exactly screaming to the screen, but I did yell into the kitchen where my son was preparing dinner. I had a 500 and it had a 68010 and 3 toggle switches through holes in the back of the case. They enabled me to switch the sound filter on or off, switch between the original KS 1.3 and 2.0 (some software didn't work with the new) and switch between 512k or 1M of chip mem. I think you prefer your cases stock and without holes, but these switches allowed me to put the machine back to stock configuration with just the flick of a switch (or 3). I really enjoyed this episode. Not just because of the recognisable, similar configuration of the machine, but also the experience of having a problem suddenly disappear. I've had it happen on the A500 a few times, where I would want to show a problem to my knowledgeable neighbour, just to discover I wasn't able to reproduce it.

I don't "like" videos very often, but then I do they are THIS cool.

Accidentally opening the on-screen keyboard multiple times really delivers a message. Of course using the mouse for a youtube video does make a lot of sense, but that menu literally craves to use the F-keys. Never mind, just my OCD kicking in, great content as always, thank you :)

I saw this video from start till end and saw next additional steps: 1) Deoxidizing Agnes chip socket :) 2) Re-soldering Jumper J1 (which looks like peace of...) 3) Re-soldering memories (which also could have bad contacts) By the way, this is the best video ever, when device restored by magic ;))) Thank you for the same content, nice to see you :)

Very real video. You could have shotgunned a bunch of parts and when it started working you could say it was parts that you replaced, but no. You honestly showed what really happened. Perhaps an old cold solder joint was causing the issue. Your channel is entertaining because we see your logic troubleshooting response. Love the whole real story you present!

While I feel bad for Adrian going through all the trouble to fix a broken system that wasn't really broken, it does make for a great video for us to enjoy.

This was a wild ride that had me gripped throughout and laughing by the end when it just worked after you returned the jumper settings to how they were before after hitting the pads with some desolder braid to clean up any potential bridging issues.

11:05 Strong Bad: "🎵 Ohh, Gary--I hope this e-mail's from Gary 🎶"

Entirely worth the watch, you demonstrated your debug process and came to the correct result, I learned a lot from it. I still have my 500 I bought new in '89, also an A1200 & 3000, all way ahead of their time......

A use I have for the RTC is knowing what date save games were made at, e.g. with Captive or Black Dawn Rebirth, I can see immediately what my last save position was when I come back to play again. So... yep, clock is important for non-productivity too! Also the 68010 solves some noise issues on the bus - this was necessary on my A1000 to use the Parceiro II, for example. (Also the speed bump is nice!)

@DbugII

Жыл бұрын

Agreed with the clock, it's also very nice when you are trying to develop (or debug) software natively on the machine, having the actual proper time stamp of when a particular executable, log file, etc... was generated is super useful. Really sucks when everything is January 1st 1970 (or whatever the Amiga OS equivalent is)

the Amiga is a living machine, it does what it pleases and loves to mess around with us. always has been... still is 😁

You gave me a hint as to what might be wrong with my Amiga 2000 in which I changed most of the RAM and all of the buffers/latches. So your video was definitely useful (and fun as always!).

I haven't watched the video yet, but when I was in college I worked part time for my step father in his Commodore factory authorized repair shop. One of the most common problems was a poorly seated Agnus chip. Symptoms varied, but the test was simple. We'd disconnect it from power, and then drop about a foot onto one particular well-padded office chair. Then plug it back in and power it back up. If it suddenly started working, we'd go through a proper chip reseat. Until we figured it out, this issue was a huge headache. Sometimes the machines would suddenly work the moment we put them on the bench, only to fail when they got home. I don't know if the problem was the engineering tolerances on the socket, but as I recall, Commodore eventually started using a redesigned socket that included a clip to hold the chip into place.

This definitely did NOT feel unsatisfying. Watching you work is always a pleasure regardless of the conclusion at the end. It's like watching Sherlock Holmes solving a mystery. And this one had ex-tra tickling drama! :D (Did anyone get that?)

I thought about the heat too. Using hot air a large-ish part of the PCB warmed up (and flexed) - including the bus transceivers I believe. That might have fixed it more or less permanently! Not useless video, still enjoyed watching it in its entirety! :)

The Amiga was just depressed and all the attention and care made it happy again.

It’s your aura which does the magic 😂

I needed to see this video! When I dive into an electronics issue, it's usually with a fairly decent plan of attack. Then things often seem go weird in ways amazingly unique to me. Almost, as punishment for attempting the repair. Watching someone miles above my abilities also wonder aloud why some things didn't , then did, then didn't, then did work gave me strange but great comfort. The memory of this video shall help me endeavor to persevere the next time I am in electronics dispair!

Adrian, you said you didn't do anything to it, but you actually did. Any rework of the board, from soldering, desoldering, Deoxit on the socket, all contributed a factor. Just because you don't know the cause nor cure doesn't mean there was not a valid issue you corrected. Sometimes you just have to accept it, and appreciate exercising your troubleshooting skills along the way. Also, make a note of this one, and if you come back to it months or years from now and see it acting weird you'll say "Ah Ha! Gotcha!" and dig back in to it again.

It's always frustrating to have problems "magically" fix themselves. You end up learning nothing, and have to worry about it coming back later. After seeing the schematic, I remembered the mod to use the trap door board as chip RAM. I modded my A500 rev 5 for that when I finally got an expansion board. I learned the hard way that those cheap PLCC extractors with the metal claws can easily take a chunk of skin off your finger if it slips. I had that happen when I was trying to pull a chip off some board. It slipped while I was squeezing it really hard, and my middle finger went right into the two metal claws. Not only was I bleeding and cursing, but the chip was still firmly stuck in place with nothing but some scratches at the corners.

The original sockets were trash. You said it yourself that they are not good. You replaced them with good ones and Bam working computer. Also the deoxit does wonders. I can't tell you how many things I have sprayed back to life with this miraculous liquid. PS: These are the best kind of videos. It teaches people how to track faults in machines like this. Even if it looks like nothing was really done it's still a teaching opportunity. Great video

The Amiga itself is an amazing machine. Way ahead of it's time. I had the Amiga 500 with A590 side ram / hard drive expansion card. It was populated with 1.5 megs of ram in the expansion card. I've upgraded my Angus to Fat 1MB Agnus and cut the trace next to the chip. I don't recall ever having to mess with any other traces. So with the side expansion card and A501 ram card in total my Amiga 500 had 2.5 megs of ram. The funny story though I actually spilled milk on my A590 which caused the computer to shut itself off. I quickly dried the computer and A590. Then took the cover off to dry everything inside of it I could. I let it sit for a few hours to make sure it's completely dry. When I turned it on it booted up just fine BUT it lost some RAM. Instead of 2.5 megs of ram it went down to like 2.1 megs. Since I didn't have the proper tools to figure out which RAM died I left it as is. It kept on working for years without issues. It's truly amazing how it "self healed" after that oops with a glass of milk! 🤣

I think that Amiga 500 felt all the love and attention it was getting after a long time and decided to make a little effort.

Apart from the mini instruction cache (small speed improvement in tight loops), the most important 68010 feature for the Amiga was the introduction of the Vector Base Register. This allowed relocating the vector table, making it much easier to intercept/override system or game interrupts and exceptions. Some special apps used this trick for easier runtime debugging, or taking control of a running game, apply cheat codes, and then resume normal execution.

Ha, ha, love it! Definitely wouldn't complain about a computer that works! Ps. I was very happy to get two of three Commodore plus/4 machines working by simply swapping around chips until I had two good sets. That made me feel like a fool as I'd not been able to use any of them for twenty-five years - yet, very pleasing!

Thank you for still posting videos like this. I'm always learning from you.

Flounder?. May I have 10,000 marbles, please? Thank you Adrian for keeping it interesting still!

I am so glad you published this video. This has been my experience with my Fairlight build. Indeed this is like my life.. I actually feel better watching this :) Thank you .. I really liked your searching and trouble shooting process.. 10/10 ! :) Failure is a success !

Not useless! I found your video very useful. These things happen and it's really helpful to see. I think the bad solder joint/dendrites situation other commenters mention sounds highly plausible. The main thing is you got there in the end!

Just before you started the desoldering process, I saw JP2 by the ROM socket and something clicked in my head about that being something to do with the RAM, but I couldn't recall what it was for, or why I remembered it, presumably a youtube video from the past, cos my 500 Plus had 1MB onboard plus 512k in the expansion slot (still have that somewhere), so never had to do those mods... :P

I like videos like this. Makes me feel way better when doing nothing fixes things for me as well.

I once was helping some friends diagnose a faulty memory expansion on an MSX, and we started following traces. "Point A should be connected to Point B" - OK "Point B should be connected to Point C" - OK "Points A and C should be connected" - Having tested the above, it should be logical to assume (like we did) that they were. But they weren't. There wasn't a single component between the points. Just direct, and very visible, traces. Somewhere, we still don't know where, there was a hairline fracture messing up the connection. A jumper wire fixed the problem, and the memory expansion worked.

This was actually a very helpful video. Amiga systems can be the worst to diagnose issues. I have owned nearly every model and sub-model. It was an art to get all the options working at times. Thank you for sharing a video that some diagnosticians may have not uploaded.

Great video. We’ve all gone down the wrong line of thinking until we got to understand the board better. .PS. The first thing I did when fixing a computer was to reseat all socketed chips. Fixed many faults.

Mate we all saw the thing refused to start, things changed when you cleaned up the jumpers, they were probably both dodgy. Seeing you freak out at problems like this reminded me of my tech days of absolute frustration as well but I don't think I had a machine self heal. For me it was always a fight to the end. There was good info in your repair that I didn't know about Amigas.

I encountered this when I got a PiStorm accelerator a few years back for my Amiga A500, when Raspberry Pis could be readily bought, and quickly found I needed 1Mb of chip RAM to run some games on it. The accelerator itself has its own fast RAM, but a lot of software needs chip RAM to run. So I popped in a RAM expansion card and converted it into chip RAM via the modifications to the jumpers Adrian had to deal with. When I watched the video I was yelling at the screen 'CHECK THE JUMPERS'. It's often I know more about a computer than Adrian does :)

Just you going through the processes of figuring out things is great learning knowledge for me. Plus I moved directly from the C64 to the PC platform so just seeing you explore Amigas is an educational experience. Thank you for all you do!

Damn You , a whole hour if Amiga on a Sunday ,, and I got the 'kids' this week

Adrian, don't sweat it. I've worked on host of systems over the past 25-years, & every now and then, weird crap can/will happen.

Sometimes this happens to all of us enthusiast. These machines are somewhat sentimental. Thanks for sharing this video which confirms your authenticity.

Back when I worked on these, I even saw the PLCC tool pull the top off of a chip. I used a set of spring removal tools instead. You are able to work the tips below the chip body at the corners and gently pry the chip up.

"It worked the whole time?" - We've all been there, wondering if the device under test is gaslighting us or we missed a logical step in the diagnosis process. (And subbed!)

Watching Adrain work and learn is the joy we all experience figuring things out. Excellent video so entertaining!

36:97 for future - address error means that a location in ram is responding at more than one address. That is, writing to one location can corrupt another location. It could be an open address line, bad chip select, or even a bad ram chip.

The DiagROM has a more extensive memory test that you can run from the menus. The one that runs when you power on is a quick test.

@retrolabo

Жыл бұрын

Yeah the diagrom got a little bit of unwarranted bad press here. The initial test is minimal just to detect some memory and try to run (which is a good design!) Then you can always trigger deeper memory test yourself from the menu

@jason50146

Жыл бұрын

Using the comprehensive RAM tests, I have experienced Diagrom passing RAM that was bad. Amiga Test Kit does a better job with RAM. Diagrom is a fantastic tool, but this is just a fact. Others have reported this, as well.

The jumper looks pretty bad before fixing. Maybe all three pins were connected ? Leading to that strange behaviour ?

@awilliams1701

Жыл бұрын

I thought maybe they weren't connected at all.

Great video! I've got a couple A500 systems I bought a year or so ago I need to resume troubleshooting. Both died within 20 minutes of being powered on. This vid has inspired me to pull them out of storage and look at them again. Thanks Adrian.

I once had an A500 I named Christine. Brother smashed it up with a hammer... It was back new-like the next day. Every now and then.. It would play an old bit of music through workbench... something from the early 1980s...

@48:27: "I lost my train of though" .... :) Must have been a long day! Entertaining video to watch!! And, surely, over the years, you have encountered computer problems that get solved by disassembling the hardware only to find that everything works once you put the exact same parts back together. In other words, something was slightly loose somewhere, and by taking it apart and then reassembling it, surprisingly, everything works. Who knows in this specific case, but something to consider. You took out the ram, and put it back, etc. :)

Thanks for sharing this buddy. As commented before, we all have been there. Problem solving is a process that sometime look like us throwing stuff at the issue until something we understand stick. The most frustrating case are the one that resolved themselves without you understanding why. I hate those, I don't learn anything and when it happen again, still clueless 😄

Always interesting to see Adrian just fiddling about with retro tech regardless of the end result!

My Amiga 500 one healed itself too. At some point the blitter seemed to be doing random things in games (without crashing the machine). The machine was less than a year old and my dad and I were about to bring it back to the shop for repair when we tested it one last time. Suddenly the problem was gone. The only thing we did was move it around a bit. The problem never returned and recently I started up my old Amiga and it still worked fine.

The pin that detects the presence of the memory card brought up some memories. Long ago when I had A500, I cut that leg of the socket on the motherboard and connected two wires on it that I then connected to a switch I installed on the back of the case. This switch allowed me to easily disable the card when I wanted to run some badly coded software from, A1000 era that thought all memory is chip memory.

I love how this is both in depth and learning-as-you go, I learned a lot thank you! Such videos are never a waste! As has been said below I'm thinking too it was either a short with the Agnes jumper or a bad solder with the memory chips. Possibly even both? You tinkering with them fixed it. As shorts on computers go they can cause any kind of illogical behaviour which this seemed to be a lot so I'd say a short on the jumper would be most probable perhaps. Think several said below they actually saw all 3 pads connected. On that thought it's fortunate computers started using jumper caps eventually, lols.

I've been waiting forever for Adrian to make a KZread video for an amiga!

I did this work for years. What you had happen was not uncommon for me. Mostly Commodore gear that would do it, but the occasional Apple and PC would have the same thing. Its funny to me that all these years later just how many PC problems can be traced back to dirty machines just like back in the day. Tear them open, clean them all up and re-assemble. Voila, working machine. On the other hand, the C=128 was a bit of a bear at times, but compared to some of the stuff today, I really miss these things.

Thank you, Adrian. Appreciate watching your process, even with such puzzlers.

X-Files and Fringe have taught me flashing screens of red and green are a form of mind control hack. Miss me with that, Amiga!