|
Question for you! How sensitive to static discharge damage would you say modern computer components are? I still use a band when assembling or messing with them, after having it drilled into me bootcamp style when I was a repairtech in the late 90's. I'd assume modern components were more prone to potential damage, but do they design boards and such now so that it is less of an issue? Subscribed to thread, I've enjoyed reading it and need an old Amiga for the man cave. Thanks. GRINDCORE MEGGIDO fucked around with this message at 13:54 on Mar 18, 2015 |
#
?
Mar 18, 2015 13:51
|
|
|
|
| # ? Apr 23, 2024 14:52 |
|
|
wipeout posted:Question for you! How sensitive to static discharge damage would you say modern computer components are? This is actually a complicated question and the best answer I can give is "it depends". ESD is a very real thing that gets largely ignored by a lot of people because you don't hear of electronic components suddenly dying in spades because of wrong handling. I assume most people who put their own PCs together at home don't follow proper protocol regarding ESD and still, nothing seems to happen. To answer your first question, with smaller and more delicate structures on the dies in more modern ICs, ESD damage *in theory* is *generally* caused more easily. The problem with ESD damage is where people have a problem wrapping their mind around, is that ESD damage to an electrical component isn't having a spark jump over and then everything is broken. ESD damage can be a lot more insidious. I think we need to define this a bit better. I mean, we all know electrostatic discharge. We all got shocked that way one time or another, I think every person knows this phenomenon and I am pretty sure it was already known by humans long before electricity was a thing. The thing to understand is that we have a higher resistance to this than ICs have. We tend not to even feel these kinds of electrostatic pulses when they are not in excess of a couple of thousand volts, while there are lots of devices that can already be damaged by electrostatic pulses in the category beginning with 20V. You do not even feel or see those, but they are there. Another problem with electrostatic discharges is that they don't necessarily outright kill components right then and there, but can damage them and even change their electrical characteristics in a way where they might fail days, weeks and months after the event that caused the initial damage. In private hobbyist environments, it is very difficult to make a connection between an ESD event and the failure of a component, in companies like where you worked such things can often statistically be noticed by a heightened failure rate and component returns. Of course, stuff like that costs money and angers customers so companies tend to be very strict about such things. (which is the right thing to do) That all being said, most ICs have some form of built-in ESD protection that proofs them a bit against being damaged too easily somewhat. Different components on PCBs can also catch ESD events to a certain degree and help to avoid damage. Ports that are exposed to the "outside world" are usually protected extra well by clamping diodes and some such. I mean, think about it - Mainboard Manufacturers and companies like that are aware that the people buying their components will often be not people who will then proceed to put them together in an largely ESD-free electronics lab with all the right equipment, no, their customers are people who will put together their new PC on the kitchen table. If, as an example, a company like Asus wouldn't invest into safety precautions against damage of that kind, word on the street wouldn't be "be careful with Asus mainboards and follow proper ESD safety precautions", it would be "don't buy Asus boards, they are unreliable and break often". And even then, I am sure still lots of mainboards die of ESD related things in 2015. As an example, the Commodore engineers were a bit naive there with their 8 bit computers, there were no protections on the outside ports of those computers. Traces led directly from the outside ports to all the important ICs. This resulted in people breaking their computers often just by touching the outside ports while the computer was running - no joke. A very common failure mode of the C64 is ESD-related damage of one of the CIAs or even the SID because of that very reason. With the Amiga, things were already laid out a lot better in that regard, damage of that kind is almost unheard of with these computers. I hope that answers the question of ESD somewhat for many. It's always a pity that companies drill these things into their employees, but often don't really explain why. The wristband you already mentioned is an important and inexpensive part everyone dealing with stuff like this should have, I would also add an ESD Mat to work on and if possible, always be barefoot and work in a room with a stone floor in which the air isn't too dry. These are far from perfect measures that would not be "good enough" for a company environment, but I would advice them of being the most practically attainable for the hobbyist who might be reading this. Sweevo posted:There's also an updated version of the chip programmer software which cleans up the layout a bit, and fixes all the typos and Engrish. I gotta give this a shot sometime. But honestly, the software was already pretty good in it's original form. If you want to see terrible Software with low-cost burners, look at the Genius line of devices. That stuff even fucks windows up. Police Automaton fucked around with this message at 12:18 on Mar 19, 2015 |
|
#
?
Mar 19, 2015 12:16
|
|
|
I came into possession of a big box of jumbled ICs. I was thinking of looking through them. Are there any valuable ICs to look for? Or perhaps, ones that would be useful to you, Police Automaton? I don't care to mail some, if you needed them. Of course, as-is. I don't really have a way of testing them. It came out of a place a dude was doing gold scrap recovery, and I would prefer something that is useful be used, rather than scrapped. I will protect against ESD if I look through them.
|
|
#
?
Apr 7, 2015 21:26
|
|
|
There must have been tens of thousands of different models of ICs produced and most are worthless. You would have to be more specific I'm afraid...
|
|
#
?
Apr 7, 2015 22:43
|
|
|
Johnny Aztec posted:I came into possession of a big box of jumbled ICs. For starters, look up the markings for CPUs commonly used in old computers, and also for chips that are actually RAM chips which may still fucntion (many older computers accepted RAM as sets of chips plugged into the board, rather than modern "card style" setups where the RAM chips are grouped on a board that plugs in.
|
|
#
?
Apr 8, 2015 00:07
|
|
|
Johnny, thanks for the offer. If you have trouble sorting it out you can PM me (or post pictures here) and we can go through what you have there. I am always interested in spares or even just some odd ICs and know always people who need something or other for one of their systems. If I think you can make a killing selling this stuff on eBay, I will tell you so, too. I have literally probably (I am not even sure) hundreds of MOS 6581/2s for example (knowledgeable people will know what this means) and could probably crash the eBay prices for some custom commodore chips if I ever decided to. I don't really care about making money in this regard. I usually don't give them away anymore to people I do not know as I usually end up finding this stuff on eBay a week later. I hope to crash the prices one day if they get particularly terrible in the $$$ range. I'll then start selling some custom ICs for five bucks a piece. I have promised this to myself already years ago, so beware scalpers! Here for example, I had this lying around in an IC-Tube (don't use anti-static IC-transport tubes for permanent storage, [in fact, don't use them at all] I am just re-organizing still and literally didn't know where to put them) that's also why the surface for the picture is an ESD Bag (which is also terrible):  (sorry for the crummy picture quality, just wanted to make a point) I'm not even sure what the current market worth of these is. Just saw Haynie selling a bunch for lots of cash. I meant to update the thread several times with some of the cards I brought home with me but then never quite found quite the time for a lengthy post, I also got stuck actually playing some modern games for a change in the bit of spare time I have these days. I also wanted to talk about AmigaOS and two particular things which were very special for their time, namely AREXX and the inherent support for Multiscreen-setups, and that's probably what's coming next. Also lots and lots of Hardware porn, especially one TIGA-Grahpics board. I just didn't really have the time. Sorry again! Didn't forget about you people, it will probably happen this week.
|
|
#
?
Apr 8, 2015 00:44
|
|
|
so how did you come across so many SIDs? Do they work? Apart from high prices the market seems to be awash with seconds (or fakes...) with broken filters.
|
|
#
?
Apr 8, 2015 01:17
|
|
|
All my SIDs work. (If I didn't destroy them by handling them wrongly) These particular ones from the Picture are probably from returned/recycled C64s. So, second hand SIDs. They usually got returned because of a dead PLA or CIA. I also have brand new ones which were on stock by authorized warranty repair locations of Commodore, mostly there I have Amiga-related stock. (I have a big tray of Lisas, for example) This stuff just gravitated towards me in the last years, it was not hard to get in germany which was always pretty central for all things Commodore. People didn't start to act crazy about this stuff until fairly recently and ten to 15 years ago, all of this was pretty much worthless junk to most people. I don't have any current sources for spare parts like this. Only thing I don't have any good stock of is everything C16/Plus/4 and C64 PLAs. I wanted to get the arrays used as PLA (which is identical to the Signetics 82s100) in the C64 and program them myself with a fitting programmer (a programmer who can program these usually cost in excess of 700$+) but neither the programmer nor the ICs are easy to come by these days and I just sort of scrapped the entire idea. Chances are that the world is never going to need all these spare parts, at least I certainly hope so.
|
|
#
?
Apr 8, 2015 01:33
|
|
|
Just picking off the top of the pile, I found a Zilog Z0840004PSC CPU. A beefy 4 MHz! This guy ran a one man E-waste gold recovery operation. I picked up all the computer/electronics left behind(and theres a lot of it!) I haven't even gone through a tenth of it yet. This guy gives e-waste a bad name. The place was just a wreck. poo poo thrown every where. Open buckets of HCL acid. Just completely unprofessional. Edit: Hrmmm, if I am going to go through these, I should buy a IC tester. Any recommendations?
|
|
#
?
Apr 8, 2015 01:39
|
|
|
Police Automaton posted:All my SIDs work. (If I didn't destroy them by handling them wrongly) These particular ones from the Picture are probably from returned/recycled C64s. So, second hand SIDs. They usually got returned because of a dead PLA or CIA. I also have brand new ones which were on stock by authorized warranty repair locations of Commodore, mostly there I have Amiga-related stock. (I have a big tray of Lisas, for example) This stuff just gravitated towards me in the last years, it was not hard to get in germany which was always pretty central for all things Commodore. People didn't start to act crazy about this stuff until fairly recently and ten to 15 years ago, all of this was pretty much worthless junk to most people. There are SID test programs if you have a C64 with a socketed SID, I guess testing them all would be quite a lot of work. I've never need to try any of them but there are several c64 PLA clones using modern parts, I don't know if the C16/Plus 4 logic is known.
|
|
#
?
Apr 8, 2015 01:48
|
|
|
Johnny Aztec posted:Edit: Hrmmm, if I am going to go through these, I should buy a IC tester. Any recommendations? Like it has been said, I am afraid you need to be a bit more specific. There are no general IC testers which can just test everything there is. That being said, I would even assume that a lot of that stuff probably works if it wasn't actually physically damaged. I would first narrow down what exactly you have there and in what amounts and then sort into categories what's worth putting more effort into and what probably isn't. Try to break that one big job into many smaller ones and all that. the wizards beard posted:There are SID test programs if you have a C64 with a socketed SID, I guess testing them all would be quite a lot of work. Actually I had a setup with a ZIF-Socket to test lots of these and I have only come across very few broken ones yet, yes even the filters. I just eventually gave up and didn't even touch the factory-new ones. Many moons ago I took an older Rev. C64 board and socketed it entirely to have an easy testing platform. I also made a few other experiments with it (back then, your average breadbin C64 would cost about 2-5 bucks on the flea market if it was working) The most common failure mode of SIDs I know of is one voice dying. Broken filters come directly after that. After that you get what I like to call the "catastrophic" failures, like SIDs lying dead on the bus and blocking the entire computer from working, all three voices completely dead etc.. When you listen very carefully, you can almost always safely tell if a SID works without a testing program by the sound the computer makes on turning on. (you have to put volume way up for this) If the data noise when the C64 is booted up and sitting at the screen has a certain tone, the SID is working. If it's somehow "uneven", louder or distorted from how it normally sounds or even not audible, the SID is broken. This is not really scientific and not always true, but in my experience is pretty spot on basically every time. I have one very interesting-sounding SID here that does play fine but everything is off-key. If you only used it for old C64 games with their bleeps and bloops you would never notice there was anything wrong with it. Some tunes do sound pretty horrific with it though. I once listened to that little portal tune that was in the cracktro of that Pirates! one-disk release and if I had to put a name to the tune, I would probably call it "aftermath" or "nuclear nightmare" or something. Regarding the PLA only ever use PLA/PLD based replacements. Many people have the misconception it can be replaced by a truth table inside an EPROM and this is the most popular (because it's cheap) replacement for a broken PLA. To be fair, this seems to be reasonable at first look, a PLA is basically a fixed AND-Array and a (E)PROM is basically a programmable AND array. The problem like so often with all things in electronics, is timing. (In electronics, timing truly is everything) It doesn't seem that way but timing is extremely critical the older the revision of C64 you are looking at as these machines are put together on very tight and touchy tolerances. PLA/PLD timing is very tightly defined and reliable, while EPROM timing can and will vary with the internal structure of the EPROM. EPROMs also have the tendency to have their output lines floating (I wanted to explain what "floating" means in my EGA post that never came, I am sorry!) in an undefined way for several ns when the address lines switch, while with a PLA/PLD they all are set pretty much at once. These short, undefined delays could lead in the worst case scenario to situations where the fake-PLA gives two devices at once permission to talk on the Bus, for example one of the ROMs and the VIC. Then we have bus contention and the internal bus drivers of two ICs duking it out with each other. Always bad and should always, always be avoided as it will at least cause ICs to heat up more than they would normally, but it is not as bad for things that are strictly NMOS like the ICs in your normal C64 as they are by internal structure fairly short-circuit proof and can "survive" those few ns of bus contention until the fake-PLA "settles" and tells one device to shut up. But - you add a device with CMOS technology to the expansion port (some random cartridge) that have enough power to pull your devices against the +5V and this can fairly certainly kill the old NMOS ICs with the currents that suddenly are flowing. (I admit, this would be a rather special case but stranger things have happened, and who can tell why an ~30 year old IC breaks?) Now imagine, all of this happens again and again.. in your normal bog standard C64 about two million times - a second. 2 Mhz Bus! EPROMs also absolutely love to glitch on their outputs when the input logic level voltages shift. So in conclusion, this all will lead to instability in the best case and milisecond-spanning race/bus contention issues in the worst case which might actually damage hardware or at least cause ICs to heat up more than they normally would. EPROMs generally *can* work when the timing is perfect and you hit all the narrow windows you have to hit, but this is absolutely dependent on the specific C64 with it's specific ICs and the specific EPROM in question and the individual device tolerances of the very specific parts in front of you. There are even some EPROMs that don't glitch but there isn't any norm or rule to this. I can't generally recommend EPROM-based PLA replacements. You usually see the crashes with EPROM-PLAs especially in demos which do fancy and clever (and very timing-critical) things with the C64s chipset. Very old revision C64 mainboards even have stability problems with the original PLAs and pretty much need to be hand-fitted to components inside certain timing tolerances. The much higher integrated and more modern C64c/C64 II doesn't know any of these problems and their highly integrated PLAs break very rarely. (which is good, as there are basically no spare parts) Man, I went way more into detail than I wanted to and I am sorry I don't explain all the physics behind it as I didn't really say a lot, but yes I also know a few things about the C64. Always thought it was an interesting little machine, not as powerful as an Amiga but quite interesting. EDIT: Fixed a bizarre sentence Police Automaton fucked around with this message at 03:35 on Apr 8, 2015 |
|
#
?
Apr 8, 2015 03:30
|
|
|
Thank you for the advice, Police. I'll lay them out soon as I can and figure out what I got. I also need to google up a bunch of info on ICs. I feel like I'm walking in a classroom halfway through a semester. Lots of new info. Glancing through the pile, I saw one that looked different. DEC 23-001C7-AA Googled it, and an electronics website has it listed for 70 bucks. So, I definetly think I should get a tester and sort through them.
|
|
#
?
Apr 8, 2015 05:55
|
|
|
Johnny Aztec posted:Thank you for the advice, Police. I'll lay them out soon as I can and figure out what I got. The biggest problem I think you have is that all the stuff you have there might as well be defective and this might be the reason it ended up in that pile to begin with. I highly doubt it as this is material recycling and usually the condition of the ICs doesn't matter for them but certainly possible. As I said, you will not be able to get a tester that will test all your chips there, the testers you are thinking about just test ICs of 74xx, CD40xx etc. families, maybe a few transistors and also op-amps and that's about it, which will not be helpful for you as such ICs usually are only worth mere cents new to begin with. If you really want to sell anything your best bet is to sell "as-is" instead of coming up with testing schemes. For things like CPUs/FPUs etc. testing them for defects and really 100% ensuring function might be hard to "next to impossible" without an appropriate testing suite, hardware setup and possibly IR-Camera. Specifically Motorola series 68020/030 are very good at "hiding" their faults, common failure mode is heat-related malfunction in damaged ICs. The only way to test for this quickly for example is to film them with an expensive IR-Cam and noting if they heat up too much in the first few minutes of operating, another way is having them do heavy computational stuff for a few hours. Pretty difficult if you have lots of them. Not all that different for other such highly complicated ICs. The big prices for most of this stuff is usually paid by people who try to keep legacy hardware running for industrial applications, which I highly doubt will be the case regarding a PDP-11. Rest goes to weirdos like me, so don't get your hopes too high. Police Automaton fucked around with this message at 21:40 on Apr 8, 2015 |
|
#
?
Apr 8, 2015 21:36
|
|
|
Dave Jones did a teardown of a late 80's vintage Sharp X68000: https://www.youtube.com/watch?v=W40qGkp-mEU
|
|
#
?
May 24, 2015 14:58
|
|
|
Man that dude talks as much as me. He has this struggle I had with my two-mile long post to convey all the information you would need to convey for such topics. Interesting and expensive-looking machine, I saw one sold here in germany around new years for a double-digit number if I recall correctly. It was one of those very lucky eBay auctions. I was not interested, I kind of am after this video. As of why they didn't roll all the generic logic into custom stuff - I'd say price. That computer for the price he named probably already wasn't a huge breadwinner. It looks way too expensive for the named price. Like he said, very well engineered, I agree 100% there. It's a pity he didn't look into the power supply, I love power supply porn. I doubt the caps were inherently "dodgy" though. They're almost 30 years old and were working in a difficult environment circuit wise. They're sort of "allowed" to fail in 2015 IMHO. Sort-of-fake-EDIT: Writing actually made me think about the TIGA card and it turns out I took pictures of it after I got it and promptly forgot I did. Enjoy! First, here is the accelerator. I posted this image already but there you have an example for an 16 Mhz 68k. The 68k could also be paired with an FPU (68881 in this picture) but this was rarely done as the need for floating point calculations just wasn't that high. Also observe the non-ESD surface I put this on.  Here is the TIGA card. Those were sort of 2D-Accelerator cards with a fast graphics processor very much in the style of later accelerators. They never could quite make it in the general non-professional market as they were just too drat expensive. Later on, VGA (the competing standard) chips with hard-wired accelerated functions managed to at least be just as fast if not faster as them while being significantly cheaper which sort of killed the entire platform off. These cards also often needed to be paired with a proper graphics card as they were not compatible to IBM compatible BIOSes. Some cards had (quite slow compared to "proper" VGA cards) VGA framebuffers, some did not. The interesting thing about the TIGA standard is that it's neither color depth nor resolution dependent. Just add RAM! I think I never posted these pictures because quite frankly, they are not really good. I would need a proper camera and a proper lightning setup to make pictures of my Hardware and if I would run a serious blog I'd consider it, but for just posting occasionally on SA it's sort of not worth the investment, both the time investment to learn everything about it and the money investment.       Isn't it a beauty though. You would also find the TMS34010 in lots and lots of arcade boards. I could talk more about it, but really just watch this video. That man will tell you everything interesting about the chip and some more. Contrary to me he's also wearing a pretty snappy suit and if that doesn't mean anything to you then I don't know what does.
|
|
#
?
May 24, 2015 17:08
|
|
|
my c64 has a blown SID i think. I'm gonna have to check it with a scope sometime but yeah, I would gladly pay $$$ if you're anywhere in Illinois.
|
|
#
?
May 25, 2015 07:02
|
|
|
Except I'm in germany. Do you have a scope? Usually somebody who has a scope also has a soldering station and can use it. If it is actually the SID or not I can't tell you, I would need more information for that. What are the Sympthons that make you think it is the SID? It's not unlikely the SID is broken but it does not have to be it. Usually, Commodore socketed this particular IC, so for a replacement you probably do not have to solder. That being said, Commodore used really, really cheap singlesided spring sockets for the C64 and I have seen over the years several cases where the ICs managed to pop out of the socket partitially, causing all kinds of weird sympthoms. Simply reseating the ICs usually fixed the issue (In bad cases ther socket was so done for that it needed to be replaced with a new one) so I would first check if it isn't that. To tell you anything more I would need good pictures of the Mainboard and the ASSY number written in one corner of it. (You can't miss it, it's pretty big, located either in the lower left or lower right corner, depending on the revision of mainboard)
|
|
#
?
May 25, 2015 13:37
|
|
|
The Apple III had similar issues with chips becoming unseated and famously the official fix was to pick it up 3 inches from the desk and dropping it. This would re-seat the chips but it's a bit of a strange thing to ask your customers to do.
|
|
#
?
May 25, 2015 14:51
|
|
|
Anyone know anything about transputers? I've found An Unusual Thing in the trash at work. (click for big)   
|
|
#
?
Jun 14, 2015 04:04
|
|
|
Does the underside have wire connections? I have no idea what it would be called, but I want to see what wires connect to what. The thing is covered in flip flops and I want to know if they help with the two CPUs or the RAM banks. Those IMSTs are pretty good. According to the documentation, the G20s could get up to 30MHz for 15 MIPS. The 386DX ran at 2.15 Dhrystone MIPS. I don't know if the two correlated. It also came with some kind of Cache, 4kbs worth. But you had to manually manage it. You could either try to manually do what Cache would do or just treat it like fast storage that you can iterate. It was also little-endian at least, so you don't need to worry about that at least. It would have been pretty complicated to program for and if I were a betting man, they probably didn't sell well.
|
|
#
?
Jun 16, 2015 10:52
|
|
|
Oh yes Transputers. Another failed technology. The idea behind transputing was not only parallel processing in hardware, not much unlike how we do it today, but to use the processors basically as generic hardware for everything. Hardware I/O for example with the Hard disk? Fronted by a transputer. Graphics processors? Transputers. CPU? Transputers. In theory, and with the proper OS, you could scale the usage of the transputer cores for the task at hand. Even expand one computer with several others you could plug into the transputer network to add to the processing power. (Like connect six machines together via networking, having the same OS run across all six machines, but being able to split those machines up again into individual computers or smaller groups when needed) The whole idea was a simple instruction set and very seamless scalability. So in that theoretical example network of ideally engineered Transputing machines you would only need one OS and boot that OS one time from one machine and the other machines wouldn't be anything else than hardware expansions, until you need them for individual usage. This was a very exciting concept in the late 80s and so completely alien to how people generally understood computers at that time. It really seemed to be the future of computing for a bit. Like so very often, the downside was price. They were very expensive compared to "conventional" CPUs and quite frankly by themselves not a lot faster. In around 1988 - 1989 - ~early 90s hardware development started exploding and quickly there were very fast integrated solutions for things like graphics and CPUs also got faster and faster so the entire thing became uninteresting very quickly for the consumer- and even professional commercial market. It would have been a lot more interesting if they would have been very cheap, and you could just stick 200 of them together for a project. But in the end, it would have been cheaper to just use some high-end conventional CPUs and develop software for distributed computing for the task. Commodore played around with the concept a bit and as far as I am aware there were prototype boards for the Amiga 2000, manufactured by Commodore Braunschweig and some experimental OS running with Transputer Boards codenamed "Helios", but like many Commodore projects, this never went anywhere. There were also some adverts in german computer magazines for transputer boards from the german company XPert, but quite unsurprisingly they never materialized. I think there were actual transputing boards for the Amiga that actually got sold, but they would be exceedingly rare. I have no idea what spread they had in the PC world at that time but considering most people never heard of them I would think "not much". I could imagine they ended up getting serious usage for a few years in some cutting-edge scientific applications where the government footed the bill and the price tag was not very relevant, or like in that article some very high end graphics stuff. I actually did see transputer boards like the above here and there on eBay but the prices were often obscene and not worth footing just to have something to play around with.
|
|
#
?
Jun 16, 2015 12:39
|
|
|
|
| # ? Apr 23, 2024 14:52 |
|
|
Oh also there's this: http://woodtv.com/2015/06/11/1980s-computer-controls-grps-heat-and-ac/ (It's an Amiga 2000)
|
|
#
?
Jun 16, 2015 22:06
|
|