A flat headed "punch tool" can also be inserted in the hole once it is drilled and you can just lever the head of the rivet off without having to drill all the way through and risk damaging the metal.
While that would work, if the rivet is spinning, a small angle grinder can also work, to cut the head of the pop rivet off of the base, and/or give you a point onto which you can hold the bugger with a basic screwdriver while drilling. I am not a fan of using pressure for this, as that is not often effective. A drill bit about 75% the size of a pop rivet surface dimple works; and if you want to be fancy, there are specific pop/blind rivet removal tools.
Hi not Apple related but fixed my first motherboard today which had a missing trace due to varta corrosion. One bodge wire and its all good - thanks Adrian for the advice and giving people the confidence to repair these things.
Definitely a larger bit helps. Almost but not quite the diameter of the head. The main thing is to drill the head off the stem, then just use a pliers to pull off the stem from the back after you separate the previously riveted parts.
Exactly - just a large enough drill bit to demolish the external face. The rest of the rivet may fall out inside, but you're typically drilling it out to open it, so just grab it from inside once you're in. Trying to be gentle and get enough of it drilled out through the hole so you can coax it out... maybe if you're drilling it for some other reason than to open it and you don't want a loose rivet floating around the internals of something, otherwise it's just a quick "zip" and you're done.
16:16 For viewers that are outside the US and have 230V, these Astec power supply could be easily converted from 110V to 230V by removing the black jumper wire next to the small transformer. It is labeled on the PCB which says "115V select".
Next time try a LARGE drill bit. The pointy bit will center on the inside and the larger portion will eat up the exterior flange of your rivet. Oh, and you did on the 2nd rivet :D
No, they were an early 'cover our assets' measure to prevent electrocution by hobbyists used to poking around inside their machines. Later PSU revisions removed the rivets and put a gigantic 'high voltage' type caution sticker on the PSU top case. Take it from a technician who slipped and got 50KV down me right arm, actually cooking me muscles while working on a Tempest vector machine, it is not fun. I was lucky, and I was not intending to be anywhere near those components. As I recall, someone walked in, asked me a question, I tried to sit up and answer them, then...bzzrrt.
@@markcummings150 "Yes they are a tamper indicator." With the confidence you wrote that ... where was that confidence when you decided to NOT show us the source of this remarkable conclusion? Then "those days" ... is 1980 for you the ancient past?[2] Are you joking? BTW "ancient": Riveted joints have been used in mechanical construction and jewelery making for thousands of years since ancient civilizations[1]. Not only were rivets (as well as metal drill bits) widely and freely available in the 1980's, but no one (including the people who certify at the FCC) would call these riveted joints tamper-proof. If you do research, you will find topics like "water proof", "cheap and tight connections", "perfect for sheet metal connections", but not tamper indication. Even your mentioned "blob of paint" would be more suitable for such an application if it contains something like a seal imprint. Sorry about the rant (which is actually: FACTS), but the grin I had reading David's comment stuck in my throat after reading your comment. [1] Why not at this point a conclusion that is not from fairy tale land and actually makes sense?: Rivets came and were used BEFORE screws, including the Egyptians 5000 years ago ... [2] Sorry, but that sounds like a small child who cannot judge time intervals. The year 1980 is 130 to 230 years AFTER the period we commonly refer to as the "Industrial Revolution". You can't really be serious about your claim about the rivets, can you?:) Further reading: Ken Shirriff's blog about the apple power supply, Wikipedia, "History of Rivets" or "A brief History of Rivets", etc.
@@markcummings150 So no citations or a source. Instead evasive maneuvers and nothing of that addresses any of my arguments or questions. So we can assume I got the important facts of my last comment right, you're okay with that, and we can safely ignore your nonsense. Thank you for the quick reply! P.S.: Maybe talk to someone at the company Loctite? To explain the purpose of what you call "paint" lies exactly how in my responsibility? Or to provide proofs for the claims of others? That isn't even funny ...
The capacitor plague around the year 2000 is a real thing. A lot of Chinese manufacturers got hold of parts of the blueprints for one of the top notch capacitor brands. Those Chinese manufacturers missed a few critical details though, which made the capacitors leak over a relatively short timespan. The flimsy capacitors ended up in a lot of devices back in the days, due to their low cost in comparison to the actual brands.
It started in around 1991, tbh, particularly with SMD electrolytic caps. Some stuff from 88-89 was also awful, but as the 90s marched on the component quality went straight to h*ll.
@@MagesGuild and it went on into the mid 2000’s. Especially with the United Cemicon KZG series which were used in a lot of motherboards from 2000-2006.
I do a fair amount of SMPS repair at work, and I tend to agree with you on the older caps being better. That being said, if they have drifted too far off value , or you have too much noise on the output, or are having weird glitches, replace those caps.
12:35 I bought solder stuff when my RadioShack was going, and a roll of solder wick was one of the things. I just thought wicks were all bad, but then I see people online using them fine, and then you say about quality wick and I am now going to upgrade. Thank you!
On another note, I have plans to try restoring DC-51 switches by adding an acid bath to the process. Don't worry, as I will be using switches that fail even with extreme use of DeOxit baths and mechanical hammering over hours. If this proves to be a success, I will post a video on it, and Datanatics switch restoration. If not, no harm done. I have a chem lab here, primarily for metals and plastics, and I regularly work with mild to strong acids, and have the right equipment to do this. In theory, unless the internal plates are warped, this may allow them to make proper contact again. The key issue is a design flaw and how the plates build up corrosion, but I am unsure on what plastics are used internally and their reaction to the acids. is On a positive note, I have about a full KB worth of dead DC-51s, and this would be a one switch per solution experiment, so I have a few dozen tries to restore totally dead DC-51s. If this works, I will share it with all of you, to keep these old machines alive and intact. If it fails, I will try to do a postmortem and explain why. I am not monetised, I do not do this for profit--retired--only for community beneficence. If you are either not suited/equipped with handling caustic solutions, are not familiar with chemistry, or generally do not know how to handle acids, or plastics, or lack the proper safety equipment, please do not tinker with these concepts. As a man who has had his lungs burnt by HCL fumes, accidentally, I urge you that this is very dangerous stuff.
As for drilling 'pop' rivets it seems natural to align the drill bit perpendicular to the rivet, as if drilling through the barrel of the rivet, but that just spins the rivet in it's hole; far better is to use a little oversized bit and begin tilting the drill so that is is NOT perpendicular. That way the bit will begin cutting away the 'head' and, at the same time, slowly rotate the rivet to cut all the way around, quickly and easily removing the 'head'. I discovered this many years ago by accident; try it, you'll like it!
Excellent as always. FWVLIW: I've tended to replace the guts of Apple II PSUs with modern Meanwell units that have most of the required voltages - Yes, the -5v needs to be added by soldering a regulator to one of the existing rails as the model that has all the other voltages doesn't have that, but it's not a big deal.* Why? Because they are more reliable and efficient than even a serviced original PSU. Also they have much more overhead for suppling upgrades. *The Apple II draws so little current on -5v you don't need to worry about bolting a heatsink to it, just find a suitable spot that won't lead to shorts. No, the new PSU won't line up with the original standoff locations but that's an easy fix - cut a robust piece of packaging plastic to size, screw the new PSU to that, and then screw the whole thing to the original standoff locations. Secure and will ensure no unintended shorts. My IIe Platinum is running an Apple2Pi adaptor of my own design that takes it's own 5v & GND supply from a modded and upgraded PSU with a Pi4 strapped to it running a custom install of my own making: Quite apart from the utility of Apple2Pi my computer does a darn good job of pretending to be a stupidly fast Stealth GS and runs RetroPie as well as giving me a functional Linux (boo!) machine. I modified a broken Apple II mouse for USB / optical / two buttons. All-in-all it's my go-to retro box. PS A solder wick tip: If you have crappy wick but a tin of solid flux paste, drag the wick over the paste. Now you have much better wick... or you could just buy decent stuff in the first place as Adrian did!
What a load of crap, those Astec PSU's have worked to 40 years.... with a little TLC and a recap they will last another 40 years. Meanwell PSU's do not have a 40 year track record so there is no evidence to support your reliability claim. If fact I'd really doubt they will last that long because we all know that modern devices are not built to last like the old school stuff was.
Must of been an interesting time to be a computer manufacturer. I'm sure at this point they had NO IDEA how long this stuff would be used. Much different than today.
@Adrian's Digital Basement ][ Please do not toss, or fail to return the Datanetics KB to the owner. That KB is a very special and hard to obtain item that proves a pedigree of 1979 or earlier to the system. They are incredibly tedious to repair or restore, but it is one of the few things I would be willing to come out of retirement to do for someone. Note also that you cannot reliably disassemble and functionally reassemble the DC-51 keyswitches, because of how they were designed and assembled. The best first option is to soak each switch in deoxit red, bang on them a hundred plus times, then soak each in deoxit fader lube (green) and repeat. If the matrix chip is faulty, r&r that first. They are still available, but growing scarcer. I have restored a few Datanetics KBs, and in the end, I always had to source spare DC-51s. Some switches simply cannot be repaired. They are also prone to splitting open on solder/desolder, (or even just with age/use), so I advise giving each one a tiny dab of super glue on its seams ere removing or installing it to prevent the thermal transfer splitting them open. In fact, I advise putting a tiny dab of super glue on the seam of good keys, so that they do not crack open during use, which is one of their failure points. Take it from someone who had their reset key crack in half at random: Seal those 45 year old plastic seams. Should that KB ever be for sale, I would buy it as-is for spares for my early machines. Note also that the keys themselves are entirely different to keys from a later ][ (or ][+). The keycaps are taller, have a different angle for the connection to the switch, are a slightly different colour, and a slightly different lettering style. All these little esoteric Apple ][ facts.
If you can't see the bottom of a radial cap then you don't know if it is good. The only comfort you should take in seeing that the leads haven't corroded is that you got there in time. I've only worked on two Apple II power supplies and I've already seen one cap that had crystals forming on the bottom that would have been a disaster eventually.
I go straight for a bigger drill bit for rivets - something about 90% of the diameter of the whole rivet head. Then lots of force downwards and the head is usually off in only a few revolutions...
That PSU has been swapped in the past. Note its serial number and style: A system from 1979 would either have the silver PSU used in base ][s until that point, or the early gold PSU with (I forget if it is a RIFA or a fuse) cover on the back. These are the types of PSUs used in the series: Type I: Silver, white label with red text , soft feeling switch, no fuse cover. 1977-79; possibly through 1980 to deplete stock. p/n A2M001. No ASTEC markings, possibly made by Apple internally, but I am unsure at present. Wire colours include the white with orange stripe and similar, with the Type II. (I strongly bel;ieve that this was TelCom wire, based on the colours.) Type II: Gold, hard press power switch, could have hand edited p/n 605-5701 (or with '1' crossed out and marked *in hand* as -5703 for very late 1979-80, maybe up to early 1981?). Wire colours are blue, white, orange, white with orange stripe, white with green stripe, black, black. Serial Number reset; mine is No. 001592. Has cover plate above three prong power connector to cover fuse or RIFA (IDR which). Used 1979?, 1980 to early 1981? (Note, this is from memory,, and I am old.) Type ?: Never seen one, but it is possible that a 605-5702 exists. The sheer fact that I have a 5701 with the 1 slashed out by felt tip and marked 5703 may indicate that 5702 was skipped for some reason, but I have no evidence for that. Type III: Later ASTEC gold PSU, no cover for fuse/RIFA; same switch as Type II, but wire colours changes to green, yellow, orange, blue, black, black. p/n printed as 605-5703. Used on 1981 and later ][+ machines, including RFI. Type IV: Used on the //e, returned to silver case, made by ASTEC, now has caution stickers about electricity. p/n 825-0510 (possibly -A or -B, etc.), 1982 to 1993. Type V: //gs power supply. Connector changed, and internals re-arranged; housing changed for lock-in elements instead of screws. Used 1986 to 1992. All of these are electrically compatible across the entire range from the ][ to the //gs. Another thing to note is that until 1979, and introduction of the ][+, PSU serial numbers tend to track very closely to the computer serial. My ][ has a PSU and a system serial both in the 5000 range; but they do not match up perfectly. That said, a 600K serial gap is a dead giveaway that the PSU was changed out in the past.
If Wikipedia isn't lying to me then the designer of the Apple II PSU was Rod Holt (Apple Employee #5) and is ~87 years old! I wonder if anyone has interviewed him about the design? 😀
Adrian, As for the rivets you can use a metal hole taper bit to cut the head off then use a punch and push the rivet through. Mason HP-2 Rivet kit sold by Snap-on, Mac (Mac-ko) and other tool truck sold a kit that had a special tapered removal bit. I have had my Snap-On version for 40 years and it has the drill bit. The newer version does not come with the bit no longer I guess. I scowered the Amazon, eBay, InterWebs in the middle of this video and apparently that bit is no longer available in the kit (or anywhere) but the case still has the molded area for the bit. Like I said mine is 40 years old... LOL I just love watching you work on stuff. I don't use any of the apple stuff but your tidbit of information is awesome...Thanks! LLAP 🖖
My Apple IIGS power supply didn’t work, I took it apart, and saw the fuse was blown, ordered a new one, and got it fixed. Good thing I noticed that before I ordered a power supply replacement kit
Ooh yeah, I hate those cable grommets! Definitely need to pick up a pair of those grommet pliers for myself. I still haven't got around to finishing off building the replacement for the power supply in my Apple II clone. I finished making a bracket to mount a TFX power supply, and I got the replacement power connector, but I still need to get a crimper since my current one doesn't have a section large enough for the contacts.
Good repair job Adrian. Given the amount of kit you work on Im suprised you dont leave a note in marker pen on the covers to say what parts were replaced and date you worked on them.
Oh actually not seen here I stuck a P-Touch label on the outside of the PSU with a date and RIFA's replaced -- as yes, good to make note otherwise I'll forget!
Some PSUs that my previous employer designed/built used these rivet-screws. In the factory we had a tool to remove them - a slender hexagonal shaft which you would insert into the central hole of the rivet. It would bite into the rivet body at six points and allow you to twist and pull the rivet out without any drilling or debris.
Great video. I'm not an expert on the safety caps but I think it should be pointed out that there are X and Y types that reflect the intended use and should be replaced in kind and definitely not with a standard cap. EDIT - meant to say the RIFA safety caps
When dealing with a rivet that small I would use a dremmel to grind the head off, then use a small punch to drive the rivet through. Or a dill bit bigger than the hole. And stop once the head is removed.
I’ve sometimes had decent results removing rivets using a heavy-duty set of flush nippers (the regular ones that you’d use to clip component leads aren’t strong enough).
angle the drill bit a bit if the rivet spins. It will gnaw away at it fairly quickly. The greater the angle, the slower the rivet will spin relative to the drill and the faster it will gnaw it away at increased risk of sliding off the head and damaging the surrounding material.
So I'm watching a plane restoration series where some rivets needed to be removed to remove part of the wing that was getting replaced. You know what the fancy tool was that they used to remove the multitudes of rivets from that wing? Exactly what you're using right there. They liked it just about as well as you seem to.
The rivet hole on the original power supply is 1/8”. Use a 3/16s or a 1/4th inch to remove the head of the pop rivet. Don’t try to drill it out vertically! Just lean the drill bit over enough to bite into the rivet “head” and chew it off. It’ll come off easily and you’ll have two pieces of the rivets. When you are done replacing the caps, put a new pop rivet into the hole on each side. No one needs to know you had to dig inside the power supply. BTW, yes, the Apple ][ was the first personal computer to use a switching power supply because it was cheaper than the equivalent linear supply. The Apple ][ is a power hog designed to support all 8 of the expansion slots. I know because at the height of my use of the computer I had all 8 slots full and it got REALLY hot. Reason why I added the fan!
Another reason what you say about caps is true is thus: electrolytes didn’t push as many limits back then, the electrons are packed-in less dense. So, to achieve high performance, you need very high purity and high surface-area. And that’s directly what increased the cost, as well as being responsible for their larger sizes. Essentially, it’s akin to how larger floppy disks are more reliable than smaller ones, as each sector is physically bigger. People who replace caps all the time marvel at how small modern caps are, but that’s also ultimately just a tighter margin for functionality. It’s no wonder plenty of HUGE filtering caps from the 30s-50s still work great with higher performance than modern ones, once you start looking at it that way. Of course they first have to not be punctured, or boiled, or burst.
Most of the apple2 PSU I have come across in the last few years seem to need a cap or two. Almost all of them have the filter cap problem, so I immediately replace it every time.
When I get back to work I am going to assemble a collection of retro computers. A Commodore 64 of course, an Apple II (probably a IIgs) and a 486 DOS PC. Those are all PC's that I've had or used in the 1980's and 1990's.
my opinion with an aging electronics always throw a UV light over circuit boards to rule out the leaking caps, whilst it is common for them blow up and out for a tell tale leak it is what happens when they fail from the leg end of the cap which might not be safe to keep in play.. with a sub 40 year old power supply I would err on the side of caution and do a proper and complete cap replacement since you haven't bothered to uv light it as when caps leak you may not see how bad a leak is..
The 2000s was a bad time for caps, but I think good modem ones will last. They started listing storage lifetimes for them now, as well as operational hours.
I was taught to hold the screwdriver in the center of the magnetizer and spin it. after I lost my original one I would just keep a hard drive magnet stuck to the bench leg and spin my driver tip on it.
i usually use a larger drill bit with a good conical tip. drill until the head separates from the shank. it usually leaves very little damage behind. also be sure to make sure you didnt get any metal in the electronics. clean everything thoroughly to avoid undesired shorts.
if you buy cheap caps, you get a cheap outcome. There are very high quality caps being made now, spend a couple extra dollars and don't cheap out. The electrolytic caps in those supplies are 40+ years old, they may be ok today but I guarantee they are going to bite you sooner rather than later, and you can't just tell by visually inspecting them.
Hey Adrian, could you someday discuss power supply safety? I am decidedly NOT an electrician and have avoided going into power supplies and monitors because of being spooked by voltages in capacitors. It's not that I don't know what I'm doing, theoretically at least, but the popping and arcing of discharging across caps (with a screwdriver, that is) makes me jump so bad that I can't stand even doing that, much less imagining getting bitten by one on accident. I also have an inordinate fear of thunderstorms, so it is just me. Thanks. I am considering looking around for an old Apple 2 to adopt and restore after watching your vids... we had one when I was a kid, but I never had the mentoring to understand the magic of the hardware or the software in these old machines.
Those Astec PSU's have bleed resistors across the mains side caps so they dont hold charge for long. If in doubt use your multimeter in DC volt mode measure between the exposed metal on the top of the largest caps, if none read more than a few volts you are good to go.
My only advice on rivets is to chock up on the drill bit so only a quarter inch of drill is exposed. I freaked out a bit when you ran half the length of that bit into the case!
Just a thought, maybe before connecting a motherboard to the PSU, connect a lamp to the 12V and/or 5V rails and let the PSU drive the lamps, then check the volts, just in case something else is wrong. That protects the motherboard from damage from PSU faults.
These ASTEC PSUs do not a load to be tested. You can test the output from the bench with basic equipment, so connecting it to a mainboard isn't required to check voltages. If you need to check it for current under load, that's when you hook it up, as this can indeed fall off, but I would check the +/- 12VDC and 5VDC rails beforehand.
With rivets Apple already had the first idea of how to design things in such a way that they cannot be repaired 😏 Today with component spare part whitelists in the firmware, it works perfectly.
The "better tool" for removing rivets is just a slightly larger bit so you can detach the flange. You're still going to hate doing it. :) It's what you ended up doing.
These power supplies are tough. I got one to play with as a young kid and I abused the hell out of it. I liked the spark and coil squeak that came with shorting it out lol. Probably did that thousands of times. Lid is full of plasma craters (pains me to write that, sorry little psu!). Always came right back up and asked for more. One day the RIFA did its thing (back in the 90s) and I thought I'd killed it - nope, just need to take it out. Can confirm the smell stuck in my mind.
The Magic smoke smell sticks with you I can tell you that. I smelled it a few times back in electronics class in high school. Years later the power supply in my main PC let out the smoke and I immediately recognized the smell and pulled the plug. The power supply didn’t make it, but the pc survived.
I do seem to recall a few ASTEC ones in my collection (1981 Made in England), I have others that are manufactured to BS415 by BSR (UK). They are also open frame (they use the computer case as a lid) with a captured mains-in cable. They are slightly smaller in length than an Apple PSU and it also has a voltage out connector to power all the peripherals that came out for the BBC. Rated for 220~240V @ 50 or 60Hz there is provision on the PCB to change the input voltage to 115V. Also the mains cables are connected via easy to remove spade terminals with the earth connection screwed in place, a slightly more expensive design choice.
That "little cap" is rarely the reason a PSU wont start up, it is more often that little feedback transformer has an open circuit winding. The small cap issue is more common with dynacomp PSU's not the Astec design.
I have an Apple ][ from 1977 (Sn 2964). The 12 volt rail in the power supply is fine but the system doesn’t boot. No biggie as I’m sure a little diagnostics on the PS and the mother board. I bought this without the case (because they weren’t available yet) and learned it’s capabilities as a naked computer leaned up against my bedroom wall (actually a sailboat). Once I got the case about three months later I assembled it together (I had been running with two linear power supplies “borrowed” from work). I still have that system from when I bought it piece meal from the Byte Shop in Fort Lauderdale FL. Believe it or not, the original $555 I had to get a personal loan to buy it from my bank. Once I had the case and power supply it looked like every other Apple ][ until you pop the lid. I have the language card as you mentioned, the color block and some mods I can’t remember. I have a RGB board that worked with RGB monitors from CAT scanners at the time plus “clone” of Microsoft’s SoftPC that I wire wrapped and it worked just fine. I had to add a fan to the system to cool down all eight of the cards in the slots. I added it on the right side and milled slots that look factory which probably destroys the value of the computer. Right now I’m just trying to bring it up in it’s nearly original form without all the bells and whistles The power supply works so I suspect a good disassembly cleaning and reassembly will bring it right back. Of all of the computers I’ve owned over the years (I’m 66) I have a special place in my heart for the Apple ][. I was president of the the “Apple Computer Enjoyment Society (ACES)” in south Florida for many years and enjoyed that part of my life. Thank you for all of the work you are doing on this system. All I need now are some boot up disks and I’ll be good to go!
Very neat. I remember the time when you could buy the system w/o a case, then the solid side case for a very short time, ere the vented case was available. Have you documented this machine on the Applefritter forums? Wait, is this the one with the ventless case and fan? Insofar as disks, I still have DOS 3.2.x diskettes that I could copy for you if you want those, along with the DOS 3.3 Developer disk, which AFAIK is not different to the commercial DOS 3.3 disk, but I have never done a checksum on it. I do not believe that I have versions < 3.2, which is a 13-sector disk. I do not have a 13-sector card now--we upgraded everything to soft 16, and the Rev 0 mainboards to the 6-colour Rev 4-- but I can always use some Muffin tools to make it work. My ][ is serial 5000-ish. I do not have it memorised.
Removing rivets while using pressure from the bottom is OK, you just have to be sure you don't bend the metal the rivet is in. Too much pressure, and you could have bent the case of the power supply. I forget what I was working on, but I've bent aluminum cases (much weaker than steel) while doing exactly this type of thing.
I agree. The metal of the case shell is far softer than the rivet. Not only can you bend it, but you can also easily mar it. (It's gold anodised aluminium, in this specific case.)
If you're wandering why a lot of stuff from the early to mid 00's had really bad capacitors, read up on the "great capacitor plague". Apparently the story goes that "a scientist working in the Rubycon Corporation in Japan stole a mis-copied formula for capacitors' electrolytes. He then took the faulty formula to the Luminous Town Electric company in China, where he had previously been employed. In the same year, the scientist's staff left China, stealing again the mis-copied formula and moving to Taiwan, where they would have created their own company, producing capacitors and propagating even more of this faulty formula of capacitor electrolytes." It's a fascinating story.
There’s was an interesting article in Business Insider that makes reference to one Bob Holt as the Apple Engineer who designed the power supply… It’s an interesting read. EXCLUSIVE: Interview With Apple's First CEO Michael Scott Jay Yarow May 24, 2011, 11:57 AM I’d provide a link but it would seem that is forbidden.
Yep. In the UK in the 70s and 80s they were mostly made by a company called Heyco, and one would be often referred to as a "Heyco grommet" - the removal tool being a pair of "Heyco pliers". I have no idea if that was just a UK thing, or if Heyco were a multi-national.
Use a step bit for the rivets instead of twist bit. That way you can get only the head/flange and it won't have a chance of grabbing and pulling into the hole. What happened on your first one is you drilled out the mushroomed portion on the inside and not the flange on the outside.
In the UK the Apple II didn't sell very well because it was expensive and we already had our own home-grown equivalent in the BBC Micro, so I've only ever owned one - a non-enhanced IIe. It stopped working at one point, I had to replace a capacitor and a resistor in the PSU to fix it. No diagnosis required, it was a standard fault and I just had to look it up online. Of course, this was after the RIFA caps exploded and made the room stink for a couple of weeks. Somewhere I still have a Mac IIci with a power supply that won't switch on, as far as I'm aware it's a similar fault.
Oh the Mac IIci has a typical problem where the SMD caps ruin the section of the board that handles the soft-turn-on feature on it. So common issues is you turn it on and it'll turn right off, or you get nothing at all. One way to tell is jump start the PSU (similar to an ATX.) I actually have one IIci PSU where I replaced the guts with an ATX PSU and just use a little toggle switch on the back to turn the computer on/off manually.
Ehhhh, yes and no. The ][ didn't sell effectively because it was NTSC. I recall needing to set up an entire NTSC lab back in the day to work on US and Japanese stuff. There were PAL modded ][ and ][+ machines by 1979, and then the Europlus, which did sell relatively well, and let's not forget the ITT 2020. ;) The actual drawback in the UK was that other system cost a lot less, and software for them was primarily on inexpensive cassettes, rather than on diskettes. The strength of the Speccy was in price, not in power. The BBC Micro was not available until '81, so it isn't a properly fair comparison. It was also very heavily adopted by the educational system of the day, and was a rather bold fusion between companies that would later become adversarial in some form or another, but for a home system, it never was able to overshadow the Spectrum, Vic-20, or C64. That said, the gigantic initial investment cost of £400 versus over £800 for the Apple ][, was a major factor.
When I was 8 and my mom bought an Apple //e for home, I do remember seeing that brass box inside the case and the warning to touch it before touching anything else in the computer to discharge from static! :)
Indeed (although in the //e, it was silver), they added an electrocution hazard warning, but they also removed the rivet. ;) it was replaced with a fifth screw, thus ten total, rather than eight.
@@MagesGuild I have a //e myself and it's gold - and also my kid's memories confirm that. Any chance the European version was different? Or maybe a different revision? And yes, mine doesn't have rivets either.
@@tony359 Ah, yes. I have been discussing NTSC stuff, as even in the UK we used a lot of that, for the general Apple market was stateside, and if it worked on a machine in NTSC spec it also worked on a PAL system, but the inverse was not always true...thank you ITT. The Euro //e did have its own PS}U for ~240VAC, and some were gold anodised aluminium, but I believe that also changed to silver in the late years. Either that, or it was silver and changed back to gold. IDR which as I was in charge of making stuff for and on NTSC machines, and only testing it on PAL machines, and examining the PSUs wan't important at the time. I know what I know based on keeping (taking home) every blasted machine and accessory that the company for which I worked was simply going to toss into the bins.
@@tony359 If you have an NTSC machine with a gold PSU, I would love to see a photo of that, as I have never seen that unless someone repaired it and put an older ASTEC brick into it; unless perhaps the very early 'green key' run had those in 1983.
Get you a pair of horseshoe nail cutters also called end nippers. Get the Diamond brand, not the Hazzard fraught junk. They will cut the top right off the rivets in about a second.
Larger drills can be used (should be used) as long as the "web" fits in the hole. All you need to do is to cut the face of the rivet off of its shaft so the shaft can be pushed through the hole.
Why not? in 40 years of repairing them I've never seen an Apple2 PSU kill the motherboard. Then add the fact that all he did was replace the rifa caps and there is even less of it harming the motherboard.
Not that it really matters but I'm not sure this was the original power supply for this machine . They had sequential numbered power supplies around this time and my Apple II (A Rev 0 Serial number 5500) has PS number 5414 stamped on it, but mine also does have the rivets holding it closed.
Not counting RIFA filter caps, the 220 uf in the middle is only one that had failed on my various Apple II PSU. And that was obvious from the outside, the label fell off when I touched it.
The "glue" your finding on the screws is a security measure so that Apple can tell if someone's worked on the device. It's painted over the screw head and onto the PCB so that, if it moves, doesn't line up, or doesn't match at all, it becomes obvious that someone's worked on it.
Those were not seals so that Apple could tell if somebody had been into the power supply. By drilling the rivet out a technician could tell you been in there. If you notice the treated screws were near the power transformer. The real reason is the transformer sends out micro vibrations that will eventually loosen the screws. Same thing happens when capacitors are operated at high frequency. That's why you often see hot melt glue and other adhesives used around the base of capacitors in switching power supplies. Dampens the micro vibrations and prevents cracking of the solder around the capacitor leads resulting in a condition known as a witch's hat. Heat and not enough solder can also result in a witch's hat joint.
my IIe doesnt have any rifas, it has the filter caps in the iec socket like you mentioned so i changed it, i took the old one apart to see what was in it and its the disk shaped caps. but it seems more had the rifas on board than not.
Hi, not apple related but crt monitors.😅 I just purchased a Clover Electronics CCTV monitor that is NOS and was wondering what is the best way to send a video signal from a PC to it? It has two BNCs one is Hi assuming it’s high impedance and one 75 ohm. 2 RCAs and A + B inputs for audio that are also RCAs. Additionally, it’s got an S-Video input that did not work when sending a signal from a Playstation 3. I’m a regular viewer of your channel and have learned a lot about diagnosing CRTs and finding faulty components. However, upon close inspection it looks like everything is in good shape and no leaky components. Also the monitor has a Samsung A34KQV42X01 tube and OSD that doesn’t do anything when pressed only blue screen with lines that can be removed when adjusting the Focus and Screen knob’s internally. It also got a Voltage Adjustment knob that I didn’t mess with yet. Thank you! 🙏🏼
If measuring AC noise levels on a DC supply, it's generally a good idea to place a capacitor in series with the meter to block the DC voltage component. That helps ensure that the RMS measurement is correct, and may be essential for measuring low level noise on higher DC voltages. Also, very brave to carry out the basic voltage test with a motherboard. I would not have taken the risk, and would have used a few power resistors to lightly load the power supply and test it for basic functionality before connecting more expensive parts.
I was gthinking exactly the same ... could have paper-cliped it. I just put the meter on AC .... I think if RMS is done properly you shouldnt have issues. A cap in series makes sense if you want to be sure. Some meters are a bit confused when AC+DC is present. It depend on the meter and how RMS is some. Most of the cheapo ones is SW based RMS up to 1Khz. Probably more flexible, but poor bandwidth. And the one with 20Khz BW or more use the traditional heat to DC converters, probably decoupled as you mentioned
I had two ceramic disc capacitors blow right next to the main power switch. It was one of those power delivery stations that would sit under a monitor, with a power switch for each peripheral. My guess was a power spike caused the ceramic dis caps to over heat, and the plates shorted inside it. It was resetting breakers, and all the switches passed a continuity check. Crazypoart about it was the caps looked good under visual inspection. They sparked when the power was safely applied using another power strip.
Those were probably MOV's (metal oxide varistors). Power spikes will degrade them to the point they short. They are actually intended to short high voltage transients to ground or across the line so they don't make it to your computer and peripherals. They only work so many times then they are junk. Open up any cheapy "surge protector" and you'll find those cap looking MOV's. In your case they were probably there for surge protection. You could just cut them out and plug the unit into an external surge protector.
If they atre small and blue, those are hv parasitica caps, but if they are more in a form of disc and big (like a dime) are MOVs. MOVs are there to catch high voltage transitns and pop the fuse while protecting the PSU. You can operate it without them, but next time a high voltage spike is present, then it will go directly to the switching circutry and in many cases, to the controller chip (for those design where the controller chip live on the "hot" side"). This is why even when they are a POS, generic ATX with controller on the cold side are way more robust. But 80 plus ones .... are not that strong in terms of voltage transients (IMHO)
Regarding removing rivets, you want a big drill to drill the head away but not mark the metal up. In my experience a wide diameter drill works best. 😁
Yep just a big enough "bit"
Drill bit should match the diameter of the rivets head.
And spin slowly with steady pressure.
A flat headed "punch tool" can also be inserted in the hole once it is drilled and you can just lever the head of the rivet off without having to drill all the way through and risk damaging the metal.
While that would work, if the rivet is spinning, a small angle grinder can also work, to cut the head of the pop rivet off of the base, and/or give you a point onto which you can hold the bugger with a basic screwdriver while drilling.
I am not a fan of using pressure for this, as that is not often effective. A drill bit about 75% the size of a pop rivet surface dimple works; and if you want to be fancy, there are specific pop/blind rivet removal tools.
Hi not Apple related but fixed my first motherboard today which had a missing trace due to varta corrosion. One bodge wire and its all good - thanks Adrian for the advice and giving people the confidence to repair these things.
That's awesome! I agree with ya, Adrian is doing some awesome work showing folks how to resurrect old tech!
Definitely a larger bit helps. Almost but not quite the diameter of the head. The main thing is to drill the head off the stem, then just use a pliers to pull off the stem from the back after you separate the previously riveted parts.
Exactly - just a large enough drill bit to demolish the external face. The rest of the rivet may fall out inside, but you're typically drilling it out to open it, so just grab it from inside once you're in. Trying to be gentle and get enough of it drilled out through the hole so you can coax it out... maybe if you're drilling it for some other reason than to open it and you don't want a loose rivet floating around the internals of something, otherwise it's just a quick "zip" and you're done.
16:16 For viewers that are outside the US and have 230V, these Astec power supply could be easily converted from 110V to 230V by removing the black jumper wire next to the small transformer. It is labeled on the PCB which says "115V select".
Next time try a LARGE drill bit. The pointy bit will center on the inside and the larger portion will eat up the exterior flange of your rivet.
Oh, and you did on the 2nd rivet :D
Those pop-rivets were an early version of Apple's 'no right to repair' policy!
I’ve removed my comments due to the trolls. You can work out who that is.
No, they were an early 'cover our assets' measure to prevent electrocution by hobbyists used to poking around inside their machines. Later PSU revisions removed the rivets and put a gigantic 'high voltage' type caution sticker on the PSU top case.
Take it from a technician who slipped and got 50KV down me right arm, actually cooking me muscles while working on a Tempest vector machine, it is not fun. I was lucky, and I was not intending to be anywhere near those components. As I recall, someone walked in, asked me a question, I tried to sit up and answer them, then...bzzrrt.
@@markcummings150 "Yes they are a tamper indicator." With the confidence you wrote that ... where was that confidence when you decided to NOT show us the source of this remarkable conclusion? Then "those days" ... is 1980 for you the ancient past?[2] Are you joking? BTW "ancient": Riveted joints have been used in mechanical construction and jewelery making for thousands of years since ancient civilizations[1]. Not only were rivets (as well as metal drill bits) widely and freely available in the 1980's, but no one (including the people who certify at the FCC) would call these riveted joints tamper-proof. If you do research, you will find topics like "water proof", "cheap and tight connections", "perfect for sheet metal connections", but not tamper indication. Even your mentioned "blob of paint" would be more suitable for such an application if it contains something like a seal imprint.
Sorry about the rant (which is actually: FACTS), but the grin I had reading David's comment stuck in my throat after reading your comment.
[1] Why not at this point a conclusion that is not from fairy tale land and actually makes sense?: Rivets came and were used BEFORE screws, including the Egyptians 5000 years ago ...
[2] Sorry, but that sounds like a small child who cannot judge time intervals. The year 1980 is 130 to 230 years AFTER the period we commonly refer to as the "Industrial Revolution". You can't really be serious about your claim about the rivets, can you?:)
Further reading: Ken Shirriff's blog about the apple power supply, Wikipedia, "History of Rivets" or "A brief History of Rivets", etc.
@@markcummings150 So no citations or a source. Instead evasive maneuvers and nothing of that addresses any of my arguments or questions. So we can assume I got the important facts of my last comment right, you're okay with that, and we can safely ignore your nonsense. Thank you for the quick reply!
P.S.: Maybe talk to someone at the company Loctite? To explain the purpose of what you call "paint" lies exactly how in my responsibility? Or to provide proofs for the claims of others? That isn't even funny ...
From my time as an aircraft technician, use a drill the same size as the head. Once you break thru, the head will spin so you know the rivet is free.
The capacitor plague around the year 2000 is a real thing. A lot of Chinese manufacturers got hold of parts of the blueprints for one of the top notch capacitor brands. Those Chinese manufacturers missed a few critical details though, which made the capacitors leak over a relatively short timespan. The flimsy capacitors ended up in a lot of devices back in the days, due to their low cost in comparison to the actual brands.
It started in around 1991, tbh, particularly with SMD electrolytic caps. Some stuff from 88-89 was also awful, but as the 90s marched on the component quality went straight to h*ll.
@@MagesGuild and it went on into the mid 2000’s. Especially with the United Cemicon KZG series which were used in a lot of motherboards from 2000-2006.
I do a fair amount of SMPS repair at work, and I tend to agree with you on the older caps being better. That being said, if they have drifted too far off value , or you have too much noise on the output, or are having weird glitches, replace those caps.
when removing rivets, use a Brad point drill bit quite a bit larger than the hole the rivet is in.
Apple doesn't want you repairing ANY of their products ! That's what really grinds my gears about apple !
Hey this video gave me the confidence to change out the RIFA caps myself on my IIe power supply! Thanks for doing it so well.
12:35 I bought solder stuff when my RadioShack was going, and a roll of solder wick was one of the things. I just thought wicks were all bad, but then I see people online using them fine, and then you say about quality wick and I am now going to upgrade. Thank you!
On another note, I have plans to try restoring DC-51 switches by adding an acid bath to the process. Don't worry, as I will be using switches that fail even with extreme use of DeOxit baths and mechanical hammering over hours. If this proves to be a success, I will post a video on it, and Datanatics switch restoration. If not, no harm done.
I have a chem lab here, primarily for metals and plastics, and I regularly work with mild to strong acids, and have the right equipment to do this. In theory, unless the internal plates are warped, this may allow them to make proper contact again. The key issue is a design flaw and how the plates build up corrosion, but I am unsure on what plastics are used internally and their reaction to the acids. is
On a positive note, I have about a full KB worth of dead DC-51s, and this would be a one switch per solution experiment, so I have a few dozen tries to restore totally dead DC-51s.
If this works, I will share it with all of you, to keep these old machines alive and intact. If it fails, I will try to do a postmortem and explain why. I am not monetised, I do not do this for profit--retired--only for community beneficence.
If you are either not suited/equipped with handling caustic solutions, are not familiar with chemistry, or generally do not know how to handle acids, or plastics, or lack the proper safety equipment, please do not tinker with these concepts. As a man who has had his lungs burnt by HCL fumes, accidentally, I urge you that this is very dangerous stuff.
As for drilling 'pop' rivets it seems natural to align the drill bit perpendicular to the rivet, as if drilling through the barrel of the rivet, but that just spins the rivet in it's hole; far better is to use a little oversized bit and begin tilting the drill so that is is NOT perpendicular. That way the bit will begin cutting away the 'head' and, at the same time, slowly rotate the rivet to cut all the way around, quickly and easily removing the 'head'. I discovered this many years ago by accident; try it, you'll like it!
Excellent as always.
FWVLIW: I've tended to replace the guts of Apple II PSUs with modern Meanwell units that have most of the required voltages - Yes, the -5v needs to be added by soldering a regulator to one of the existing rails as the model that has all the other voltages doesn't have that, but it's not a big deal.*
Why? Because they are more reliable and efficient than even a serviced original PSU. Also they have much more overhead for suppling upgrades.
*The Apple II draws so little current on -5v you don't need to worry about bolting a heatsink to it, just find a suitable spot that won't lead to shorts.
No, the new PSU won't line up with the original standoff locations but that's an easy fix - cut a robust piece of packaging plastic to size, screw the new PSU to that, and then screw the whole thing to the original standoff locations. Secure and will ensure no unintended shorts.
My IIe Platinum is running an Apple2Pi adaptor of my own design that takes it's own 5v & GND supply from a modded and upgraded PSU with a Pi4 strapped to it running a custom install of my own making:
Quite apart from the utility of Apple2Pi my computer does a darn good job of pretending to be a stupidly fast Stealth GS and runs RetroPie as well as giving me a functional Linux (boo!) machine. I modified a broken Apple II mouse for USB / optical / two buttons. All-in-all it's my go-to retro box.
PS A solder wick tip:
If you have crappy wick but a tin of solid flux paste, drag the wick over the paste. Now you have much better wick... or you could just buy decent stuff in the first place as Adrian did!
What a load of crap, those Astec PSU's have worked to 40 years.... with a little TLC and a recap they will last another 40 years. Meanwell PSU's do not have a 40 year track record so there is no evidence to support your reliability claim. If fact I'd really doubt they will last that long because we all know that modern devices are not built to last like the old school stuff was.
Must of been an interesting time to be a computer manufacturer. I'm sure at this point they had NO IDEA how long this stuff would be used. Much different than today.
@Adrian's Digital Basement ][ Please do not toss, or fail to return the Datanetics KB to the owner. That KB is a very special and hard to obtain item that proves a pedigree of 1979 or earlier to the system. They are incredibly tedious to repair or restore, but it is one of the few things I would be willing to come out of retirement to do for someone.
Note also that you cannot reliably disassemble and functionally reassemble the DC-51 keyswitches, because of how they were designed and assembled. The best first option is to soak each switch in deoxit red, bang on them a hundred plus times, then soak each in deoxit fader lube (green) and repeat.
If the matrix chip is faulty, r&r that first. They are still available, but growing scarcer.
I have restored a few Datanetics KBs, and in the end, I always had to source spare DC-51s. Some switches simply cannot be repaired. They are also prone to splitting open on solder/desolder, (or even just with age/use), so I advise giving each one a tiny dab of super glue on its seams ere removing or installing it to prevent the thermal transfer splitting them open.
In fact, I advise putting a tiny dab of super glue on the seam of good keys, so that they do not crack open during use, which is one of their failure points. Take it from someone who had their reset key crack in half at random: Seal those 45 year old plastic seams.
Should that KB ever be for sale, I would buy it as-is for spares for my early machines.
Note also that the keys themselves are entirely different to keys from a later ][ (or ][+). The keycaps are taller, have a different angle for the connection to the switch, are a slightly different colour, and a slightly different lettering style. All these little esoteric Apple ][ facts.
If you can't see the bottom of a radial cap then you don't know if it is good. The only comfort you should take in seeing that the leads haven't corroded is that you got there in time. I've only worked on two Apple II power supplies and I've already seen one cap that had crystals forming on the bottom that would have been a disaster eventually.
I go straight for a bigger drill bit for rivets - something about 90% of the diameter of the whole rivet head. Then lots of force downwards and the head is usually off in only a few revolutions...
That PSU has been swapped in the past. Note its serial number and style: A system from 1979 would either have the silver PSU used in base ][s until that point, or the early gold PSU with (I forget if it is a RIFA or a fuse) cover on the back.
These are the types of PSUs used in the series:
Type I: Silver, white label with red text , soft feeling switch, no fuse cover. 1977-79; possibly through 1980 to deplete stock. p/n A2M001. No ASTEC markings, possibly made by Apple internally, but I am unsure at present. Wire colours include the white with orange stripe and similar, with the Type II. (I strongly bel;ieve that this was TelCom wire, based on the colours.)
Type II: Gold, hard press power switch, could have hand edited p/n 605-5701 (or with '1' crossed out and marked *in hand* as -5703 for very late 1979-80, maybe up to early 1981?). Wire colours are blue, white, orange, white with orange stripe, white with green stripe, black, black. Serial Number reset; mine is No. 001592. Has cover plate above three prong power connector to cover fuse or RIFA (IDR which).
Used 1979?, 1980 to early 1981? (Note, this is from memory,, and I am old.)
Type ?: Never seen one, but it is possible that a 605-5702 exists. The sheer fact that I have a 5701 with the 1 slashed out by felt tip and marked 5703 may indicate that 5702 was skipped for some reason, but I have no evidence for that.
Type III: Later ASTEC gold PSU, no cover for fuse/RIFA; same switch as Type II, but wire colours changes to green, yellow, orange, blue, black, black. p/n printed as 605-5703. Used on 1981 and later ][+ machines, including RFI.
Type IV: Used on the //e, returned to silver case, made by ASTEC, now has caution stickers about electricity. p/n 825-0510 (possibly -A or -B, etc.), 1982 to 1993.
Type V: //gs power supply. Connector changed, and internals re-arranged; housing changed for lock-in elements instead of screws. Used 1986 to 1992.
All of these are electrically compatible across the entire range from the ][ to the //gs.
Another thing to note is that until 1979, and introduction of the ][+, PSU serial numbers tend to track very closely to the computer serial. My ][ has a PSU and a system serial both in the 5000 range; but they do not match up perfectly. That said, a 600K serial gap is a dead giveaway that the PSU was changed out in the past.
11:52 "Fancy leg bending"
I struggle with those moves too 🤣🤣🤣😂😂😂
If Wikipedia isn't lying to me then the designer of the Apple II PSU was Rod Holt (Apple Employee #5) and is ~87 years old! I wonder if anyone has interviewed him about the design? 😀
I'd certainly like to know!
Adrian, As for the rivets you can use a metal hole taper bit to cut the head off then use a punch and push the rivet through. Mason HP-2 Rivet kit sold by Snap-on, Mac (Mac-ko) and other tool truck sold a kit that had a special tapered removal bit. I have had my Snap-On version for 40 years and it has the drill bit. The newer version does not come with the bit no longer I guess. I scowered the Amazon, eBay, InterWebs in the middle of this video and apparently that bit is no longer available in the kit (or anywhere) but the case still has the molded area for the bit. Like I said mine is 40 years old... LOL
I just love watching you work on stuff. I don't use any of the apple stuff but your tidbit of information is awesome...Thanks!
LLAP 🖖
Another nicely done repair and functional check! Testing post-repair is so important to continued health of the gear.
dont forget alot of Caps of the 2000s had the capacitor plague as it was dubbed, where alot of runs used water for caps instead of electrolytes
My Apple IIGS power supply didn’t work, I took it apart, and saw the fuse was blown, ordered a new one, and got it fixed. Good thing I noticed that before I ordered a power supply replacement kit
i was going to laugh when you going to turn it on but you did remember that reefer need be replaced lol awesome to remember that lol
Ooh yeah, I hate those cable grommets! Definitely need to pick up a pair of those grommet pliers for myself.
I still haven't got around to finishing off building the replacement for the power supply in my Apple II clone. I finished making a bracket to mount a TFX power supply, and I got the replacement power connector, but I still need to get a crimper since my current one doesn't have a section large enough for the contacts.
:shrug: The grommet used here is a rotate and slide style IIRC. I have never required a special took to remove or insert them.
This machinist says your rivet-removal technique is OK. Maybe not perfect, but it does the trick and doesn't damage the part significantly.🙂
Good repair job Adrian. Given the amount of kit you work on Im suprised you dont leave a note in marker pen on the covers to say what parts were replaced and date you worked on them.
Oh actually not seen here I stuck a P-Touch label on the outside of the PSU with a date and RIFA's replaced -- as yes, good to make note otherwise I'll forget!
@@adriansdigitalbasement2 2:56🙂
Stepped drill bit work great with rivets.
Some PSUs that my previous employer designed/built used these rivet-screws. In the factory we had a tool to remove them - a slender hexagonal shaft which you would insert into the central hole of the rivet. It would bite into the rivet body at six points and allow you to twist and pull the rivet out without any drilling or debris.
Great video. I'm not an expert on the safety caps but I think it should be pointed out that there are X and Y types that reflect the intended use and should be replaced in kind and definitely not with a standard cap. EDIT - meant to say the RIFA safety caps
Yes. NEVER use standard capacitors to replace X/Y rated capacitors.
When dealing with a rivet that small I would use a dremmel to grind the head off, then use a small punch to drive the rivet through. Or a dill bit bigger than the hole. And stop once the head is removed.
I’ve sometimes had decent results removing rivets using a heavy-duty set of flush nippers (the regular ones that you’d use to clip component leads aren’t strong enough).
angle the drill bit a bit if the rivet spins. It will gnaw away at it fairly quickly. The greater the angle, the slower the rivet will spin relative to the drill and the faster it will gnaw it away at increased risk of sliding off the head and damaging the surrounding material.
So I'm watching a plane restoration series where some rivets needed to be removed to remove part of the wing that was getting replaced. You know what the fancy tool was that they used to remove the multitudes of rivets from that wing? Exactly what you're using right there. They liked it just about as well as you seem to.
The rivet hole on the original power supply is 1/8”. Use a 3/16s or a 1/4th inch to remove the head of the pop rivet. Don’t try to drill it out vertically! Just lean the drill bit over enough to bite into the rivet “head” and chew it off. It’ll come off easily and you’ll have two pieces of the rivets. When you are done replacing the caps, put a new pop rivet into the hole on each side. No one needs to know you had to dig inside the power supply. BTW, yes, the Apple ][ was the first personal computer to use a switching power supply because it was cheaper than the equivalent linear supply. The Apple ][ is a power hog designed to support all 8 of the expansion slots. I know because at the height of my use of the computer I had all 8 slots full and it got REALLY hot. Reason why I added the fan!
Another reason what you say about caps is true is thus: electrolytes didn’t push as many limits back then, the electrons are packed-in less dense. So, to achieve high performance, you need very high purity and high surface-area. And that’s directly what increased the cost, as well as being responsible for their larger sizes.
Essentially, it’s akin to how larger floppy disks are more reliable than smaller ones, as each sector is physically bigger.
People who replace caps all the time marvel at how small modern caps are, but that’s also ultimately just a tighter margin for functionality. It’s no wonder plenty of HUGE filtering caps from the 30s-50s still work great with higher performance than modern ones, once you start looking at it that way. Of course they first have to not be punctured, or boiled, or burst.
Most of the apple2 PSU I have come across in the last few years seem to need a cap or two. Almost all of them have the filter cap problem, so I immediately replace it every time.
When I get back to work I am going to assemble a collection of retro computers. A Commodore 64 of course, an Apple II (probably a IIgs) and a 486 DOS PC. Those are all PC's that I've had or used in the 1980's and 1990's.
my opinion with an aging electronics always throw a UV light over circuit boards to rule out the leaking caps, whilst it is common for them blow up and out for a tell tale leak it is what happens when they fail from the leg end of the cap which might not be safe to keep in play..
with a sub 40 year old power supply I would err on the side of caution and do a proper and complete cap replacement since you haven't bothered to uv light it as when caps leak you may not see how bad a leak is..
"The surface is marred up a bit but what are you gonna do?"
Put tape around it lol
I never think of that until the deed is done either
The 2000s was a bad time for caps, but I think good modem ones will last. They started listing storage lifetimes for them now, as well as operational hours.
I was taught to hold the screwdriver in the center of the magnetizer and spin it.
after I lost my original one I would just keep a hard drive magnet stuck to the bench leg and spin my driver tip on it.
i usually use a larger drill bit with a good conical tip. drill until the head separates from the shank. it usually leaves very little damage behind. also be sure to make sure you didnt get any metal in the electronics. clean everything thoroughly to avoid undesired shorts.
very nice older stuff always works better
if you buy cheap caps, you get a cheap outcome. There are very high quality caps being made now, spend a couple extra dollars and don't cheap out. The electrolytic caps in those supplies are 40+ years old, they may be ok today but I guarantee they are going to bite you sooner rather than later, and you can't just tell by visually inspecting them.
Hey Adrian, could you someday discuss power supply safety? I am decidedly NOT an electrician and have avoided going into power supplies and monitors because of being spooked by voltages in capacitors. It's not that I don't know what I'm doing, theoretically at least, but the popping and arcing of discharging across caps (with a screwdriver, that is) makes me jump so bad that I can't stand even doing that, much less imagining getting bitten by one on accident. I also have an inordinate fear of thunderstorms, so it is just me. Thanks. I am considering looking around for an old Apple 2 to adopt and restore after watching your vids... we had one when I was a kid, but I never had the mentoring to understand the magic of the hardware or the software in these old machines.
Those Astec PSU's have bleed resistors across the mains side caps so they dont hold charge for long. If in doubt use your multimeter in DC volt mode measure between the exposed metal on the top of the largest caps, if none read more than a few volts you are good to go.
Heathkit used to call that little black grommetty thing a "Strain Relief", and that's what i believe they're still called ;)
My only advice on rivets is to chock up on the drill bit so only a quarter inch of drill is exposed. I freaked out a bit when you ran half the length of that bit into the case!
Nothing in that part of the case but air, so long as you shake out any metal shavings there wont be a problem.
Just a thought, maybe before connecting a motherboard to the PSU, connect a lamp to the 12V and/or 5V rails and let the PSU drive the lamps, then check the volts, just in case something else is wrong. That protects the motherboard from damage from PSU faults.
These ASTEC PSUs do not a load to be tested. You can test the output from the bench with basic equipment, so connecting it to a mainboard isn't required to check voltages. If you need to check it for current under load, that's when you hook it up, as this can indeed fall off, but I would check the +/- 12VDC and 5VDC rails beforehand.
First six or so minutes of this video...absolutely riveting!
Those old PSU were workhorses. Thanks for the repair vid.
Put a little button magnet on the stem of the screwdriver, easier than rubbing the screwdriver against a magnet.
Nice QC Pass sticker. 🙂
With rivets Apple already had the first idea of how to design things in such a way that they cannot be repaired 😏 Today with component spare part whitelists in the firmware, it works perfectly.
The "better tool" for removing rivets is just a slightly larger bit so you can detach the flange. You're still going to hate doing it. :)
It's what you ended up doing.
These power supplies are tough. I got one to play with as a young kid and I abused the hell out of it. I liked the spark and coil squeak that came with shorting it out lol. Probably did that thousands of times. Lid is full of plasma craters (pains me to write that, sorry little psu!). Always came right back up and asked for more. One day the RIFA did its thing (back in the 90s) and I thought I'd killed it - nope, just need to take it out. Can confirm the smell stuck in my mind.
The Magic smoke smell sticks with you I can tell you that. I smelled it a few times back in electronics class in high school. Years later the power supply in my main PC let out the smoke and I immediately recognized the smell and pulled the plug. The power supply didn’t make it, but the pc survived.
This Apple PSU actually looks somewhat similar in design to the BBC Micro power supply that ASTEC also made. Probably very close.
I do seem to recall a few ASTEC ones in my collection (1981 Made in England), I have others that are manufactured to BS415 by BSR (UK). They are also open frame (they use the computer case as a lid) with a captured mains-in cable. They are slightly smaller in length than an Apple PSU and it also has a voltage out connector to power all the peripherals that came out for the BBC. Rated for 220~240V @ 50 or 60Hz there is provision on the PCB to change the input voltage to 115V. Also the mains cables are connected via easy to remove spade terminals with the earth connection screwed in place, a slightly more expensive design choice.
That "little cap" is rarely the reason a PSU wont start up, it is more often that little feedback transformer has an open circuit winding. The small cap issue is more common with dynacomp PSU's not the Astec design.
I have an Apple ][ from 1977 (Sn 2964). The 12 volt rail in the power supply is fine but the system doesn’t boot. No biggie as I’m sure a little diagnostics on the PS and the mother board. I bought this without the case (because they weren’t available yet) and learned it’s capabilities as a naked computer leaned up against my bedroom wall (actually a sailboat). Once I got the case about three months later I assembled it together (I had been running with two linear power supplies “borrowed” from work). I still have that system from when I bought it piece meal from the Byte Shop in Fort Lauderdale FL. Believe it or not, the original $555 I had to get a personal loan to buy it from my bank. Once I had the case and power supply it looked like every other Apple ][ until you pop the lid. I have the language card as you mentioned, the color block and some mods I can’t remember. I have a RGB board that worked with RGB monitors from CAT scanners at the time plus “clone” of Microsoft’s SoftPC that I wire wrapped and it worked just fine. I had to add a fan to the system to cool down all eight of the cards in the slots. I added it on the right side and milled slots that look factory which probably destroys the value of the computer. Right now I’m just trying to bring it up in it’s nearly original form without all the bells and whistles The power supply works so I suspect a good disassembly cleaning and reassembly will bring it right back. Of all of the computers I’ve owned over the years (I’m 66) I have a special place in my heart for the Apple ][. I was president of the the “Apple Computer Enjoyment Society (ACES)” in south Florida for many years and enjoyed that part of my life. Thank you for all of the work you are doing on this system. All I need now are some boot up disks and I’ll be good to go!
Very neat. I remember the time when you could buy the system w/o a case, then the solid side case for a very short time, ere the vented case was available. Have you documented this machine on the Applefritter forums?
Wait, is this the one with the ventless case and fan?
Insofar as disks, I still have DOS 3.2.x diskettes that I could copy for you if you want those, along with the DOS 3.3 Developer disk, which AFAIK is not different to the commercial DOS 3.3 disk, but I have never done a checksum on it. I do not believe that I have versions < 3.2, which is a 13-sector disk.
I do not have a 13-sector card now--we upgraded everything to soft 16, and the Rev 0 mainboards to the 6-colour Rev 4-- but I can always use some Muffin tools to make it work.
My ][ is serial 5000-ish. I do not have it memorised.
Removing rivets while using pressure from the bottom is OK, you just have to be sure you don't bend the metal the rivet is in. Too much pressure, and you could have bent the case of the power supply. I forget what I was working on, but I've bent aluminum cases (much weaker than steel) while doing exactly this type of thing.
I agree. The metal of the case shell is far softer than the rivet. Not only can you bend it, but you can also easily mar it. (It's gold anodised aluminium, in this specific case.)
If you're wandering why a lot of stuff from the early to mid 00's had really bad capacitors, read up on the "great capacitor plague". Apparently the story goes that "a scientist working in the Rubycon Corporation in Japan stole a mis-copied formula for capacitors' electrolytes. He then took the faulty formula to the Luminous Town Electric company in China, where he had previously been employed. In the same year, the scientist's staff left China, stealing again the mis-copied formula and moving to Taiwan, where they would have created their own company, producing capacitors and propagating even more of this faulty formula of capacitor electrolytes." It's a fascinating story.
There’s was an interesting article in Business Insider that makes reference to one Bob Holt as the Apple Engineer who designed the power supply… It’s an interesting read.
EXCLUSIVE: Interview With Apple's First CEO Michael Scott
Jay Yarow
May 24, 2011, 11:57 AM
I’d provide a link but it would seem that is forbidden.
Nothing beats and Apple II power supply!
that thing you was taking it is called a cable cleat and they were cable cleat pliers you was using to get it out Adrian
Yep. In the UK in the 70s and 80s they were mostly made by a company called Heyco, and one would be often referred to as a "Heyco grommet" - the removal tool being a pair of "Heyco pliers". I have no idea if that was just a UK thing, or if Heyco were a multi-national.
19:27 - that reminded me of Marcus Brigstocke sample, I Don't Smoke The Reefer...
Use a step bit for the rivets instead of twist bit. That way you can get only the head/flange and it won't have a chance of grabbing and pulling into the hole. What happened on your first one is you drilled out the mushroomed portion on the inside and not the flange on the outside.
Or use a very large bit that will eat the top of the rivet before it touches the sheet metal.
Reefa Madness
I usually get a sharp chisel and cut the head off the rivet, can be done without and damage if you are careful
In the UK the Apple II didn't sell very well because it was expensive and we already had our own home-grown equivalent in the BBC Micro, so I've only ever owned one - a non-enhanced IIe. It stopped working at one point, I had to replace a capacitor and a resistor in the PSU to fix it. No diagnosis required, it was a standard fault and I just had to look it up online. Of course, this was after the RIFA caps exploded and made the room stink for a couple of weeks.
Somewhere I still have a Mac IIci with a power supply that won't switch on, as far as I'm aware it's a similar fault.
Oh the Mac IIci has a typical problem where the SMD caps ruin the section of the board that handles the soft-turn-on feature on it. So common issues is you turn it on and it'll turn right off, or you get nothing at all. One way to tell is jump start the PSU (similar to an ATX.) I actually have one IIci PSU where I replaced the guts with an ATX PSU and just use a little toggle switch on the back to turn the computer on/off manually.
Kind of sorry I sold my Apple II some 15 or so years back, OTOH I probably would have lost it in our house fire anyway.
Ehhhh, yes and no. The ][ didn't sell effectively because it was NTSC. I recall needing to set up an entire NTSC lab back in the day to work on US and Japanese stuff. There were PAL modded ][ and ][+ machines by 1979, and then the Europlus, which did sell relatively well, and let's not forget the ITT 2020. ;)
The actual drawback in the UK was that other system cost a lot less, and software for them was primarily on inexpensive cassettes, rather than on diskettes. The strength of the Speccy was in price, not in power.
The BBC Micro was not available until '81, so it isn't a properly fair comparison. It was also very heavily adopted by the educational system of the day, and was a rather bold fusion between companies that would later become adversarial in some form or another, but for a home system, it never was able to overshadow the Spectrum, Vic-20, or C64.
That said, the gigantic initial investment cost of £400 versus over £800 for the Apple ][, was a major factor.
When I was 8 and my mom bought an Apple //e for home, I do remember seeing that brass box inside the case and the warning to touch it before touching anything else in the computer to discharge from static! :)
Indeed (although in the //e, it was silver), they added an electrocution hazard warning, but they also removed the rivet. ;)
it was replaced with a fifth screw, thus ten total, rather than eight.
@@MagesGuild I have a //e myself and it's gold - and also my kid's memories confirm that. Any chance the European version was different? Or maybe a different revision? And yes, mine doesn't have rivets either.
@@tony359 Ah, yes. I have been discussing NTSC stuff, as even in the UK we used a lot of that, for the general Apple market was stateside, and if it worked on a machine in NTSC spec it also worked on a PAL system, but the inverse was not always true...thank you ITT.
The Euro //e did have its own PS}U for ~240VAC, and some were gold anodised aluminium, but I believe that also changed to silver in the late years. Either that, or it was silver and changed back to gold. IDR which as I was in charge of making stuff for and on NTSC machines, and only testing it on PAL machines, and examining the PSUs wan't important at the time.
I know what I know based on keeping (taking home) every blasted machine and accessory that the company for which I worked was simply going to toss into the bins.
@@tony359 If you have an NTSC machine with a gold PSU, I would love to see a photo of that, as I have never seen that unless someone repaired it and put an older ASTEC brick into it; unless perhaps the very early 'green key' run had those in 1983.
@@MagesGuild mine is PAL, I’m in the UK 🙂
This video was.... (clears throat)...
RIVETING!!
Get you a pair of horseshoe nail cutters also called end nippers. Get the Diamond brand, not the Hazzard fraught junk. They will cut the top right off the rivets in about a second.
Larger drills can be used (should be used) as long as the "web" fits in the hole. All you need to do is to cut the face of the rivet off of its shaft so the shaft can be pushed through the hole.
I can't believe Adrian tested it on a MB 🤯
Why not? in 40 years of repairing them I've never seen an Apple2 PSU kill the motherboard. Then add the fact that all he did was replace the rifa caps and there is even less of it harming the motherboard.
@@kazdean
Because there are less risky methods❕
Adrian now often uses a light bulb❕
@@electron-1979from your other comment is clear to me that you dont know what you are talking about.
To drill out a rivet, use a drill the same size as the rivet, turn speed on drill very slow, push gently, job done...
Not that it really matters but I'm not sure this was the original power supply for this machine . They had sequential numbered power supplies around this time and my Apple II (A Rev 0 Serial number 5500) has PS number 5414 stamped on it, but mine also does have the rivets holding it closed.
Rivets, nice warranty protection...
Not counting RIFA filter caps, the 220 uf in the middle is only one that had failed on my various Apple II PSU. And that was obvious from the outside, the label fell off when I touched it.
The "glue" your finding on the screws is a security measure so that Apple can tell if someone's worked on the device. It's painted over the screw head and onto the PCB so that, if it moves, doesn't line up, or doesn't match at all, it becomes obvious that someone's worked on it.
Those were not seals so that Apple could tell if somebody had been into the power supply. By drilling the rivet out a technician could tell you been in there. If you notice the treated screws were near the power transformer. The real reason is the transformer sends out micro vibrations that will eventually loosen the screws. Same thing happens when capacitors are operated at high frequency. That's why you often see hot melt glue and other adhesives used around the base of capacitors in switching power supplies. Dampens the micro vibrations and prevents cracking of the solder around the capacitor leads resulting in a condition known as a witch's hat. Heat and not enough solder can also result in a witch's hat joint.
A larger bit works better just slightly larger than the hole because it will just cut off the head and the rest can be just pushy through.
my IIe doesnt have any rifas, it has the filter caps in the iec socket like you mentioned so i changed it, i took the old one apart to see what was in it and its the disk shaped caps. but it seems more had the rifas on board than not.
You can't unscrew the screw right away "CHATTERBOX🤣🤣🤣🤣
Well there was a kind of capacitor plague in the early 2000s I remember ordering several that were essentially no good
Still funny to me that this channel is often more analog than digital :)
if you use a hammer and tap you can push the rivet through
If wick doesn't work great, or even if it's good just add a little flux and it'll be like a solder vac.
Hi Adrian, Great video! I have the 230V version, and want to convert it to 115V. Would you know where to get the needed cable jumper? Thx!
Didn't see you blow/vac out the case for shavings. Knowing your thoroughness you probably did. 😁
I remember opening the metal casing trying to fix one of these as a kid, and shocking the shit out of myself with the stored capacitor charge.
they have a bit designed to remove rivets, search for automotive bits..
Love the II 😊
Hi, not apple related but crt monitors.😅 I just purchased a Clover Electronics CCTV monitor that is NOS and was wondering what is the best way to send a video signal from a PC to it? It has two BNCs one is Hi assuming it’s high impedance and one 75 ohm. 2 RCAs and A + B inputs for audio that are also RCAs. Additionally, it’s got an S-Video input that did not work when sending a signal from a Playstation 3. I’m a regular viewer of your channel and have learned a lot about diagnosing CRTs and finding faulty components. However, upon close inspection it looks like everything is in good shape and no leaky components. Also the monitor has a Samsung A34KQV42X01 tube and OSD that doesn’t do anything when pressed only blue screen with lines that can be removed when adjusting the Focus and Screen knob’s internally. It also got a Voltage Adjustment knob that I didn’t mess with yet. Thank you! 🙏🏼
If measuring AC noise levels on a DC supply, it's generally a good idea to place a capacitor in series with the meter to block the DC voltage component. That helps ensure that the RMS measurement is correct, and may be essential for measuring low level noise on higher DC voltages. Also, very brave to carry out the basic voltage test with a motherboard. I would not have taken the risk, and would have used a few power resistors to lightly load the power supply and test it for basic functionality before connecting more expensive parts.
I was gthinking exactly the same ... could have paper-cliped it. I just put the meter on AC .... I think if RMS is done properly you shouldnt have issues. A cap in series makes sense if you want to be sure. Some meters are a bit confused when AC+DC is present. It depend on the meter and how RMS is some. Most of the cheapo ones is SW based RMS up to 1Khz. Probably more flexible, but poor bandwidth. And the one with 20Khz BW or more use the traditional heat to DC converters, probably decoupled as you mentioned
@@38911bytefree Exactly - some meters will get it right, some not, and in a few cases the meter could be damaged if the offset voltage is too high.
I had two ceramic disc capacitors blow right next to the main power switch. It was one of those power delivery stations that would sit under a monitor, with a power switch for each peripheral. My guess was a power spike caused the ceramic dis caps to over heat, and the plates shorted inside it. It was resetting breakers, and all the switches passed a continuity check.
Crazypoart about it was the caps looked good under visual inspection. They sparked when the power was safely applied using another power strip.
Those were probably MOV's (metal oxide varistors). Power spikes will degrade them to the point they short. They are actually intended to short high voltage transients to ground or across the line so they don't make it to your computer and peripherals. They only work so many times then they are junk. Open up any cheapy "surge protector" and you'll find those cap looking MOV's. In your case they were probably there for surge protection. You could just cut them out and plug the unit into an external surge protector.
You may find they were metal oxide varistors and they failed due to a power spike - which is what theyre for :)
If they atre small and blue, those are hv parasitica caps, but if they are more in a form of disc and big (like a dime) are MOVs. MOVs are there to catch high voltage transitns and pop the fuse while protecting the PSU. You can operate it without them, but next time a high voltage spike is present, then it will go directly to the switching circutry and in many cases, to the controller chip (for those design where the controller chip live on the "hot" side"). This is why even when they are a POS, generic ATX with controller on the cold side are way more robust. But 80 plus ones .... are not that strong in terms of voltage transients (IMHO)
It's a thing of beauty.
That smaller cap you pointed out is severely swollen and should be replaced.
Apple hired Rod Holt to design the power supply. Woz wasn’t interested and didn’t get that excited by analog circuits.