https://www.righto.com/2012/02/apple-didnt-revolutionize-power.html
The new biography Steve Jobs contains a remarkable claim about the power supply of the Apple II and its designer Rod Holt:
Instead of a conventional linear power supply, Holt built one like those used in oscilloscopes. It switched the power on and off not sixty times per second, but thousands of times; this allowed it to store the power for far less time, and thus throw off less heat. "That switching power supply was as revolutionary as the Apple II logic board was," Jobs later said. "Rod doesn't get a lot of credit for this in the history books but he should. Every computer now uses switching power supplies, and they all rip off Rod Holt's design."
I found it amazing to think that computers now use power supplies based on the Apple II's design, so I did some investigation. It turns out that Apple's power supply was not revolutionary, either in the concept of using a switching power supply for computers or in the specific design of the power supply. Modern computer power supplies are totally different and do not rip off anything from Rod Holt's design. It turns out that Steve Jobs was making his customary claim that everyone is stealing Apple's revolutionary technology, entirely contrary to the facts.
The history of switching power supplies turns out to be pretty interesting. While most people view the power supply as a boring metal box, there's actually a lot of technological development behind it. There was, in fact, a revolution in power supplies in the late 1960s through the mid 1970s as switching power supplies took over from simple but inefficient linear power supplies, but this was a few years before the Apple II came out in 1977. The credit for this revolution should go to advances in semiconductor technology, specifically improvements in switching transistors, and then innovative ICs to control switching power supplies.
(Score: 2) by looorg on Friday March 22 2024, @12:22AM (5 children)
A 12 year old story? Still interesting. Question is if true why didn't Apple sue over it? Also less then a decade later they had swapped to using power supply units from Hitatchi and Sony or Aztec -- the once from Sony was the good once as I recall it.
(Score: 3, Informative) by JoeMerchant on Friday March 22 2024, @01:28PM
We tried to do a lot of things with analog ICs in the very early 1990s that were tantalizingly just out of reach. Unstable, noise floor too high, etc.
We revisited the ideas in the late 1990s and, with the newly available higher performing parts, the concepts were realized like child's play.
🌻🌻 [google.com]
(Score: 2) by owl on Friday March 22 2024, @02:55PM (2 children)
Ten Thousand [soylentnews.org]
(Score: 2) by janrinok on Friday March 22 2024, @03:52PM (1 child)
I am not interested in knowing who people are or where they live. My interest starts and stops at our servers.
(Score: 2) by owl on Friday March 22 2024, @09:57PM
That is very weird... Let's try that again.
Ten Thousand [xkcd.com]
(Score: 2) by driverless on Saturday March 23 2024, @11:58AM
I first heard this claim in, I think, Fire in the Valley, which was published in 1984, so it's been around for at least forty years. Found it a bit odd even then, the use of a SMPS wasn't game-changing, it was just cheaper and lighter than a more conventional linear supply.
(Score: -1, Flamebait) by Frosty Piss on Friday March 22 2024, @12:55AM (1 child)
More Apple Hate Shit. No one else was doing it, but since it was Apple, it was just hype. So much hate here.
(Score: 5, Informative) by sjames on Friday March 22 2024, @02:28AM
Nobody other than IBM, HP, DG, and DEC at least.
IBM even did it using vacuum tubes before transistors were a thing.
RTFA. It wounds like Apple's power supply was a decent one, but far from the first.
(Score: 2, Redundant) by Tork on Friday March 22 2024, @01:20AM (15 children)
I have a question: Was this the sorta thing where a bazillion companies moved relatively fast and Apple was first, or was it something where Apple did it first and the rest jumped in after seeing that it works? If it's the latter Apple certainly does deserve credit of some sort for taking the risk.
I have a general distrust of articles like this because Apple says "innovate" a lot and some people think they heard "invent". That means we get lots of articles in the spirit of taking Apple down a peg. Unfortunately that means some articles have about as much value as Apple's cringey marketing puffery. They did spin the ad-counter at the green site, though. 🙄
🏳️🌈 Proud Ally 🏳️🌈
(Score: 3, Interesting) by SomeGuy on Friday March 22 2024, @02:10AM (13 children)
The point of the article is they weren't the first. But I'd still go out on a limb and say that it was probably "revolutionary" just getting it in there for Apple, the Apple II, and users of small microcomputers, even if others were doing it first.
At the time, Apple could have gone with a linear power supply, and then what? It would have been bulkier and hotter, and possibly not quite as popular as competition adopted newer power supplies.
TFA didn't really list what other similar machines were using. I believe the Commodore Pet was using a linear power supply. I forget what the TRS-80 model 1 used, other than it had a large external power brick.
(Score: 5, Informative) by sjames on Friday March 22 2024, @02:34AM (12 children)
I don't know for sure about the PET, but the C64 used a standard linear IC. They tended to fail early because the whole thing was fully potted, heatsink and all. I know because I rember chiseling away at it to expose the wiring so I could solder in a replacement from Rat Shack.
(Score: 2) by RS3 on Friday March 22 2024, @04:37AM (9 children)
Oh, you're more determined than me. Not sure if you know this trick, but most potting stuff will rapidly deteriorate if you put it in boiling water. It'll get sort of brittle, crumbly. Gotta do it hot, chip away, boil it some more. It's no fun, but there's that big dose of satisfaction. :)
I didn't know they potted the C64 PS. Really dumb. I wonder why? Maybe to reduce shock hazard, being a home / consumer product, maybe they feared the things getting wet?
Or maybe it got too hot, so they potted it to keep the heat from burning people? :) (that was sarcasm for anyone who didn't get it...)
(Score: 4, Insightful) by owl on Friday March 22 2024, @05:08AM (7 children)
The normal reasons for potting are typically vibration resistance and/or environmental exposure resistance (i.e., exposure to rain/etc.).
But as a C64 was not typically intended for a high vibration environment, nor for use outdoors exposed to the elements, the potting was likely either 1) Commodore got a good deal from some sub-ontractor on PSU's, and the PSU's just happened to be potted or 2) they were trying to hide something (either poor engineering, or they or the supplier thought they had some 'secret sauce' that had to be kept hidden).
(Score: 0) by Anonymous Coward on Friday March 22 2024, @07:44AM (4 children)
Keep us experimenters out of it so they don't get sued by ignoramus with tools. If one did succeed, they have plausible deniability that they did make it difficult to kill yourself tampering with stuff ignoramuses shouldn't be messing with.
I don't know of anyone here I would classify as such, but I have seen my share of people who have the tools to do something, but not the wisdom to do it right. I am sure we have all seen it.
(Score: 2) by RS3 on Friday March 22 2024, @02:29PM
I get what you're saying, but many of us are often like explorers into the unknown. We don't know what we don't know. I'm fairly cautious by nature, but I've done a lot of things that could have resulted in great bodily harm. Some kids have to touch the hot stove before they (we?) truly learn what "hot" means. Obviously the hope is you learn from the experience.
(Score: 2) by owl on Friday March 22 2024, @02:49PM (2 children)
That's possible, and the C64 might just predate the plastering of "no user serviceable components inside, must be opened only by trained service personnel" stickers on everything in an attempt to remove said liability. So that could be another reason for potting it, the "liability sticker" craze had not started yet.
(Score: 3, Informative) by looorg on Friday March 22 2024, @06:12PM (1 child)
Could be but I don't know. It's sort of then like that cardboard foil they had inside to prevent interference but all it did was to create a heat trap for the components out of which none had heatsinks or a fan etc. A lot of those chips became real hot after a while.
Considering that the wedge PSU was all potted I don't see what a trained professional could do about it either. The trick about them and the super brick Amiga once was that you had to cut off the bottom plate that was fused plastic. Then you took the PSU and banged it a few times in the floor or some hard surface so the brick loosened and fell out or was removed or didn't stick to the covers anymore. Then you either put a new and better PSU in the old case, cause the Wedge case and super-brick case both look nice and are classics. Or you had to start to excavate the epoxy blob -- chisel and hammer and hope for enough cracks and so forth that you could eventually uncover it. I'm not sure how else you are going to do it, perhaps there is some chemicals you could dissolve the epoxy with but I never saw or had access to any such things.
(Score: 2) by ChrisMaple on Sunday March 24 2024, @04:04AM
Hot acetone will attack epoxy. I wouldn't use it myself or recommend its use, due to fire hazard. Also keep in mind that epoxy is one of the common plastics used for packaging semiconductors, so whatever chemically attacks the potting also attacks the semiconductor packages.
(Score: 2) by RS3 on Friday March 22 2024, @02:44PM (1 child)
My dad was an EE and I was always into electronics, from a very young age. I worked for several years before going to college. At my first (real) job one role was PC board "stuffing" and soldering. Then the boss had me use a buffing wheel (and "rouge" polishing compound) to buff all the printing off the ICs. I thought that was kind of dumb. I understand they wanted to protect their design, but frankly the design was a huge kluge.
It was a small electromechanical thing that was used in industrial / factories. Anyone could see and copy the mechanicals, and the electronics were functionally trivial. I was convinced that part of their strategy was to keep it way overly complicated to discourage copying, and to make people think the thing was worth the ridiculous price.
There were primarily 2 PC boards, one having 20 or so 14 and 16 pin ICs. The other had a few ICs, but also 120V stuff to energize a motor and a few solenoids. The fun part was: the boards were not keyed nor labeled. Customers would accidentally swap the 2 boards and 120V would blow holes in many of the ICs. Not repairable of course.
(Score: 2) by sjames on Friday March 22 2024, @04:31PM
It's amazing how often I have been able to guess some company's "sooper sekrit design" by naming the most common obvious way to do something.
When I hear sales types going on about secret designs, I mostly just flash back to the old Calgon commercials from the '70s "Ancient Chinese secret!"
(Score: 2) by sjames on Saturday March 23 2024, @10:20PM
It was the determination of a teen with more time than money and a computer that wouldn't turn on.
(Score: 3, Informative) by looorg on Friday March 22 2024, @12:28PM
As far as I can remember for the C64 it was always potted. The first and old one, the brown wedge shape, those were always potted. I don't know if it was for heat reasons or they needed something with a bit of weight to it. After all not everyone had a monitor of their own at the time so you perhaps used the family TV so there was a lot of connecting and disconnecting and perhaps moving things around. So nice to have something with some weight to it. I recall opening it up after it had failed and it had basically boiled itself, there was bubbles in the epoxy and black charred marks. I'm a bit vague on the last chain of the production if they had moved to the half brick PSU, I have seen C64s that doesn't have the wedge PSU then and it was unpotted, at that time I had mainly moved on to the Amiga. Not counting the various after market PSU:s that came about.
The other weirdness is that some Diskdrives had the PSU internal and then it wasn't potted. But if you had an external one it was potted.
It was a bit more inconsistent for the Amiga, the all in one models as the A1000/2000/3000/4000 all had internal PSU like any old desktop. But the initial big super heavy brick was potted. But not all of them, I think for the last produced A500+ with KS1.3 etc they had moved away from it and did not pot them. Or they just found some cases later on and used them to house the PSU and wasn't potting anymore. So it's usually a way of telling how old or when in the production cycle it was made, late light - early heavy. They did also pot the half brick PSU (brown and black coloured once) used for the A600/1200, but stopped with that to eventually.
My memory of it now is that the non-potted once just felt so flimsy. People preferred the heavy once so when you moved from 500 to 1200 you kept the 500 PSU if you could. But the light once you could just push them over and they were a bit of a pain. After all the old one could be placed on the floor or under the desk and you could power it on and off with your foot but the light one just moved around and you had to sort of trap it before you could flip the switch.
(Score: 2) by driverless on Saturday March 23 2024, @12:00PM
It wasn't just that, the C64 was built in the cheapest, nastiest way possible (Tramiel was notorious thoughout the entire industry for his cost-cutting), so it's not a good example of reliability. I remember a site I worked on at the time that received a shipment of C64s, seeing the pile of DOA out-of-the-box or failed shortly after first powering on units stacked in a corner for replacement.
(Score: 3, Interesting) by Mykl on Friday March 22 2024, @04:18AM
Agree. Look at portable music players and smartphones. Apple were far from the first to enter those markets, but they absolutely defined what the market looks like today.
(Score: 4, Informative) by bzipitidoo on Friday March 22 2024, @03:25AM (2 children)
There was another revolution in power supplies in the early 2000s, the 80plus project to make desktop computer power supplies at least 80% efficient. I got what at the time was a good power supply for a Pentium 4 computer that I built myself. On the box it even bragged about how efficient it was: 70%. More typical was the vicinity of 60%. Yeah, before any of the power from the wall outlet ever reaches the motherboard, 40% is already turned into heat and lost. So when these more efficient power supplies came on the market, I gave one a try. Replaced that 70% efficient one with an 82% efficient. Checked the numbers with a power meter, and they were right on the money. The desktop PC used 151W with the old power supply, 129W with the new.
But if you really want to save power, you should go for a low power computer. Laptops, for instance. Typically, those things use at most 30W. A Raspberry Pi is 6W.
(Score: 4, Interesting) by ledow on Friday March 22 2024, @08:52AM (1 child)
I am slowly building up a small solar setup in the UK (which, as you might imagine, isn't exactly blessed with sunshine).
It's separate from the rest of the house and as it grows, I put more load onto it and then expand again, on a very casual hobby basis.
In doing so, I dug out all the Raspberry Pis I had in the house (having been a tester for the early models).
It's now running a 19" rack of Raspberry Pis, one for each main job, with USB peripherals and hats, my router and other little bits and bobs. All stuff that's left on 24/7 and so would contribute quite a bit to the electricity bill.
It literally consumes less than my laptop, which itself is the only "main" computer in use and only on when I'm using it.
And it means that it can all run off solar / batteries and has done when I've had extended power cuts. The efficiency of such machines is amazing nowadays, and the processing power delivered is far in excess of what I actually need. Currently it's driving two DVB-T2 recording machines, a Plex, a Wiki, a Nextcloud instance, FlightAware, multiple SDR radios, HomeAssistant with some 50+ devices (all of them either battery or very low power themselves, and using the SDR to pick up even more), multiple cameras (via PoE), a VPN, the main internet connection, the Wifi, and storage (an SSD connected to each Pi).
It consumes about 80W, 100W at peak if I'm at home and watching things. And for that I have so many GHz, GB and TB respectively, that I have no need for anything else running in the house - there's spare capacity on everything to run even more services and I've not hit any temperature, power or shortages yet, just by spreading it out across a few Pis.
And the whole house - everything -only uses 7KWh a day on average, once you take into account cooking, washing, etc.
People see my setup and say "That must cost a fortune to run", and it doesn't cost anywhere near as much as their fancy TV or their sound system or even their laptop to run it all.
Tech has really come on. In my storage cupboard, I have an old 1.2GHz PC that I haven't turned on years, with a no-moving-parts heatpipe PSU that - for its time - was incredible. 270W. That one PC draws more than all the above draws, and isn't as powerful and the PSU not as efficient. Hell, even the drives in it would need separate power supplies to boot them up now to get the data off them if I wanted to do it via my Pis.
As someone who works in IT, people expect me to have huge rigs pulling stupendous amounts of power - and I have a rack full of things churning 24/7 generating zero noise (not one fan needed in all the above!), and a gaming laptop that runs literally any game I throw at it, plus a HTC Vive Pro VR, without issue. And put together they take less than some people's GPU alone.
Power draw, noise, efficiency, heat, ... these are all issues that modern power supplies and low-power devices make possible. I'm not even doing that in the most efficient way - the solar is charging batteries (2KWh, more than enough to get through the day for all the above) which are driving an inverter which then powers lot of individual power adaptors for everything. If I wanted, I could probably go entirely DC and cut some 20-40% of that out. And in fact even my PoE cameras are DC solar -> AC inverter -> 48V AC adaptor -> PoE Switch -> PoE Splitter -> 9V -> Camera because the models I use aren't directly PoE compatible. If I was bring 12V to everything, I could probably cut the losses in even this system in half.
But it's not worth it when it all runs "for free" all day long, and will ride out even an 8-hour complete outage with me happily gaming away online and watching my entertainment without issue. It actually got to the point where I slowed down my solar project because... well... it runs everything I need to and I don't want to get into converting the entire house electrics yet.
There are no bragging rights in having - as I've seen some people have - nearly 1KW in their gaming PC.
Even servers nowadays. In work, I have something like 6KW of servers running 24/7 and stupendous amounts of switching and PoE. But actually for my home projects and hobbies, I keep coming back to Pis.
In my old age, I'd much rather be able to power literally everything I ever want to use myself than plugging in KW of PCs, TVs, boxes, just to watch TV or browse the web.
(Score: 3, Informative) by JoeMerchant on Friday March 22 2024, @01:37PM
I put a Pi Pico W in the yard on solar power. I bought two 5W solar panels for the project box because, you know: tree shade, clouds, night-time, non-optimal angles during the day, battery charge/discharge inefficiencies, etc. Plus, two panels forms the whole roof-side, a single panel wouldn't have covered the project box completely.
The Pico is connected via WiFi to a router approximately 80 meters away, inside a (granted, wood frame) building. it runs a web-server 24-7 on the default power management settings, when I call for a page from it it responds within less than 2 seconds.
The solar panel to battery charge IC I bought doesn't quite handle the output from both panels in series, it was easiest to rewire it to have a single panel instead of putting them in parallel, so I went with that at first, figuring I would have to reconfigure in winter due to the shorter days... nope. The only problem that setup has had after nearly 2 full years of deployment is occasional over-temp conditions in the worst of summer heat, the box it is in bakes to up over 140F and sometimes that shuts down something in the system and it needs a reset. It runs from August through May without a single restart, on a single 5W solar panel charging a smallish battery.
🌻🌻 [google.com]
(Score: 4, Informative) by anubi on Friday March 22 2024, @07:33AM (5 children)
They used a switching power supply to derive both horizontal deflection and high voltage for the Cathode Ray Tube anode.
The horizontal output tube is driven as a switch. The resulting rectangular pulse is integrated by the highly inductive load of the transformer and deflection yoke to form the sawtooth current waveform necessary for the linear horizontal sweep, while the flyback pulse from the collapsing magnetic field developed enough high voltage DC to drive the CRT.
Even before that, car radios used a mechanical switcher to power the high plate voltages needed by vacuum tubes. It was known as a vibrator. It was notorious for it's EMI generation.
The idea of making AC out of DC so it can be easily transformed goes back to Nikola Tesla's days...
"Prove all things; hold fast that which is good." [KJV: I Thessalonians 5:21]
(Score: 2) by RS3 on Friday March 22 2024, @02:51PM (2 children)
Being an electronics enthusiast I remember reading about, and probably seeing somewhere, "motor-generator" sets that sometimes used an AC motor to drive a DC generator, or AC-DC (brush-type) motor to drive an AC generator, and other combinations.
(Score: 2) by kazzie on Friday March 22 2024, @05:32PM
Sound like it might have been the Ward Leonard system [wikipedia.org].
(Score: 2) by ChrisMaple on Sunday March 24 2024, @04:21AM
Motor-generators were also used to generate 400 Hz from 60 Hz for testing avionics, etc..
(Score: 2) by DannyB on Friday March 22 2024, @09:03PM (1 child)
As late as the late 1970s, in high school, I got my hands on one of those! And it actually worked.
It was a Motorola brand. Ran on 12 volts, negative ground. You could just barely hear the vibrator. As soon as the tubes warmed up the radio worked. It didn't even need any real antenna to get some stations, out in the middle of a small town nowhere.
I was simultaneously horrified by primitive tech but awed by the ingenuity of designing something like this and making it actually work so well, and built so ruggedly that it could survive beyond the lifetime of an automobile it was installed in.
Fact: We get heavier as we age due to more information in our heads. When no more will fit it accumulates as fat.
(Score: 2) by driverless on Saturday March 23 2024, @12:06PM
You can still get them now, or at least semiconductor-based equivalents, there's an entire (admittedly small) subculture of folks designing substitutes that replicate the required properties of the mechanical originals [siliconchip.com.au] while fitting inside the same enclosures.
(Score: 2) by VLM on Friday March 22 2024, @01:35PM (1 child)
Its fun to try and research actual "first" for switching power supplies.
In about five minutes work, this is as far as I've gotten. Back in the old days one way to get "a hundred or so" volts for a post-1922 car radio was a dynamotor which is a 12 volt electric motor (like your cooling fan) connected to a high voltage DC generator. Was popular for a long time in mobile radio because you could put a flywheel on the shaft LOL to smooth out AM transmission voice peaks (as AM was popular in mobile radio up until the 60s/70s) A lower power alternative that was popular for cars was a vibrator supply using a little moving reed that buzzed and a transformer. I'd argue that's a switching power supply, just using copper electrodes instead of silicon switching. Before 1944 tung-sol was selling the classic 0Z4 cold-cathode rectifier marketed specifically for this application because I had no difficulty finding an old Tung-Sol datasheet dated 1944 for that tube. Engineering data sheets greatly preceded modern microcontroller sensors LOL. Anyway, car radios had switching power supplies sometime between the first Chevy car radio in 1922 (why no celebration in 2022? Did I miss it?) and 1944 when Tung-Sol was making and marketing specialized rectifier tubes for the application. So the ceiling on switching power supplies would seem to be 1944 although I'm sure it's FAR earlier and I'm sure people were doing crazy motor speed control tricks back in the 1800s doing stuff with mercury vapor rectifiers and stuff like that.
(Score: 2) by RS3 on Friday March 22 2024, @03:08PM
Ah yes, the ignitron (also [britannica.com]https://en.wikipedia.org/wiki/Ignitron) [wikipedia.org] but not back into the 1800s.
Not enough time to research it- people certainly were messing with mercury vapor light in the 1800s, but I haven't the time to find out when they first figured out how to trigger the on state before the ignitron invention in the 1930s.
I saw a yt vid where someone restored a very old motor that had a movable brush assembly. You could move the assembly to vary the motor's speed, including reversing. At zero RPM the motor growled a good bit, as to be expected, so probably not good to keep it energized in the zero position. Point is they were figuring that stuff out in very ingenious ways.
I have a very old "Ediphone"- 1890 or so that has a speed-regulated motor. It has a simple flyweight mechanism with a spinning ring that pushes on a contact. If the weights fly out too far, they pull on the ring which pulls open the contacts and interrupts current to the motor. Yes, it sparks constantly, but that was totally normal and acceptable in those days. It's surprisingly stable / consistent. The motor is inside a thick metal box, so you barely hear it when running. Yes, you can open it and put your fingers in.
(Score: 2) by Rich on Friday March 22 2024, @02:02PM (2 children)
Rod also didn't arrive on a chopper, as portrayed in the Jobs movie, just because someone interpreted "he rode bikes" wrongly:
The switching supply helped with enough excess power to supply the expansion slots and it definitely saved cost on the disk drives. The Disk II margins brought in the money that made the Mac.
(Score: 3, Interesting) by DannyB on Friday March 22 2024, @09:12PM (1 child)
The original Mac power supply had barely a milliwatt of spare capacity. For such great software technology, I was surprised they couldn't have had just a teeny bit beefier power supply.
The TRS-80 had a large external power brick, with the key word being large (and hot) as someone else already pointed out above.
An obscure mini computer [trailing-edge.com] I used in college had two separate power supplies inside its cabinet. (I was privileged enough to have access to the computer room and could open the cabinet back door and look but not touch.) Each of the two power supplies was 5 volts, 150 amps. No multiple voltages. Just 5 volts. The main board was wire wrap with a lot of recognizable 7400 series chips and some others I did not recognize.
Fact: We get heavier as we age due to more information in our heads. When no more will fit it accumulates as fat.
(Score: 3, Insightful) by owl on Friday March 22 2024, @10:12PM
After 1983's $9,995 price [wikipedia.org] for the Lisa ($31,141.65 today scaled for inflation [1]), and the resulting lackluster sales that resulted from such a very high price, Apple was most likely trying to cost cut everywhere in the Mac to lower the price as much as possible, knowing that the lower the price, the more likely it will sell vs. competitors. And the PSU likely got cut to the bone along with everything else.
Note that this was the era of the IBM PC (released 1981 for $1,565, equivalent to $5,342.80 today) so while they likely felt they could get a price premium over the PC for having a GUI on their system, they likely also knew full well they could not get a 5.8X price premium over the IBM PC.
Inflation Calculator [usinflationcalculator.com]