I recently acquiired a complete Apple IIc system to add to my growing collection of 8-bit computers. A few years back, I had acquired a Apple IIe. I had read back then that I should not trust the old poewr supply in the IIe, so I sent it off to be updated. Now that I have the brick power supply that comes with the Apple IIc, I'm asking if the same recommendation applies to the IIc power supply.
Should I replace the Apple IIc power supply? I can get a new one form 8bitclassics for $26 plus tax and shipping.
You should not replace it if it is working, just like you should not have replaced the power supply of the Apple IIe if it was working. Simply recapping would have been enough.
I sent the Apple IIe PSU to ReActiveMicro to have it rebuilt per Henry's strong recommencation. I don't have the skills to do such things.
Thank you for your prompt reply!
The //c power brick is just an analog supply and if it's working you should leave it alone. The power converter inside the //c is very reliable and rarely needs any attention. So there's nothing really to do at this point.
Henry has a little bit of a vested interest in power supply rebuilding. I have replaced a number of power supply boards with the ones he sells, so it isn't like I'm against Henry... But for most //e power supplies that are working, the only thing I recommend doing every time is replacing the RIFA caps. Those are known to blow, and it is smelly and messy when they do. And sometimes it will take something else in the power supply out when it does. Most of the time it is just the fuse but not always. Anyway, RIFA cap replacement is easy and much cheaper than replacing the whole board. When I have the power supply open to replace the RIFA caps I normally look for any signs that any of the elecrolytic caps are bulging or leaking. If they show any signs then I will replace them. Otherwise I normally just leave them alone.
But opinions on these things vary...
I have several old computers here and their original power supply adapters etc, everytime I turn one of them on I expect it to blow up in my face, this includes everything from IBM PCs to Atari computers, most experts say that any electrical component that deals with high voltage and is decades old should not be used and or replaced.
I think if any of these old things are only used while in the company of an actual person who can act if something goes wrong is most likely safe, however leaving any of them on for long periods of time where none of it being monitored is basically asking for trouble.
Is a fire worth the risk for the price of replacing an old power supply is what it all comes down too.
A few months ago, I plugged in a power adapter for an Atari 1200XL, the thing is over 30 years old, did it get warm, no, it didn't blow up either, however would I leave this stuff on and leave the house and come back hours later F**K NO!
There was an interesting guy who sold replacement OEM power supplies and other accessories for the //c on eBay some years ago.
I believe he was a professor somewhere in New Mexico.
If you are interested in what he has to offer, let me know here and by PM (so I will be alerted by email) and I will track down his information.
If the power supply assemblies formerly were not a fire risk, and now are a fire risk, a scientific question to ask would be: in what way did they change so as to enable fires?
We can rule out defects in design or construction, since none of that changed. Either some part of the assembly providing safety escaped, or some foreign matter reducing safety got in, or one or more of the physical materials remained but changed in form.
I don't think anyone will say that any substantial part of the assembly escaped—although drying out of wet aluminum electrolytic caps could be considered escape of water vapor.
Foreign matter in the form of dust or insects could have got in. This is easily visible and (usually) easy to clean away.
The most interesting part of the problem is: what materials could have changed in their physical form?
Everyone is familiar with metal articles that corroded, so that is one suspect; however, it is not immediately obvious how corroded metal can start a fire. If the wires carrying line AC power were corroded to the point that only a few of the original strands conducted, they could overheat and potentially create an arc. This would be pretty easy to spot on a visual inspection.
Another common transformation happens inside the line EMI suppression capacitors if they are the metallized paper type (RIFA brand). Over time the outer case becomes crazed, then cracks form, and humidity leaking in causes the layers of paper to swell up and make short circuits. This can cause sparks and smoke, and it seems prudent not to leave equipment with these old RIFAs running unattended. In some cases a RIFA is connected to the line before the switch, which means it can burn even when the equipment is switched off. Even if you don't see any RIFAs, you may have a similar problem, since they were used inside Schaffner filter modules and possibly others.
As softwarejanitor said, removing these EMI filter capacitors is quite easy. You don't even need to fit replacements for the system to work: they just prevent the emission of interference.
There aren't any other materials that change over time in a way that supports an electrical fire. Sure, a power transistor or diode may reach the end of its life and blow up, but that doesn't start a class C fire either. Electrical fires need to have a conductive path that can carry a large current despite the destruction of whatever was originally there by flames. Circuit boards are designed with special attention to spacing (called clearance and creepage values) to make this "tracking" fire, as it is called, not be possible. Power supplies are usually single-sided for this reason: it halves the opportunity for tracking. They also commonly have cut-outs around primary circuits, which improves creepage. Finally, the provision and continued use of the proper type and rating of fuse will prevent fire from starting in case of a fault.
Stephen Buggie. He's still around. I have several of his iic supplies. They work well. And he is .. interesting. Expect many post cards about Gallup, NM, as well as other ephemera in your package. Kinda fun if anything.
As for Henry and his replacement supplies - I've replaced all my GS and //e supplies with the RM version.. Still have the originals of course, but I'm not always in the mood to do a recap when a drop-in will do. They work fine, but yes, so will a recap in many situations.
Thank you to everyone that replied. The general consensus is to not replace it. So that is what I'll do.
The additonal thing that prompted me to ask the question is that with the Apple IIc that I recentyl acquired, I also acquired several Commodore computers (VIC-20, C64, and C128) which came with a recommendation from many Commodore users to replace the brick power supplies because when they fail they inject higher voltage into the computers and cause damage. The brick design was part of my concern.
Over-voltage from a failing power supply unit certainly merits consideration.
Regulated power supplies designs are in two broad categories or types: Linear and Switching. They differ in how they transfer energy from the primary (line voltage) to secondary side.
A linear supply uses an iron step-down transformer to convert line voltage to a low AC voltage. It then uses a bridge rectifier to convert that to a DC voltage, some capacitors to smooth it out, and a linear regulator to lower it to the supply voltage needed by the equipment. There can also be several linear regulators in parallel to supply different supply rails.
One of the problems with linear regulators is that they waste more power as heat the more the connected equipment uses. Another is that the only thing between input and output voltages in most cases is the pass transistor, which can fail collector-emitter short. When it does, the smoothed DC voltage from the rectified AC is let through to the PSU output. This is the issue with all the Commodore power supplies.
The other category, Switching supplies, have a completly different design, with no iron transformer, instead having a lighter ferrite transformer in between primary and secondary stages. The full line voltage is rectified, producing up to 350VDC. That is then switched by a MOSFET at high frequency through the primary windings of the transformer, with several stepped-down ouput windings on the secondary side, where the low voltages are again rectified to DC. If the switching transistor fails short, the lighter ferrite transformer saturates and usually blows a fuse. The transformer isn't big enough to convert at 60 Hz, so the output voltages never rise above normal levels.
For a long time, linear supplies were preferred because they were simple to design and build and had lower noise. But all Apple computers (except the Apple I!) used switchers.
Just to add to that: the C64 PSU is literally the example for the cheapest and worst possible power supply design. It's bascially made up of the least possible number of components (a transformer, 4 diodes, a single voltage regulator and a capacitor). And it's lacking any failsafe protection. Its main component, the voltage regulator, is constantly driven close to its maximum rating. That's why the C64 power bricks got super hot. And, as robespierre explained, the typical failure mode of the voltage regulator is to go shortcircuit when it dies. This will instantly expose the machine to higher voltage and kill many components. So, yes, these are well-known to be terrible. Rightfully many "experts" recommend to avoid these - and use a modern replacement. The simple C64 PSU design has already killed countless machines.
The PSU of the Apple is completely different matter. It's switching PSU is not just a much more advanced and efficient design. It also has levels of failsafe protection. It has a "crowbar circuit" for overvoltage protection. This will instantly kill the output rail if overvoltage was detected. And this will also cause the primary side to shut off, so no more power is delivered. And it has overcurrent protection (shutting off the primary side when the output rails are undervoltage). It's also a much safer design when considering failure modes: when capacitors degrade, or the main switching transistor dies, then this PSU design will normally just become weak or stop working.
I know, it's an endless discussion concerning whether to replace or restore PSUs. But in any case, advice for the C64 PSUs is completely unrelated when considering the Apple II PSUs. It's almost like comparing a two-stroke carburettor engine with an (admittedly early design) of a FADEC controlled turbine engine. :)
My guess the ones for Atari computers were probably just as bad, although I haven't taken one apart to see the level of quality or lack there of. I have most Atari models on hand, most haven't been on in three decades plus.
The above about the Apple II switching power supply design is pretty much spot on. I've rarely ever seen one that ever produced enough over voltage to cause a problem. Worst was maybe almost 15V on the 12V outputs and almost 7V on the 5V ouputs. That was one power supply and I'd consider it an anomaly. Virtually all of them I've seen fail were significantly under voltage, less than 10V on the 12V outputs and less than 4V on the 5V outputs. The most often failure I've seen is no output at all. Sometimes they'll click click click and show brief blips of usually significantly low voltage. Not to say I've never seen a motherboard that I suspected a failing power supply might have damaged... but I think in general it's fairly rare.
Here is my Apple //c "portable" that I used at the Cannes Film Festival in 1988 with a 1200 baud modem, now powered by the Steven Buggie power cord, as it weighs so much less than the Apple brick. Also, when new SLA batteries are installed into the power box on the left, this works on the beach, too (with proper sun screen!).
Apple ::c Prairie Portable Power Pack.jpg