I recently bought an Apple /// (256k model), Monitor ///, Disk /// external drive, manuals, and software. All of the units are cosmetically in amazing condition (no yellowing, no keys missing, etc). However, the /// itself doesn't work. What's strange is when I first turned it on, the computer started perfectly (it said something on the screen, maybe "retry"? I can't remember). After turning it off and coming back to it only about five minutes later, I flipped the switch only to find a screen full of garbled, flashing pixels. The internal drive's light comes on for a split second before turning back off, and the speaker emits a faint buzzing noise that changes pitch slightly when the pixels on the screen do (about three times a second). Sometimes when I turn on the machine, the screen is dark and the speaker is quiet until I hold control-reset; other times it goes straight to the garbled screen and buzzing noise. I tested the /// with another monitor I knew was working correctly, and the same flashing pattern was displayed.
I then set out to try to fix the problem. I took off the top case and found a heavily corroded battery pack clamped to the case (for the clock add-on). I removed the pack and disconnected it from the motherboard, since I wasn't planning on using the clock anyways. I examined the power supply, which is in full working order. I then took out the motherboard and everything seemed fine at first (no corrosion from the batteries in the clock battery pack had touched the motherboard, luckily). No capacitors on the motherboard or power supply looked like they were bulging or exploded. However, I noticed after a while that some chips on the board (around 7-8 of them) had a sort of black tarnish-like material coating the legs which could be scraped off by a screwdriver. I then took the time and painstakingly removed every chip on the motherboard, cleaned the legs with an eraser (or, if it was a tarnished chip, an emery board), dusted off the socket and reseated them. I also cleaned and reseated all of the chips on the RAM daughterboard (and the daughterboard itself). After doing this, I plugged everything back in and tested the system. Nothing had changed since my last attempt, sadly.
I'm at a bit of a loss of what to do next here. I'd really like to get this /// up and running soon, so if anyone has any ideas of what's going on or how to fix it, please let me know!
Pictures available upon request.
2 of the most common mistakes while extracting the chips, cleaning them and reseating are:
while reinserting them by flipping one of the chips aroung 180 degrees and by that inserting a chip wrong side
around .... so first visual inspection before turning on power switch again should have been to examine
if chips are located in correct position in the socket.... ( by inspecting the notches or maring at the chip....
and second ehile reinserting the chip it´s a very common fault that one of the pins might have bent below the chip
and by that that pin has not slipped into the socket.... so inspecting the correct seating of each pin of the chip
to be really inserted into the socket and not bending and slipping below the chip is second task to carry out before powering up again....
in case you have missed this 2 important visual inspections ( maybe with use of additional magnifying glass )
a power up might damage a chip or at least lead to malfunction....
and a third problem is the correct replugging the RAMboard on top of the mainboard.... have the contact pins been cleaned carefully ?
The mistake explained ( garbaged pixels ) is very common to malfunction of the RAM....
the black covering of the pins was indicating a high grade of oxidated pins...
i do hope you performed the correct procedures of preventing static harm to the chips while cleaning...
specialy the RAM chips are very sensitive to static voltages...
( see wikipedia about infos to static volatge and harm protection against static voltage damage )
i guess another member will carry out this task of debugging, cause i had in former days an Apple III system....
but nowadays ( due to the fact that i have sold that system 5 years ago ) i´m rather more specialist to Apple II,
while other members still having an Apple III in their collection even now.... )
besides i would advise to get again running battery in the system... otherwise you will have to correct
system time and date after every startup of the computer within the SOS....
You might benefit from referencing the Apple III/III Plus Technical Procedure documents. They might contain some diagnostic procedures. They can be found at the following link:
Thanks for the reply, speedyG.
I've rechecked the board, and all of the chips are inserted correctly with all legs in their proper sockets. Yes, I did clean the RAM board's contacts and protect the chips from static while doing so.
I've done a little more reading about the problem and I agree that one or more of the RAM chips has probably failed. If this is the case, how would I know which one has failed? I'm away from home for a few days so I can't check, but do any of the Apple ][ series computers have the same RAM chips that I could swap out and test (I own a ][+, a couple IIes, and a IIc)?
As for the battery issue, I agree it's a hassle to reset the clock every time the computer starts up, so once I get it working, that'll be my next step in the repair process.
first of all :
one of the most complete online sources related to the Apple III is:
I´d recommend to download from there the shematics and the technotes.....
as far as i remember there have been 2 very different versions of the RAMBoard
availiable for the Apple III one Version for 5 Volt and one version for 12 Volt.
You should research the availiable material to identify the correct version you have in your computer.....
( also by use of the shematics and the technotes )
and you should scan the technotes for the common known problems with the RAMboards....
another thing to scan along the documents is the key-combo used for the internal diagnostics....
as far as i remember, there was a keycombination availiable at the startup of the Apple III that invoked
a simple memory test to identify faulty chips.... This should be mentioned either in the User manual or in one of the technotes.... that should be first point to start after identifying the kind of RAMboard you have....
and by identifying the kind of RAMboard on your computer you might be able to detect which kind of RAM is used at your RAM board and decide if that computer is able to run with only one row of RAM at the board present...
( i.e. if that computer has 512 kB it should be able to startup with only 256 kB present )
BUT KEEP CARE ! because of different kinds of voltages used at 4116 and 4164 chips
you shall NEVER mix chips at the RAM board !
If there are 4116 then only use alternate other 4116 and
if the RAMboard contains 4164 chips then of course only alternate 4164 chips may be used...
anyhow the important thing to know if i remember correct :The Apple III requires minimum of 256 kB to bootup SOS correct.....
this will at least show within which limits the first tasks of searching the malfunction will have to be executed...
and if i remember correct also at the mentioned site there is a variety of system testing disks availiable....
that would be the next step after getting the system that far up to boot from disk....
I've looked over the board and here's what I found:
The memory board is a 5 volt model. The Apple /// is configured for 256k of RAM.
The memory board has two rows of 16 chips installed, for a total of 32 chips.
The top row of RAM chips are all labeled the same: MSM3764-20RS, and below that, 23106. They are black chips with an Apple logo on them. A quick web search identifies these chips as 64k dynamic RAM chips.
The bottom row of RAM chips are labeled as follows: they are black chips manufactured by NEC, and they have a small Apple logo on them as well. The top row of the chip has a number that follows the format 82**EK***, and below that, it says D4164C. A quick web search identifies these chips as dynamic RAM chips, and I assume they are 64k as well.
If I remove the RAM board altogether and try to boot the system, there is no more buzzing noise and the screen is just solid green (as expected).
If I remove the top row of RAM, there is no more buzzing noise and the screen is solid green.
If I remove the bottom row of RAM, the symptoms don't change (garbled screen, rhythmic buzzing).
My thought is that the system only needs the top row of RAM to function, and that one of the chips on the top row has gone faulty which is causing my issue. Does this seem like a reasonable idea?
I've heard before that the key combo for internal diagnostics is to hold down control and the open Apple key and press reset. However, this doesn't do anything different to the system; it just acts like control-reset was pressed. I'll keep scanning the documents and see if I can find it. Thanks for the source!
i just do now a few math:
8 chips of 4164 chips are a 64 kb Bank of a byte and 16 chips add up to 128 kB.
I do agree that at least one entire row must be present ( 128 kb ), but i maybe wrong
that 128 kB won´t be enogh for booting SOS from Disk ... anyhow i recommend to check this out by the documents availiable at the given source.... in the very early days i believe some few models have been sold with only 128 kB - but again ... i´m not sure about that....
I hope another member will overtake this task, who has more current experience with the Apple III - cause i just explained before: under normal circumstances i´m off the Apple III´s for several years now.... causing to forget things known in former days....
switching the chips from one row to the other is in normal case a pretty good chance to detect a bad RAM chip....
and if you get a startup that at least enables the "magic keystrokes" to invoke the internal RAMtest
you will be one giant leap ahead....
some of the former RAMdisk cards in the Apple II days also contained that kind of 4164 chips ( like the Saturn 128 kb Ramdisk card and several similar copy cats...)
the 4164 chips have also been used at several 80 ´cards of the IIe .... But you again must keep care:
some of those cards used 8 chips for a RAMbank ( which would be the correct ones )
and other cards used only 2 chips ( in general labeled as 4464 chips meaning that the chip had
instead of one bit databus a 4 bit databus ( which may not be used ! )
and of course you may not use chips labeled as 41256 !
the coding of -25 tell the chips to have 250 miliseconds timingframe....
so you should use also as replacement chips at least that fast
( or in other words -25 or -15 will fit but - 3 will be too slow.....
because Apple performed Chipstests in the factory before adding to them a small Apple Logo
it may be assumed that several manufacturers also delivered -25 chips that did not fullfill the needs of Apple....
therefor if not being able to test timing of the chips it´s strongly recommended
to use in case of replacement chips labeled as -15 equal to 150 nanoseconds timingframe
( which have been the fastest among the availiable sold bunch...)
if labeled -15 you may expect them to fullfill surely the demands of the chips labeled for the
lower speed setup as -25...
I tested with my A3 with just the one row of ram chips installed, in row D on the 5V ram board. It will boot ok using the sos1.3 system utils disk.
I also tested with no ram chips installed and I get a blank screen to start with, then once I press reset I get a white screen. (I have a color monitor connected). So this is similiar to what you see.
I would swap them over like Speedg suggests, put the other row of chips into row D and just run that row and see what happens.
The open-apple control reset takes you straight into the monitor bypassing the diagnostics. Without pressing the open apple, the diagnostics run, then the A3 will boot.
I tried swapping the chips over to row D from the other row, and the same problems persisted.
I then swapped each chip that was in row D into an Apple //e that uses the same RAM, one by one. The //e worked every single time without a hitch, so it appears (at first glance) that all of the chips that were in row D are indeed working correctly...
It's been a while, but I've discovered a couple more things that may be the cause of my ///'s malfunction.
I tested the voltages coming out the PSU the other day again, and noticed something strange. The 12V line is stable, but the 5V line was going nuts and jumping around. I looked around the PSU again, and noticed capacitor C10 was tipped slightly to its side. I desoldered it, and noticed that the bottom was bulging slightly. It's rated 10V, 220 µf, but after a testing with my capacitance meter, it read in at almost 350 µf! Yikes!
I've ordered a replacement, which should be here soon. Hopefully this will take care of the unstable 5V line, and possibly even resurrect the system. I'll keep this thread updated with what I find next (I know there are others with issues similar to mine, so hopefully this can be of some use to them).
If the capacitor has any leakage, it will not test properly for capacitance. The capacitance function on most multimeters are pretty useless for finding problems in switching power supplies. An ESR meter is very helpful in finding problem caps as is a tester that will measure leakage at the caps operating voltage. There are two types of leakage in a cap, one is the visible puss oozing out of it. The other not visible leakage is a resistance in parallel with the cap. An Apple III power supply will probably have several bad caps by now.
Heat and bad airflow ventilation was very common problem at all Apple III models
resulting also from use of very large wire/ceramic resistors limiting current to
the additional RAM board. And heat allways results to faster aging of electrolytic
capacitors by dryout or leakage.
So that probably won´t be the only cap to request for replacement....
I certainly agree that this won't be the only capacitor that will need replacing (I've already replaced two of the filter caps because they were cracking). After I check the voltages after replacing C10, I'm probably going to replace the rest soon just to be safe.
Can a modification be made to the /// to incorporate a fan by any chance (and has anyone ever done this)? I'd think that in a machine plagued by so many heat-related issues, someone would have come up with a mod for this by now...
unfortunatly i´ve sold my /// 4 years ago because i have limited my collection to be "II-series only"
few years ago.
But i can explain, what i had done with mine while using it:
Untighten all screws of the bottom plate and insert
at each screw one or two waschers of 2mm to 3 mm thickness
and screw back the plate to the casing. If done correct there will by a surrounding
airgap along the entire groundplate that permit air to enter the case.
Then try to get a cylindric ventilator and fit a case to the interior rear
of the Slot area.
I´ll later add a sketch drawing to explain better:
that will suck the hot air out of the case blowing it outside at the rearside.
By this method you will ensure that the entire inside of the case will get less or better cooling.
Some problems to solve:
find ventilators that operate at 12 volt.
fitting a rearblade that permits the air to be blown outside of the case and keeping the other
slots at the rear closed so that that hot air can´t turn back inside the case.
2 examples display solutions for PC..... schould be something similar like that....
find a ventilation system that is not too loud.....
maybe ( in my former Apple i didn´t have a regulation dependent to temperature ) it´s good idea
to add electronic regulation.... the target should be a complete airflow regulation....
The mystery continues...
My replacement capacitor came in today, and I soldered it in the circuit. Same capacitance, same voltage, double-checked polarity. I plugged it in, and the III displayed the same symptoms as before. When I turned on the supply when not plugged into the III (to measure voltages), the ground fault interruptor in the room I was working in went off, and I apparently also blew the fuse in the PSU (checked later, there was no path between the two ends). Even stranger, when I checked the capacitance of the other cap I ordered (I ordered 2 of the same kind), it read 320 µf instead of its rated 220 µf! Is there something I'm missing here?
Also, another thing to note: two components by the switch labeled L1 and L2 look a little brown on the inside (they're wrapped in what looks like a rubber covering). Should this be a cause of concern?
I can post some pictures of my board if that helps.
By the way, thanks so much for the fan info Speedy - *if* this computer ever works, I'll definitely look more into it!
Pictures from the supply may be rather more usefull at the moment.
description sounds to me like a shortcut in the primary section of the psu
recalling that the psu was disconnected from the board, when the test and incident happened.
Here's some pictures:
Hope these help!
If your 5 volt output is not stable, you should try to load test the supply to confirm it's working by itself. The switching supply needs to have a light load on it before you can test it. With the power supply disconnected from the motherboard, take an old and preferably noisy hard drive and wire the power connections on the drive to the +5, +12, and the common terminals of the power supply. This should allow the power supply to start-up. Take your meter and conform that you have the proper voltages on the output of the power supply. The next step will require you to connect additional load to the +5 volt and common terminals (I keep the hard drive connected so I can hear if it stops spinning). I use a 1.5 ohm 50 watt load resistor and make sure that the +5 volts output remains steady. If the supply starts chirping or you hear the hard drive stop, then you know you have a problem with the supply. If the 5 volt test passes, I do the same thing with the +12 volt output, but use a 5 ohm resistor. I usually don't load test the two negative supplies, but I do test them. If the testing confirms that the supply has a problem then you might want to go ahead and replace the remaining capacitors and verify that you don't have any shorted diodes or open resistors. If you find that the supply passed the test, but the 5 volt is not stable when the motherboard is reconnected, then you probably have a shorted chip or decoupling cap on the motherboard and that can be difficult to find. I have a small infrared thermometer pen that I probe each chip with as the shorted one will get much hotter than the others. I don't know your comfort level with all this, the power supply has lethal voltages that are exposed and easy to touch while the supply is open, so this may be a job for someone else. Was this sold to you as a working computer? Do you have any recourse with the seller?
thanks for the pictures.
At the moment ( i´m still fixed with another deadline task ) some brief and short reply:
I´d recommend to view the pages at:
at the first page the 3rd picture is the circuitplan of your supply, and it´s quite good idea
to get some info´s by reading about basics at that pages....
Second please also recognize the serious points mentioned in previous posting and at my pages.
For very short explenation the following picture that explains the Tooth and Tail side of the Tiger,
when attempting the supply ( "toothside" w. lethal voltages present that can bite - and "tailside"
with generaly non lethal voltages ):
Important remark: This powersupply makes links to sections by use of the case as "electrical wire" !
This means that the sections are linked at safty aspects to a kind of "physical grounding" by use of the
screwjoints and if that links are missing that may cause malfunction of the supply due to missing
links of that groundings as voltage reference ! This joints must have some kind of wired connection
for correct operation of the supply also while testing !
In this picture i marked up some common known issues. Check them out !
You can check this parts with powersupply not powered up:
... as a last resort, I could lend you a 128 k 5 V RAM board from my ///. The board is not 100% working, as it fails the Confidence RAM test. However, it is good enough to let the machine boot and run SOS.
If you think the effort would be worthwhile, do not hesitate to get in touch: "therealwozniak(at)gmail.com"
Thanks for your help!
Here's some things I checked:
The safety resistor looks fine, without any burns. I desoldered the four big caps in the primary section (250V, 100 µf), and noticed that under each, there's an oily film. I think that these caps may have leaked slightly, so I'm going to replace them, along with some others, and see where that gets me. Also, I don't know if I mentioned it before, but in the pictures, I've already replaced the capacitor at C10, along with the two blue filter caps.
Just to be sure I'm doing things right, when you replace a capacitor, the capacitances ideally should match, and the new voltage should equal or be greater than the old, right? I keep seeing references to tolerance, ripple current, etc. as well; do I need to take these into account when buying replacements?
Also, about the fuse: the board says 2.75A/125V, and that's what the old fuse was labeled. I'm assuming I should replace it with an identical fuse, not a 240V?
first the fuse: 125 Volt shall serve right..... while repairing the higher voltage often turns out to be usefull,
but it´s not a requirement..... you may stay well with 125 Volt.
About the capacitors: and kind of oily residue beneath indicates leakage and in such
case a replacement is strong recommended.
About the replacement values:
It´s recommended to replace with same values....
the problem with higher voltage limits: It permits the capacitor to load up
voltage to the limits and in several cases that might exceed the limitations of
the other parts in neighborhood area.... the next higher limitation at capacitors
after 250 Volt is 350 Volt. And tolerance of most capacitors is in general 20 %.
And if 250 Volt capacitors are "in line" the doubled voltage then won´t be 500 Volt but instead 700 Volt.
Some switching transistors have limitations to 250 Volt or 325 Volt.
The one in this powersupply isn´t labeled, so precise limits are unknown.
I'm not interested in a posting war, but your statement about the voltage rating on replacement capacitors is not correct. Good design practice is to use a capacitor with a voltage rating that is 2 to 3 times the voltage that would normally across the capacitor. The rating is a measure of the voltage that can safely applied across the capacitor before internal damage occurs. The only problem that you might encounter with capacitors with a higher voltage ratting is that they are typically physically larger and may not fit in the space you have to work with. The following is a good write-up on the subject:
Take note that a capacitor's voltage rating is not the voltage that the capacitor will charge up to, but only the maximum amount of voltage that a capacitor should be exposed to and can store safely. For the capacitor to charge up to the desired voltage, the circuit designer must design the circuit specificially for the capacitor to charge up to that voltage. A capacitor may have a 50-volt rating but it will not charge up to 50 volts unless it is fed 50 volts from a DC power source. The voltage rating is only the maximum voltage that a capacitor should be exposed to, not the voltage that the capacitor will charge up to. A capacitor will only charge to a specific voltage level if fed that level of voltage from a DC power source.
Keep in mind that a good rule for choosing the voltage ratings for capacitors is not to choose the exact voltage rating that the power supply will supply it. It is normally recommended to give a good amount of room when choosing the voltage rating of a capacitor. Meaning, if you want a capacitor to hold 25 volts, don't choose exactly a 25 volt-rated capacitor. Leave some room for a safety margin just in case the power supply voltage ever increased due to any reasons. If you measured the voltage of a 9V battery supply, you would notice that it reads above 9 volts when it's new and has full life. If you used an exact 9-volt rated capacitor, it would be exposed to a higher voltage than the maximum specified voltage (the voltage rating). Usually, in a case such as this, it shouldn't be a problem, but nevertheless, it's a good safety margin and engineering practice to do this. You can't really go wrong choosing a higher voltage-rated capacitor than the voltage that the power supply will supply it, but you can definitely go wrong choosing a lower voltage-rated capacitor than the voltage that it will be exposed to. If you charge up a capacitor with a lower voltage rating than the voltage that the power supply will supply it, you risk the chance of the capacitor exploding and becoming defective and unusable. So don't expose a capacitor to a higher voltage than its voltage rating. The voltage rating is the maximum voltage that a capacitor is meant to be exposed to and can store. Some say a good engineering practice is to choose a capacitor that has double the voltage rating than the power supply voltage you will use to charge it. So if a capacitor is going to be exposed to 25 volts, to be on the safe side, it's best to use a 50 volt-rated capacitor.
Also, note that the voltage rating of a capacitor is also referred to at times as the working voltage or maximum working voltage (of the capacitor). So when seeing the (maximum) working voltage specification on a datasheet, this value refers to the maximum continouous voltage that a capacitor can withstand without becoming damaged.
i won´t argue against the facts you explained.....
in fact for electronic design throughout the past 25 years in most applications your explenations are correct.....
and i even by myself keep according to the same behaviour....
but unfortunatly the explenations don´t fit in all cases....
ancient Apple powersupply design is one of the cases where you better not
followup such rules !
In fact at most models of Apple II and Apple III the design of the powersupplies are rather poor not to say even bad.....
It starts that the limits of the powersupplies have been made for empty not expanded systems and
you will face very fast limitations if you start expanding your system beyond the very basic expansion
of 2 or 3 expansioncards,
causing lots of heat problems an instable powersupllies running into their limits....
also the cooling is often a severe problem - at Apple III even desastrous because the design
of the case does not support realy good airflow and causes heat accumulation.... specially the Apple III is realy famous for such trouble.... in former days there even a bunch of jokes were passed by:
"whats the most expensive way to heat a room in wintertime.... using the Apple III as surrogate for the stove...."
it´s also a reason that the electrolytic capacitors age faster than usual due to massive heat exposure....
next problem is that Apple always performed strict policy of cheapest materialsupplier and if there
was a choice to select better device with better reserve limitation instead the cheaper device with
less reserve was selected.... This unfortuntly also affects electronic design of that supplies....
capacitors in that old powersupplies are often by design not only chosen for electric buffering of
electric voltages and capacities but also by wanted effects to operation frequency in conjunction
with transformers that have been designed for operation in very narrow limits ( so the capacity quite
often is not only related to filter away specific frequencies but sometimes even for generating
loopback frequencies as target to lower price of other parts.
yes in a specific way Woz even can be claimed to be one of the innovators of early switching powersupplies
with higher effeciency, but he unfortunatly didn´t care much about the later ongoing development in that
specific section of electronics and missed quite a lot of feedback from other engineers who specialized
to that topic....
the bunch of aspects i have treated up till yet in the pages i have made about powersupplies is only
very small view to that topic....
but please you may believe me, that "oversizing" capacitors with higher limits in old Apple powersupplies is only at very specific parts a good idea.... and i really also don´t want to startup here academic analysis with several dozends of pages limited to that topic only.... cause that would fullfill what i have mentioned in other threads as "hijacking a thread".... and i want to avoid such habits ( at least i don´t want to perform such hijacking )....
just like i expleined at the beginning of this posting.... to the most of your explenations i agree 100%
if related to general habit in electronics....