I've recently acquired a PAL IIe with Disk II card and extended 80 column card and for an hour after arrival it was working fine including a self test, writing floppies with ADTPro, etc - it then froze when running ADTPro and hasn't worked since. A list of what I have done so far:
1. Power cycle after initial freeze (didn't respond to soft reset) resulted in a white screen, no beep.
2. Removal of 80 column and Disk II card resulted in beep and command prompt, froze shortly afterwards. Reboot results in screen of garbage.
3. Removal of power for a bit gets me back to 2. for a short while (~10 seconds) - sometimes it will start the self test if I do it but will never complete and will freeze at a random point.
4. If I reinsert the expansion cards it will get me back to 2. again for a short while, but will get the white screen instead of garbage on reset.
The random behaviour coupled with it sort of recovering after a power off led me to suspect the power supply - the RIFA filter caps were replaced before I got it, but there was a faint smell of burning when I opened it up after the initial freeze. Nothing seems to be visibly burned/bulging/leaking except for dark spots next to the large resistor adjacent to the transformer on the low voltage side of the PSU. Voltages coming from the PSU are all fine according to the multimeter.
Before concentrating on the PSU I put a multimeter in continuity mode across a few capacitors with C9 and C10 registering a beep (not really sure what I'm doing here so that might be normal).
It all points to PSU to me but would welcome any input!
Your 2. and 3. lookes just like what my apple II board did just recently.
After I replaced 2 electrolytic capacitors on the main board it worked again.
The capacitors looked absolutely fine, but when measuring the resistance across them, I saw that one of them had a resistance close to 0 ohm. Almost a dead short.
Of course your problem could be a faulty capacitor in the psu.
A capacitor may look fine, but be faulty anyway.
Generally a solid freezeup points to a bad clock signal causing the CPU to halt.
If your screen image remains stable then something weird is going on after the clock signal is divided.
If the screen image flickers or blanks momentarily then the clock generator circuit itself is probably at fault.
Thanks for the responses - I couldn't leave it alone so I desoldered one leg of C9 and got no evidence of a short, and the short on C10 was also gone. I resoldered it and no beeping any more on either capacitor (no idea what was going on here - likely user error).
I plugged it in to test, immediately got a screen of beige coloured garbage (stable not flickering) and heard nasty electrical arcing sounds from the low voltage end of the PSU so quickly switched it off again. I'll inspect/replace the PSU before doing more testing.
I've rigged up an ATX PSU for testing and I'm getting the same results as with the dodgy PSU, so it looks like it isn't that:
IMG_0185D.jpeg
First thing to do is test ROM and RAM. Is this a later //e with the self-test in ROM?
It does have the self test in ROM (and it passed when I first got the machine) but it doesn't respond to any keyboard input at the moment - can I access it another way? I've had a prod of all the ICs and none of them are hot and reversed the position of all the RAM chips with no change so far.
If I remove the CPU and power on I get a screen with black and white vertical bars and the odd coloured square, and similar results if I remove some RAM chips (white screen with coloured squares).
"It does have the self test in ROM (and it passed when I first got the machine) but it doesn't respond to any keyboard input at the moment"
Yes. Unplugging the keyboard from the motherboard will automatically boot into self-test on power-up by default.
"If I remove the CPU and power on I get a screen with black and white vertical bars and the odd coloured square,"
That's normal as far as I know. Without the CPU the machine won't execute any code.
"if I remove some RAM chips (white screen with coloured squares)."
Generally speaking, that's expected behaviour. //e's will behave unexpectedly when missing RAM.
I would direct you to A2Online as they had a great troubleshooting page, but sadly it's offline at the moment. However, I did find this:
https://vintageapple.org/apple_ii/pdf/Apple_II_Plus_IIe_Troubleshooting_&_Repair_Guide_1984.pdf
Hopefully, it will be helpful.
Hm - there is no difference when powering on with the keyboard connected or not at the moment (it did boot into self test without the keyboard connnected a couple of days ago but crashed within a few seconds).
OK! Found the A2Online repaire guide (at least an archive of it)
https://www.google.com/url?sa=t&rct=j&q=&esrc=s&source=web&cd=&ved=2ahUKEwjN_tXCnYvwAhWEsp4KHf2ADysQFjAPegQIChAD&url=http%3A%2F%2Fwww.retrowiki.es%2Fdownload%2Ffile.php%3Fid%3D20004243%26sid%3D8d367fef35b7318937246636cf28702c&usg=AOvVaw2ZbhppPU3_igP_qVva3ECz
Following on from this I ordered some RAM from ebay and after some swapping about it's alive again (turns out one of my original chips and two of the replacements were bad). I'm now going to recap the PSU as I suspect its misbehaviour is due to leaky electrolytics, but would rather replace the guts with something new - I am looking at various options for this including the Reactive Micro one (but would rather spend less money!).
The memory likely didn't fail because of a flakey power supply. It failed because it's old and unreliable. And you discovered that when you found bad RAM amongst your purchased replacements.
If you want to spend less money then leave the power supply alone (except maybe the RIFA branded filter capacitor).
It's highly unlikely that the capacitors in the power supply are bad, especially if it's an Astec supply. Those capacitors are properly specified, high quality components that operate well within their temperature and voltage ratings.
I agree ref. the failing RAM not being related to the PSU. The filter caps have already been done but it was making popping noises and burning smells which I thought might be related to the electrolytics - having opened it up and removed a few caps they look fine with no leakage as you suggest. I think I will replace the innards of it with a Meanwell or similar.
UPDATE - I am an idiot and had clearly nudged the adjuster on R64. I turned this and everything is now OK.
Continuing the fun with this machine I have been messing about with it and all of the diagnostics on the Apple IIe diag disk pass, however I am getting some funky colour output (this is a PAL system). The black colour on a standard screen seems tinted red, and in graphics mode it's blue instead of black (it does this to a less noticable extent on my CRT also - Karateka's background is dark blue for example). Most colour issues I have found seem to be related to NTSC machines or involve a complete lack of colour, here everything seems fine except the black signal, so I am a bit confused.
IMG_0196D.jpeg
IMG_0197D.jpeg
What PAL display are you using? If it is newer than the mid 90s, this is expected. The Apple ][ 'PAL' colour spec is pretty far off from the proper PAL spec, and it worked on old analogue CRTs, but on anything newer, it likely will not.
The red and blue shaded pics above were on a mid-2000s LCD, the problem was still there but much less noticable on my mid-90s CRTs (garbled text image earlier in the thread). Adjusting the signal has resolved it on both.
I'll keep updating this thread as my tinkering continues in case it's useful to anyone. I sourced a flex-ATX PSU from an old PC, replaced the capacitors, and removed unnecessary cables so had quite a neat replacement PSU unit. This worked OK for a bit, then the red screen issue as seen above returned much worse than before - tweaking the pot brought it back into line, but when powered on the next day my IIe will only now run in black and white. The colour killer switch has never done anything on this machine - I cleaned the pots, fiddled with the switch, reverted to the other ATX PSU and nothing changed.
The swimming around the edge of graphics as typical with PAL IIes is completely missing (80-column text now readable!) so I think the TCA 650 demodulator chip has failed, with the previous colour changes symptomatic of it being on the way out. I've ordered a new one and will see what happens.
I've had another look at my motherboard while waiting for the new chip to arrive in order to see what's going on with the colour killer switch. As previously mentioned it hasn't done anything since I got it.
A multimeter across it doesn't achieve continuity on either position and resistance doesn't seem to change when the position changes. It seems to connect ground to a point on the PCB (C77) which on mine is linked with two bodge wires (factory? they look old) between it, a pin on the LS74 at position C15, and C117.
Is this normal for a PAL IIe (pic attached)?
IMG_0254D.jpeg
There is a blue component at the position of "OS1". I cannot recognize that.
Regards
Ralf
That's the colour switch on early boards (at least that's what googling has told me).
I got a new TCA 650 chip and installed it, which has restored colour to my IIe. The switch at OS2 still doesn't do anything, but there is a change in behaviour with the new chip which I guess is normal - with the old chip before it died, I had the "running water" effect and colour fringing over everything all the time, including white text, which made 80-column mode basically useless. Now the computer seems to sense when colour is needed, so I get nice clear black and white text and colour graphics, but there is a fraction of a second of black and white graphics when it switches modes (seen when moving between rooms in Prince of Persia).
I had a IIe in the late 80s/early 90s but it had an RGB monitor so the peculiarities of the composite output are new to me, but I though I would document it as there doesn't seem to be a lot of info out there about the PAL version of this computer. It seems to be fully functional again with the exception of the big ATX PSU hanging out of the back so I will see what breaks next.
For deep information about the Apple //e PAL video signal, see Noel's video.