General Category > LisaList2
And so the Lisa adventure begins
rayarachelian:
--- Quote from: blusnowkitty on April 11, 2020, 06:42:26 pm ---The completely dead power supply I can usually get it to click a bit and power off instantly.
--- End quote ---
That's a classic symptom of bad caps right there.
--- Quote from: blusnowkitty on April 11, 2020, 06:42:26 pm ---In both cases it usually takes several hours for whatever's inside to allow me another power-on attempt. Very strange... But, can't do any real testing 'til I get everything recapped.
On another note, Ray, I blame you for getting me into Lisas in the first place. I was watching LisaEm's progress for a year or two prior to the very first working build of the emulator :D
--- End quote ---
Indeed, fix the caps, that's the first step.
Oh, and uh, you're welcome.
patrick:
Check the 5 V setting pot (R29, 500R) on the power supply board. If the slider loses contact with the cermet track, the output voltages will increase until they trip the crowbar thyristor. This can happen once after warming up (power supply simply switches off) or frequently (power supply makes clicking noises).
As mentioned above, the golden RIFA capacitors usually need to be replaced. Use X2 resp Y2 parts to comply with your local safety regulations. The power supply also works without these parts, but EMI will be above CE resp. FCC limits.
Of course, electrolytic capacitors can fail after some time, just like any other component. But with THT electrolytes from the 80s there is definitely no need to replace them prematurely. The quality of these parts is much better than "consumer grade" components you can buy today. If something needs to be replaced after all, use components from well-known manufacturers (like Nichicon, Panasonic, Nippon Chemicon, Elna...), buy them from a reliable dealer who can give the exact type designation (no Chinese mail order!), and check the specifications in the manufacturer's data sheet. In this way you will achieve a service life of about 10 - 20 years again.
(SMD electrolytic capacitors from the '90s are a different issue. Some series, like the ones used in the Mac II family, in Tektronix testgear or Sony Hifi equipment, tend to spill and must be replaced now!)
Elcaps from the '80s and before lose their dielectric oxide during storage. This results in an increased leakage current, but can be cured by running it at its rated voltage for some time. Parts from the '90s and later have oxidizing electrolytes. They do not degrade during storage, but if they leak damage to the PCB tracks will occur.
To force the power supply on, connect the ON signal to +5Vstby. On the PSU connector ON is pin X and +5Vstby is pin 20. For standalone operation you will need some load at +5 V and +12 V, an automotive headlight bulb works fine for the 5 V rail, and a 5 W lamp is sufficient for the 12 volts. To build a test rig, get a "JAMMA Arcade" board edge connector from ebay and cut it to the required length.
To force power-on from the Lisa I/O board, bend U9F pin 24 out of its socket (or remove the COPS421 completely) and connect U7F pin 11 to the standby voltage (U9F pin 9).
One electrolytic capacitor that frequently fails is C8 on the CPU board. If it has an increased leakage current, the reset generator hangs and with it your Lisa, she shows a random pattern on the screen. The capacitor can be replaced or, if you want to leave the board in its original state, it can be regenerated by connecting it to its rated voltage for a few hours with a 10k current limiting resistor.
blusnowkitty:
--- Quote from: patrick on April 12, 2020, 08:38:20 am ----snip-
--- End quote ---
I'll check it out and I'll give that thing a healthy dose of Deoxit once I get my capacitors in. And don't worry, I've specced out exact replacements for both the analog board and power supplies, all of which are high-temperature Nichicons. No generic capacitors here! I agree with your point on not replacing capacitors prematurely, but these are getting on for nearly 40 years old and with Jobs hating fans I'm sure they've had a pretty rough life... Probably guaranteed that they've had a rough life if that burned up trimmer coil is anything to go by.
Interestingly, my COP421 chip is actually soldered in place on the I/O board and not socketed. I have been able to test the power supplies as far as putting out the correct standby voltage and I can verify that voltage is getting to the COP421. During some additional probing last night I found that both PWRSW and ON in the schematic aren't making it to the COP421 at all so I'm guessing that's one reason why this thing is so hesitant to power on. Time for more patch wires!
Am I correct in thinking that the COP421 is pinned out like so?
--- Code: ---|--------------|
|24 13|
| COP421 |
|1 12|
|--------------|
--- End code ---
blusnowkitty:
Capacitors came in today. I recapped the analog board and one of my two power supplies. Plugged everything back in and... absolutely nothing. Case interlocks were bypassed; I made a quick modification to the IO board to bypass the COP421 and still nothing.Switching out for the one I haven't recapped but was able to power the system before, and it's also dead. Currently the only things connected are the CRT and its electronics, the I/O board, and the CPU board. RAM, LisaLite, and Sony 400k drive are all disconnected.
At this point I'm not sure if it's the I/O board or that both power supplies have failed simultaneously, but with as much corrosion and rework as I've had to do to the I/O board I'm leaning on the former. I've put in for a Sapient repro so I'll be fitting that with my recapped supply when it comes in and I'll see what happens.
patrick:
--- Quote from: blusnowkitty on April 15, 2020, 04:21:13 pm ---Capacitors came in today. I recapped the analog board and one of my two power supplies. Plugged everything back in and... absolutely nothing. Case interlocks were bypassed; I made a quick modification to the IO board to bypass the COP421 and still nothing.Switching out for the one I haven't recapped but was able to power the system before, and it's also dead.
--- End quote ---
Before spending money on replacement boards, make sure that the PSU will start when ON is applied and that it puts out the proper voltages.
Make a test fixture and check the power supplies alone before putting them into the Lisa. Connect automotive lightbulbs to the +5V and +12V rails to provide some load and visible indication. Connect +5VSEN (N) to the +5V rail. Apply power and check if +5VSTBY (20) is there. Then connect ON (X) to +5VSTBY and check the other voltages for amplitude and ripple. If excessive ripple is present, one or more capacitors need to be replaced.
If one voltage is shorted (e.g. by an broken rectifier) you will see the others starting up for a few milliseconds before shutting down.
Tap the voltage alignment pot (R29) inside the PSU and check if the voltages remain stable. If this causes spikes, replace the pot.
Corrosion usually affects the mainboard, too. It might be necessary to replace the I/O board connector, and of course to check the PCB tracks itself. Connectors might look good but have pins rotten off on the bottom. This is not visible until you desolder the connector.
Navigation
[0] Message Index
[#] Next page
[*] Previous page
Go to full version