Extreme Screen Flickering Issues on Mac SE/30

[davidg5678]

I have a Macintosh SE that occasionally has a massive amount of screen artifacts. (See video below) Horizontal lines randomly run across the screen in no recognizable pattern. There are random flashes in all parts of the screen. The interface is semi-visible and the computer still works. Sometimes the computer boots with no issues at all and then eventually the problem appears after using the computer.  Other times I have turned on the computer, seen the artifacts, used it for a while and watched them disappear.

So far I have replaced the motherboard with a cleaned and recapped SE30 board, recapped the analog board, and installed a new fan. None of these solutions have changed the issue. This probably isn't motherboard related as it occurred with two separate boards (both SE and SE30 with almost no corrosion). I plan to recap the PSU but have not gotten around to it. There were no visible cracks in the analog board or video board solder joints.

Does anyone know what the issue may be?

Mac Screen Issue.MOV

 

[apm]

Whatever it is, I'm interested to find out because it's very similar to the lines I ended up with when I did some resolution changing experiments on the SE/30. When I drove the flyback at one frequency (22kHz) and the yoke at another (31kHz), I got horizontal lines like this, though more regularly spaced there.

Two things stand out:

(1) The lines are a change in brightness, not actually a change in the white/black video signal. Black pixels are still black, white pixels are white to varying shades.

(2) The brightness changes are accompanied by slight spatial distortion -- each line starts in a slightly different place depending on its brightness, just like the image in the thread above.

I'd guess it's from changes in one of the CRT neck voltages, other than the video signal itself. Try turning the Cutoff and Focus trimmers back and forth a little bit, in case one of them has a dirty or loose connection. Check that the solder joints on the connector from analog board to CRT board are all intact.

Then check the caps and solder joints on the CRT neck board itself (carefully -- the neck is fragile!). There are only ceramic caps on there, most of them .01uF, 1kV rating, but it's not impossible one could have failed and needs to be swapped.

 

[techknight]

did you resolder the analog board? 

Also, try heating the power supply caps and see if it clears up. any noise on the 12V sweep rail will do this too. 

 

[davidg5678]

I have adjusted the cutoff and focus potentiometers previously with no success; I don't think they are the issue. I tested the solder joints of the connector with my multimeter and they all seem to be working. The neck board solder joints look to be functional, but I don't know how to check the neck board capacitors. I have some basic multimeters but I am not sure how to do this. Can someone explain the process? I have partially resoldered the analog board but it can't hurt to do the entire thing. I will also reflow the neck board. Would it be better to replace all the power supply capacitors first or after reflowing the old ones? What makes this difficult is the fact that the issue is intermittent -I need lots of testing to figure out if my attempt works.

 

[apm]

If you haven't already done it, start by reflowing the connectors on the analog board to the PSU and the CRT neck board, even if they look okay. Better yet, remove the old solder and add some new solder in its place.

If you have another SE, you could swap power supplies to see if it changes anything. You might be able to see if there's noise on the 12V line by connecting your multimeter on AC (not DC) voltage mode. In DC mode, you should see +12V. In AC mode, you should see zero if the line is totally stable. I don't know what the normal level of fluctuation is, but I'd be suspicious of anything above about 0.2V.

Don't probe the CRT neck while the power is on unless you have specialised high-voltage probes. There can be voltages there of more than 500V. For the neck capacitors, I'd start by reflowing them and the wires to the connector, and then if nothing else works, replacing them with similarly rated (1kV) caps. But I'd guess a bad connection or a bad electrolytic cap is more likely than a failed ceramic cap, which to be honest I haven't come across (yet).

Edit: this is a simple one, but there's always the chance that the connection to the CRT itself is loose on one of the pins. Gently (!) removing and reattaching the neck board could reseat it.

 

[techknight]

if its intermittent, its likely bad solder joints. 

you cant "test" bad solder joints with a multimeter unless they are grossly broken and that would be plainly visible to the eye. throw that idea in the garbage and never look back to it ever again in all of eternity. 

You need to resolder the board with fresh solder, and then if that doesnt fix it (it should), you need freeze spray, and a soldering iron and start going after the transistors, capacitors, and ICs testing for thermal sensitivities and it can likely trigger the symptom. Especially if it goes away or gets started as things warm up. 

 

[davidg5678]

I have just finished redoing all of the solder joints on both the analog board and neck board.  All of the connections on the neck board are continuous. I will recap the PSU tomorrow; I hope it doesn't come to replacing ICs.

 

[davidg5678]

It turned out that my Macintosh came with an Astec branded PSU that uses non-standard capacitors and is incredibly difficult to take apart and recap. After finding and buying a very reasonably priced and working Sony CR-44 power supply on eBay, I will be able to rule out whether or not the PSU is the problem. I planned to recap my PSU anyway and this should make things much easier. I already have all of the required capacitors and the supply should come this Tuesday. I will provide an update around then. Thanks for all the excellent suggestions!

 

[davidg5678]

I have recapped the PSU and have not seen any screen issues so far. Although I have no idea where to find a replacement for the massive 400 Volt 620uf capacitor. The original Mac SE (not 30) chassis does not seem to want to close with the new CR-44 PSU installed. Is there any way I can modify the PSU or chassis to fix this? Is this a known problem with putting SE/30 parts into a normal Mac SE (not SuperDrive or dual floppy)? I have also noticed that there is also a very tight clearance between the power connector on the SE/30 motherboard and the video/power wire. I hope that replacing the power supply will finally end all of the issues I have been experiencing. Only time will tell.

 

[68krazy]

Check out my SE/30 thread to see a potential solution to the filter capacitor problem in the Sony CR-44 PSU:


Full Recap Thread

 

[davidg5678]

Thank your helpful link to the thread! After reading it, I realized I had made the exact same mistake in mounting my new fan; it was mounted too far forward and was the real reason the case would not close. Have you experienced any issues with the slightly different capacitor rating? Are there any risks that may result from not using the same capacitance? As long as it is safe for my SE/30, I will take a look at Digi-Key today and buy this (493-8392-ND) 420 Volt 680uf Capacitor. 

 

[68krazy]

I haven't had any issues, and we're a few months in with lots of hours and power cycles put on the machine since the recap. If you're really worried about it, it would probably be best to wait for the opinion of someone who understands electronics theory (techknight, et. al.). But 60uf is less than a 10% difference from the original, and I believe that's well within tolerance for a line filter capacitor.

 

[apm]

Using a larger capacitance in this situation should be completely fine, as long as the voltage rating is sufficient (which it is).