Thursday, July 3, 2014

Day 1

Interestingly, none of the electrolytic caps on the power supply board were bulged or leaky: compare that to a supply from the turn of the century...
They should still probably be replaced, and while I might do that eventually, I'm more interested at the moment in leveling progress breadth-first, and getting a usable workstation going.

When initially powering the system on, as can be seen from the photos, I opted to run the power supply with no load first to make sure the levels were reasonable. Since the supply is designed to power the system board, it is not surprising the levels are a bit low with no load: once I put the system board back in (it plugs into a non-removable backplane in the front of the system case that is riveted into place, and provides the connector to which the power supply connects) the levels went up to spec, so I closed everything up and did the load test, of course forgetting to take pictures then.

There is a NORM/DIAG switch on the system board, which, when in DIAG position, runs diagnostics and outputs the result to the serial port A. If you press a key in time after the diagnostics complete, you get a monitor prompt, and can view/edit memory, CPU registers, set breakpoints, run additional diagnostics, attempt to boot from a given device, etc. I chose the "Compare" memory test and "loop forever, halt on error" and let it run through 50 passes while I caught up on some media consumption.

---

Using my TeleVideo 9065 serial terminal is quite annoying because some of the keys are very difficult to press, so you have to strike it quite hard for the keypress to be picked up. Of course, if I get a working CRT attached to the bwtwo, I can use the Sun 3's keyboard (which might be just as bad), so I thought, I'll clean both of these keyboards, and that will give me a more pleasant experiance later. Especially since, if I am intending to program for the bwtwo, having a comfortable-to-type-on serial terminal would be indispensable.

Well, opening the TeleVideo keyboard was no trouble: two phillips-head screws, and some plastic tabs (that I managed to avoid breaking), but the Sun 3 keyboard needed a longneck Torx T10. I have the Torx lineup in removable bits, but because of how the screwhole is recessed, that only let you get the bottom half of the screws out, so I ended up wasting my daylight obtaining the pack of Torx drivers at the hardware store (I was hoping to retr0brite the keyboards if I had been able to open them in time). Inside, the keyboard is attached to the bottom piece of plastic using T15 screws, so not even the same as the enclosure.

That reminds me of another point I didn't bring attention to yet: to remove the system board from the Sun 3/50 requires a hex key. The plastic exterior, and the power supply cover use Phillips. And the power supply fan uses slotted, so where's the consistency Sun? Slotted, Phillips, Hex, Torx? Pick one type, or two even. Oh well, now that I have the right tool it doesn't really matter I guess.

I spent more time than I'd like to admit searching various cardboard boxes, but I did manage to find an optical mousepad. It's Mouse Systems brand, just like the mouse (I didn't get one with this system, so this is just something I had packed away from other gear), so it ought to work, however it is scuffed up pretty badly, so it might be annoying to use. I also found a VGA plug from one of those SuperSocket 7 motherboards with onboard video, which had to plug in via a header since AT didn't provision for onboard video like ATX does. Nearby I also found a DE-9 male plug intended for use as a serial port, so I should be able to make the physical connections to the bwtwo and the VGA, leaving me to spend the rest of today's time reading about video signalling standards, and rustling up pinout diagrams.

For level conversion, it seems like a Schottky diode has a voltage drop in the range I need (>0.1V, but not enough to entirely block the signal) for the video signal pin, and so I dug one of those out from the uncountable array of cardboard boxes also.

Probably the way to proceed would be to only attach the green V- and V+ in the VGA connector initially, using the Schottky in the correct orientation to fix up the levels, and of course hook up the sync lines (which are TTL), and their ground, and see if I can get a picture on one of my VGA CRTs (of course selecting one that can do the desired mode). Supposing that works, I might fix up the resistance and try and drive all three RGB lines using the one signal for a pasty white like it would have been on the original monitor. Unless of course the green looks too cool to change.

More photos to come (not right away since its sleepy time by now).

No comments:

Post a Comment