malachykidd

I'll take two, if not too late.

The keyboard on this unit is a Mitsumi. I've fixed one Mitsumi previously: a 1200XL which-- I now realize-- went far more smoothly than normal. I'd read about the difficulties (particularly the silicone layer sticking to the PCB, especially around the screw holes), and I had some of that with the 1200XL, but that came off fairly easily with only one tear in the silicone near the connector. This 800 was an entirely different headache. I initially tried to get the whole thing off, starting at one corner and working it loose by stages, but it fought me ever step of the way, and I wound up with lots of tears around the screw holes. Finally, I resolved to just get the edge around the connector loose, since it seemed the connector was the only real problem and I didn't want to unnecessarily damage the membrane. I managed to do that, with more tearing at the screw holes flanking the connector, removed the adhesive strip and stuck-on adhesive, and cleaned the PCB connector with acetone followed by 91% IPA. Fortunately, the adhesive didn't stick to the mylar layer and none of the traces were torn (unlike on the 1200XL), I assume because the 800 used carbon-over-metal for the traces, and the adhesive couldn't stick to the carbon. Reassembled, loaded SALT, and tested everything fine. The only problem now is with the START key, which is a bit intermittent when pressed. I imagine contact cleaner would suffice to clean it, but I'm not going to touch it until 1) I've verified whether that will help (or at least not harm), and 2) I've finished all the other work I've lined up on other computers. Thank you for the input, folks.

That's exactly it. For example, I have a very nice 800, which came in its box and bag, with boxes for the TV switcher and power supply (but not the actual components), which was purportedly well stored for a couple decades. The case is in excellent condition, but when I powered it on I discovered that only four or five keys respond when pressed. I opened it briefly last night, to take a quick look at the interior condition (two very small dust bunnies, but nothing else I could see), and I'll open it again to see if I can diagnose the keyboard. As I wrote earlier, I'm inclined to at least get it into working condition, which might require reworking solder joints, replacing bad ICs and/or caps, etc, but I was curious about the wider opinion of the community regarding repairs to a system intended for infrequent-- if any-- use.

Are you using distilled water? I imagine the chlorine/chloramine and/or dissolved solids in tap water might cause problems for the resin or carbon in the ink. Have you tried Circuit Scribe's Conductive Ink Pen? It looks like it's more or less the same base as Bare Conductive-- water and "resins"-- but Circuit Scribe uses silver particles instead of carbon. It's in a ball-point pen format, but you don't have to press hard to get it to flow. I've only used it on two keyboard mylars so far, but it's worked really well for them-- just lightly touch the tip of the pen to the mylar and the ink flows.

My second 800 (NTSC), which I finally unboxed: 472889 053 (AW 457740 2/1)

When I decided to pull my Atari computers out of storage, and to expand my collection, I planned that I would purchase at least two of each model: one to use, with whatever repairs and modifications I considered appropriate (with some case modifications to accommodate new ports and such, but without completely destroying the machines); and one to keep as close to stock as possible, with whatever boxes, manuals, accessories, and packaging as I could get. With regard to that second set of machines, how "as close to stock" is appropriate for a collectable computer? I generally lean toward making whatever repairs are necessary for the system to operate as intended: repairing or replacing the keyboard membranes, replacing defective chips and capacitors, and replacing or repairing (or at least bagging with a DANGER DO NOT USE note-- I'm looking at you, Ingots) defective power supplies. But are such repairs even appropriate for a collectable, especially for a system which may spend another ten or more years in its box? On the one hand, everything deteriorates eventually: paintings, cars, buildings, etc. What good is a vintage 1950s auto with original everything-- hoses, tires, and all-- if you can't so much as start the engine because all the soft, organic bits will disintegrate immediately? On the other hand, that auto has all the original bits! As I said, I'm leaning toward making necessary repairs, but I'm curious how the wider community feels about that for the strictly collectable hardware?

83S AW 34306 183 (AW 502470 4/29 -- I think; difficult to tell if that's a 7 or not)

I just received my first 1200XL, bought on eBay from a gentleman in Placerville, CA. Made in Taiwan Serial: 72R DA 20354 233 Motherboard label: DA 183 REV. A

The new video converter arrived and... I have color! I tested with the 130XE using S-Video directly and RF->VCR->S-Video. I'll test more tonight and/or tomorrow, then get down to the business of cleaning up the computers and mapping out modifications. TL;DR: cheap video converters can have compatibility issues with old 8-bit computers (and presumably consoles), so you might have to buy a replacement.

The Startech VID2HDCON (there's also a VID2HDCON2, which is less expensive and-- I think-- less capable). It has more options for HDMI output resolutions: 1080p, 720p, 576p, and 480p (the latter two are auto-sensing on a single switch selector), versus just 1080p and 720p on the Tendak. Apparently will handle both NTSC and PAL, and it has an "output mode" switch to set whether the output is a monitor or television (we'll see how that works, but I think it controls the aspect ratio). Still inputs composite and S-Video, but audio-in is via 3.5mm instead of RCA.

My TV supports analog RF, but it requires channels be found via a scan-- I can't enter an arbitrary channel number-- and it didn't detect the 130XE when connected via the antenna connector. I used the same cable and RCA-to-F-Type adapter as previously on the workbench, so I know they pass at least some signal and the RF modulator works, but the TV didn't register it. Not a big deal for most 8-bit computers, and I might be able to get the scan to work using a different device, like the VCR. I then connected to the TV's composite input using the monitor cable, and that was a complete success: white(ish) READY prompt on a blue background. So, absolute worst case, I can use the television and composite to work on the computers. Yeah. I know chroma is connected on the 130XE because I traced the chroma pin on the monitor jack (and I believe chroma was connected by default on all 130XEs, unlike the earlier models). Color should also always be present on composite connections, as that's mixed on the motherboard. Anyway, I expect to receive the new video converter today-- hopefully it plays nice with the Ataris.

That's the one. I'm using HDMI directly from the video converter to the monitor's HDMI port, and tested at both 1080p and 720p. I suppose I could dig out a HDMI-to-DP adapter to test with, but I doubt it would change things. I forgot to mention that I also tested with a second monitor-- also BenQ, but a different model (EX3501R)-- using a different HDMI cable, but same effect.

The converter is NTSC, as are the 600XL and 130XE (no PAL components are the motherboards).

I occurs to me to ask: how many LEDs for the backlight and how many mA do they draw at full brightness? I don't expect they'll tax even an original power supply, but it might be good to know for a heavily modified computer if they're consuming an extra, say, 100 mA.

I recently decided to unbox and expand my Atari 8-bit collection. Lacking a suitable CRT television or monitor, I bought a BenQ GW2480 LCD monitor to use with a Tendak composite/s-video to HDMI converter I already had. I also moved an old SVHS VCR onto the workbench, to use with a 600XL and any computers without functional monitor ports, and I bought a pair of Atari monitor-to-RCA & S-Video cables from TBA. When I connected a 130XE to the video converter with a monitor cable using S-Video, I was disappointed to see a B&W READY prompt. I had the same result with composite and via the RF modulator and VCR, I tried the other monitor cable (which wouldn't affect RF anyway), and adjusting R38 did nothing. (As an aside, I checked the voltage off pin 17 of the GTIA, per Sams Computerfacts, and saw a range of 0V to 8V, suggesting R38 is fine.) The video converter displays color bars when no input signal is detected, I successfully displayed an old Android media streamer via composite, and the VCR's menu has a blue background over S-Video and composite, so there's no inherent problem with color between the converter's inputs and the monitor. I then checked an 800 (composite, from the monitor port, and RF) and 600XL (RF) with the same results: B&W only. I've had the 130XE for more than a decade, but the 800 and 600XL I just bought from separate sellers, and the 600XL had photos (purportedly) of the computer displaying in color on a television. It's possible all three computers now have problems with their color circuitry, but that is very unlikely, and suggests a compatibility issue with the HDMI video converter. I opened the Tendak converter and it's all surface mount components-- no potentiometers, dipswitches, or other configurables. I've ordered a new unit (different manufacturer), and will report back. Have any of you experienced this?

It looks like microcontroller access to the PBI has been done before, for example with the UnoCart: https://github.com/robinhedwards/UnoCart There have been a couple threads on Atari Age over the years discussing it-- and the difficulty of accessing the bus between 6502 clock cycles (aka "racing the bus"):

My 3rd grade class had an Apple II, but somehow I never got to use it. At nine or ten years old (either 4th or 5th grade), I was given a lightly used Atari 800XL for Christmas. I learnt Atari BASIC, then a bit of 6502 assembly (MAC65). A few years later, I was given an XT compatible PC and I moved on to Microsoft BASIC. I don't remember if I bought Borland's Turbo C for the XT or the 486 I received in high school, then learnt x86 assembly, Pascal, Visual Basic, Delphi, Java, and so on.

Question: has anyone made-- or considered making-- an optoisolator interface for SIO, to prevent damage between devices? I know they're available for RS-232, but I don't know if they'd work for SIO.

Fab multiple boards onto a master board, with the sub-boards separated by drill slots? Then you could snap the sub-boards apart. I've never used OSH Park, so I don't know how much they'd charge for that, but it looks like they offer it.