Project: Atari 2600 Repair/Restoration

On the bench are two Atari 2600 video game systems that were given to me by one of the members of the local Slack community I hang out in. They’d been sitting in storage for the past few decades and didn’t seem to be working so he asked if anyone wanted them before they went out to the dumpster. I, of course, said I’d take them.

They’re of two different vintages based on the serial numbers, and came with a number of controllers and cartridges. For having been in storage, they looked in decent shape, although the older one was obviously sitting out in the open more than the other newer one.

The older Atari 2600 system.  SN 82227274
The older Atari 2600 system. SN 82227274

The 2600 isn’t too difficult to disassemble, and there’s already a pretty good Atari 2600 teardown on the iFixit site. This particular unit was made for Atari by Dimerco Electronic Corp in Taiwan and has serial number 82227274 according to the sticker on the back. I was able to brush off a lot of the dirt on this one, but it could still use a good cleaning.

Inside looked relatively clean with some tarnish and corrosion on the RF shield. No obvious issues with any of the components. From the silkscreen, this is a Rev 14 board.

Atari 2600 main board
Atari 2600 main board

Underneath the RF shield are the brains of the system: the MOS 6532 RAM-I/O-Timer (RIOT) chip just below the cartridge connector, the MOS 6507 CPU, and the Atari television interface adapter (TIA) chip at the bottom.

Underneath the RF shield; MOS 6532 RIOT chip (top), MOS 6507 CPU (middle), TIA chip (bottom)
Underneath the RF shield; MOS 6532 RIOT chip (top), MOS 6507 CPU (middle), TIA chip (bottom)

The RCA jack to the right is where the video signal goes out to the TV.

Initial testing of the two systems didn’t show anything on the screen. I ordered some RCA-F connector adapters (no idea these even existed until I started looking for them) to replace the RF modulator, and was able to get some images on the screen with one of the cartridges inserted. There was no response to any of the switches or controllers though, so more troubleshooting is going to be required.

Atari game on the television
It lives…sort of

I’ll need to do some more research on repairing and refurbishing these. I’ve found some potentially useful threads on the AtariAge Atari 2600 forum, so I’ll dig around there some more.

Project: Heathkit GC-1092A digital clock

Starting to tackle some of the projects that have been stacking up on the workbench. First one up is a Heathkit GC-1092A digital clock (serial number 00433). This was part of a collection from KB4NNM (SK) that was donated to the club earlier this year that’s been sitting on the workbench waiting for me to have time to get to. Decode Systems has a very useful page on Heathkit clocks.

Heathkit GC-1092A digital clock
Heathkit GC-1092A digital clock

Aside from a bit of dust and a few scratches and scrapes on the case, it’s in pretty decent condition. The clock works well, although the numbers flicker sometimes, especially when handling the clock. Haven’t tested the alarm capabilities yet. Setting the clock and alarm is done with a bank of switches on the bottom of the clock. A helpful sticker shows how to set things with the switches.

Setting switches for the Heathkit GC-1092A clock
Setting switches for the Heathkit GC-1092A clock

Removing 4 easily accessible screws from the bottom lets you take the top of the case off, revealing a fairly densely populated (for through-hole components anyway) circuit board and a big chunky transformer. Sitting in center stage is the brains of the operation, a MOSTEK MK5017AA clock chip (the white ceramic IC package in the center). There’s some corrosion on the heat spreader for the IC, but operation doesn’t seem to be affected.

Heathkit GC-1092A main circuit board and transformer
Heathkit GC-1092A main circuit board and transformer

The underside of the board has two large filtering capacitors, and the setting switches and speaker are attached to the case below the board.

Below the main board of the Heathkit GC-1092A clock
Below the main board of the Heathkit GC-1092A clock

The front panel contains the 3 2-digit 7-segment display units (Beckman SP-352). There are also a couple of 555 timer chips on the board, but I haven’t explored the schematic enough to know what they’re doing. Probably something related to the alarm function.

Heathkit GC-1092A display circuit board
Heathkit GC-1092A display circuit board

I’ll need to do some more research and studying the schematic to see if I can figure out the display flickering problem. Closing this clock back up and putting it back on the pile for now.

I did a thing

Today, I successfully defended my PhD dissertation! Now I am Almost PhD!

Now and then
Then (Jan 9 2019) and now (Nov 12 2020)

The only thing left to do now is make sure the requisite paperwork gets all the signatures and turned in to the Graduate School, and to finish up writing the dissertation. The dissertation is pretty much completed now. Just a few more modifications to make and I’ll have another draft to send to my advisor for feedback. I’ve got about three weeks before the submission deadline, but hope to have it ready to submit before that. Then the final step will be the graduation ceremony in mid-December.

Inside some Radcal Accu-kV sensors

The kV sensors in one of my Radcal 9000 kits failed calibration, and unfortunately Radcal no longer has spare detector modules available to rebuild the sensors anymore, so I had them just recalibrate the ion chambers and send everything back.

I use the Radcal kits primarily for making fluoroscopy exposure measurements and the Accu-kV meter and kV sensors don’t really get much use these days. I have other meters that I used for x-ray tube voltage and exposure measurements, so losing the Accu-kV sensors isn’t a big deal.

Since I’m taking them out of service anyway, I thought I’d crack the sensors open to see what’s in them. I only had to undo a few screws to get the cover off

Under the cover is a stepped copper filter that attenuates the x-ray beam by different amounts. The ratio of attenuation through the different filter thicknesses is used to calculate the x-ray tube voltage. The filters are attached to a block of lead that blocks x-rays from getting to the rest of the sensor. A little bit of wiggling and gentle prying let me lift the block out to look at the insides.

The sensor module itself fits snugly into the lead block and is held in place by a brass bar screwed into the lead. The circuit boards contain a couple of AD822 op amps and supporting components that take the signal from the sensor module and send it to the 4082 meter.

kV sensor module
kV sensor module

The kV sensor module itself appears unremarkable. There’s a white plastic 4 x 6 x 40 mm bar glued to the black carrier board. I have a vague memory of the 40×5 kV sensors being photodiode type detectors, so the white plastic would probably be some kind of scintillator material, and there would be some photodiodes underneath. Not positive about that though, so I’ll have to do a bit of digging to find out.

The 40×5-MO mammography kV sensor is similarly constructed, and aside from having to undo a few more screws, came apart pretty easily.

The sensor module in the mammography sensor fits into a brass block, and the stepped filters are much thinner (possibly aluminum?). The sensor module itself is virtually identical to its 40×5-W counterpart.

When I get some spare time, I’ll get some x-ray images of the sensor modules to see what’s in them. Then I’ll put them back together and they’ll become part of my museum collection.

Update: Here’s an x-ray image of the detector modules. The row of pin headers is in the middle, and the square blocks are the individual detectors.

X-ray of the 40x5-W and 40x5-MO detector modules
X-ray of the 40×5-W and 40×5-MO detector modules