The descriptions are there under "media info", but that box is collapsed
by default. Once you open it, it stays open as you navigate.
Caution: long post ahead.
The Scientific Atlanta box is a "cable box" from the early 80s or so.
It will have two coaxial connectors on the back - one for the feed
from the cable TV company and another that outputs to your TV set on
channel 3 or 4. It may also have a convenience outlet; this was so you
could plug the TV into it and still only need one wall outlet. The big
knob selects the channel and the small knob shifts the tuning frequency
(I think) a little for the best picture.
The way cable TV worked back then was that several analog TV channels
would be sent down the same cable on different carrier frequencies.
From an end-user point of view, it was like having an antenna that
picked up 30 or 40 channels. However, they deliberately did not use all
the same frequencies that TVs of the era could tune - partly because it
was easier for them to use lower frequencies than broadcast TV did, and
partly to prevent people from hooking a TV straight to the cable and
getting all the channels. These boxes changed the cable TV frequencies
to something the TV could accept. They were supposed to be supplied by
the cable TV company only; under no circumstances were you to order one
from a PO box in the back of Radio-Electronics or Popular Mechanics
magazines and hook it up yourself, because that would be illegal.
Under one scheme, channels 2-13 were the same frequencies for broadcast
and cable, and 14 was the first channel that was on a different
frequency for cable. For a long, long time, a lot of US cable systems
had HBO on channel 14. One way DRM (ARM?) was implemented by
deliberately injecting interference on the premium channels; if you paid
for one of those channels, a tech would visit your house and install a
notch filter at the pole that removed the interference. Again, under no
circumstances were you to order your own notch filters...
Sometimes you could use the "fine tune" knob to get far enough away from
the interference to get a picture, but you usually also lost the color
when you did that. If you had an old B&W TV hooked up to it, this
didn't matter. The TV audio would still have a series of beeps or be
otherwise garbled; for that you would hook an FM tuner to the cable TV
line, because they would rebroadcast the audio on FM frequencies so you
could get stereo sound, which TVs of the time didn't have. (You had to
tune both the cable box and the FM tuner when you changed channels.) I
can neither confirm nor deny that Skinemax^WCinemax was viewable in this
manner, particularly on Friday nights. [0]
The second picture (IMG_0697) has some EPROMs, a HardCard, and some
kind of backplane card. EPROMs were how you did firmware before there
was flash memory. Depending on era, they varied in capacity from maybe
2 KByte on up. When you figured out what the firmware should be, you
would put the EPROM into a programmer, which was often hooked to a PC
parallel port. The programmer used higher-than-usual voltages to blow
or not blow little chemical fuses inside the EPROM. If you screwed up
the firmware, you peeled off the sticker on the top of the EPROM,
exposing a little clear window, underneath which was the chip. If you
had money you then put the chip in a box with a fluorescent tube that
emitted UV light for 15 minutes or so. If you didn't have money you
then put the chip outside in the sun for a few weeks. This "reset" all
the chemical fuses so you could program them again.
Your gadget would have one or more sockets for the EPROMs, depending on
how much space was needed. If there was more than one socket, sometimes
chip 1 had (say) bytes 1-4096 and chip 2 had (say) bytes 4097-8192.
Sometimes chip 1 had bytes 1, 3, 5, 7, ..., 8191, and chip 2 had bytes
2, 4, 6, 8, ..., 8192. I swear I am not making this up. Firmware
upgrades happened by unplugging the old chips and plugging in the new
chips. Oops, not all the pins went in the socket, try again. Okay, now
it won't boot... oh, it's rotated 180 degrees in the socket, unplug it
and try again. Oops, one of the pins broke off...
The HardCard was a "simple" way to add a hard drive to a system; you
didn't have to string cables around from the controller card to the
disk drive itself. It was also a space-saving trick. The original IBM
PC had two full-height drive bays (full height is twice as tall as a
modern desktop CD or DVD drive), and these often each had a full-height
5.25" floppy drive in them. People didn't want to give up one floppy
drive for a hard drive, so a HardCard allowed the hard drive to live in
one of the expansion slots.
I am not as sure what specifically the backplane card is for, other than
to say that some industrial computers were built that way. Rather than
having a motherboard with CPU, support chips, and several connectors for
expansion cards, you had a "dumb" backplane that just carried signals
between card connectors and maybe provided power. The CPU was on a card
that plugged into one slot, and other peripherals plugged into other
slots. These got used (among other places) for automated test
equipment; you might buy a backplane, plug a CPU card in, plug in a
digital input card, a digital output card, an analog input card, etc,
depending on what inputs and outputs you needed. There were/are several
standards for this; VME is one but many others exist.
The third picture (IMG_0696) has probably a CGA or possibly EGA video
card (the big one) and a memory expansion card (the smaller one). I
remember CGA cards having composite video output (the RCA connector) but
I don't remember if EGA cards did. CGA to a dedicated computer monitor
could do 80 column text, but to a TV set could only do about 40 columns
due to bandwidth limitations.
The memory expansion card was designed to allow you to add memory to a
PC that didn't have sockets for more chips on the motherboard. Those
three rows of sockets at the back are for the memory chips; often the
same card would be sold with different numbers of sockets populated for
different prices. Each chip was often 1 bit wide by thousands of bits
deep, so you had to install them in sets of 8 or 9 (parity) at a time.
Early in the game, the expansion cards were not standard, and you had
to check that each application could talk to the specific card you had.
Later on, there were standards, including LIM (Lotus-Intel-Microsoft).
(Intel and Microsoft are obvious. Lotus was there because business
people that had never owned or used a computer before would go out and
buy one just to run the Lotus 1-2-3 spreadsheet.)
The fourth picture (IMG_0695) I am less sure about... at this point I
would go with Ralph Wade Phillips' ID as telecom-y things for handling
T1 circuits.
The fifth picture (IMG_0694) is probably a 10 or 20 megabyte hard drive,
full height, with its controller card. If I remember correctly, it was
possible to run two hard drives off of one controller card. I remember
for sure that one of the cables had the same connectors as a
contemporary floppy drive cable, but was wired differently; using a
floppy cable on a hard drive was a quick way to fry things. This
would have gone in an 8088 or 8086 PC, or *maybe* a 286.
The controller card had a lot of influence over the analog signal being
presented to the magnetic heads in the disks. Early disks used what was
called Modulated Frequency Modulation (MFM) encoding. Later on, you
could buy a different controller card that could do what was called Run
Length Limited (RLL) encoding. [1] The reason you would do this is
that it gave you more space on the same hard drive. The drive
manufacturers warned that putting an RLL card on an MFM drive would
curve your spine, warp your mind, and cause you to have ugly babies;
they wanted you buy an official RLL-compatible drive at official RLL-
compatible prices. People used RLL controller cards on their MFM
drives anyway, and it worked. Usually.
The sixth picture (IMG_0693) I am again less sure about; I'll go with
Ralph's idea about telecom-y things. Gargling "Paragon Networks
Transmaster" yields
http://www.at2.com/downloads/documents/all_others/transmaster.pdf
, copyright 2001.
You might know this already, but sometimes you can date things by the
date codes on the ICs. There are many schemes, but a common one is a
4-digit number by itself, which decodes as YYWW. YY = year % 100 and
WW = week-of-year. 8306 would be mid-February 1983 and 9551 would be
a few days before Yule 1995.
Pardon me, I have to go see how Spinrite is coming along.
Matt Roberds
[0] Why do you think the net was born?
[1] These were the popular names. The technical differences are
probably more subtle.