Museum of the Broadcast Television Camera

E-mail Received, 14 May 2020

RCA Advanced Development Group -- CCD camera development


Mark writes:-
I worked on a few CCD cameras at RCA Broadcast in the mid-1970s.  Sid Bendell was my boss in the Advanced Development group of the TV camera department from 1972-76 (at which point I followed the TK-76 developmental camera into the product development phase & group).
I recall us making at least 3, perhaps more, CCD cameras as new imagers became available.  The first two, I'm pretty sure, used Fairchild CCDs of low resolution.  The first one could not be run at NTSC scan rates, so we ran it at whatever the manufacturer suggested, and displayed the image on a Tektronix 524 (probably) oscilloscope, set up for X-Y scanning.  The next one was much better, but it only had about 320 lines, so we arranged it so that it clocked the lines out at field rate so we could display it on a regular monochrome monitor.  The vertical resolution was, of course, halved.  My boss was trying to figure a "quick & dirty" way to make it interlace, so I suggested using a speaker to move a mirror at field rate to optically interlace it.  This worked, at least well enough for measurement purposes. 
By this time, RCA Princeton Labs and the semiconductor division were making their first TV-rate frame-transfer CCDs (these have two complete raster fields of CCDs, one of which is just to store and read out the image created by the other).  The clocking requirements were very complex -- multi-phase drive pulses for the upper and lower raster areas' H and V clocks, plus the transfer sections and output section.  The early ones were extremely unstable, so every clock signal had both an amplitude and offset control which had to be constantly adjusted.  We built a large control panel with pots for every voltage, and a switched digital voltmeter to record all of them.  Alas, we learned that a setting that made some kind of picture ("how many fingers am I holding up?" was our repeated question) on one day would not necessarily work on the next.  At some point my boss, in frustration at the variability, sprayed a chip with clear Krylon spray paint (the chips were not sealed, just mounted on a big DIP header).  To our delight, that made it much more repeatable and stable.  Princeton didn't much like our solution, but the next batches of CCDs were much better behaved.
They made pretty rapid improvements in the imagers and we were able to get workable pictures at NTSC rates, but not yet good enough noise or resolution. So my boss got us a TK-42 camera that had been a trade-in, and we stripped it down and put 3 RCA frame-transfer CCD imagers in the place of the '42's original vidicons in the RGB color channels.  The luminance channel's original Image Orthicon tube was replaced by a 30mm Plumbicon from a TK-44.  So, we built what we think was the first CCD color camera, sort of.  It did work, just barely well enough to encourage further investigations. At this point we started working on what became the TK-76, so CCD work took a back burner for us. 
But Princeton continued and eventually I think produced a very small 3-CCD color camera, which the Advanced Development group polished up and took to NAB (1977 or '78?).  I looked in on it as they were getting it ready: it was in a sealed box with a viewing port on the side, so you could see the camera -- they didn't want anyone looking too closely at it (the Japanese had taken a lot of careful measurements and photos of the TK-76 at NAB, and produced their copies the next year).  The camera was very insensitive for some reason, so this big wooden box had several kilowatts of quartz lighting pointed at a target in the end of the box.  I was asked to adjust one of the lights once, using a broom handle to poke it into position, and the intensity caused the wood to start smoking!  A huge fan attempted to keep the box and its contents cool with a high rate of flow-through air.  It was a kludge of the highest order, but it DID make a semi-reasonable color picture.  I'm pretty sure it was shown, if at all,  only to select audiences.
That work, though, did lead to the CCD-1 camera eventually, though I had departed RCA by that time. 
Thanks for the opportunity to remember some of this stuff!
Mark Nelson