These are scans of the notebooks of K2RNF, Wendell Anderson, the first radio amateur to receive pictures from a satellite.

I have started scanning his notebooks that tell the story of this build. Although I was small at the time, I remember parts of it. Now that I am reading his notebooks, I better understand what he was doing.

Darkroom setup and first contact

November 23, 1963

Dad had already been listening to satellites since Sputnik. This project was to get retina-quality pictures from weather satellites. The calculations on page 226 describe the limits of eye and describe how there were to be no visible scan lines on a 5" by 7" picture held at 20" viewing distance. The picture needed to be recorded on film, and he already had darkroom gear, but he had no enlarger to project the image from the negative onto the photographic paper. He bought the enlarger and set up a new darkroom in the basement of our new house on Paul Drive. I watched him paint the basement windows black, and helped him find light leaks. These notes also show a first pass at detecting satellite signals in our new house.

Modifying the SX-25

January 18, 1964

He ended up making a new front end, a new IF, and improving the noise floor of the SX-25 receiver. The note from February 16 includes the word 'lashup' which was used at the time to mean what we now call 'kludge'. These notes are from early ideas about the project. There is exploration of various motor options, so it appears that he knew that he might need to move the film. He also tested the idea of running a phono stylus backwards to make an actuator.

IF design

August 22, 1964

These are the designs schematics for the 9.6 MHz FM IF and the horizontal sweep circuitry. There are notes about interference from a local municipal airport, which appears to have been operating at the image frequency of his receiver. It was a challenge to get rid of this, and is referred to in later notes as UNICOM. There are also Nimbus orbit details, including a description of 'high noon orbit'.

First Pictures Received from Nimbus I

August 29, 1964

The first pictures from the satellite were taken. The scheme was to take a time-lapse exposure of a Heathkit oscilloscope that had X, Y and intensity (Z) inputs. During a slow line scan, the displayed dot would traverse the row, and slowly move down the screen. At this time he knew very little about the format of the signal. He only realized that he had received some hints of a picture two days later, and he wrote "Can see both horizontal and vertical sync bars. Don't know which end is north or east. May be upside down and backwards. Black may also be white." By September 1, he had a picture that matched a Nimbus picture. He also improved the front-end and made a custom IF for the SX-25 during the project. The September 5 entry shows his design improvement of adding a cascode tube to the input and adding shielding. At this point the antenna was on the roof, stacked over our TV antenna.

Tape recorder

September 6, 1964

I remember that the project became serious after Dad had bought the most expensive piece of gear for the project, a stereo 1/4" reel-to-reel tape recorder with 7" reels. It was a Miranda, and cost $199.50. It was full of 9-pin tubes, and still sounds great 50 years later. Even with the new tape recorder, there were still problems with using the Heathkit scope as a slow-scan light source.

The argon lamp

September 17, 1964

With the new tape recorder, Dad began recording the satellite passes. He kept improving the radio for the next pass while also working on the signal processing to get pictures from the baseband signal, which had been recorded on tape. Folded between page 283 and 284 was a scrap of paper of the type that he used for many calculations. This type of paper is called a spreadsheet. The first pictures were on October 6, 1964, but were not of sufficient quality. The problem was hum in the Heathkit scope. Dad took home a Tektronix 535 oscilloscope home from work and tried it. Even using the better scope, the pictures were still not of the quality that he was looking for. This brings up a question. Why was he doing this project, and what were his goals? His story was that he was trying to get an article in the Scientific American Amateur Scientist column. He also wanted to know how good a quality picture could be made from the satellite's signal. His drive for retina image quality led him to abandon the oscilloscope approach and use an argon lamp. The lamp fed a microscope eyepiece to create a small blue dot focussed on film mounted on a cylinder. The spinning provided the fast axis, and a lead screw provided the slow axis. A tape recorder head and a permanent magnet provided speed feedback. The synchronous motors he used are now usually called stepper motors.

Driving the argon lamp

November 30, 1964

These are the schematics of the initial argon lamp drive design.

Documentation and simplification

January 23, 1965

In preparation for submitting to the Amateur Scientist, the project documentation is consolidated here. While preparing the documentation, Dad simplified the circuitry and worked on the 2400Hz demodulator. He still used the abbreviation cps (cycles per second).

Separating carrier recovery and data demodulation

February 5, 1965

New circuitry separates the carrier recovery from the data demodulation circuitry. The recovered carrier is fed to a free-running 4800Hz oscillator, with the idea that the oscillator will keep running if the signal fades. The circuit requires a 10V Zener diode, which was a rather exotic part at the time. In the schematic in the QST article, it has been replaced by a voltage divider that has a potentiometer.

Matching the lamp drive circuit to the photographic film

February 14, 1965

Many measurements and calculations were involved in matching the demodulator boxcar filter output to the film.

QST plans

March 13, 1965

Some mechanical improvements. Also plans are revealed for an article in QST, which Dad wrote instead of an article for the Amateur Scientist. The clock recovery circuit is changed to a phase locked loop design, although it is not yet working.

Antenna and Radio Improvements

May 9, 1965

The phase locked loop is debugged. The antenna is still on the TV rotor. Better signals were recorded by taking the tape recorder to Hightstown, or perhaps by receiving a reel of tape from Goddard. In any case, it became clear that the radio needed improvement. The block diagram of the system on page 73 is showing more complexity. There is a small project to build a strobe light synchronized to the demodulation. The purpose for this is to help debug the mechanical design.

Another antenna design

June 19, 1965

This shows a new antenna design. This is still not the final design. There is also an analysis of using the 1N1763 diode as a varactor.

More notebook entries to come later after I get them scanned.