The Practical Wireless "Tele-Tennis" build.
...34 years after the original publication
by Grant Searle
Last update: 15th January 2013
Original magazine article construction publication date : July 1974 - November 1974 (5 parts)
Actual build date : February 2008
The arcade version of TV Tennis - Pong - hit the arcades in 1972.
In 1976 the AY-3-8500 "Pong in a chip" was introduced, and the tennis game, along with football, squash, solo practice and shooting games became very popular (and affordable) in the home.
In between these dates there were a couple of "analogue" TV games released in magazines and some commercial ones were made. The Practical Wireless "Tele Tennis" was one such magazine construction. This particular one was covered in detail over a total of 5 months (!) with part of its construction being shown each month. This is built in a highly modular form, and each module could be tested before the next month's article was released.
The 6 board modules are as follows:
Following on from the 5 parts were some additional modules produced by the original author and also some other authors, such as on-screen scoring, sound effects and variable ball speed/bounce.
I wanted to build this for many years, and finally had the opportunity to do so in 2008.
What is shown here is the original version (with some small enhancements, as described below).
Practical Wireless Tele-Tennis magazine articles
I own all the magazines that contain the construction articles for this game, along with the follow-on projects.
The publisher of Practical Wireless has very kindly granted me permission to publish a PDF document on this web site containing my scans of the complete construction articles. This PDF also includes the follow on articles from later issues, including the "Points Arising", "Sound Effects Unit", "On Screen Scoring" and "Ball Speed Control", entries from the "Letters" pages and the "reprint" advertisement.
The main construction articles are over 5 issues. The first is the introduction, the following three give details on building 2 boards and the final one contains the complete inter-wiring and setup.
The complete PDF is in the 5 links below (in a split ZIP file). Total size approx 16 MB, split over several files (because this web hosting service annoyingly keeps on deleting large files with no warning!)
Part1, Part2, Part3, Part4, Part 5 - ensure you download ALL five parts to the same location then open the "PW-TeleTennis.zip" file (winzip etc). If your unzipper supports zip spanning then it will re-combine properly and show one PDF containing all pages.
Note: If the above links fail, you can also get it in a single ZIP file from my SkyDrive here.
Published with the permission of the copyright
holder PW Publishing Limited
The construction details are shown in the PDF file above. What follows here are pictures of my build using these magazine articles.
The PCB layouts were published (two per month) in the magazine in parts 2,3 and 4. These were shown "full size". I used these to produce my own boards.
Six photo-etch boards were cut to size and etched using the foil pattern supplied in the magazine series.
Here they are laid out, top view. The tracks are visible through the fibre-glass PCBs.
The lower boards, left-to-right are "Board A", "Board B", "Board C". The upper row are "Board D", "Board E" and "Board F". The gap on the top right is where the transformer will sit.
Here is the view from underneath.
Holes were then drilled (mostly 0.8mm for the ICs and small components) for all of the components and for the mounting holes.
Each board was then soldered individually, and once all components were mounted, the boards were mounted onto a frame to allow the interconnections to be made.
I wasn't going to use a TV modulator so this is not installed. Instead, I connected the composite video output straight to the output connector on the back of the case (as described in the magazine articles). This can then be connected to the Video input of a modern TV.
Additionally I used "74LS" ICs instead of the higher power "74" series, so R66 on the power supply board is not needed so replaced with a wire link. R67 and R68 on the same board were not needed (to filter the supplies to prevent power fluctuations affecting the modulator) and were also replaced with direct wire links.
The original prototype in the magazine articles used aluminium extrusions to mount the boards, but I wanted to use wood, as I intended to make a wooden console case for the game.
These mounts are made from wooden batten. The centre of the boards are mounted on two battens firmly stuck together. Small screws are then used to mount the PCBs onto the frame.
Solder pins are used at each board connection point, and solid core wire is then wrapped and routed between the pins.
Once the interconnections were complete, short lengths of flexible wire were used to temporarily connect the two sliders and three switches to the board to allow it to be tested before mounting into the case. A bench power supply was connected across the smoothing capacitor and the composite video output connected to the video input of a monitor/tv.
With the game now fully functional, the case was then constructed.
This was made to match the original prototype as far as possible, however, I wanted an insulated case, so I used wood instead of aluminium. The external dimensions match the original as far as possible, and is as slim as can be made without the aluminium heatsink touching the top panel.
The case was made with MDF and used batten in the corners for strengthening. Two strips of batten at each side of the top was added to support the front panel. The bottom of the case was covered with textured black sticky-backed plastic.
Here is the case with the MDF front panel in place.
Here is the case with the PCB frame and transformer (not yet wired up) placed inside. Holes were then drilled in the back for the Video output connector (not using a modulator) and the mains input.
Once the front panel had been cut to hold the sliders and switches, it was then covered with wood-effect sticky backed plastic to match the original.
The pictures from the magazine were rotated, stretched and skewed so that I could clearly see the "PW" logo, and using the flattened view of the top of the case I was able to approximate it's size to match the original. Similarly, the sizing and positioning of the other labels were approximated.
The PW logo was re-drawn using "Visio", along with the labels.
Here is my front panel.
Here is the completed machine, internal view
Here is the outside view
Some modifications were made as I was not happy with parts of the operation of the circuit.
1. Speed linearity correction
The ramps used in the 555 timers are not linear, and start rapidly then level-off as the voltage climbs. This ramp voltage is used as a comparator with the linear sweep voltages for the horizontal and vertical ramps. Using the original circuit, this non-linear ramp in the 555 resulted in an inconsistent ball speed as it moved across the screen. It would slow down when approaching one of the sides and also along one of the edges. It would then speed up again as it moved in the opposite direction. This clearly gave an unfair advantage to one of the players, as one player would always have a slower ball than the other. This is rectified (very easily) by ensuring the 555 timer charging ramp voltage is linear. This is achieved using a constant-current source for the capacitor charge, and required two transistors, two diodes and an extra resistor to do this for both horizontal and vertical movements.
2. Expanding the usable screen
Diodes (x3) added between pins 6 and 7 of the two 555s controlling the ball allowed the threshold voltage of the 555s to be lowered. This allows the trigger to occur much earlier in the sweep (if needed) so the ball is then able to use the higher part and left hand side of the screen. Without this modification it was necessary to lower the top border of the playfield otherwise the ball would never hit it. Similarly, the left hand side needed to be brought in. This resulted in a smaller playfield than could have been used. With this modification the complete TV screen can be used.
Both modifications can be seen here.
Once this change was made the ball moved in a very constant speed across the screen, as it would be in the arcade / bought versions of the game.
Here is a screenshot of the complete playfield. The ball is stationary (pressing the "ball boy" button).
Here you can see the ball bouncing off the bottom of the screen. The slow camera exposure shows several consecutive positions of the ball.
Click here to see my eMail address and main index page