80-Bus News


March–April 1983, Volume 2, Issue 2

Page 20 of 55


John R. Hunt

I was interested in W Van Malderen’s article in Vol.1 Issue.3: it confirmed what I had thought for some time, that with appropriate circuitry data rates significantly higher than Nascom’s standard 1200 and 2400 Baud are possible.

I would like to point out a possible problem with the Biphase (Manchester II) coding used, namely that the recorded data for a stream of ones and a stream of zeros are ambiguous, as the following sketch shows.

Thus, when the tape is replayed, the ‘blank’ section before the data starts can be decoded randomly as mark or space, and in the latter case the first data character will be missed. This does not matter for tapes recorded in Nas-Sys format as a block of mulls is output before the data, and these will guarantee synchronisation of the decoder. For other file structures (eg Hisoft Fascal 4?) reliable data recovery will not be achieved.

There is a related Biphase coding method, Biphase-M, which does not suffer from this deficiency, and is also unaffected by possible signal inversion in the recorder. I refer you to an article in the Feb 1982 issue of Wireless World for a discussion of this coding method along with circuits for coder and decoder. Apart from the addition of a phase-locked loop for clock recovery, these are only a little more complex than W Van Malderen’s circuit. I intend to experiment with this when time permits.

The article in Wireless World also points out the desirability of phase equilisatian of the signal on playback for reliable operation. Based on the Wireless World circiut I have built an equiliser which I have installed permanently between the ouput of my portable recorder and Nascom. The result has been a great improvement in the reliability of reading data tapes, particularly for tapes recorded on other machines. Most significantly, error free reading can now be achieved over a very wide range of volume control settings. The circuit is powered from the 12V supply available on the serial

This is an OCR’d version of the scanned page and likely contains recognition errors.

Page 20 of 55