by Carl Evans
Message on the Medium
The ATARI 410 Program Recorder, sometimes called the cassette drive, has a mixed reputation. Many users have had exasperating problems with it that have driven some of them straight to diskettes. Still, the 410 is a popular device, especially among ATARI users with tight budgets. The 410 can be more efficient and reliable if you learn something about its quirks.
The kind and quality of tape used with the 410 are important factors contributing to successful use. You might assume that the best quality, highest priced tapes, perhaps the so-called "computer" tapes, are what you should use, but that is not necessarily so. A good brand of music tape is usually sufficient.
You should avoid really cheap tapes, or tapes with strange brand names, and also avoid the "chromium dioxide" tapes, and "digital data- recording" tapes advertised for use with some computers. Short tapes are often better than long ones, and there are some cassettes out now with as little as three minutes of tape on them. Really long tapes, like the C-90 and C-120, are unnecessary and have special problems that disqualify them for computer use.
Cassette tape is made of a tough, flexible plastic, coated with a magnetic material, usually ferrous oxide. When you record a program on a tape, the recorder generates a magnetic field that leaves a magnetic trace at a certain spot on the coating. The tape is pulled past the recording heads of the recorder at a constant speed, leaving a series of magnetized traces. The strength of the traces depends on many things, but the nature of the coating, its formulation and the evenness of its distribution are important.
In general, the price and quality of a tape will reflect these factors. Very expensive tapes will have highly responsive coatings securely bonded to the plastic in very carefully controlled thicknesses. Medium priced tapes, like music tapes, sacrifice some of this quality. Cheap tapes have inferior coatings and quality control, and are really suitable only for non-critical purposes.
The ATARI computers record data on the tape using a method known as frequency shift keying, or FSK. This means that the computer actually generates sound frequencies and sends them to the recorder. The frequencies generated by the ATARI are very precise at 3995 Hz (for zero) and 5327 Hz (for one). The 410 recieves these signals from the computer and leaves the appropriate magnetic traces on tape. This means that the trace on the tape is approximately the same as generated by the computer, but not quite. Variation within components of the 410 may throw the frequency off, or the speed of the capstan motor may vary enough to change the frequency of the trace. The amount of electrical energy to the recording heads may fluctuate too, so the traces written by the 410 may stray somewhat from their intended values.
Although the ATARI computer writes sound frequencies to the tape, it does not respond to these frequencies when reading from tape. The 410 actually listens to the tape, interprets the frequencies as either low or high, and passes a voltage to the computer. Zero voltage equals zero, and approximately 5 volts equals one. The computer does not hear the frequencies, it reads voltage changes coming from the 410. This is the crux of the problem. If the 410 does not properly interpret the frequencies and pass the right voltage, the computer gets bad data.
Tape can be responsible for these errors. Look at Figure 1. In this normalized figure we find 3995 Hz and 5327 Hz as they might appear on an oscilliscope. Pretend that they represent two bits of data on tape. The amplitude (height) of each is equal, indicating the traces have equal signal strength. Tapes prepared by the ATARI Program Exchange (APEX) come close to this ideal. Tapes recorded by other vendors, or by the 410 itself most often do not. They look more like the waveforms in Figure 2, where the higher frequency has less amplitude, or signal strength. This loss of strength for higher frequencies is called attenuation and is a normal condition.
If attenuation causes the higher frequency to be recorded at a strength lower than the threshold for recognition by the 410, or the components of the 410 are not operating up to their specifications, the 410 will fail to recognize a one, and "drop" a bit. The computer is monitoring the load, and if it senses an improper sequence it will issue a loading error message. But the bad bit(s) may be in data statements about which the computer cannot judge, and so some problems won't show up till later.
Highly responsive computer tape might seem a solution to this problem, but in fact it catches us going the other way. Being highly responsive, it does not attenuate certain non-signals, like the switching current that makes a "spike" every time the recorder is ; started. Spikes are often interpreted by the 410 as signals, so a one is passed to the computer when none is supposed to be there. This, of course, gives a bad load. If you have this problem, you can get around it by positioning your tape on CLOADS so you start just past the point where the leader tone begins. Other intermittant electrical noises can creep onto high grade tape in a low grade recorder, and that's why medium grade tapes are better with the 410. They will tend to attenuate spurious signals to strengths below threshold.
The problem with chromium dioxide tape is that it is very abrasive to recording heads and the 410 is not built to take it. Beware of companies offering software for the ATARI recorded on chrome tape. Head wear is another cause for distortion and loss of signal.
Other things that can cause the 410 to misread a signal include "fade out" and distortion. Fade out is the loss of strength of a trace on tape due to thin spots, perhaps caused by loss of coating from repeated playing. Music buffs claim they can hear this loss after about fifty plays. That is about the point at which fade out begins to affect data too. After about fifty reads, medium grade tape may lose enough coating to lower the trace strength below threshold for higher frequencies. Running tape on Fast Forward or Reverse does not damage the coating since the heads are not engaged.
Sometimes distortion results from the physical stretching of tape. If a tape is stretched, a high frequency signal will become lower, and may not be recognized by the circuitry as a high signal. Stretching takes place most often near the beginning and end of tapes, near the hubs. Allow plenty of leader to reduce this problem. Thin tape stretches easier than thick tape, and thin tape is used with long tapes, so avoid any tape longer than 30 minutes per side.
Good practice suggests using short tapes of good quality from a manufacturer whose products give satisfaction. I usually put only one program on a tape and repeat the save a couple of times on that tape. I might also save it with other programs on a longer tape for archival purposes. Keep your tapes away from magnetic forces and environmental hazards.
This covers the main points of the tape medium as they pertain to the ATARI system. Next month we will look at the 410 machine itself. Your comments and questions are invited. Please send them to ANTIC for my attention.
Carl Evans is an electronics engineer who earned both BSEE and MSEE at Georgia Institute of Technology. He now works at Aerojet Electronics Systems, and was formerly with General Dynamics. In 1979 he bought a TRS-80 and used only tape for mass storage. When lured to the ATARI 800 in 1980 he continued to use tape with the 410 Program Recorder. He has written many programs for the ATARI, including several tape utilities, and has formed a new company, VER VAN Software, to market some of them. ANTIC readers should be pleased to learn that Carl will be coordinating this department of the magazine.