Andy Honeybone begins the year with renewed determination, vigour and curiosity, not to mention the screwdriver his aunt gave him for Christmas. In the first of a series of investigative tech-type-features, he studies the workings of the mild mannered tape recorder. There's more to it than you sync.

The inventor of the transistor, William Shockley, is said to carry a cassette tape recorder with him at all times — firstly, because he likes to refer to it as a product which would not be possible without his invention and secondly, because he uses it to record his every publicly spoken word. Without wanting to risk litigation, let's just say his views are considered somewhat radical and his tapes are at least a document of what he actually said.

Of course tape recorders existed long before transistors were invented and, if my memory serves me correctly, the concept was understood even before valve amplification existed. Magnetic recording tape has seen the most change over the years having first existed as metal wire, progressed to unstable, oxide-coated cellulose acetate, survived the plastics technologies of PVC and Mylar (polyester) and most recently acquired the full-circle treatment of metal-alloy formulations.

Tape is Coated with extremely fine particles which can be oriented by a magnetic field so they represent a sound 'image'. The recording head is an electromagnet which provides a magnetic field from the highly amplified input signal which 'orders' the particles. The passage of the tape over the playback head converts the magnetic flux on the tape into a minute electrical signal which requires additional amplification to bring it up to a useful level.

The buzzwords of tape recording such as bias and equalisation arise from the fact that the basic record and playback scheme outlined above is far from linear and has trouble coping with all but a very narrow range of midband frequencies. To get an optimally flat response it is necessary to boost the treble during record and then try and flatten things off by some hefty playback EQ.

The voltage from the replay head is very low — in the order of one millivolt for a 0 VU (volume units) signal on a cassette deck. This puts great demands on the noise performance of the replay amplifier.

Another problem is preventing the playback head from becoming magnetised by currents flowing back from the replay amplifier when the machine is turned on. Coupling capacitors are out of the question, so the replay amplifier is direct coupled through the head to a biasing voltage rather than referencing the head to ground.

If the head becomes magnetised it will act as a weak erase head and skim off treble every time the tape is played. The classic way of demagnetising (de-gaussing) is to use a purpose built tool (a rod wrapped with many turns of copper wire) which, when plugged into the mains, produces an alternating magnetic field. The tip of this evil instrument is brought close to the face of the head for a few seconds and then is slowly drawn away so that the magnetic flux dwindles to nothing. The result is a de-magnetised head but, to be honest, I've never been able to tell any difference — rather like changing the oil in a car. The construction of cassette decks makes it all but impossible to get at the heads and so TDK make a battery powered degausser which is housed in a cassette shell.

The main job of the replay amp is to equalise the signal from the head. There are three equalisation settings currently adopted for the main cassette tape types — ferric, chrome and metal.

Normal position (ferric) requires playback through a 120 microsecond (upper) time-constant filter whereas chrome and pseudo chrome tapes need 70 microseconds. These settings modify the low pass characteristic which corrects the problem of the replay head output increasing with higher frequencies. As the time constant is reduced, so the cut-off slope moves toward higher frequencies.

Unfortunately, the output from the head stops rising at around 8kHz and starts to plummet downwards. There are several reasons for this miserable state of affairs and the first is down to tape speed. Faster tape speeds make more magnetic material available to catch a rapidly changing signal and so the high frequency response is extended. The other gremlin is the width of the gap of the playback head, because when it is the same length as the recorded wavelength there is no difference in magnetic potential across the gap and that means no signal.

The record amplifier has to boost the treble like crazy to complement the head and replay characteristics. One fly in the ointment is the problem of avoiding high frequency tape saturation, because metal and chrome type tapes have a capacity for treble information which would grossly overload the ferric variety. The solution is in the provision of the equalisation settings which, by further low pass filtering, allow all the different tape types to be accommodated.

An example of high frequency overload that you may have already experienced occurs when some drum machines are direct injected. The white noise cymbal circuits push out so much treble that either the tape or the input amplifier will saturate and produce a distorted hard edge to the replayed sound.

The record head is an inductor and, by its nature, exhibits a rising impedance to an increasing frequency. To overcome this further attempt to rob us of a bit of treble, the head should be driven at a constant current which is independent of frequency. As this is a bit difficult to do, a voltage drive is used and a large value swamping resistor is placed in series with the head to approximate the ideal.

The erase head is another electromagnet which is driven by a high frequency oscillator in the range 60-100kHz. The drive should have a sine waveform as this gives the cleanest erasure and the action is much as for the head de-gaussing tool described earlier. Cheap portable recorders have used permanent magnets or DC electromagnets for the erase head because of cost and the problems involved in producing the large signals required from low voltage power supplies. Some designs utilise the replay power amplifier as the bias oscillator and this can lead to instability arising from the complex switching required. Again on the subject of cost-cutting, the record and replay amplifiers may be one and the same with multipole switches changing the sense of inputs, outputs and filter response. The record and playback heads may also be amalgamated which prevents their individual optimisation and off-tape monitoring but on multitracks allows the simul-sync track bouncing we know and love.

Bias is a word which covers a multitude of sins. It refers to a high frequency signal which is injected at the recording head with the aim of improving frequency response and reducing distortion and modulation noise. Anyone who has struggled to set up a tape machine wil appreciate that all these factors interact and a compromise is all that can be expected. The best bias level for a 1kHz tone will not be right for a 10kHz signal. Further, the bias level which gives a low modulation noise is unlikely to give a low distortion figure. The bias signal is a trimmed down version of the erase voltage and, not surprisingly, if the bias is set too high it will erase the high frequencies as they are being put on the tape. Hence optimisation is the name of the game and just to pull the rug from underneath the unwary, the optimum varies for each brand of tape.

Many machines are now set up for the pseudo-chrome, cobalt-doped, ferric (TDK SA — type II) tapes and aren't too happy with anything else. Some decks allow you to alter the bias to fine tune to your favourite brand but this requires either a knowledge of your tape's preference or some test gear. The addition of bias can cause intermodulation distortion and saturation if it is allowed to infiltrate the record amplifier and to prevent this, a bias trap is interposed which is yet another adjustment which has to be made.

Tape recorder hardware has its own problems. The tape is pulled past the heads by the pressure of a rubber pinch wheel against the rotating capstan. Oxide from the tape becomes deposited on the pinch wheel and deteriorates the stability of the transport if not regularly cleaned off.

The heads also pick up detached oxide and this prevents proper tape contact which reduces the high frequency response. Regular swabbing (every 10 hours) with an alcohol-based cleaning solution is recommended (by the manufacturers of alcohol-based cleaning solutions). The setting of the head is also crucial to recording quality. The zenith adjustment ensures the tape is flat on the face of the head and the azimuth adjustment sets the angle of the head gap to the tape. The azimuth is set by adjusting for maximum output from a test tape of a pre-recorded 10kHz sine tone. There are about 20 electrical adjustments and a similar number of mechanical checks to be made when servicing a domestic tape machine. An oscilloscope, signal generator and test tape are the minimum tools you'll need and a stiff drink is also a good idea to revive yourself when you discover how much the spare parts cost. Economically, it is a better bet to buy a cheap but no frills model as a replacement if your machine is in such a bad way that you're actually thinking of having it looked at. Anyone want to buy a used Sony reel-to-reel?