The most essential piece of equipment you have is also the most easily neglected, your tape machine. It is one of the few pieces of studio equipment that needs regular attention.

Your workhorse plods on and will take a great deal of abuse before displaying faults due to poor or non existent maintenance, but one day you may find the best take of your recording career needs to be done again and a near magical performance is lost forever. To avoid this, the most basic attention you should undertake is regular cleaning and demagnetisation of the heads, tape guides and any other metal in the tape path (see Figure 1).

Whether you have a reel to reel or cassette recorder these are simple processes and are carried out frequently, often daily, in professional studios where a conscientious engineer is on hand. Perhaps you need not go to such lengths, but regular (at least daily) cleaning and demagnetisation each 50 to 100 hours use would be a reasonable target. Both maintenance procedures are to help ensure a good performance. Head and guide cleaning reduces the chance of sudden signal loss when recording due to oxide from the recording tape which builds up on these components breaking away causing momentary loss of contact between head and tape and you'll probably be unaware of this occurrence until playback. By using tape from less reputable manufacturers, you're more likely to encounter problems but all tape sheds some oxide and you do run across the odd bad batch. I found one particular batch of tape stock from a good source shed so much oxide that it was impossible to run the tape for more than 30 seconds without complete signal loss. That was replaced by the supplier but the inconvenience is inestimable.

Assuming you are now convinced that this is going to be worthwhile, a trip to your local chemist for a packet of cotton buds, preferably with wooden not plastic sticks, and a small bottle of isopropyl alcohol is first on the agenda. There are several types of fluid available categorised as 'tape head cleaner' but I have found some brands suspect since they have been known to cause rust on heads, which suggests a high water content. A drop or two on a cotton bud will remove tape oxide easily, the cotton bud ensuring that the heads or other metal parts are not scratched. Use a new bud each time. If you have to clean the heads too often it is probably caused by an inferior tape or worn parts. Tape guides should have a round profile, and although a degree of wear is unavoidable if yours have obvious flats they will act as scrapers and damage the tape. Some guides can be rotated to present unworn portion to the tape. If they're fixed the only solution is replacement.

Demagnetisation

Regular demagnetisation is crucial to maintain the high frequency performance of your recorder and to avoid damage to recordings already made which can be adversely affected by playing back such tapes on a heavily magnetised machine. You should encounter no difficulty with the procedure... but a dire warning; ensure no recording tape is within several feet of the area in which you're operating a demagnetising tool. I have not yet erased a recording unwittingly, but that may well be due to excessive precaution on my part since I tend to regard 12ft as a safe operating distance. I understand that 3ft is generally thought sufficient. You could experiment I suppose, but perhaps with an obsolete recording.

The tool generates a magnetic field of its own and some are designed for use upon cassette machines and as such may be inadequate for open reel recorders especially on a multitrack with larger head size. Recommended are those made by Teac or Ferrograph. No metal to metal contact should take place due to the possibility of scratching polished surfaces, especially heads, and if you have any uncovered metal portions on the tool you should cover them with something like insulating tape. The degree of vibration generated by the tools transformer can overcome such self control as you think is enough to avoid contact. The process is as easy as falling off a log when suffering an attack of vertigo. With mains power disconnected from your recorder, pass the tip of the 'demagger' over and around all heads and guides without passing it over the moving coil meters (ie. those with needles) since their sensitivity can be affected by it. The demagnetiser should be switched on at least three feet from the machine and moved slowly towards it and then slowly away from it when the process is complete. When the demagger is a couple of feet away from the recorder you may turn it off. It is important not to turn the device on too near the recorder as this can actually magnetise the heads to an extent that is difficult to rectify.

Standards

If the previous two items are regarded as essential regular maintenance, the rest can be considered less so, especially if you keep all your recordings 'in house'. You may have recently bought a secondhand machine or have had one a very long time without it being checked over, and if your tapes go to other studios, or reel copies to places such as radio stations, they must be accurately recorded to standards. To speak of standards may sound like coming the 'old soldier' but all will be revealed as eminent sense.

My own experiences and subsequent disappointment has been with workshops who undertake tape machine service but either return it in much the same condition they received it, or worse, fail to have it ready when you call to collect it as arranged. If you have to travel a great distance and then discover that it's still not delivering a signal on one channel much frustration and anger is caused. Perhaps you have a determined desire to become more independent of such cowboys who charge for such non-service. Before the letters column fills with the responses of aggrieved service agents I should say that there are many helpful and efficient ones out there who will not only complete good work at modest prices, but also give excellent advice for free. If you find one such as this hang on to them.

The idea of delving amongst your machine's circuit boards may start your hands shaking in either gleeful anticipation or nervous trepidation, but be assured that if you can read a VU meter, have grasped the fundamentals of electricity and have a small amount of determination it shouldn't prove too difficult. My own beginnings in this field were initiated by necessity and a need for the convenience of having as little down time as possible. It has led to a greater understanding of, and sympathy with, my hard working recorders. The basics covered here relate primarily to open reel machines, but generally apply to cassette decks also and if you are concerned about making a pigs ear of your machine, perhaps your initial attempt should be on an older, less important, machine.

The first extra item of equipment you must have before you can get to grips with any serious alignment of your machine is a test tape. This needs to be of a specific format to suit your machine and, to an extent, your own needs or desires. The tape consists of a number of accurately recorded tones at set frequencies at 0VU referenced to a particular level, ie. 200 nWb/m or 320 nWb/m. The expression nWb/m need not concern you too deeply but it is a measurement of flux level or signal strength (actually an abbreviation of nanoWebbers per metre). The test tape can be supplied by agents for your machine's spares or test tape specialists, the latter needing to know both this level (nWb/m) and the EQ (equalisation) standard. The manufacturers manual will list these parameters, but as a guide Revox A77 and B77 Machines operate at 257 nWb/m and Teac A3340 and 3440 at 185 nWb/m. Tape machines being produced now tend to have the ability to operate at higher levels, principally to take advantage of improvements in recording tape which can handle such elevated levels without difficulty. Revox and Teac are normally supplied form the factory set up for NAB equalisation, the other EQ being known as IEC or CCIR but is the same curve.

Cries of 'does EQ matter?' break out across the land. Well, since the signal you apply to your tape recorder has a degree of compensation applied to it before it is committed to tape, the same degree of compensation but in exactly the opposite manner must be applied as it comes off the tape for you to hear it in the way you did when you recorded it. So the answer is yes. The principal reason for keeping to such EQ standards is that if you transfer a tape recorded on one machine, it needs to be played back elsewhere with similar performance, even if it's within your own studio. Recordings can sound particularly strange around the upper frequencies at the wrong EQ, and in some circumstances this may be a definite improvement. But if your masters are carefully made you may be very unhappy to hear them with a large amount of top cut or boost as they will when played at the wrong EQ curve. The final criteria for your test tape format are track format and speed. Full track is the most flexible since it will be usable on several formats, the only drawback being the setting of low frequency response of a recorder (if variable) but this can be undertaken when setting up the record electronics and therefore not a real problem. To satisfy the curious mind a further word on EQ; the NAB (National Association of Broadcasts) is an essentially American standard, although the music industry in Europe tend to use it. The IEC (International Electrotechnical Commission) is the same standard as CCIR (Comite Consultant International des Radiocommunications) and the European standard generally used within the broadcasting industry on this side of the Atlantic. The EQ of your machine can be changed relatively easily, a matter of a few components on each record and replay card for each channel and you may discover that your secondhand bargain has been changed over to a non-standard standard!

You have the test tape but you'll additionally need a means of measuring the output of the machine. A digital voltmeter will suffice, or a tape recorder test set which will set you back a couple of hundred pounds or so, but has the advantage of an inbuilt oscillator and allows measurement of aspects of tape machines not covered here. An oscillator is essential and needs a variable output level and frequency. A neat way of avoiding expense here is to use a synthesiser capable of single oscillator operation, providing you can calculate the frequencies of the various notes/tones. Before attempting to undertake any of what follows, the tape path must have been cleaned and demagnetised. Be sure that you do not switch into record while you have the test tape on the transport, it will be an expensive error.

Going through the Motions

Follow the line up procedures in the manual if you have one, but otherwise seek the 10kHz reference tone on the test tape, select the AC volts range on your voltmeter or test set, connect it to the output of channel one and set any output level control to maximum. You are about to check the azimuth alignment of the play head (see Figure 2) which has to be at a perfect right angle to the tape. If this is marginally out, there will be a loss of high frequency performance. On a stereo machine this causes a muddy image and on multitracks creates general phase errors. Other head alignments such as tilt, tangency and height need to be completed when a head is changed and need not concern us now. Start the test tape in play mode and whilst reading the voltage, around 1 to 2v and probably less than this, turn the screw which varies the azimuth of the play head very slowly. A very small movement will produce a change of reading, and what you are looking for is a distinct peak of voltage. It should be very sharp rising and falling rapidly either side of this peak. If it doesn't and you find it remains at a maximum whilst you turn the screw (a plateau) there may well be a fault in the head and you should consult a specialist. To set the azimuth on machines with more than 2-tracks it is a good idea to compare the readings from the two outer tracks, which should be nearly identical, and compare them with the inner tracks and adjust them for the best compromise to minimise phase errors. When you're happy that the azimuth is correct, secure the screw with a drop of nail varnish (in your favourite colour!) or cellulose paint.

If you have a sync facility on your machine you can set the azimuth of the record head now, since this is the active head in sync mode. Simply repeat the procedure as defined for the play head whilst each channel is in sync. If there's no sync mode you will have to set the record head azimuth after the playback electronics have been aligned.

Next set the replay level of each channel. At this point you can make a choice about the levels at which you wish to record, and decide whether it's advantageous to operate at elevated levels. If you don't use a noise reduction system it can be useful to drive the tape harder, leading to a better signal to noise ratio and subjectively a lower level of tape hiss in relation to the recorded signal. This, however, is a trade-off against headroom levels and the margin at which the tape can handle no more signal without distorting (clipping) or excessive frequency squashing (compression) taking place. In some situations, particularly music recording, this compression can be desirable to a degree but can become obtrusive and you should be guided by the tape manufacturers' suggestions and specifications, and your ears. If you do use noise reduction, little can be gained by elevated levels and you're probably better off lining up to standard recorder specification. The portion of the test tape with a tone of between 500Hz-1kHz should be run and an output level monitored, on each channel. Revox A77 and B77 output should be 0.775V (775mV) when the meters read 0VU; on a Teac 0VU is -8dB relative to this and you should read 0.308V (308mV) with the output pots set at position '8'. If any channel deviates from these readings, the trimpot controlling the replay level which can be found on the channel card should be adjusted. I run most of my machines at the elevated level of 320 nWb/m, which on a machine normally set up to run at 257 nWb/m represents a difference of approximately 2dB (see Figure 3). Assuming the use of a test tape with 257 nWb/m the replay levels are set so that the VU meters read -2dB. Having set these levels do not change the settings of the output controls since they are references for all the following measurements.

Having set the replay chain, wind off the test tape and store it with the tail out. The reason behind this is to make it necessary to run through the complete tape before it's next used. This helps reduce 'print through', a pre- or post-echo which would affect the reliability of your test reel. Many professionals store tapes tail out for this reason, and if tapes are to be stored for long periods should ideally be wound through periodically.

Record Heads and Amplifiers

To deal with these, you'll need a fresh reel of your favourite tape. Apply a signal of 500Hz-1kHz to your machine inputs (paralleled for ease if you wish) and set its level to give an output 10dB lower than you measured as your reference (remember?) but without touching any output level controls. -10dB is approximately ³/₁₀ths of 0dB, so that 1V at -10dB is around 300mV. The examples previously mentioned will be 245mV where 0VU was 775mV, and 97.5mV where 0VU was 308mV. Keeping the same level of signal change the frequency from your source to 10-12kHz and whilst watching the voltage at the output set the machine running in record. You are again seeking a defined maximum of the output level as you slowly turn the azimuth adjustment screw of the record head but this time will experience a delay of the change in the response to it because of the distance between the record and replay heads. Observe the same procedure as for the play head and don't forget to apply the drop of sealer to the screw when you have finished.

Another factor in optimising the performance of a tape recorder is bias level, and although most tapes within a type such as 'standard' or 'long play', tend to be compatible with each other, it is generally advisable to set bias levels with the tape you normally use. Each channel needs setting individually and the same signal used to set the record head azimuth (-10dB of the reference level) can be used. Run the transport in record and set the bias trimmers (not bias traps) to minimum and increase the levels, watching the machine's VU meters, until they reach a peak and begin to drop back whilst turning the adjustment in the same direction. Recorder manufacturers and tape manufacturers each have their own recommendations as to how much the 'over drop' should be, but turn the pot or trimmer until the drop is 3dB at 7½ips and 2dB at 15ips. This is a compromise which affects the high frequency performance of your machine, but since this is compensated for in a later adjustment, it should be satisfactory.

Select your tone source for around 500Hz and set the input level controls to read 0VU at the input. Then put the tape in record mode and adjust each channel's record level trimpot to read 0VU on the meters, whilst monitoring the output. Your machine is now operating at whatever levels you chose when setting up in relation to your test tape. High frequency performance is adjusted next and is effected by applying a 12kHz signal to the input at 0VU and with the machine in record. Whilst monitoring the signal off tape adjust each channel's high frequency trimpot for 0dB reading on the meters. Some recorders also have low frequency trimpots which are adjusted using a 100Hz signal and the same procedure as for high frequency. As a final check sweep the oscillator or tone source across its frequency range at 0VU on the input. 0VU should be present at the output throughout the range, with a slight deviation either side of this level, depending on the tape machine's specification. If it remains at 0VU, don't sell your recorder because it's unique! More usual is up to 3dB variation either side of that 0VU (±3dB) over the range 50Hz-20kHz but ±2dB is more acceptable.

The essential electronics are now lined up, and if you wish, and especially if your tapes leave your studio, the speed at which the recorder is running should be checked. Most machines have some means of varying the speed internally, apart from any deck mounted pitch change or remote varispeed control. You'll need some means of counting the frequency from a test tape. It's a rather hefty investment if you only use it to check your own machine's performance, but is also an accurate way of knowing how close your hoped for 15ips actually is. Reading a 10kHz tone from your test tape, any deviation from this is the degree of error; ie. 9,900Hz is a 1% deviation. Most machines specify a deviation of 0.5% as being within tolerance, some higher. It's possible to use a guitar or synth tuner for this; because they are frequency counters but at specified frequencies. If you have a calculator handy you can work out the various frequencies at each specified note relative to 440Hz and can therefore read those pitches. The test tape tones will probably not coincide with these. However, if you have a friend with a recently serviced machine, you can record a guitar or synth note on that, (accurately tuned with the tuner) and by playing it back on your machine you'll be able to read any speed deviation. Of course if your friend's service agent is one of those mentioned earlier, and the speed is not correctly set, you'll be setting your machine to have the same error. There is a means of checking the speed by another method, and it can provide an idea of deviation. Assuming you can run at 15ips notionally, every minute you should use 75 feet of tape. Accurately measuring a length of recording tape (groan) of as many minutes worth that you can bear, leader each end so you will see the start and finish points clearly. Using a stopwatch time it through the machine. Any tape left over after your calculated running time suggests slow speed, calculable in relation to the original length, and if the tape goes through before you reach the reckoned running time it's too fast. Ingeniously tedious.

Basic lineup is now complete. It's time to close the doors and go down the pub to celebrate. If you've found your first attempt at such a service an anxious experience, will you believe me if I say it will become a more relaxed adventure next time? As you develop confidence it will be more speedily accomplished and you may be encouraged to delve more deeply into the innards of your machine. If you have completed your first line up with only this article as guidance and your appetite is whetted, you should have a look at the makers manual and perhaps get to grips with erase levels, bias leakage and so on.

But that's another story.