Care & Repair
Tape Machine Line-Up
Richard Dean looks at tape machine line-up.
Looking after your gear once you've got it is something that is all too often neglected in many studios, large and small alike. In the first of this occasional series, Richard Dean delves into the intricacies of keeping your tape equipment in good — nay — perfect working order.
Probably the biggest knock to demo studios up and down the country in recent years has been the development of relatively cheap, high quality stereo and multitrack tape machines.
These shining examples of technology-in-the-home are the musician's dream. At last he is free to spread his talents over a range of instruments and work away at compositions of breathtaking originality.
When our imaginary hero puts this theory into practice, however, he can be disappointed. The first thing he discovers is that keeping rhythm with himself is not as easy as he had assumed. The second difficulty is maintaining a correct mix, especially when using the 'sound on sound' or 'track bouncing' method of recording. But the third and most crushing close encounter is with the quality limitations of the machine, again particularly marked when track bouncing.
Before we're all moved to tears, let's stop here for a moment. It is a fact of recording life that the most popular form of musician-operated recording, that is track bouncing, carries with it the inherent characteristic of progressive degradation of initially recorded material, in the form of a signal-to-noise ratio drop of 3dB on every good transfer. Other faults are also multiplied on each transfer. Multitrack recording doesn't initially suffer from the same cumulative problems, but as you start to track bounce to get some room for overdubs, you suffer the same shortcomings.
Before you give your tape machine to a jumble sale, remember the point I am making is that there are limitations to be reckoned with, but these can be minimised. The key phrase is correct machine alignment, the elements of which I'll outline in this article.
Possibly the most neglected requirement of good recording is the scrupulous removal of dirt. Intimate contact between the tape and the heads, guides, capstan, and pinch roller is essential. For this reason the tape path, that is all surfaces bearing on the tape during play, must be kept clean. Use cotton buds (Q Tips or equivalent) dipped in cleaner solution (Ampex 087-007, Radiospares S54-838, or denatured/isopropyl alcohol) for best results. Otari suggest tape path cleaning every eight hours, which sounds about right. A related factor to consider is the environment. Although modern machines are tolerant to temperature variations, they don't like extremes of dust, smoke, and damp; keep the room ventilated.
After a period of recording and playback, a residual magnetism is retained by the heads and induced in the tape path components. These stray magnetic fields are undesirable as they can increase background noise, partially erase tapes, and increase distortion. Removing these fields is a simple matter with a good head demagnetiser. The tape machine should be switched off, tape removed, and the demagnetiser brought into close contact with guides, heads, capstan, tape deflecting arms, rollers — in fact anything potentially magnetic in the tape path. This should be done after every eight hours of use. Before switching off, always bring the demagnetiser away slowly from the machine, to a distance of about arm's length, or you could end up with far more residual magnetism than when you started!
Now on to the alignment proper. For tape heads to perform correctly the adjustments shown in Figure 1 must be optimised.
Height can be adjusted visually. Tape is threaded into the tape path and play mode selected. The tape edges should be equidistant from the top and bottom shield plates on the head. Alternatively, the head pole pieces can be observed by using transparent tape; they should be centralised across the tape width. On some ¼in tape machines, it's difficult to observe the heads, in which case a test-tape must be used. For ¼in tape 4-track machines, the test tape has a signal recorded across its entire width, with track 3 accurately erased. The playback head is then adjusted so that track 3 has a minimum output. For ¼in tape 2-track machines, the test-tape has a signal recorded in the middle of the tape, ie in the guard band. A correctly adjusted playback head will not reproduce the signal. Where a sync output is available, the record head can be similarly aligned. If not, you'll have to record a tone from oscillator, disc (or — in Europe at least — vhf radio in the small hours) and adjust the head until maximum signal is obtained from the playback head, monitored simultaneously off-tape. Erase heads are less critical, and are commonly fixed, the only way of adjusting height being by lifting or lowering the entire tape path. Visual alignment is quite adequate.
Wrap affects high frequency response, and is shown correctly adjusted in Figure 1. The centre-line of the head must be at an equal angle to tape either side of the head. It is checked with the 16KHz section of a test tape. The playback output is monitored on the machine's output meters and the tape tension increased by lightly touching the supply spool. The level should not increase by more than 1dB when the wrap angle is correct. The record head can be checked similarly, either by listening to the test tape via the sync output, or, if sync is not available, by recording a high frequency tone and monitoring level changes on the playback head. The erase head must also be adjusted for wrap angle, this being correct when maximum erasure is achieved.
Zenith is parallelism of the tape to the tape head surface. A correct zenith adjustment ensures even contact between tape and tape head, and consequently even head wear. An incorrect zenith adjustment creates an uneven wear pattern on the head over a period of time and simply leads to premature head replacement. It's a common sight in repair workshops up and down the country, and can be easily avoided. Simply smear some ink on the head face and put the machine in to play mode for a couple of minutes. Adjust the head until even wear is observed. Repeat for all tape heads.
Azimuth is critical and should be checked regularly. It is the perpendicularity of the head gaps with respect to the tape movement, and particularly affects high frequency response and stereo image. Here a test tape must be used; if there's one test tape you should have, it's this one. The tape contains either a full track high frequency tone (around 10KHz) or full track white noise. With both types, the tape is played in mono and the playback head adjusted to give maximum output. When using 10KHz test tapes, it's all too easy to line up on peak a few wavelengths away from the correct (and highest output) peak. So the head has to be roughly right before using this method. On multi track machines, always start with the two middle tracks and work outwards in pairs. This will progressively 'trim' your initial judgement and reduce the risk of lining up on the wrong peak. For listening only, the white noise tape is the easiest to use; you will clearly hear the 'phasing' effect as you tilt the head and different frequencies cancel. Where there is the most 'top', where the high frequencies peak, you've got the optimum position. Turning up the treble control of your amplifier accentuates the effect. If the record head has a sync output, the process can be repeated. If not, record a high frequency tone or the 'white noise' from an off-station vhf tuner, and adjust the record head for a similar peak on playback, monitored simultaneously. Note however that the effective azimuth position will vary with record bias — so it may be an idea to set the bias first if you have no sync facility on the machine. Erase head azimuth is not so critical, visual alignment should suffice if indeed adjustment is possible.
Before describing alignment procedures for the electronics of the machine, it is perhaps a good idea to outline the basic principles involved, to give the adjustments a bit (and probably only a small bit!) more meaning.
The tape recording process has certain inherent weaknesses. To optimise the process, electronics interfacing with the tape heads are designed to shape the sound in a way that the tape system can use most efficiently. If you connected an ordinary flat frequency response amplifier with no additional electronics to a record head, even assuming that level and impedance matching was correct, the replayed result would sound awful! The signal input to the record head must be treated in two ways before satisfactory recording is achieved, equalisation and bias.
Any tape system stores low frequencies at a higher level than high frequencies, particularly at low speeds, before distortion due to overload occurs. So bass and treble frequencies are respectively boosted and cut during record, and cut and boosted by the same amount during replay. The overall system frequency response is therefore flat. The process is called tape equalisation. A typical recording curve is shown in Fig. 2.
Recording standards authorities differ in their opinions of the optimum equalisation, so they each produce their own standard curves. The two in most common use are NAB and IEC, but there are others. Any given curve is defined by the roll-off points (the frequencies at which the level has dropped by 3dB); or, more commonly, the time constants of the circuits producing them (expressed in microseconds). You will need a test-tape to check equalisation and Table 1 should assist you in selecting the right one.
If fed straight into a record head, the waveform of an audio signal would be distorted, due to the non-linear relationship between signal input and tape output inherent in any tape system (see Fig. 3). A high frequency voltage, referred to as bias, is added to the audio signal, which has the effect of restricting it to the straight (linear) sections of the signal input/tape output curve. In this way waveform distortion is prevented.
So much for the classroom. Now let's get on with some lining up! The playback level is set by using the 1KHz reference level section on a test tape, monitored on the output VUs or an external AC millivoltmeter.
But, like so many things in the recording standards world, it's not as easy as its sounds. In fact, this is an area of confusing, conflicting articles, books, and reports that has left grown men whimpering 'why don't they leave me alone' in its wake. So I'll summarise ruthlessly. Here goes.
In selecting and using a test tape, two important reference levels in the recording system have to be chosen. The first one is recording flux density, measured in nanoWebers per metre (nWb/m). The second is line level. This is the level at which signals are conveyed between equipment, and must be standardised. The original line level defined by the GPO is 1mW into 600Ω. This is designated 0dBm. Most studios today operate at +4dBm, and their tape machine output meters are set up, if calibrated in VU, to read 0VU at a line level of +4dBm, (normally 0VU equals 0dBm). This also applies to most semi-pro machines with output meters, in which case the playback level controls must be adjusted to produce 0VU, when playing the 1KHz test tape. In any event, line levels can be checked with an external ac millivoltmeter (meter should read +4dBm). Now back to the first quantity, recording flux density: I'll put it in perspective by talking a bit about tape generally.
Modern high output tapes will store a peak of about 520 nWb/m flux density before 3% total harmonic distortion (THD) occurs. 3% is considered to be the highest value of THD permissible. So all that a reference flux density level does is to define an operating level. But here's the tricky bit. The operating level must be high enough to modulate the tape satisfactorily, and at the same time allow enough room below the 3% THD point ('headroom') for peaks. Standards authorities and manufacturers have come up with the following flux levels: 185 nWb/m (Ampex), 250 nWb/m (MRL 'elevated' level), and 320 nWb/m (DIN reference level). There are others; these are just a selection, and are compared in Fig. 4. The DIN level is likely to be adopted by the recording industry universally in time, but at present, many standards are in use. Ampex level is sometimes used with modern tapes as a working level, but more often it is used as a reference, eg 'Ampex level +4dB'.
If your test-tape has the wrong flux density, you can still use it. For instance, if you had an Ampex test-tape but you wanted to line up for DIN level, you would simply adjust the output level to read -5dB on VUs, or 5dB down from the normal level as read on a millivoltmeter (see Fig. 4).
Having set playback output at 1KHz, you can now check that other frequencies on the test-tape give the same output. Where playback frequency response 'trim' controls are provided in the machine, tweak these for the flattest response. Before adjusting for top, check tape head azimuth and cleanness. If trim-pots are not provided, and the deviation from flatness is serious, you probably have a fault. You could use the eq on the mixing desk channel, and make a note of the correction made for easy repeatability, but this is a poor second to finding out what's wrong and should only be regarded as a temporary solution. Expect a rise of around 3dB at 50 Hz when using a full track test-tape on a stereo or multitrack machine. This due to 'fringe' effects from the normally blank guard bands. Some full-track test-tapes are compensated at low frequencies to minimise this effect, but find out first. DIN test-tapes make life difficult by providing frequency response check bands at 10 or even 20 dB below their standard level, in which case you will almost certainly need an external AC millivoltmeter.
Now we come to the recording adjustments, the first of which is the bias level previously outlined. A most important adjustment, this, as it effects high frequency response and distortion. Bias is set by recording a 10 KHz signal and adjusting the 'bias adjust' pot in the machine, initially for a peak on the output VUs. The bias is then increased until the level drops by a certain number of dB. Note that on Revox A77 machines the bias adjust pots are labelled 'oscillator'. The number of dB drop from peak is commonly called overbias, and is shown graphically in Fig. 5. The amount of overbias depends on the tape speed and type of tape used. Table 2 gives some overbias values commonly used. The bias procedure must be carried out at a recording level of about -5dB to avoid high frequency compression effects.
Now you can set the recording level to comply with the reference flux density selected, whether it be Ampex, MRL, DIN, or whatever. As with the playback level, 1 KHz is used as the reference tone. Record the tone from an oscillator (some machines have an oscillator built-in), and adjust the record level controls until 0VU appears on the output meters (or +4 dBm or so, according to operating level, on an external millivoltmeter scale). Disregard any record-level meters; these will have to be recalibrated so that they read 0VU when output meters read 0VU. If for some reason this recalibration is not possible, it's not essential if the level going into the machine is metered at some other point, for instance the output stage of a mixer.
For this test you need an oscillator capable of producing the same frequencies as the test tape. The frequencies are recorded in turn, at 0VU, on to tape. The output readings are then compared with those achieved with the test tape. Correction can be made with the machine record eq trimpots, or if serious, eq controls on the output channels of the mixing desk. Again, before adjusting for top, check tape head azimuth and cleanness.
So that's it. The story of how one man, with the odds stacked high against him, came out smiling at the end of the day with a correctly aligned tape machine. This article is not intended to be a complete guide to tape machine alignment, but rather an introduction to routine maintenance, and an attempt to clear up the misconceptions so abundant among tape machine users.
Feature by Richard Dean
Previous article in this issue:
Next article in this issue:
mu:zines is the result of thousands of hours of effort, and will require many thousands more going forward to reach our goals of getting all this content online.
If you value this resource, you can support this project - it really helps!