You are here

Analogue Tape Machine Maintenance

Feature | Tips & Tricks By Hugh Robjohns
Published May 1997

Regular maintenance can not only ensure that your analogue recorder gives years of reliable service, but can also immeasurably improve the quality of the recordings you make with it. Hugh Robjohns retires to the test department and meets the challenge head‑on...

Analogue tape recorders, particularly open‑reel machines, are valuable commodities in an increasingly digital world, so it's worth taking good care of them. At a basic level, routine cleaning and simple preventative maintenance will ensure that your machine remains reliable and sounds good, potential problems being identified long before they do any damage — either to the machine itself or, more importantly, to your master tapes.

The more advanced aspects of maintenance, such as re‑aligning the heads, adjusting bias levels and so on, should be approached rather more cautiously, however. To perform these tasks accurately, you'll need test tapes, measuring equipment, and ideally, the technical manual for the machine in question. Whilst the information in this article is intended to provide some background understanding of the various principles and techniques involved in tape machine maintenance, it does not cover the specifics of any particular machines. And before you go any further, it's advisable to read through my 'Analysing Analogue' article on how analogue tape machines work, starting on page 122 of this issue, for some of the terms and concepts involved.

If you're uncertain of your technical ability to work on a machine, don't do it! Better to get someone who knows what they're doing to re‑align a slightly off‑tune machine than to wreck the thing — or yourself! Always remember that there are lethal voltages and hazards within any electronic equipment. Something else to think about is that professional machines are generally a lot easier to work on than semi‑pro ones. There are a number of reasons for this, but most of them relate to the cost of the original machine. Professional machines generally give better access to the adjustment controls, and these rarely interact with each other. On the other hand, many semi‑pro machines, being built to a much tighter budget, often hide controls away in operationally awkward places.


The easiest maintenance routine, but one that's frequently overlooked, is cleaning and demagnetising the tape path. This should be done on a regular basis, but exactly how often will depend on how much usage your machine sees — anything from once a day to once a fortnight, perhaps. If the machine is unlikely to be used for a while, I would clean it before putting it away, and probably again before using it after storage.

The most commonly used cleaning tools are cotton buds or swabs and a little isopropyl alcohol (extremely flammable) or Colclene spray (contains CFCs). Personally, I prefer to use alcohol, but it can be difficult to acquire sometimes (try the chemists). An alternative to either of the above is Chemi‑Swabs, which are medical swabs pre‑soaked in isopropyl alcohol and individually sealed into foil sachets. Although expensive, these are very easy to use, avoid any risk of solvent spillage and minimise the potential fire risk considerably compared with a bottle of alcohol thrown in the tool box! The swab is also far less likely to leave fibres in the works than the cheap and cheerful cotton bud.

Always remember that there are lethal voltages and hazards within any electronic equipment.

The idea of cleaning the tape path is to remove the fine dust which is naturally shed from tape, plus any residue from the tape lubricants and the chinagraph marks left on the heads from editing. Only ever use the soft tip of cotton buds on the heads, guides and rollers, and use the cleaning solvent very sparingly. If the solvent drips into the bearings on rotating guides it could cause premature bearing wear by attacking the greases, and if it encounters anything hot or any area where there are small sparks, the result could be quite dramatic! Occasionally, you may encounter stubborn deposits but refrain from attacking them with force — use patience and perhaps a little more alcohol instead.

Make sure that you clean the entire surface of rotating rollers (not always as easy as it sounds) and, as you clean each element of the tape path, examine the components for any signs of wear, such as flat spots or sloppy bearings. The latter could increase flutter and mechanical noise, and the former will lead to tape scoring or edge damage (both of which cause drop‑outs).

There are two schools of thought when it comes to cleaning the pinch‑roller and other rubber‑faced guides. One is that isopropyl alcohol does no significant harm, and the other is that it strips the natural oils from the rubber, speeding up its ageing process and leaving it glassy and brittle. Personally, I have never had a problem with sparing applications of alcohol to release grime, but if you are concerned, try using either a little pure water, or a very dilute solution of a mild detergent.


Once the tape path is clean, it can be demagnetised. The idea here is to ensure that nothing the tape comes into contact with is magnetised in any way. If it was, the recording would suffer from an increase in background noise and partial erasure of the higher frequencies. With modern machines, the increased use of non‑ferrous (non‑magnetic) materials reduces the likelihood of magnetised tape paths considerably; however, the heads will still benefit from an occasional demagnetisation.

There are a number of easy‑to‑use devices on the market these days to perform demagnetisation, but take the trouble to read the operating instructions before use — incorrect operation can leave the heads and/or guides strongly magnetised, so be careful! The aim is to gradually introduce a strong alternating magnetic field at one end of the (empty) tape path, then move it slowly and smoothly along the path close to all guides, rollers and heads, before allowing the magnetic field to fall very slowly back to zero, at which point the guides and heads will retain no magnetic flux of their own.

Some demagnetisers have electronic circuitry which does the ramp‑up and down automatically, but most types require the operator to physically move the device slowly towards and past the guides and heads to produce the necessary build‑up and ramp‑down. If the demagnetiser is turned on or off too close to the machine, the sudden creation or collapse of its magnetic field will magnetise all manner of things!

Take great care not to let the demagnetiser's probe touch the heads or guides, because the inherent vibration within the device could cause scratches in the same way as a mechanical etching tool (most modern demagnetisers have a soft plastic cap on the probe to prevent such accidental damage). Also make sure that the demagnetiser stays well away from any kind of moving‑coil meter (VUs or PPMs, for example) because it will happily demagnetise their internal magnets, causing the meters to under‑read!


Mechanical maintenance of the machine is normally restricted to turning fixed guides, lubricating bearings and linkages, or adjusting mechanical spool brakes.

Often, the fixed guides in a tape path are located with a central clamping screw. If the guide is starting to show any signs of developing a flat surface, it may be possible to loosen the screw and rotate the guide so that a round surface is presented to the tape. Some engineers rotate the guide as a matter of course once every couple of months or so (depending on the machine's usage) so that the guide wears evenly around its entire circumference and flat surfaces never arise. Worn guides are the biggest culprit when it comes to excessive oxide shedding from tapes, or drop‑outs caused by scoring and edge damage.

The mechanical bearings and linkages within a tape machine should be lubricated occasionally, ideally with a light machine oil but, once again, make sure you use the oil sparingly — a drop will go a very long way. While you're inside the machine, check out the condition of any rubber belts or drive wheels. Cracked or glassy surfaces and evidence of rubber powder deposits are all indications that replacements are needed.

One of the most common mechanical problems with tape machines — particularly multitrack machines — is the efficiency of the brakes. Some modern machines control the spooling motors to provide both the drive and braking actions, the mechanical brakes only being used once the reels have stopped turning. Older machines usually rely on the friction brakes for all braking effort, and if the machine is taking a long time to reach a stop after the tape has been fully rewound, or if it's throwing loops, the brakes should be adjusted. There's usually a fairly simple and obvious mechanism which sets the mechanical brake tension, but be careful not to overdo it — the spool hubs should be free to rotate when the brake solenoids are drawn in, and any drag from the brakes will only cause premature wear in the friction linings.

Replay Alignment

It is essential to set up the replay side of the machine before attempting to adjust the record side, as the replay alignment is referenced to a test tape (see the 'Test Tapes' box eleswhere in this article), while the record side is set against the newly aligned replay conditions. If the machine uses any form of noise reduction, make sure it is switched out before attempting any of the line‑up processes described below.

The first thing to do is to set the operating level for the machine by using the full‑width calibration level portion of the test tape. This establishes the mid‑frequency gain of the system accurately, regardless of minor head positioning or equalisation errors, and acts as the reference base for pretty much everything else. Note that all settings must be repeated individually for each channel and at each operational speed, which can take a considerable time.

The satisfaction of knowing you've set your machine up to perform superbly well is hard to beat!

Setting the replay gain is largely a matter of personal choice, but will also depend on the capabilities of the machine and the level of compatibility you want with other machines. Most modern tapes can accommodate flux levels well up into the 1200 or 1400nWb/m region, but the heads of older machines will probably saturate before this. In general, the higher the peak level, the better the signal‑to‑noise ratio, but the greater the harmonic distortion. You may prefer to set a lower peak operating level on a high‑output tape, to take advantage of better transient distortion performance, for example. If you want to match the operating level of another studio, find out what peak flux they record to and make the appropriate calculations to refer it back to your own test tape reference level.

The next thing is to align the replay head — its azimuth, height, wrap and zenith (see the 'Analysing Analogue' feature starting on page 122 of this issue for a diagram illustrating these). It's rare to have to adjust anything other than the azimuth, but it's as well to check the others anyway. Before you tweak anything, examine the mechanics of the head mounting and make absolutely sure you know what screws adjust which aspects. The azimuth adjuster is usually fairly obvious, and may appear to be the only adjuster available. If the head has a pronounced worn or flat area, it may need replacement or re‑lapping (re‑profiling), and in this case realigning the machine is pointless until the head has been replaced.

Head height can be adjusted by sight, but is best adjusted using a special portion of the test tape. If you do not have access to a test tape, do not attempt to adjust the head height.

Azimuth is adjusted by comparing the replayed phase of a test tone from top and bottom tape tracks simultaneously, normally with the aid of an oscilloscope. Start with a 1kHz tone and make sure that the two oscilloscope traces (one from each track output) are vertically aligned, adjusting the azimuth set screw on the side of the head block as necessary. Next, use a 10kHz test tone and fine‑tune the alignment. If you do not have an oscilloscope, the alignment can be done (albeit rather more crudely) by mixing the two outputs together at uniform gains, but in opposite polarities. Any phase shifts between the channels will prevent their complete cancellation, so adjust the azimuth for minimum output level. Again, start with 1kHz and then use 10kHz for fine tuning.

Wrap and Zenith can be checked by marking the head with a special non‑permanent dye and running a blank tape over the head. The dye will be removed from the contact area and the shape of this will indicate if the head is mis‑aligned. If the wrap is wrong, the gap will not lie in the centre of the cleaned area, and if the zenith is incorrect, the clean area will be trapezoid in shape rather than rectangular. It's extremely rare to have to make these adjustments, by the way, unless the head has been replaced.

If the tape machine has the ability to replay from its record head (Sync, Sel‑rep or overdub mode), you can check the mechanical alignment of the record head in the same way as described for the replay head.

Once the head has been aligned correctly, the replay equalisation can be set. The test tape will have a number of spot‑frequency tones — normally about 20dB below the reference level to avoid tape saturation effects. The aim is to establish the flattest possible replay frequency response and there will normally be adjustments for the HF and LF ends of the spectrum. Only adjust the HF end at this stage; the LF end will be done after the record side of the machine has been set up. The LF equalisation cannot be done using the test tape because head bumps (inherent frequency‑response variations) and fringe effects from the full‑width test recordings (ie. crosstalk from the nominal guard‑band areas) can give very misleading results. Ideally, all signal levels should be measured with a proper audio test set, but the tape machine's own VU meters or a mixer's PPMs can be used instead.

At this stage, the complete replay‑line up has been completed (bar the replay bass equalisation) and it is usually a good idea to quickly check the reference levels and sweep frequency responses once more before moving on to the record side of the machine. If you plan to use the machine at different speeds, you'll need to repeat the electrical alignments for the other speeds — there should be separate adjusters for each speed. Take care not to accidentally re‑adjust a control for a speed which has already been set up.

Record & Bias Alignment

Bias affects record equalisation and level to some degree, so it is usually the first thing to be adjusted, although it may be a good idea to set the record drive for roughly unity gain through the machine at 1kHz first, just to make life easier. Make sure you load the machine up with a brand new reel of the tape type you plan to use, as bias is very specific to tape types and speeds.

If you are working on a three‑head machine, record adjustments can be made in 'real‑time' as their effects can be monitored directly from the replay head as the tape runs past. However, on a two‑head machine it's a very tedious case of making an adjustment, winding the tape back, measuring the change, making another adjustment, and so on. Rather you than me — especially as you need to repeat the electrical alignment for each speed on the machine too.

...a decent analogue open‑reel recorder should be able to give even the sexiest digital toy a decent run for its money when properly set up.

The easiest way to adjust bias is to monitor the output level of a 1kHz tone (input to the machine roughly 10dB below the expected peak operating level), as the bias is slowly increased from its lowest setting. The level of the 1kHz tone should rise steadily at the output until a slight plateau is reached, after which it will start to fall: set the bias initially to the position which gives maximum output. Next, input a 10kHz tone and fine‑tune the bias for a maximum output level again, noting what this level is. Then increase the bias slightly more, so that the level at 10kHz falls by either 1.5, 3 or 6dB relative to the noted peak level, the amount depending on tape speed (for 30, 15 and 7.5ips respectively). You may find you can get better results with slight variations of my suggested over‑bias values: the exact numbers are dependent on the specific tape, and the tape's manufacturer will be able to provide ideal specifications.

Once bias has been set, the record head alignment can be adjusted (if necessary). The recording is not actually made in the head gap, but by the edge of the magnetic field, and this will change in size as the bias is adjusted. Consequently, record head azimuth can only be set once the bias level is fixed. Record a 10kHz tone (at about 10dB below peak level) and adjust the record head's mechanical azimuth for accurate phasing between edge tracks when viewed on the replayed outputs. Record head height can also be adjusted now, by maximising and matching the output level from the two outside edge tracks.

Next, the record level and equalisation can be set. The record level was roughly set at the start, but it can now be accurately aligned so that the machine has unity gain between input and output when recording (ie. the input level is the same as the output level). Adjusting the record equalisation is a case of trying to achieve the flattest record‑replay frequency response, and normally only a high‑frequency equaliser control is provided. Record a 1kHz tone about 15dB below peak level and note its level on the output as a reference. Next, increase the record frequency to 5kHz, 10kHz, and 15kHz, adjusting the HF equaliser to get the flattest possible response between the three spot frequencies. A final check of the top‑end quality of the machine can be made by sweeping the oscillator over the range 1kHz to 20kHz and noting the delightfully flat and extended frequency response!

Next, re‑tune the oscillator to the low‑frequency region (say between 30 and 200Hz) and adjust the Replay low‑frequency equaliser to get the smoothest low‑frequency response. This is very unlikely to be really flat because of the inherent problems of head bumps, or woodles, but you should be able to achieve a reasonably flat response, and certainly within the tolerances given by the manufacturer.

Odds And Ends

The erase head and signal level rarely needs adjustment because its settings are not particularly critical. However, since you're now knee‑deep in the machine, you might as well check it out. The azimuth and head height should be checked visually, and the erasure performance checked by trying to erase a peak level recording. If the level of erasure is not as good as expected, try adjusting the erase level before altering the head height.

Finally, the tape speed should be checked, and if necessary (and possible), adjusted. Ideally, a frequency counter would be used to measure the pitch of a calibrated tone recorded on the test tape, but good results can also be achieved by timing a long recording made on a known good machine (and whose time is accurately logged), or a strobe leader tape could be used. If the machine uses a synchronous capstan motor, there is probably little that can be adjusted, but if a servo system is employed, there will be a speed adjuster somewhere in the works.


The maintenance and alignment of an analogue recorder is reasonably logical and straightforward, but demands meticulous attention, a little thought, a lot of time, and a few specific items of test gear to really do the job properly. If the whole thing fazes you, don't worry about it — just enjoy using the machine and call in an expert when you think it needs re‑alignment or other maintenance (many of the reputable service centres advertise in Sound On Sound's Classified section, towards the back of the magazine). If you feel comfortable with the ideas and concepts, and have access to the necessary hardware and (ideally) test tapes, have a go. The satisfaction of knowing you've set your machine up to perform superbly well is hard to beat! But please remember that there are very serious hazards involved when working on any electrical equipment and, particularly with semi‑pro machines, it is often necessary to remove the machine's covers to gain access to the appropriate adjusters. Always make sure someone is nearby while you are working (to identify the charred remains, if nothing else...) and arrange for the mains switch or plug to be within easy and obvious reach.

The basic procedures described above apply equally well to multitracks, 2‑tracks and even cassette machines, and I have often been amazed at the improvements made possible by a little judicious re‑alignment now and again. It is all too easy to ignore falling performance levels until they become desperately bad, but a decent analogue open‑reel recorder should be able to give even the sexiest digital toy a decent run for its money when properly set up.

Test Tapes

Many of the electrical adjustments in this article are best performed with the assistance of a decent test tape — but the bad news is that a good test tape is going to cost the thick end of £100 (they are precision tools, after all). Some adjustments can be made reasonably well by other devious and less expensive means, but if you're seriously keen to maintain your analogue tape recorder in tip‑top condition, a test tape (or two) is really an essential tool.

Make sure that the test tape is of the correct equalisation, speed and level standard for your machine. Test tapes are normally designed to be used at one speed only (so you may require two different tapes if you use both speed settings on your machine). The equalisation standards are normally referred to as IEC (also CCIR or DIN), NAB (sometimes called IEC‑2), and AES. The last one only applies to high‑speed (30 inches per second), 2‑inch multitrack machines, but the other two standards cover the full range of tape widths. A broad generalisation would be that most European‑made recorders conform to the IEC equalisation standards and most American or Japanese machines use the NAB standard. However, it is important to use the right tape, so always check with your tape machine's manual for the correct specifications.

The recorded signal strength on the tape is defined in terms of magnetic flux, measured in Webers. There are a number of common level standards (185, 200, 250, 320 and 510nWb/m), the most common today probably being the 250 and 320nWb/m forms. Unfortunately, there are two accepted, but different, methods of defining reference flux levels, with a 10% disparity between them! One is the ANSI system, and the other is the DIN system, the latter being 10% higher than the former. The discrepancy could affect the precision of your replay level calibration, but only by about 1dB.

In practice, the reference flux level on the tape is used as a base point from which the maximum operating level can be established, in much the same way as a line‑up tone is used to convey expected peak programme levels. (Note that most test tapes record calibration sections across the full width of the tape, not in defined recording tracks). The basic procedure is to adjust the replay gain so that the test tape's reference signal aligns to a known level relative to the wanted peak operating flux level.

As an example, let's assume that I want to record tapes with a peak flux of 1000nWb/m (a fairly typical figure with modern tapes). If my test tape has a reference flux of 320nWb/m, the formula I need to use is:

test tape level below peak level (in dBs)

= 20 log (test tape flux/wanted peak flux)

In this case, 20 log (320/1000) is ‑9.9dB, so the replay gain would be adjusted to set the reference level tone from the test tape 9.9dB below peak operating level. Assuming that the studio's normal peak signal level is +8dBu, the test tone should end up replaying at ‑1.9dBu, which would be about PPM 3‑half or about ‑6VU.

It is essential to take great care of test tapes, since they are the only true reference tool you have. Try to avoid stopping the transport in the middle of recorded test signals, and never wind the tape against the heads. When storing the tape, make sure it has been wound neatly, with no blocking (raised sections within the spool) and keep it away from sunlight and heaters or radiators.

Calibration Tones

Tape machine alignment is a complex and time‑consuming business, and there are very few points of reference. Consequently, it's vital that before every new recording, a sequence of calibration signals is recorded on the tape so that when it is replayed on a different machine, or on your own machine in the years to come, the correct levels and frequency response can be obtained.

As an absolute minimum, you should record your reference level (0dBu) tone at 1kHz for about 30 seconds, followed by a 10kHz tone (usually at ‑8dBu). The 10kHz tone can be used to align the replay head azimuth as well as the replay equalisation, and if you're really keen, you could also put down a low‑frequency tone (say 63Hz at ‑8dBu) to align the LF replay equalisation.

All this effort would be completely wasted if you didn't mark the tape box accurately with which tones have been provided, and at what levels. The latter point is complicated by the need to refer to both your studio operating levels, and the flux level laid down on the tape but, assuming we'd aligned the machine based on a peak recording level of 1000nWb/m, as described earlier, the tape label might read something like this:


1kHz @ zero level 398nWb/m for 30 seconds

Peak recorded level = 8dB above reference (= 1000nWb/m)

10kHz @ ‑8dB for 30 seconds

63 Hz @ ‑8dB for 30 seconds

With this little lot, an engineer has everything needed to align the replay head, set replay equalisation, and match the tape levels to the studio operating levels, so that peaks on tape are within the rest of the system's capabilities.

SOS covered tape recorder line‑up in step‑by‑step form way back in September and October 1986, and the novice to this particular form of recreation might find those articles worth a look. The issues in question are now out of print, but photocopies can be obtained for £2 per article from SOS Mail Order, Media House, Trafalgar Way, Bar Hill, Cambridge CB3 8SQ.