Are home-made 78 rpm records possible?
This web-page is coordinated with a YouTube video showing the cutting lathe in action.
This allows much more detail than could be shoehorned into a video, and at the
same time, a video of it working is a good adjunct to the web-page.
Link to YouTube Video: http://youtu.be/T7nKa_aHyR8
Well, as we said in the preamble on the index page, yes they are, only given that you have a disc cutting machine, or can make one. We have had several of them, but that was well over 30 years ago. Recently we got a small, old one and have been having great fun with it, though intermittent frustrations tend to occur. Hence this web-page, which may be of use to others who wish to read about, or even follow the path to the home-made 78. Many might think it an eccentric and whimsical path to follow; indeed, I think that myself. But – if you are a 78 collector, just think of the exquisite pleasure of having that old, scarce & delectable side a friend emailed you as an mp3, in the form of a 78 rpm disc to take its place in your collection! The possibilities are very great; one can easily obtain circular peel-off labels on which you can print your own replicas of your favourite labels and place them on your new 78s. Perhaps better still, you can create your very own record label for your most treasured sides.
None of the discs in the image above ever existed; they are all old labels, scanned and re-titled, except for the one at bottom right, which is my own label – or at least, the first of them, for I doubt that the temptation to create still more esoteric ones will go away…
So – where do we start? This page has three short articles, taking one aspect of the project at a time. Let us start with the type of blank disc we will need. After all, if we do not have a blank disc we cannot make a record. We have tried to make our own blanks, but without success. Still, our efforts may at least afford you some amusement.
Part I. A Chamber of Horrors: Attempts to make recording blanks.
Summer 2014. Normally, blank recording discs look like the one above. They are round of course, and black – or at least very dark in colour. They consist of an aluminium disc coated with a lacquer, mostly of cellulose nitrate, a plasticizer (many different ones have been used over the years) and a black dye. The coating is very, very smooth with a mirror finish which above, is reflecting the desk lamp, the underside of a record deck, a loudspeaker, some wallpaper and the lens hood of the camera. They are usually used with a chisel-type sapphire cutter, but a carefully made metal one will also work quite well. All being well, the groove is practically noiseless. Indeed, for many decades, such discs have been used as master records, rather than the wax blanks which they supplanted about 1950. In short, they are really great discs. But as far as the amateur is concerned, they have one severe drawback: they are expensive. When our interest in making 78s returned a few months ago, the obvious thing was to make discs with a cheap alternative coating. One obvious substitute stood out: gelatine. Not many people know this, but when cellulose nitrate discs were developed in this country by Cecil Watts, various other coatings were also under development. One of them, made (or imported) by the V.G. Company, had a glass disc coated with a chemically treated gelatine solution. Gelatine is cheap, and also has the great advantage that it is non-inflammable (cellulose nitrate (a.k.a. gun-cotton) and the solvents used to make the solution are all very highly inflammable.) The downside, is that gelatine is water soluble, so you must keep your gelatine-coated discs dry.
We took some unimportant old 6" aluminium home-recording blanks, and began trials with solutions of gelatine of different strengths. We tried dipping them, or pouring the solution on the discs. Alas, the results were never encouraging. Adhesion was poor, even though we roughened the surface with emery paper and were always careful to clean & degrease the discs before coating. Fountain pen ink was used to dye the mixture. Glycerine was used as a plasticizer, and a photographic wetting agent added to the mixture. Solids, such as china clay, carbon black and lamp black were added to give the coating some ‘body’, but without success. We tried spraying the discs instead of dipping them, as Cecil Watts had originally done. Unfortunately, if you spray gelatine through an air-brush, it forms globules – probably due to a sudden cooling effect when the pressure drops.
In the end, we abandoned our attempts to coat discs, and turned instead to a completely new medium, which many people have already successfully used for making their own disc records, as YouTube amply attests, though not usually at 78 rpm. Polycarbonate plastic is fairly cheap, and while you cannot cut a grove in it, it is perfectly possible to ‘plough a furrow’ in it, or in other words ‘emboss’ the surface – exactly as was done in the 1920s and early 1930s on thealuminium discs used by the various types of home recording device that appeared in those years. Such a disc is shown above, with a label.
CONCLUSION: The blank discs shall be 10 inch diameter polycarbonate plastic.
Part II. “Cutting” Styli for making the records.
October 2014. The word “Cutting” is in inverted commas because it is not a correct description. No material is removed from the polycarbonate disc; a groove or furrow is simply impressed in it. It’s exactly like ploughing a field, even to the extent that the furrow on polycarbonate will have a ridge on either side of it.
Above is a radial section through the surface of a polycarbonate disc. The thin grey line is the original surface level. Five furrows are shown. Each furrow has a ridge on either side, and between each furrow and its attendant ridges is the original surface of the blank disc. There are at least two drawbacks to ploughing a groove in this way. 1. Each groove occupies quite a bit of space, because the ridges have to be taken into consideration. Therefore there is a limit to the fineness of the groove pitch on a disc of this type. On a lacquer disc, which merely has a cut groove with no ridges, we can make them much closer together. 2. The red ball represents a playback stylus. The ball with the tick is in the correct place, sitting in a furrow, where it can pick up the sound in the groove walls. But it is possible for the stylus to take up a position between two adjacent ridges, as shown by the ball with the cross. There is no modulation on the outside of the ridges, so to all intents & purposes, the record will be silent! This problem is commonly encountered when trying to play back old 1920s & 1930s aluminium home recorded discs. Indeed, some of those old discs have been furrowed in such a way, that our modern styli have a distinct preference to ride between adjacent ridges rather than the groove itself; very frustrating. Even using a lightweight moving iron or needle armature pick-up with a miniature steel needle may not help very much, as I have found out to my annoyance. Aluminium discs have other vices, but fortunately we don’t have to go into those here!
The correct type of point to plough a furrow in polycarbonate seems to be pretty much the same as was used to plough one in aluminium. That is, a blunt cone that is not too sharp. The first ‘cutters’ I made were quite wrong; we copied those used on lacquer discs, which (in their simplest form) have a vertical v-shaped chisel facet at the front, with two upwardly raked relief faces at the back. While some erratic success was achieved with these, it was not until my friend & colleague Mike Thomas, co-worker on this project since its inception some months ago, pointed out the successes achieved by the use of a conical point, as demonstrated on a number of YouTube videos by several different workers. Belatedly, I looked at a set of ‘cutters’ supplied with an early 1930s Ekco‘Radiocorder’ home recording device in my collection. This apparatus embossed on aluminium discs, and my example had never been used, though it had much deteriorated in damp storage conditions. So the packet of ‘cutters’ had never been opened.
We opened it. They were all rusty, but happily, one alone had a bright point. It was scanned at 3600 dpi, and gave the following result.
This was very interesting, because 84° was almost exactly the same point angle as chisel-face sapphire cutters for lacquers, belonging to Mike Thomas. If eighty-four was a Magic Number (it is, after all, twice ‘forty-two’, which as is widely known, is the Secret of Life, the Universe and Everything 8^) ), then we would adopt an 84° cone as a new standard. This proved to be a wise course. (On-line research showed that in the late 1940s, 50s &c., the recommended point angle became larger – up to 89°; but this was the era of microgroove, vinyl, and stereo., in which we have no interest.)
What about the material of the ‘cutters’? The shank diameter of a typical ‘cutter’ was, traditionally, one-sixteenth of an inch, which is so close to 1.5mm that almost nobody can tell the difference. Development work was done on grinding 1.5mm silver steel in a small lathe, against miniature abrasive discs. These discs were quite brittle, and though some of the ‘cutters’ worked OK, it was a tedious & unpredictable business. To save a long rigmarole here, we jump ahead to the final procedure which evolved. We switched to 1.5mm High Speed Steel rod, because it’s much harder, and also 2000 & 3000 grit metal 6" (15cm) grinding discs, all of which had to come from the Far East, though at very reasonable cost. Here is the final procedure.
1. Here a length of 1.5mm HSS is being sharpened to quite an acute point. HSS is so hard that it will take a very sharp point on a fine ordinary wheel.
2. The rod is aligned by the brass jig at the desired angle against a 3000 grit metal wheel, and gently ground to a cone.
3. The finished ‘cutter’ has to be parted off on the corner of a normal fine wheel; HSS is too hard to be cut with a hacksaw, or at least any of mine. Also, since the images were taken, we now use a pin-chuck for these operations, which is far more convenient – and safer!
Here is the final job. It’s been tested, & works OK, so it has had the Green Band of Approval painted on it. At 5/8", it’s a whisker short. The recommended length for cutters on the MSS lathes was 11/16"; but what’s one-sixteenth of an inch between friends? 8^) Seriously, one doesn’t want too short a cutter, otherwise it’ll start to point backwards & dig into the surface & probably bust if it’s a sapphire type. I bust one of Mike’s semi-irreplaceable sapphires in that fashion, early in the Project, but he was very tolerant about it. Happily, polycarbonate is much more forgiving than lacquer; and besides, you can always fix the external length before you screw up the cutter retaining screw – for yes, the MSS cutting head has such a screw.
CONCLUSION: The optimum ‘cutters’ are 1.5mm diameter High Speed Steel, 17.5mm long & have an ~84° cone.
PART III. The MSS ‘Utility Model’ cutting lathe.
November 2014. Here is the machine in its present state, which is very different from that in which it first saw the light of day. That would, I suspect, have been during WW2, 1939-1945. I will not say here much about the history of these machines, as I am currently writing an article on them (and the chap who developed them), for a specialist magazine, which has the policy of not printing anything that has already appeared on-line. N.B. The publication of this article has now occurred, so it will appear, in due course, on a nearby web-page, to which the link will be given here. So for the present, suffice it to say, that this was a more portable version of a disc-cutting machine that had been developed in the U.K. from the late 1920s, & brought to a good state by the early-mid 1930s. During WW2, the BBC and especially the Armed Services saw the usefulness of such machines, and had a lot of them made – probably many hundreds. These were usually a completely ‘portable’ mains-powered version in a carrying case, complete with amplifier and loudspeaker in the lid of the carrying case. They were also very heavy! I found just the chassis of one, without a motor, in the late 1960s.The disc cutting bug bit severely, and in the next few years I acquired several more cutting lathes – which is a completely different story, here irrelevant. The one above was sold to my friend Mike at least 30 years ago, and he used it successfully for some time, until it stopped working. He then acquired another, much larger & more recent one, but never ever commissioned it. Over the intervening years, perhaps fifteen or so, we often talked about cutting 78s, and moaned about the prohibitive price of lacquer discs, &c., &c. A couple of years ago I retired from work & could finally begin tinkering with projects long deferred. When the subject came up again, it was for the last time, because it was finally decided that both these long-neglected machines should be woken up, investigated, conserved, repaired & made to work again. It became something of a crusade with us. Weeks, indeed months have been spent on it, not to mention quite a bit of money, but in the end positive results have been attained. What follows concerns only the above machine, which I had re-acquired from Mike.
Here’s a rear view of the machine. The box is just cheap white 15mm contiboard, sprayed black on the outside. The motor is a new one, and was inexpensive: a 1400 rpm synchronous motor (the original was 1500 rpm) which is actually a fan motor for a commercial freezer. When it arrived, it rotated counter-clockwise, but since both end-caps were identical, it was easy to bore out the blank one & change the rotor around to give the desired clockwise rotation. A long brass drive pillar was turned, as the original motor had a long shaft & the replacement didn’t. The most radical innovation was the replacement of the original drive for the feed screw. This originally consisted of a vertical shaft passing up through the base-plate, driven by a belt from a pulley on the turntable bearing. No run-off scrolling was built into this machine, and variation of groove pitch was not possible while cutting. It was decided to fit a small stepper motor to drive the feed screw, which would enable both scrolling and also pitch variation ‘on the hoof’, if need be.
At this point, I make grateful acknowledgement to Keith Harrison, a senior member of the CLPGS*, an expert in constructing equipment for accurately playing early forms of sound recordings (especially phonograph cylinders) using modern technology. Without his help and lucid advice, I would never have been able to understand stepper motors!
Here you see the rear of the deck. All is self-explanatory. The original idler wheel was quite useless; it was a relatively small brass bush clad with a thick rubber surround. After probably 70 years, the rubber was rock-hard, polished with use and full of ‘indents’ where it has been allowed to rest against the drive bush and the turntable itself. We made a new one, following a concept I have long held, based on the superiority of modern materials. Rather than a small bush with thick rubber surround, we made a large bush of aluminium (with a central phosphor bronze bearing) and cut a groove in the edge, into which was placed a single neoprene rubber ‘O’ ring, slightly stretched. This worked perfectly well. Such ‘O’ rings are readily available in a vast range of sizes, and at negligible cost.
The essential concept of the MSS machine is very simple. A transit rod at the rear supports the transit arm, on which the cutting head is mounted, balanced on pivots. Four pairs of pivot-points are available on the cutting head, so that pressure can be varied. A spring-loaded knife-edge almost engages in the feed screw when the transit arm is lowered into the horizontal position. After the motor has been started, the lift-lower lever is slowly brought forward, and the transit arm descends a little more, lowering the cutter into the surface of the blank disc, and fully engaging the knife-edge. Simplicity itself. The transit arm is fitted with a roller at the front, which allows it to move smoothly along the front support arm, as the feed screw turns. Many other early (and indeed later) lathes had the whole weight of the cutting arm & head borne at the rear of the machine. This may have led to greater accuracy in cutting, but imposed a much greater load on the rear assembly, so that it had to be more heavily built, usually requiring two thick transit rods to share the load instead of just one.
The controls on the panel are simple. The coil in the cutting head, which is a moving-iron device, was originally of high impedance, as it was intended to be in the anode circuit of the output valve which powered it. It had a DC resistance of typically 1100 Ohms. Later versions used a low impedance coil, of DC resistance about 10 ohms, corresponding to a loudspeaker output. For simplicity of operation, a low impedance coil was wound & fitted, and is driven from a very standard sold-state amplifier, such as the NAD 3020A. Any similar amplifier will give more than adequate output. The head is obviously monaural, so just one channel of the NAD is used to drive the cutting head, while a little of the signal can be put on the other channel, to provide a monitoring facility. The stepper motor speed control is crucial: it needs to range quite widely. First, it should ‘low-range’ and offer the choice between fine, medium and coarse cut; but it should also ‘high-range’ and be much faster, in order to enable scrolling. More on this below.
The control panel hinges out, giving access to the simple circuitry inside. There is a 12 V DC stabilized power supply. Up to 1 Amp is available, but the motor consumes far less than this. The pulses to operate the micro-step board are generated by the simple NE555 timer chip. The micro-step board is set to 16 pulses per motor step for smoothest running. The stepper motor is a NEMA 17 type, which has 200 steps per revolution, so at 16:1 the board needs an input of 3.2 KHz for 1 revolution per second. This is about what the feed screw needs for recording, but obviously must go a lot faster for scrolling. So by splitting the resistance which determines the frequency into a 10 KΩ preset (on the board) and a 68 KΩ pot (on the front panel), the frequency range can be jiggered to about 2.4 KHz for fine groove, 3.5 KHz for coarse groove, and about 20 KHz for scrolling. In order to be able to set (and re-set) the cutting pitch reliably, a tiny portion of the output of the timer is taken via a 10 KΩ resistor & applied to a 0.047µF capacitor, which charges to a peak voltage. This voltage is inversely proportional to the frequency, and is rectified by a germanium diode, charging up a 1µF capacitor. The voltage on this last capacitor is indicated by the small meter on the front panel. When the lowest speed (fine cutting) has been arrived at, a 100 KΩ preset is used to show FSD on the meter. If the motor speed is increased, this voltage will drop, so we can use the meter to set our groove pitches. To scroll out, we just turn the pot fully clockwise, and the motor will speed up to about 400 rpm, giving a nice steady scroll. To stop scrolling, the pot is turned fully anticlockwise, and then the power to the motor is switched off. The ‘cutter’ stops moving & makes a locked groove.
The lathe is seen here in operation, ‘cutting’ a 10"(25cm) polycarbonate disc. The disc is held in place by a brass bush to stop slippage, though the drag is far less than that encountered when cutting a lacquer disc. In fact, the furrow can be ploughed without a securing bush on the inner half of the disc. But it is better to use the bush, as this type of turntable is usually slightly ‘dished’; i.e. very slightly lower in the middle than at the edge. This was done so that anyaluminium-based lacquer that was slightly warped would tend to flatten out. The polycarbonate discs we have had made are usually completely flat, though one or two have been slightly warped. This has had no negative effect, as the head rides up & down very well. Also – we almost forgot – before ‘cutting’, the blank disc is lightly sprayed from an aerosol of furniture polish. This is then rubbed over the surface of the disc to give a thin waxy or oily coating, which improves the adhesion of the point to the disc. All the 1920s/1930s machines that recorded on aluminum required such a coating. Light machine oil was often recommended, as was Vaseline. After recording, the polish is washed off the plastic disc, in case it might later solidify and clog the furrow.
Link to YouTube Video: http://youtu.be/T7nKa_aHyR8
* The CLPGS is, de facto, the British national society for enthusiasts of cylinder phonographs, disc gramophones, and the records played on them. It was founded in 1919, and it is for that reason that the name has never been changed to reflect its national and international status, for The City of London Phonograph and Gramophone Society is justly proud of its ninety-odd years of existence under that name! If you are an enthusiast of such machines and records, please visit the society’s website – perhaps you might wish to become a member? You will be very welcome! www.clpgs.org.uk .
Page slightly revised 14th November 2015.