.

von OTTO J. M. SMITH, Ph.D. - "Dept. of Electrical Engineering, University of California, Berkeley, Calif.
.
Describing the many superior features of this plastic as a record material.
.

.

..... is one of the better new phonograph record materials. It has very little surface noise, is durable, does not collect dust, and does not become disagreeable to listen to as it wears out.

This material was chosen after a study of the characteristics of surface noise and the wearing qualities of conventional phonograph records.

A meter was made whose reading correlated well with subjective opinions of the disagreeableness of surface noise. This was used to test a great many records of different labels and experimentally pressed records.

The results of these tests indicated that the ideal phonograph record material should have high impact strength, low dielectric constant to prevent the accumulation of dust particles in the grooves, low thermal expansion, low water absorption, and resistance to abrasive wear.

A survey of the characteristics of available plastics indicated that polyethylene would be the best material. (Comparative data with other record materials is given in Table I.)

Records were pressed from a number of different plastics, and surprisingly enough, the polyethylene records were the best.

It was discovered that measurements of output voltage from blank record grooves did not correlate at all with subjective opinions of the disagreeableness of the scratch noise.

Most of the listeners could tolerate a much greater low frequency scratch voltage than in the 5,000 to 10,000 cycle-per-second range.

The playback unit consisted of an Astatic MLP-1 crystal pick-up in a one-ounce arm, a high-frequency compensating circuit, an Altec-Lansing 74db amplifier, and an Altec-Lansing dual coaxial speaker in a sound-proof room. The scratch meter was a peak vacuum-tube voltmeter with a 0.3-second mechanical time constant.

A two-stage, tunable, high-pass filter was inserted between the speaker and the VTVM, and a series of tests were conducted on records with both low and high frequency scratch.

The filter was adjusted until the scratch meter reading correlated with the subjective listening tests. This particular setting is given in Fig. 1, and the frequency characteristics of the entire scratch meter are shown in Fig. 2.

The filter cut-off frequency was 3.000 cps, where the response was down 6db, and at 1.000 cps the slope of the cut-off was 12db per octave.

The surface noise was measured on new vinylite and shellac records of seven different manufacturers. The noise was measured at three places on the record, the outer edge, the inner edge, and half-way between.

Weighted averages of these figures for two or more new records, and in some cases four different records, are shown in Fig. 3. The superiority of vinylite over the usual shellac compositions is quite apparent.
.

The popular records were worn out by 100 plays on an automatic machine, using an LP-2 cartridge in a 2-ounce arm. A microscopic comparison of the new and worn records revealed to a certain extent the mechanism of wear. The new records had a smooth high-gloss surface in the groove, due to shellac coating the diatomaceous earth, cotton fibers, and other filler.

In a worn record, particles of material were removed, exposing to the needle the rough sides of the filler particles.

For example, an exposed 5 mil diameter cotton fibre would produce one cycle of a 5600 cpsa note. The material worn off of a shellac record was singular shaped, while that which accumulated on the vinylite record during wear looked like slivers.

The magnitude of wear and the volume of surface noise was greater in loud musical passages where the needle had high lateral velocities.

Unfortunately, shellac is too brittle to be used without fibrous fillers, though a homogeneous material would probably have less surface noise.

Figure 4 shows the results of these tests. Listening tests also indicated that the quality of the music had been changed due to wear. The reason for this was not apparent, but it may be that wear changes the shape of the groove and the wave shape of the reproduced audio signal. This would appear as distortion.

A satisfactory record material should be a homogeneous plastic, non-brittle, which could be pressed at low temperature. It would not have to be hard to have good resistance to abrasion. In fact, rubber is well known as a flexible, yet very hard-wearing material.

Vinylite becomes electrified and collects a considerable amount of room dust, so that even a new record may sound poor after being played and handled in a dusty room. A low dielectric constant is therefore desirable to minimize electrification.

Polyethylene is one of the best materials available in this respect. As Table 1 shows, it has half the dielectric constant of vinylite. It is a soft, greasy-like polymerized plastic used for high-frequency insulation and for bottles for corrosive chemicals.

In an experimental demonstration, it can be used as a lubricant. It is possible to rub the metal of a noisy sleeve hearing with solid polyethylene and leave enough residue to give an audible improvement in the lubricating action.

When used for a phonograph record, the polyethylene does not wear off as other record materials do. It only deforms as the record wears out. A good record material should have a low or negative thermal expansion to reproduce accurately the press die. Polyethylene is again one of the best available materials in this respect. It is also practically impervious to moisture, and is therefore free from warping due to either temperature or high humidity.

Many different kinds of polyethylene records were pressed. Some were

(a) solid polyethylene,
(b) polyethylene with carbon black,
(c) 3-mil poly sheet over cardlward,
(d) 3-mil poly sheet over conventional shellac, and
(e) 3-mil poly sheet over floor lineoleum.

On (d), trouble was experienced with volatile components of the shellac forming minute bubbles under the polyethylene film, which would begin to break through after about 20 plays.

On (e), very good records were obtained. Bonding to the linoleum was good, and the time required to cool in the die was quite short.

The importance of this last item is illustrated by Fig. 5, which shows the variation of profit with pressing-cycle time for a small seven-inch record plant with special design for multiple-die presses.

The total heat required to press a polyethylene record is twice that for a vinylite record. But since the former has almost twice the thermal conductivity of the latter, and since polyethylene has what appears to be a solidifying point, it can be removed from the presses in less time.

The experimental records were tested in the same manner as the commercial records, and the results are shown in Figs. 3 and 4.

The initial surface noise of polyethylene was of the same magnitude as vinylite records made by the A and B companies. The increase of surface noise with wear was also of the same order of magnitude.

On listening tests, the worn polyethylene was superior to the worn vinylite. Polyethylene wore out by losing volume at the high frequencies. The worn record was just as pleasant to listen to as the new record, though the bass was predominant.

The material is soft enough to be scratched with a fingernail. A deep scratch with a needle can be heard on playback as a "bump". The smooth contours which polyethylene takes when scratched does not produce the disagreeable noise that a scratch on a shellac record does. The dust which it colected after being left out in the room could be easily blown off.

The increasing use of polyethylene may lower the material cost until it can be considered not only for good quality records, but also for popular and children's records, and perhaps for dictating machines and home recordings.