By J.D. HIRSCH
TEST RECORDS and PHONO CARTRIDGES: Our test reports on phono cartridges refer to certain specific test records used for measurement of frequency response, channel separation, distortion, and other aspects of cartridge performance. To simplify comparisons between cartridges, we have tried to standardize our test records and procedures, but from time to time records must be deleted from or added to our active list. In many respects, measuring cartridge performance seems ridiculously easy-simply play a suitable test record and measure the electrical outputs at the cartridge's terminals. Unfortunately, it is not easy to separate the limitations of these two imperfect links in the measurement chain. Most frequency-response test records have a number of frequencies (or a continuous sweeping tone) recorded at a constant velocity over a major portion of the audio-frequency range. To limit the groove amplitude on the record to a reasonable value, it is customary to record with a constant amplitude characteristic below 500 Hz.
Sometimes the lower limit of the record is put at 500 Hz and constant-velocity recording is used throughout. Since most cartridge aberrations occur at middle and high frequencies, this is not a serious limitation in testing.
Test records designed for professional use invariably have a sweep tone that is synchronized with the drive speed of the chart of a graphic-level recorder. It is possible to use spot frequency measurements as a rough indication of cartridge response, but they lack the detail necessary for a thorough evaluation. A comprehensive series of test records has been produced by CBS Laboratories, and we-along with many others-use the CBS STR100 disc for the measurement of frequency response, separation, and output level. The sweep bands of the STR100 are synchronized with the General Radio 1521B Graphic-Level Recorder. The Danish Bruel & Kjaer test instruments are widely used throughout the world. B&K, and a number of other manufacturers, produce test records synchronized with their chart recorder. Unfortunately, however, the frequency calibration of their sweeps is not compatible with the General Radio recorder.
Since the introduction of CD-4 cartridges, there has been a need for test records useful to 50,000 Hz (50 kHz) instead of the mere 20-kHz limit of most stereo test records. The CBS STR120 has been available for some time, and we have used it for testing stereo cartridges whose response extended appreciably above 20 kHz. However, its inherent performance is not adequate for testing today's CD-4 cartridges, and we now use the JVC TRS1005 for that purpose.
The TRS1005 sweeps from 1 kHz to 50 kHz with constant recorded velocity, and it is synchronized with the B&K recorder, so we use a conversion scale with our General Radio charts to calibrate the frequency axis.
Any specifications of a cartridge's frequency response or channel separation must identify the test record used for the measurement if they are to have any meaning. No two brands of records we have used are alike. and there can even be significant differences between different pressings of the "same" record made a couple of years apart. A record can be calibrated independently of a phono cartridge by optical means, or with a variable-speed turntable, but these are somewhat involved procedures and are hardly justified except for a cartridge manufacturer. We must content ourselves with measuring the frequency response of various cartridges with the same test records, as much as possible, thus providing at least some basis for comparison if not a precise absolute measurement.
Determining the tracking ability of a cartridge also presents problems. Tracking is a function of the recorded velocity (or amplitude) as well as frequency, and we know of no test record suitable for a complete measurement of this important parameter. We have for years used two records to evaluate tracking requirements at low and middle frequencies (neither has been available for some time, however). The monophonic Cook Series 60 record has a number of low-frequency bands recorded with Fletcher-Munson compensation, so that the lowest frequency of 32.7 Hz has an extremely high amplitude of about 0.034 cm. (This level of groove modulation is easily visible with the naked eye at a distance of several feet!) When playing this high-level band, the sound is often quite distorted, even when the stylus manages to stay in the groove. A few cartridges can reproduce it without serious distortion, and others simply jump out of the groove even when used at their maximum rated force.
Between these limits, we judge the low-frequency tracking ability of a cartridge by how much vertical tracking force is required to play this record, and by the amount of audible distortion. This is essentially a test of the static compliance and peak amplitude limits of the stylus.
Another out-of-print record in our library is the Fairchild 101, which was made to demonstrate the effectiveness of anti-skating compensation when Fair child introduced an arm with the feature more than a decade ago. This record has short 1,000-Hz bursts (about one second long, to avoid overheating the recording cutter) at a level of 30 cm/sec. By watching the two outputs of a stereo cartridge on a dual-trace oscilloscope, the anti-skating compensation can easily be adjusted for symmetrical peak clipping of the waveform. As we use it for a tracking test, the vertical tracking force is in creased (up to the cartridge's maximum rating) until the output waveforms show no further improvement. The higher of the tracking forces obtained with this and the Cook record is used for our sub sequent tests and listening to the cartridge, and that is the force we recommend for its use.
Another record we sometimes use for a similar purpose is one of a series produced by the German Hi-Fi Industries Association. Among other things, it has 300-Hz bands recorded at increasing velocities up to about 20 cm/sec. By listening, one can immediately hear the beginning of mistracking and the effect of increasing the tracking force. When using these records to compare two cartridges (at the same tracking force), it is not uncommon to find one of them excel ling in one or two of these tests, but al most never in all three. Despite its lack of rigor, we find this a quick and useful way to judge the tracking ability of a phono cartridge.
A more quantitative approach is to measure the distortion of the cartridge output as a function of recorded velocity. Bearing in mind that cartridge tracking ability is a function of frequency, it is obvious that more than one kind of test is needed for this evaluation. We have used the monophonic RCA 12-5-39 record as one indicator of lower-mid-range tracking and distortion at peak velocities from about 4 cm/sec to 27 cm/sec. This record has 400- and 4,000-Hz tones, in an amplitude ratio of 4 to 1, so that the intermodulation distortion generated can be read directly with a standard IM distortion analyzer.
But this RCA 12-5-39 is a 78-rpm record (!), and this has caused some problems in its application. Often we test integrated record players that do not have the 78-rpm speed. Also, slight record warp and eccentricity can cause severe meter fluctuations at 78 rpm. At one time we tried to use the Shure TTR102, a 33 1/3-rpm record with the same recorded bands as the RCA record, but we found extremely poor correlation between the results with the two records.
Shure has developed a comprehensive tracking-test record, the TTR103, which we have been using for over a year. Un fortunately, most of the test data derived from it cannot be easily correlated with earlier tests, so that we are effectively "starting from scratch" in our cartridge tracking tests. The TTR103 has three distinct sections, which require different instrumentation, for judging different aspects of cartridge tracking performance. It has the same 400- and 4,000-Hz IM test tones described above, at velocities from 15 to 30 cm/sec. Then there are 1,000- and 1,500-Hz tones at equal amplitude, with peak velocities from 15 cm/sec to 30 cm/sec. Finally, a unique test involving shaped 10.8-kHz tone bursts, evaluated with special filters, supplies quantitative data on high frequency tracking at velocities from 15 to 30 cm/sec.
THE handful of records we have de scribed are currently used for the bulk of our cartridge testing. However, we have about thirty different kinds of test records, many of which are used as circum stances warrant. As a result, our test data and curves will not always be directly comparable to the data on a cartridge we tested a year or two earlier.
This is one price we must pay for progress, as cartridges continue to be improved and better test records are developed to measure them. However, our verbal evaluations can be considered comparable--except, of course, that "the best" in any aspect of performance three or four years ago may already be second-best today.
-------------------
Sequerra Model 1 FM Tuner; Jensen Model 15 Serenata Speaker; Superex EP-5 Stereo Headphones; Technics RS-676US Cassette Deck
----------------------
==============
Also see: