Behind the Scenes (May 1977)

In March I began my report on the remarkable new Ampex ATR-100 tape recorder with an in-depth look at the inner workings of the transport mechanism and associated controls, and the head block assembly. This report will conclude with an examination of the sophisticated electronics of the ATR-100, and the results of test measurements on performance parameters.

As noted in my description of the electronics assembly, there is a plug in the main audio board for each channel (Ampex calls them PWA...Printed Wiring Assemblies) and a PADNET (Parameter Determining Network) plugs into the rear of the PWA. On each main audio board, there is a vertical array of screwdriver-adjustable potentiometers covering such functions as Reproduce Gain, Sync Gain, Reproduce Equalization for high speed (30,15 ips) with high and low frequency adjustments, Reproduce Equalization for low speed (7 ½, 3 3/4)

with high and low frequency adjustments. Then there is Record Gain and Record Equalization for high and low speeds, and bias adjust. On the PAD NET board are two jumper pins, for high and low speeds, and according to the way in which they are positioned, they make available any pairing of the four tape speeds...30, 15, 7 1/2, 3 3/4. ATR-100 units are normally shipped from the Ampex factory set for 15, 7 1/2 ips operation, with a four speed master bias. If a speed is selected via the switch on the transport control panel for which the recorder has not been properly set up, i.e. 30 ips instead of 7 1/2 or 15 ips, a red light will glow on the "lockout" indicator and the recorder will not operate in either the Record or Playback modes.

In addition to the main audio board PWA for each channel (in my case, four PWA, since I have a four-channel half-inch unit, with a stereo head assembly as well), there is an audio control PWA. This audio control board has a master crystal oscillator, which is a multi-function device in this recorder. There are separate divisions of this crystal to avoid interaction. First we have the erase circuit with a frequency of 144 kHz, and there is four-speed master bias at a frequency of 432 kHz.

Other subdivisions of this crystal handle the transport servo control.

On this audio control board are jumper pins which can be positioned to provide two-speed dual master bias operation or four-speed master bias operation. When in the four-speed mode, a single master bias level is provided for each speed. Whatever speed is selected on the transport control panel, master bias level is automatically switched for each speed.

While it is true there are bias controls on each of the main audio boards, it is much easier to use the master bias level control to set bias on all four channels simultaneously, and ignore the individual controls. Getting back to the main audio PWA boards in the PADNET section, there is still another jumper pin which controls a function Ampex calls PURC...which stands for "pick-up recording capability." In Record mode, the erase and record heads are energized simultaneously, and since there is a 1.46 inch gap between erase and record head, there is some over-recording on unerased tape and a loud click is heard on the recording. With PURC, separate erase and bias amplifier circuits are provided and, in essence, when Record mode is entered, the erase is energized first, then after a delay (97 milliseconds at 15 ips), the bias amplifier comes on. On cessation of recording, erase circuit shuts off first, then the delay and the bias amplifier goes off. Voila! No clicks!

Level Controls

The ATR-100 as normally used is fitted with input/output modules, the number of them depending on the track configuration. Each module has a meter which may be set to operate as a standard VU meter or with EBU ballistics for peak indication. There are record and reproduce level controls with a switch to set them in manual or pre-set modes. Red "confidence" lights glow when bias and erase signals are present at the heads.

A record calibration control is used to set input monitoring level and meter indication for "off-tape" levels. A headphone jack permits monitoring with 600-ohm phones. All input and output connections are with standard XLR male and female connectors. Depending on how the connectors are wired, balanced or unbalanced line inputs and outputs can be set up. Balanced input impedance is 50 kilohms.

Variable input level can produce a maximum of plus 40 dBm. Balanced output can be 50 ohms or 600 ohms.

At 600 ohms, the maximum output level can reach an astonishing +28 dBm! Operating level of the system is 370 nWb/m at 0 VU, which is used for Ampex Grand Master tape, and is 6 dB higher than the 185 nWb/m reference level of Ampex alignment tapes. As shipped, the pre-set mode switches are adjusted for +4 dBm input and output line levels.

In my initial description of the Ampex ATR-100, I pointed out that one of the units most interesting capabilities was that it could record square waves, and thus was essentially free from phase non-linearities. The question of the audibility and the importance of phase non-linearities notwithstanding, it would seem reasonable that if these anomalies could be corrected without inordinate complexity and expense, it would certainly be worth doing. Alastair M. Heaslett, Senior Staff Engineer for Ampex and one of the principal designers of the ATR100, gave a most interesting paper at the 55th AES convention at the Waldorf, on "Phase Distortion in Audio Magnetic Recording." Mr. Heaslett points out that "anyone who has ever attempted to record square waves on an audio recorder will have seen the effect of the resulting phase non-linearity. The leading edges of the reproduced waveform have a large amount of overshoot. In addition, if the reproduce head resonance were not adequately controlled, the high frequency "ring" after the leading edges will be very evident." Mr. Heaslett goes on to note the importance of strict adherence to the equalization standards that have been established for magnetic recording, and then has this to say..."some attempts in the past have been made to provide corrections for phase non-linearities in audio recorders. However, the methods used for correction have principally been applied to the reproducing side of the system. When adjusted correctly, they certainly produce excellent square wave and transient response. This, however, is only true with the same type of tape, identically biased. In other words, this approach does not yield a recording implicitly compatible with the equalization standard. If such a recording were reproduced on our "ideal" reproducer, there would no longer be an accurate time domain representation of the recording." Mr. Heaslett then describes the special ATR-100 record equalizer which preserves the integrity of standard equalization from 3 3/4-to-30 ips, while essentially maintaining phase linearity. The equalizer is far too complex to describe here, but the important thing is, in Mr. Heaslett's words, "it has the very real advantage that a recording made using it, will play back on any conventionally equalized audio recorder with the improved phase linearity."

Machine Measurements

Okay, we've poked our noses into all the nooks and crannies of the ATR100, and now it is time to see what kind of performance such a sophisticated machine affords....) must at this point gratefully acknowledge the assistance, both morally and technically, of good friend Frank Dickinson.

Frank runs the "9 West Recording Studio" in Bloomfield, New Jersey, as well as an elaborate scientific machine shop and is considered a master in the care and feeding of Ampex and other professional tape machines. He grinds his own stainless steel spindles for his rebuilt motors, balances them with painstaking precision, and they are super smooth running and absolutely whisper-quiet. Frank loaned his fine dual-trace Tektronix scope to the cause. For distortion measurements, both total harmonic and inter modulation, we used the new Sound Technology 1700B analyzer. This is a great tool and makes measurements with an ease undreamed of a few years ago. I'm going to do a run-down on one of these units before long.

Through the courtesy of Eli Passin of Gotham Audio in New York, wow and flutter measurements were made on the ultra-sensitive EMT 424 flutter meter. I should point out that Ampex quotes their distortion figures for third harmonic, while we were measuring THD with the 1700B. I should also note that all measurements on the ATR-100 were made using Ampex Grand Master tape. Frankly, the machine's high performance is built around this tape, although I should also mention that the ATR-100 has the erase and bias frequencies and drive capabilities, plus bountiful headroom, to cope with super tapes not yet formulated. This machine should have no problems with the iron-particle tapes now under development. Just as an aside, even the fine Ampex 406 tape couldn't quite match the performance figures of Grand Master, nor could three other brands of high quality tape.

All tests were performed at 15 ips, using the two-channel stereo head assembly. Ampex spec for the ATR-100 for third harmonic distortion was 0.1 percent @ 1 kHz. We measured THD C 0 VU, C 1 kHz at 0.65 percent; same conditions at 10 kHz, 0.85 percent; at 30 kHz and minus 5 VU, the THD was 1.4 percent. Three percent THD was reached at +11 VU C 1 kHz! Inter modulation distortion was measured using the SMPTE 4-to-1 ratio of 7 kHz and 400 Hz. Ampex spec is 1.0 percent at 0 VU. We obtained 0.8 to 0.9 percent at 0 VU. Three percent IM occurred at +6.5 VU. Frequency response was less than ±2 dB from 20 Hz to 20 kHz (meets spec), was down 3 dB at 26 kHz, and down 5 dB at 30 kHz.

Wow and flutter, ANSI/DIN weighted C 3150 Hz, was an astonishing 0.015 to 0.02 percent. Signal-to-noise ratio (measured at 9 dB above the operating level of 370 nWb/m) at 15 ips, NAB, unweighted, was -71 dB. These are obviously impressive figures that place the Ampex ATR-100 in a class by itself.

Overall Impressions

On an overall basis, the ATR-100 is the easiest tape machine I have ever used. Tape threading is actually simpler than many consumer machines, and once you master the little "jiggle" which locks in the reel servos, you're in business. The controls are quite explicit and a joy to use. Tape handling is the most positive ever, but very gentle at the same time. The exceptionally good tape motion and good head wrap result in almost non-existent modulation noise, and tape playback is smooth and ultra-clean. Needless to say, recordings made on the ATR-100 are virtually mirror images of the source. A nice plus is the very quiet operation of the machine, and although used for many long hours, it never became more than warm. The Ampex 440C was and is a fine recorder, but it was the end result of an evolutionary design. The ATR-100 is in many respects a revolutionary machine that has brought analog magnetic recording to a new peak of achievement. Several engineer friends have expressed two interesting thoughts.... One said that the ATR-100 was probably built and is so constructed that it should be easy to convert to digital use; the other said that after checking out the ATR-100, he doesn't see how they could carry the analog concept much further, and it might be quite likely that the ATR100 is the last of the breed. Premature thinking? Who knows? All I know is that it is one helluva tape recorder, and I'm going to enjoy it for what it is, and leave the speculation to others!

(Source: Audio magazine, May 1977; Bert Whyte )

= = = =