equipment reviews

(Revox A77HS [high speed])

MANUFACTURER'S SPECIFICATION (Agfa PE36 at 38 cm/s). Professional mains transportable stereo recorders with Tachometer capstan motor .speed control. Tape Transport: Direct capstan drive, separate motors for fast forward and reverse wind. Wow and flutter: 0.04% maximum. Tape slip: less than 0.2%. Spool capacity: 26.5 cm, NAB hubs. Frequency response: 30 Hz - 20 kHz 1.5dB. Distortion: 2%. Equalization: NAB or IEC to order. Signal-to-noise ratio: 60 dB. Crosstalk: (1 kHz): 45 dB stereo, 60 dB mono. Tape speed, 38 and 19 cm/s. Bias frequency: 120 kHz. Inputs: 0.15 mV at 50 ohms to 6 K or 2 mV at 100 K (microphone, switchable); 2 mV at 33 K (radio); 40 mV at 1 M (auxiliary). Outputs: 2.5 V at 600 ohms (line); 1.2 V at 2.5 K (radio). Weight: 15 kg. Dimensions: 359 x 215 x 413 mm (plus spool overhang). Prices: On application. Manufacturer: Willi Studer, CH-8105 Regensdorf, Zurich, Switzerland.
Distributor: C E Hammond & Company Limited, Lamb House Church Street Chiswick London W4
To meet the demand for a basic transportable professional recorder in this country, a number of A 77 machines were modified by the distributor for 38 cm/s operation. This entailed a larger capstan diameter and alterations to the record and play equalization to meet the 50 ÁS NAB standards for both 38 and 19 cm/s. As I have access to several of these recorders, David Kirk asked me to prepare a review under the conditions suggested in his October Editorial, i.e. I was to be allowed to set the bias and equalization for optimum performance on any given type of tape. As he forecast, such tests used up a vast amount of time as test tapes, for example, had to he checked one against the other to make sure that the published responses were as accurate as possible and did not include small errors in the calibration.
It was therefore with some consternation and considerable interest that we learned, only a few days before publishing deadline, that a Swiss high speed recorder with a number of mechanical and electronic modifications had arrived in this country. This was an opportunity not to be missed and Revox Ltd. had it down to me within a few hours.
In the time available I was not able to repeat all the tests or optimize the bias and record equalization for the best possible response with the different (IEC) equalization. Nevertheless there were significant improvements in wow and flutter, signal-to-noise ratio and cross-erase which take the 77 design a further step towards perfection. The modifications to the Swiss machine listed were as follows:
(a) Metal spool carriers to eliminate static charge on metal spools and (together with b) to give slightly faster and more even tape winding.
(b) Roller bearing on left-hand tape guide to give lower friction and appreciably less wow and flutter.
(c) Damped transient absorber near left hand guide to give better starting.
(d) Wider pinch roller to reduce tape slip (my measurements on the width of the old and new pinch rollers showed them to be nearly identical, so perhaps this modification is yet to come).

Fig. 1 shows the pen recordings of cumulative wow and flutter at 38 and on the UK modified (HS77/NAB) recorder. At 19 cm/s the WHM fluttermeter reading remained steady at 0.03 % RMS for long periods. Only occasionally did it rise to 0.04%. as record and play speed fluctuations come into step for short periods. At 38 cm/s, the meter reading remained rock steady at 0.02% RMS with very occasional excursions to an absolute maximum of 0.025 %.
On the Swiss machine, the 19 cm/s reading was 0.03% and I could not coax it any higher by slipping or phasing the tape for cumulative adding of any of the wobble components. At 38 cm/s we hit an all-time low of 0.015% RMS combined wow and flutter with occasional short rises to an absolute maximum of 0.02 %.
Fig. 2 shows the responses at line output when playing 19 and 38 cm/s NAB test tapes with 50 ÁS high note pre-emphasis and 3180 ÁS bass pre-emphasis. It will be seen that playback equalization is exact within 1 dB to 10 kHz at the lower speed, and to 15 kHz at 38 cm/s. The solid curve in fig. 3 shows the electrical equalization as measured from head terminals to line output.
The HS77/lEC was tested with IEC 70 and 35 ÁS test tapes to give the play-only responses of
fig. 4. These response are level within 1 dB limits to 20 kHz at either speed with a very slight bass roll-off below 100 Hz at each speed.
The UK HS77/NAB was tested with the recommended Agfa PE 36 LP tape, bias and HF recording pre-emphasis was carefully adjusted at each speed to give the overall responses of
fig. 5 which are within the specified limits.
1 kHz 32 mM/mm reference tape level was recorded at + 2 dB on the VU-meter scale. Total harmonic distortion at this level was 0.8 %, mainly third harmonic. Unweighted system noise, with no tape passing the heads, was - 56 dB on reference tape level. Weighting to the IEC 'A' response, which matches that of the ear at low listening level, gave a reading of -63 dB.

Bulk erased tape was 61.5 dB below reference and machine-erased tape was exactly 60 dB below 32 mM/mm 1 kHz reference tape level.
All the above distortion and noise levels were identical at 19 and 38 cm/s.
The Swiss HS77/IEC recorder was tested with the identical Agfa PE36 LP tape to give the overall frequency responses of fig. 6. It was obvious that the top track bias was a bit low at 19 cm/s, resulting in the rising high note response and very slightly lower mid and low frequency responses shown by the solid curve. The bottom track was nearer the optimum bias and pre-emphasis settings for the recommended Agfa tape as shown by the dotted response at 19 cm/s.
At 38cm/s top and bottom tracks were almost identical and the response was within the specified limits from 30 to 20 kHz.
Unweighted system noise was -58 d below reference tape, level and weighting it to the IEC 'A' response gave the very low reading of -66 dB. This reading remained the same at 19 and 38 cm/s.
Bulk-erased tape noise was -62.5 dB at 19 and -64 dB at 38 cm/s. Machine-erased noise was -60.5 dB at 19 cm/s and -62 dB at 38 cm/s. These noise differences with tape speed are mainly due to the change of extreme HF response as the playback equalization is switched from 70 to 35 ÁS as shown in
fig. 4.
The 19 cm/s 1 kHz harmonic distortion at 32 mM/mm was very slightly higher than the other recorder at 0.9% and it dropped to 0.75% at 38 cm/s. This could be due to the slight under biasing of the top track on which the measurements were made, or it could be due to the 3 dB difference in equalization at 3 kHz (3rd harmonic) as again shown in
fig. 2.
There was another small difference in the two recorders which could only be detected if the machines were used for industrial or scientific purposes, such as the production of test tapes, where 0.5 to 1 dB differences in extreme high note response were significant. If a mono recording of 10 kHz was recorded on the top track only of the NAB machine and erase and bias applied to the bottom track, either by switching heads or reversing the tape and leaving the bias and erase on the top track, then the 10 kHz signal dropped by 0.5 to 1 dB whenever the other track was erased. The trouble was traced to thc erase head, as inserting a slip of paper between the erase head and the tape stopped the effect. It was also proved that the fault was not present on stereo recording.
Measurements showed that the voltage across each section of the erase head was 15 V on stereo recording,- but rose to 25 V when only one erase head was in use. . This indicates that on mono the erase head is over powered and cross coupling within the head powers the unused head with 6 to 7 V which is sufficient to just begin to erase high frequencies on the unused track.
Screwing in the ferrite core of the dummy load coil L601 dropped the volts to 20, reducing the effect slightly, but this caused a change of volts on the record head's on switching from mono to stereo which could not be tolerated as the frequency responses were markedly altered.
On the Swiss HS77/IEC the fault did not show up at all, although the same change of volts appeared across the erase heads on switching from mono to stereo I can only assume that the impedance or depth of the gap in the erase head has been altered slightly as the type number has been changed from Z740 to Z940.
The last few paragraphs, showing concern over a 0.5 to 1 dB difference in response, indicates that these HS77 machines are in an entirely different class to the average domestic recorder and that testing and adjusting these machines to such high standards is a long and time-consuming job. It is made somewhat easier on these recorders by grouping all presets for bias, meter, record equalization for each speed and track, etc., along the base of the recorder so that when it is removed from the cabinet they are fully accessible and well identified for adjustment with the recorder in a vertical operating position.
These recording instruments set new standards in wow, flutter, frequency response and signal-to-noise ratio which must ultimately benefit the design of transportable studio recorders.

(Remark: Historical information, original published approx. in 1970. Please note that the adresses and phone numbers may not be correct any longer.)

Back to the A77 MK II main page