Reel to reel analog tape recorders on stage...

Without any doubt the Revox series of reel to reel taperecorders have once been the workhorses for all contemporary music production involving electronic technology in the timespan 1956 to 1990. In the sixtees model G36 (model A was introduced in 1956!) was dominant (a very heavy, 22kg vacuum valve machine), in the seventies the A77 (produced from 1967 to 1977) model was predominant and in the eighties superseded with the B77 (introduced in 1978) and PR99 models.

For all concert uses only the 2-track type, preferably with speeds 7.5 / 15 ips (19 cm/s and 38 cm/s) were to be used. Types with 9.5 cm/s and 19 cm/s speeds were also widespread.

Taperecorders on stage were used in a broad range of applications:

• - playback of prerecorded tape material in combination with instrumental music making or performance. The term 'composition for instrument and tape' is today still used by quite many composers although the tape has long been replaced by an audio track on a computer...
• - use as an audio delay line. The minimum delay time is determined by the distance between recording and reproducing tape heads. With a minor preparation is was (and is...) possible to enlarge dis distance by pulling the tape out between these two head and feeding it through an external pulley.
• - use as echo machine: this was done by recording on one track and sending the (delayed because of the distance between recording and reproducing tape heads) reproduced signal back to the other channel, thus creating a delayed feedback system.
• - use for tape loops, using one or two machines. Stockhausen's 'Solo mit Rueckkopplung' is a good and sophisticated example. Brian Ferneybough's Time and Motion studies as well.
• - use for inverting sound and music by playing the tape backwards
• - use for pitch transposition by changing either the capstan wheels or the capstan motor speed.

Although it is perfectly possible to replace these analog machines nowadays with digital, mostly computer based, technology we are convinced that an essential part of the performance gets lost by doing so.

However, using the old technology raises problems, as well functioning reel to reel taperecorders are harder and harder to find. In almost all cases the machines to be used will be between 60 and 30 years old and will be in need for maintenance and repair. Undertaking such a task requires a good knowledge of analog electronics as well as servicing experience, skills that are becoming rare.

Without a proper set of schematics, dont even try to repair a Revox or any other professional reel to reel tape recorder. The schematics are all available for download from the Studer website.

Here is a link to the circuits for the Revox A77 model on our own site.

Case reports on repairs done by us on  Revox tape recorders:

Revox model A77, 1972 2-track
This machine didn't do anything when switched on. First we discovered that the housing has a couple of pins that insert in a plastic cap covering the mains connection. They break the mains power as soon as the machine is removed from its cabinet. So first thing we did was inserting two bolts in these holes. This at least connected the machine to the mains power. But, this was not the real problem... We checked the 21V power supply to discover this voltage was absent. Not the 630mA fuse was blown, but the bridge rectifier appeared to be open circuit. We replaced it with a 4 A type mounted on the side of the chassis, as it was to big for the original PCB. Still it wasn't working, although the 21V power voltage was found o.k. now. Further checks revealed electrolytic capacitor C307 (250uF/25V) was shorted. Resistor R313 (39 Ohms) was found to be burned. We replaced both parts. The capacitor with one rated for a somewhat higher voltage.
Therewith the job was done and the machine back in operational condition.

Revox mode A77, 1975 2-track
On this machines the lights came on, the takeup and rewind motors worked as normal but the capstan motor appeared dead. At the most, on the switch on surge, it would move a bit and then stop again. The 21V power supply was checked and found out to be o.k. and within specs. We replaced the 3.5uF AC capacitor for the motor with the result that the motor started spinning, but way too slow and without any torque. We checked the tacho sensor with the oscilloscope and this worked o.k., however as soon as we reconnected it the motor started turning at a way to fast speed... This was going to be a tedious job, as this meant we had to remove the motor control board. Checking the board made us replace all electrolytics and tantalums with modern and fresh types. The power transistor driving the recifier bridge in the motor control had a CE short. Its a 2N3583 in a cute TO66 package. Of course, this type is since long no more on the market. Looking up the specs, we decided to subsitute it with a MJE13007 high voltage type. This one comes in a standard TO220 package and thus mounting and wiring  had to undergo some changes. Still it wasn't working... We replaced the NE555 timer chip as well as the power transistor PNP driver (in the circuit specified as a BC178 but what we found soldered in was a BC308. This may have been an earlier repair... As we didn't have exact replacement parts in stock, checking the specs of the original transistor made us turn up a 2N3906. Before reassembling we checked the board's functionality by connecting a pulse generator (800Hz) to the tacho input, a lab power supply set to 21V. The board draws some 50mA. By connecting the speed switch wire (red) on and off the 21V we could verify the pin 1 output of the (obsolete and hard to replace) TBA931 reacting properly. The base drive for the power transistor changed accordingly. So far so good. We reassembled the board and fired up the machine again.
Success! It works again, the motor nicely spinning.

As we noticed a little bit op pop-corn noise on the outputs of the playback amplifiers, we replaced the following capacitors: C803 (1600uF) -here we replaced it with a 2200uF type, as 1600uF is no longer on the market; C811 (100uF), here we inserted a tantalum polymer type; C814 (10uF); C813 (25uF), replaced with 22uF. Here is a view on the board before the replacements:

Here is a view on the internal guts of the machine:

A final problem with this machine was that due to unthoughtfull treatment by someone not too familiar with fine equipment, the knob for the rotary turn-on switch was broken off. This obviously was not an electronic problem but nevertheless a pretty tedious one to solve. We had to make a new bus fitting over the 6 mm axle on the lathe in order to replace the knob.

A few general remarks.

If you Google the topic on Revox repair on the internet, very often you will encounter the practice of replacing all electrolytic and tantalum capacitors. We have serious doubts as to this practice. Modern electrolytic capacitors have a lifespan (MTBF) of only ca. 3000 hours. Moreover, in an experiment, we have removed some 30 electrolytics from machines at least 40 years old and found that only 2% of them where really out of specs. We stick to the rule 'if it aint broke, dont repair it'. If replacements are needed and the capacitor values are in the range 1uF to 100uF, it's a good -though very expensive- idea to replace them with tantalum or aluminium polymer capacitors. Smaller caps really rarely fail and for the larger ones, electrolytics are the only choice. If measuring capacitors they need to be unsoldered first and then measured both for capacitance and leakage. The Fluke 87 multimeters can handle this measurement very well and reliably. Generally leakage is the main problem with old electrolytics.

Another issue is the 2-prong mains entry socket. This way of connecting equipment is not conform to modern rules with regard to electric safety. Many repairman have changed this connector for a modern 3-pole IEC model and connecting the earth pin to the chassis. If only a single machine is used, there are no objections against doing this. However, as in the case with on stage uses or professional studio set ups with lots of gear and equipment, we would rather not have this earth connection as it may indoduce ground loops and thus hum in the system. If you are really conscerned with safety (in fact the risk is quasi non existent, as the power transformer in these Revoxes is of outstanding quality and performs well as an insulation transformer) and/or rules, a better alternative is to use a hefty insulation transformer on stage and leave all equipment floating versus earth and mains power lines.

Not too many people nowadays seem to know that the mains voltage on the European continent has slowly been risen from 220V up to the late sixtees to mostly 240V nowadays. So, we advize strongly to change the voltage selector switch (it can be seen in the picture above) accordingly. Doing so will reduce dissipation and heat production in the machine. This may very well be one of the main causes of common 21V power supply failures in the A77.

Revox model G36

This model uses vacuum tubes. The motors are much stronger than the ones used in the later (A77 and further) models. This made them very popular in applications involving long tape loops. Most often, the (mono) audio amplifier driving the internal speaker will be found to be defect. The best solution here is not to repair it at all, but to remove the ECL86 vacuum tube from its socket as well as the loudspeaker. All professional applications for such a machine use the line outputs exclusively. Moreover, by removing these components the machine draws substantially less current and also produces less heat.

Here is a direct link to the service manual for all G36 models.

Carad models R62,R53, R59

Vacuum tube full track mono machines produced in Flanders (Kuurne) between 1955 and 1971. The later models using transistor technology and stereo. These later models were rarely used in live performance however. The older full track machines are worth to be preserved. Here is one:

At the time of this writing, it is still in full working condition.

dr.Godfried-Willem Raes

Ghent, April 11th 2015

This article is part of a research project on Experimental Legacy financed by the Orpheus Institute in Ghent. www.orpheusinstituut.be

to home page Godfried-Willem Raes