Back to Glass Audio Article List

Back to my Home Page

Overhauling a Revox G36

The additional gadget seen on top of the G36 in this photo is a peak level meter that uses indicator tubes for faster response than the customary meter movements.

Tape recorders started my audio hobby 25 years ago. I bought my first tape recorder at the age of 15. It was a portable Sony with 4" reels and selectable 1 7/8 and 3 3/4 i.p.s. speeds. Since the built-in speaker was quite inadequate, I soon built my first speaker cabinet, housing a double cone speaker recycled from an old TV set. During my first student year, I then bought a real 7" reel tape deck, an Akai 4000DS. Even though the Akai at that time was a fairly decent piece of equipment, so much so that I never cared to get myself any cassette deck, I had a passion to once own a genuine Revox! During the following twenty years a couple of opportunities emerged for buying a used series 77 Revox, but somehow I always had some more urgent use for the money.

After I got hooked to tube equipment, I naturally ordered all the back issues of GA. When I read Charles King's article "Greening the Revox G36" in the very first issue of GA (0/1988), plus a follow-up in GA 1/92, the old passion was immediately revitalized. And the funniest things do happen in life. When I least expected, a G36 was offered to me, even though I hadn't been actively seeking one. The word about my tube enthusiasm had evidently spread around, and the seller had been tipped off by my colleague. I managed to get a deck in reasonably good condition for about $100! The following contains some notes of my restoration project, concentrating to the mechanical aspects.

All the functions of the deck were fully operating, except some the dial lights. The tape heads were pretty worn out albeit still usable, but the capstan shaft needed immediate replacement. My newest treasure was a two track unit, whereas my tape library, accumulated during twenty years, was in four track format. Since the tape heads needed replacement anyway, I decided to modify the unit into a four track machine. The unit had the usual problem with the older Revoxes, the tape erasure was a little weak, at least with the original erase head.

I managed to get a copy of the Service Manual plus some smaller parts from the local Studer representative at a reasonable price. They didn't have in stock the major parts, tape heads and the capstan shaft, which should have been ordered from the factory in Switzerland, and the prices were outrageous!

Luckily I noticed a small ad in GA, and was extremely happy to be able to order the remaining parts from JM Technical Arts, in Nashville TN. A complete set of four track tape heads (newer, longer lasting version) and a new capstan shaft cost $360, a very modest price indeed!

Taking the deck to pieces needed some courage, but everything was pretty easy after I found out which wires to unsolder in order to separate the upper and lower chassis. Most removable wires go to the terminal strip near the power supply capacitors. Some wires go to the selector switches S5 and S6. If you remove the switch wafers of S5, no wires from it need to be removed. The wires going to switch S6a must be unsoldered. Be very careful to make notes of the position of the different wires. Once the unit is in two pieces, all the parts are easily accessible, even changing a switch contact is relatively easy.

I couldn't resist the temptation to take apart the bulky and expensive looking capstan motor and check it for signs of wear. I was amazed at the condition of it: it looked almost new, just a hint of shine on the bearings and absolutely no slack after almost 30 years of use! I was real happy I took it apart, though, since the sealing ring of the lower bearing was crumbled to pieces, and a small amount of thick oil had escaped. The oil had penetrated between the motor case and it's mounting bracket, and none of the leak was visible from outside! The problem was easy to fix with a suitable o-ring. After this discovery I checked the reel motors as well, and they were in perfect shape.

While at it, I spray lacquered all the motor and line transformer windings and relubricated the sintered motor bearings. I did this by impregnating the felt lining between the two sintered bearings with oil. Any excess oil not absorbed in the felt dripped away, when the bearing was kept in upright position overnight. I expect this overhaul will take the motors through another 30 years of use!

I had no use for the internal power amplifier and loudspeaker, and Mr. King's article inspired me to remove the output tubes and the output transformer. The original RCA connectors were in sad shape. I removed the whole back panel and replaced the jacks with good quality gold plated connectors. I also added a removable mains connector to the back panel. Selector switch S2 was rewired to act as an output source selector for B output jacks. I cleaned all the switch contacts with contact cleaner, they really needed this treatment.

Head Games

As the new heads were 4 track models, I made the component changes listed in the schematic. The frequency of the bias oscillator is about 70 kHz, rather low by present standards. Changing the bias oscillator frequency turned out to be relatively easy by just replacing capacitor C32 (new value 680 pF). The capacitors in the bias traps, C26/C29 (250 pF) and C48/C76 (470 pF), should also be changed.

There are very few questionable component selections in the G36: I bypassed the only electrolytics in the signal path (C41/C50) with 1uF polypropylene capacitors. I don't know the material of the rest of the capacitors, but replacing all of them with polypropylenes would be a formidable task!

The Service Manual was very brief considering the alignment of four track heads, calling it a very delicate job and urging the reader to obtain the Revox alignment tape, which would also include the instructions. Considering the pricing policy, I wasn't too keen on following the advice. Instead I found out my own procedure for head alignment.

Some special tools are needed in the job: a slotted screwdriver for the locking nuts, a magnifying glass or a loupe (with a magnification about 5 to 10), a small plastic or cardboard square (I used a business card). The most important accessory is a 6" length of clear tape, the oxide coating of which has been dissolved off with acetone (or common nail polish remover). Screwdrivers and other steel tools should be demagnetized before proceeding.

Before changing the heads, it's wise to make a reference tape using the old heads. This tape can be later used for adjusting the head azimuth to be compatible with previously recorded tapes. The suitable test signal is a frequency modulated sine wave, with the frequency slowly changing between 10 and 15 kHz.

Alignment procedure

  1. The pinch roller lever is removed to get a clear view with the loupe to the head surfaces.
  2. The heads are replaced, making sure that the wires are connected right. Only the larger center nuts have to be undone when replacing the heads. After replacement, the new heads should be pretty close to correct line-up.
  3. A tape with the transparent section is threaded normally. The overall position and line-up of the heads is first checked with the clear tape.
  4. The positions of the tracks, which the recording head would produce, are carefully marked onto the tape with a narrow felt-tip permanent ink pen, taking care not to smudge or scratch the head surfaces. The reels are exchanged and the other two tracks are marked similarly. Now the track positions and guard bands between the tracks are visible. If necessary, the height of the record head is readjusted. Note that if the height is changed with the adjustment screws at the front and back edge of the head, the azimuth adjustment screws must be turned the same amount to keep the head tilt about the same.
  5. The vertical adjustment is checked with the square. The head mounting plate is used as a reference. Items 5 and 6 are repeated until all adjustments are perfect.
  6. The playback and erase heads are adjusted similarly, but using the track marks as a reference.
  7. Lock nuts are tightened lightly. The heads are demagnetized before reinstalling the pinch roller lever.
  8. A dual trace oscilloscope is connected to both playback outputs. The previously made reference tape is played and the azimuth nut (left) of the playback head is adjusted until the two sine waves have minimal phase shift between channels, regardless of the frequency. This is a fool proof method for doing the azimuth adjustment. Note that the nut at the right of the head has spring washers that let the head tilt slightly, when the left nut is turned. The right nut is only adjusted to keep the tension suitable.
  9. The same reference signal as in item 1 is recorded and the output from the playback amplifier is monitored. The azimuth nut of the record head is adjusted until the two sine waves have no phase shift between channels, regardless of the signal frequency. Now the record head azimuth have been set to precisely match the setting of the record head.
  10. All the lock nuts are tightened and everything is rechecked. Finally, a drop of paint is applied to the nuts to keep them secured, and the heads are demagnetized again.

After adjusting the bias traps at the output of the record and playback amplifiers, I set the record bias levels for the tape used, and calibrated the VU meters. Everything went smoothly except for one detail: tape erasure was no better that before. The service manual and schematics had suggested to change R118 from 1 kohms to 2.2 kohms in four track units. This lowers the supply voltage of the bias oscillator from 215V to 195 V. The easiest remedy for weak erasure was not to do this change. In case you have similar troubles with erasure, the first thing to try is to increase the bias oscillator supply voltage by decreasing the value of R118. The bias signal amplitude increases in the same proportion as the supply voltage. After this fix the erasure was satisfactory.

The sound of the machine was as expected: superb. Even my old tapes, originally recorded with solid state recorders, sounded more open with improved imaging. Test recordings proved that discarding my old solid state recorder would not be any sacrifice.

This project showed clearly, that old professional grade hifi gear is well worth fixing and restoring. Taking apart a high quality and well designed instrument just plain fun! Fixing of course requires, that you can get spare parts at a reasonable price. For the Revox fans, JM seems to be well stocked with very reasonable prices. It was uncanny, how well the instrument nearly thirty years of age had lasted. I am quite confident that this unit will serve me for years, possibly decades. I'd like to thank Mr. King warmly for his articles, because they had a vital role in making my dream come true: own a Revox (a tube one!), even after 25 years.


30 Music Square West # 156
Nashville, TN 37203
Phone: +1-615/ 244-6892
Fax: +1-615/ 254-8826

More information on tape head alignment:

Bill Vermillion on Magnetic Tape Recorder Alignment

Analog Tape Recorder Alignment

Back to Glass Audio Article List

Back to my Home Page

Copyright ©1993 Jukka Tolonen

Any feedback is highly appreciated: