Sony TC-WR645S Cassette Deck Repairs

Regulars will notice this humble site has a new look. I woke up recently to find it in a sorry state. It turned out the WordPress theme it was running is no-longer being actively maintained by its developer and updates to the WordPress core finally broke it. This site has been due a refresh for some time and it looks like matters now force my hand. This is the look we have for now, hopefully the logo will reappear in due course. Hopefully in the new year I will bring the software that runs things for us up to date and treat the site to something more than a thrown-together design. Until then…

Michael reader reached out to me some time ago to offer some kind words on my work and to ask about the possibility of repairing some of his hi-fi. I don’t cover everything I repair here but I do detail the more interesting pieces, or those for which I can offer tips to overcome issues that might trip up the DIY repairer.

Full Front View

Here’s a Sony TC-WR645S. The S in the name denotes that this is one of the few twin tape decks with Dolby S noise reduction, an impressive ‘last gasp’ for cassette that could attain signal to noise ratio levels in excess of 82dB without the aggressive side effects of Dolby C. It also has the usual features including high-speed dubbing, recording calibration and various reverse and synchronised playback modes, as well as music search, electronic counters and memory recall.

There were six issues with this particular example. Bad belts, a stuck motor, dry joints, a cracked gear, a dislodged spring and dirty switches. Both sides use a two-motor mechanism, deck B with a record facility and deck A playback only. The construction is interesting in that most of the recording circuitry is contained on the mechanism board itself. There is a good reason for this. These mechanisms were lifted from other Sony models of the time, including integrated hi-fi systems where the intension was to minimise the required external electronics. Fitting the head amplifiers, bias circuitry and record muting to the back of the mechanism also shortens signal paths considerably, and leaves the main board to handle logic, power supply, noise reduction and some of the audio input and output stage.

Full View Inside Case

It was on the main board that most of the dry joints were found, mostly on the power supply regulators located near to the mains transformer. They do use decent heatsinks which is unusual for a Sony deck of 1990s vintage, but repeated heating and cooling cycles, and the weight of the heatsinks, can break the joints over time. The larger filtering capacitors were similarly afflicted, the output RCA jacks too, and the connectors that mate with the boards of the mechanisms when the front panel is in place. These are all stress points and should be examined if the deck is behaving strangely.

Main Board

Soldering sorted it was time for the mechanisms. The front panel comes off with the removal of five screws underneath, two of which provide additional support to the mechanisms. The panel then pulls forward. Be mindful of the mating connectors which are relatively fragile.

The mechanisms are constructed to be as self-contained as possible.

Rear Of Deck B

The board comes off first after removing the three cables. There is a cable on the top connecting the reel motor, a cable on the bottom right side connecting the heads, and a flat flex cable connecting the capstan motor. Take particular care with the latter as it will break easily, and they are extremely difficult to solder back down.

Mechanism Board

The boards are held by plastic clips and a screw roughly central. The screw thread wasn’t especially cooperative. Removing the board entailed releasing the top claw and lifting up on the top edge to get the screw thread to bight. Eventually it came out and the board unhooks from the bottom to be set aside.

Side View Of Motor Housing

Two screws located above the switch board of the mechanism remove the top layer of plastic frame which holds the two motors and some gearing for the servos that change modes. With those screws removed, the hole rear section of the mechanism lifts up and unclips from the bottom, coming away to reveal the two capstan flywheels and just beneath them the large gear that drives most of the mechanism.

Motor Housing With Belt And Cogs

It was that gear that had cracked in one of these mechanisms. Removing it involved taking the two flywheels out, first removing the retaining washers from the front of the capstan shafts and then sliding them out. The gear could then be removed and replaced with a part removed from a scrap mechanism. I forgot to get precise pictures of this process, but it’s not hard to figure out.

Close Up Of Motor Cogs

Likewise the dislodged spring, which is seen on the left side of the mechanism next to where the gears of the reel drive motor would land. That spring goes over a small plastic lug just behind the switch board, and applies tension both to the actuator that releases the primary cam gear and the lever of the reel pulley. It is not uncommon for springs to fall out of this mechanism as you work on it, or if the cassette deck is handled roughly in shipping. Most of the springs are supported by tiny plastic lugs, held in place by even smaller plastic pegs usually with protrusions of a millimetre or so. There is a lot of plastic in this mechanism and a lot of tiny, delicate parts including the springs just waiting to fall out and ruin your day, so be mindful of them if your mechanism is misbehaving. Usually the mechanism simply won’t change modes at all, or will go into forward playback but refuse to reverse, fast forward or rewind.

Deck A had a stuck reel / servo motor, so I removed the motor using a Philips 00 screwdriver and applied a few drops of oil to the shaft beneath the pulley once the oil worked its way in with the help of a few jolts from a bench power supply, the motor spun freely and quietly. I oiled the other three motors in this machine for good measure. I then moved on to cleaning the tape detection switches by unclipping the switch board and spraying the contacts with switch cleaner. I didn’t photograph the removed board as it’s self explanatory, there is a clip at each end which releases the board and it simply lifts out.

Now for the belts. Getting the belts installed is a bit of a challenge. There is a small square belt which links the reel / servo motor to the reel drive pulley and a large flat belt which drives the two flywheels. You first have to set the small square belt on its pulley, ensuring it doesn’t catch on the gears. It needs to sit underneath the small gear set attached to the motor shaft. I found it easier to remove the outer gear from the motor shaft that holds these in place, allowing the gear set to be lifted off. Then position the belt on the pulley, and with it positioned roughly in the right place reinstall the gear set and the outer gear. Pay close attention to the position of the gear set, as the mechanism won’t fit back together properly if you position it incorrectly, and it won’t work even if you manage to force it.

View Of Fly Wheels With Belt Held On Pin

Then hook the capstan belt around the outer circumference of the flywheel opposite the capstan motor, over top of the second flywheel and onto the assistance pegs on the side of the mechanism body. There are two pegs, but I found it easier to use the shorter of the two and manipulate the belt into place with a hook tool once the mechanism was reassembled.

Close Up Of Belt On Pin

To reassemble, bring the motor frame over the rear of the mechanism, hook the square belt on its pulley, and set the mechanism in place. Then use a hook to grab the capstan belt through the gap, lift it from the pegs and set it over the capstan motor pulley ensuring it remains positioned on the flywheels and doesn’t get trapped between the top of the flywheel and the motor frame.

Side View Of Drive Belt In Place

Replace the screws and the board and re-connect the cables.

Motor Housing And Board Re Installed

Then reinstall the front panel. I first removed the circuit board from the front panel, cleaning the display with IPA and the switches, pots and buttons with switch cleaner. Oxidisation on these was causing erratic behaviour when buttons were pressed, and as the buttons are configured as a resistance ladder some buttons were operating the wrong function. The record level and balance knobs were also extremely stiff to turn.

With the panel back in place and the cables reconnected, you will probably need to recalibrate the speed. This is done using a 3kHz test tape, and a shorting plug to bridge two pins of a test point to put the machine in service mode. I didn’t have a plug of the right size so I used an insulated clip lead for this purpose. Bridge the pins and power up, at which point everything on the florescent display will illuminate to confirm all of the elements are working.

Insert a 3kHz test tape into deck A. There are 2 trimmers located on the PCB on the rear of the mechanism, right next to the capstan motor. There are access holes in the PCB, but more importantly there is a mains transformer and exposed live mains connections right next to where your hand will be when you make the adjustment. Cover these temporarily, before powering on, with electrical tape or a similar insulating material. No matter how experienced you are, it only takes one slip for your hair to be standing on end or your bowels to be voided.

Once they’re covered, play back the 3kHz test tape and adjust the low-speed trimmer to achieve as close to 3kHz as possible. Repeat for the high speed trimmer to achieve 6kHz, holding in the high-speed dubbing button on the front. Move back and forth between normal and high speed, adjusting the two trimmers until you are as close as you can get to 3kHz and 6kHz for the normal and high speeds respectively. Repeat for deck B. I would advise re-checking the adjustments once the belts have settled in. I put a couple of old C90s in each mechanism and allowed them to wind back and forth for half an hour to give the belts a workout. After that the speed was stable enough for a final adjustment.

Rear Of Deck A

These are new enough that you probably own’t encounter any electronic issues, and the rest of the calibration should be spot on. As they use a rotating head for the autoreverse, you might find the azimuth is a bit out. You can adjust this with a scope and a 6kHz or 10kHz test tape. But if you don’t have a scope, play back a pre-recorded tape with a mono signal, and adjust the screws to either side of the playback head until the level is the same on both channels and the treble is maximised. You’ll need to remove the doors to get at the screws, which simply slide straight up.

If you only have stereo tapes, adjust until you achieve the best stereo image with the most treble. Azimuth varies depending on the tape anyway, as in order for the azimuth to be correct the alignment of the head used to record the tape and to play it back must be identical, which is rarely the case.

There you have it. A well-featured machine even if it wasn’t Sony’s finest in terms of its construction quality. These make decent recordings, and are fine playback decks especially if you want the convenience of autoreverse. I’d probably still go for the more modern TC-WE475 or TC-WE675 for its simpler mechanics and better build quality, or an older JVC, Sony or Technics unit which are just as complex but better built. There’s no denying the quality of a Dolby S recording though, if you’re still recording cassettes.

By Ashley

I founded Audio Appraisal a few years ago and continue to regularly update it with fresh content. An avid vinyl collector and coffee addict, I can often be found at a workbench tinkering with a faulty electronic device, tweaking a turntable to extract the last bit of detail from those tiny grooves in the plastic stuff, or relaxing in front of the hi-fi with a good album. A musician, occasional producer and sound engineer, other hobbies include software programming, web development, long walks and occasional DIY. Follow @ashleycox2

Share Your Thoughts