Tascam 38

Reel-to-reel Tape Machine

Introduced in 1983 along with the models 32 and 34, the Tascam 38 was offered as a relatively low-cost 8-track tape recorder, intended for semi-professional use (i.e. serious musicians and low-budget studios). It uses 1/2" tape on up to 10.5" diameter reels, features three heads, and three brushed DC motors. Each channel has its own lighted VU meter and "Function Select" / record enable button. A "Cue" lever allows the tape lifters to be disengaged, allowing the output from the heads to be heard at all times, thus enabling precise locations on tape to be found during editing. Three "Output Select" buttons allow switching the selected output between the input, sync head, and repro head. On the back, a 12-pin WAKA remote connector allows a remote control to be attached, such as a Tascam RC-71, and a 1/4" jack allows a punch in/out pedal to be used, such as a Tascam RC-30P.

Since this is a relatively common model with a good user's manual available online, this article will focus primarily on the repairs necessary to make my unit fully functional again. My overall impression is that it is a somewhat inferior machine to my Tascam 58; the transport is sloppier and less robust, the cue/edit function is cruder, the counter counts nonstandard lengths of time instead of seconds, the build is lighter and with more plastic, etc. However, it is still capable of excellent recordings, and is built with quality components and with a high degree of serviceability scarcely found in modern equipment.

Features

  • accepts 1/2" wide tape on 10.5" max. diameter NAB-hub reels
  • 8-track format
  • 15 ips speed, with adjustment control
  • three heads: erase, record/sync, and play
  • three brushed DC motors: capstan, left reel, and right reel
  • separate "Function Select" / recording enable switches for each channel
  • digital logic control of most transport functions
  • "Cue" lever for editing
  • connections on back:
    • eight RCA jacks for channels 1-8 input
    • eight RCA jacks for channels 1-8 output
    • 12-pin WAKA for remote (such as Tascam RC-71)
    • 6-pin WAKA connector for DBX unit control signal
    • 1/4" jack for punch in/out remote (such as Tascam RC-30P)
    • Permanently attached power cord

Repairs

This example of a Tascam 38 was purchased inexpensively along with a Tascam 32, its two-track equivalent. Both machines were very heavily used by a local low-budget studio in the 80s and 90s, followed by years in a slightly damp storage locker until the studio owner's unfortunate death in February 2016. Needless to say, this machine was not in great condition when I received it in March 2016. Besides cosmetic issues (such as a thick coating of smoke residue, leftover silicone sealant used for splicing block attachment, no nameplate, etc.), both the pinch roller and the capstan belt had turned to goo, the capstan was glazed and covered with rusty patches, all but two VU meter lamps were burned out, channel 4 had no output, channels 1 and 3 had no signal in repro mode, the FF/REW functions were extremely sluggish, among other small faults. Here are some pictures of the machine's initial condition...

...and here is the list of repairs done to bring it back to good working order:

  • Thorough internal and external cleaning, as well as some cosmetic fixes. Unfortunately, the largest cosmetic issue was only partially resolved, which was the patch of silicone sealant that was left over from where a splicing block had been installed; it was reduced to a thin layer.
  • Printed a new nameplate on glossy photo paper.
  • Refurbished the reel motors.
  • Replaced all VU meter lamps.
  • Replaced the VU meter foam mounting tape.
  • Modified the clear tape holding the meter lenses on so that the meters can be opened without removing them from their mounting tape.
  • Set reel table height based on typical reels, after tape path adjustment.
  • Refurbished the cylindrical dampers for the tension arms, since the old grease had gone hard/sticky. First tried NLGI #2 synthetic grease, but found it was much too thin; settled on some "silicone differential fluid" from a hobby shop with viscosity 125,000 cSt.
  • Removed rust spots from capstan using an x-acto knife and isopropyl-alcohol-soaked swabs.
  • Resurfaced capstan to have a slightly rough finish, done by wrapping the pinch roller with 80 grit sandpaper and running the machine for a few minutes.
  • Replaced capstan belt (which was originally sticky and snapped), and relubricated capstan motor & capstan bushings.
  • Had the rubber of the pinch roller remade by Terry's Rubber Rollers.
  • Very thoroughly cleaned tape path. Especially, the tension arm rollers had some sticky-shed tape residue on them that required acetone and a little Brasso to completely remove.
  • Demagnetized the tape path and heads.
  • Fixed dead repro on channels 1 and 3; in both cases it was due to cold solder joints on the mother PCB that looked fine to the eye.
  • Fixed dead output on channel 4, caused by a broken pin of the female connector going from the mother PCB to the in/out PCB, which I luckily had an exact replacement pin for.
  • Resoldered some bypass wires across the muting relays on a few of the channel cards, which were very poorly soldered by a previous technician.
  • Rotated one of the two stationary guides of the head block (to give an unworn surface); the other was stuck.
  • Installed four new rubber feet on the bottom, since the original feet had been lost after being removed for rack mounting.
  • Aside from those thing already mentioned, adjusted/calibrated: capstan thrust clearance, brake torque, reel motor torque, pinch roller pressure, tape speed, tension arm height, sync and repro head azimuth, input levels, meter levels, reproduce levels (for 0VU = 355 nWb/m), reproduce EQ, sync levels, sync EQ, bias levels (for Pyral SM911 tape), record levels, and record EQ.

Some of these repairs are worthy of further explanation, so I will describe them below.

Firstly, the main cosmetic fix on the machine: making a replacement for the missing plastic nameplate. These often tend to be missing, due to the plastic warping outwards and the glue becoming brittle and cracking, and the fact that many owners dispose of them once they fall off. My solution was to print a copy of the original design on glossy photo paper with an inkjet printer, cut it to fit the spot, and glue it in with some PVC-E glue. A nice printable version of the design had conveniently already been made by Mike Zee of MZE-Electroarts Entertainment (see Links), but his solution of cutting a plastic piece and securing it with screws was not as appealing to me as gluing. Anyway, the result is shown below, which I'd say looks pretty good:

Another repair, partly cosmetic and partly functional, was replacing the burned-out VU meter lamps. For these, type DA513 lamps from the supplier JKL were used (see Links), which have a current rating of 55mA at 8V; with the actual lamp supply of 6V AC, the current is somewhat lower. Although slightly larger than the originals and clear instead of frosted, they look great, and with very similar illumination. I replaced all of the original lamps, including the two non-defective but probably nearly-expired ones, for consistency and to avoid having to replace any more in the near future. In the leftmost picture below, an original and new lamp are shown side-by-side, with the original on the right. The other two pictures show all new lamps.

When performing this lamp replacement, the VU meters must be pulled off of their mounting tape (which in my unit was double-sided foam tape that had decomposed), and the horizontal strips of clear tape holding the meter lenses have to be removed in order to allow the lenses to hinge upward. On each meter, I replaced these horizontal tape strips with new vertical pieces holding the lenses in place, so that they can be easily detached without having to remove the meters from their foam mounting tape again. I also did replace the mounting tape; for this, I used new "Duck" brand 3/4" wide foam tape, which may be slightly excessive for this task. When sticking the VU meters to this foam tape, I found it was necessary to use the front panel to hold the meters in proper position as they touched it, which necessitated that the meter lenses were placed in their rectangular holes and held from the front with some temporary clear tape as the front panel was moved into normal position. If you just attach the meters haphazardly, the meter lenses will never fit properly in the holes on the front panel.

Refurbishment of the reel motors was necessary to fix the issue of sluggish Fast-Forward and Rewind modes. These brushed DC motors have a tendency to accumulate shaved-off carbon brush material (which is conductive) between sections of the commutator, resulting in poorer performance due to improper energization of the rotor coils. The removal of the motors from their mounting is the trickiest part of the process, in my opinion, due to the way the back mounting bracket is designed. It also typically involves desoldering the motor from the control PCB. Once the motors are out, the refurbishment involves removing the brushes, disassembling them carefully, cleaning out the carbon powder (especially from between commutator sections), relubricating the bearings, and then putting everything back together. There is a good online tutorial about this process, which I've included in the Links section. What really needs to be remembered is to keep track of the orientation of the brushes when removing them, as well as the sections of the motor when separating them. Here are some images of the process, showing the incredible amount of carbon shavings built up in the motor, and the result of cleaning both the case and the commutator.

Another worthwhile fix to mention is that of the dead channel 4 output. I had this type of problem on both this unit and the 32, and the cause was the same: a broken pin on one of the female 0.1" polarized connectors from the "motherboard" to the output jacks. You can easily tell this, once the connector is removed, by holding it in front of a light and looking into the holes; a bad pin will show up as excess light coming from the wire side through the hole, and the broken metal piece will move around inside. It's easy to fix this, simply by cutting and stripping the wire to the broken pin, securing a new one to the wire, and installing it into the connector. Here's the broken pin alongside its replacement:

Something else important is to refurbish the cylindrical dampers for the tension arms, since the old grease tends to become hard and sticky. As mentioned in the list, I first tried NLGI #2 synthetic grease, but it was much too thin; instead I settled on some "silicone differential fluid" from a hobby shop with viscosity 125,000 cSt. I didn't take pictures of this process, but there is a good forum thread with pictures in the Links section.

The last repair I'll mention is the bypassing of the K102 muting relays on the channel boards. These relays are NEC PG1A-24 type, are apparently often defective, and are not able to be opened or otherwise repaired. On this machine, 7 of the 8 relays (one for each channel) already had the switch bypassed before I received it, either with a wire directly across the pins, or with the relay completely removed and a jumper wire inserted. The only consequence of this is that the output is always on, but it is not of much consequence. Two of the relay bypass jobs are shown in the images below, along with a picture of all 8 channel cards, just to show how they are mounted in the bottom:

Finally, here are some shots of the head and capstan area after the repairs, showing clearly the cleaned tape path, new rubber on the pinch roller, and a resurfaced capstan:

And the finished unit itself, aside from the addition of rubber feet:

Overall, this machine took a heck of a lot of work to bring up to spec, but certainly a fun sort of work. I'm quite happy and amazed with the result! In any case, if you have any questions about this machine and my repairs to it, feel free to ask.

If you notice any errors or have additional information that you would like to add, please contact me!


Last Updated: 8/17/2016