Tandberg TD 20A

Reel-to-reel Tape Machine

Made in Norway starting around 1979, the TD-20A is one of Tandberg's last models before ceasing production of tape machines. As evidenced in the owner's manual (see Links), the TD 20A was marketed mostly as a machine for the "keen amateur", rather than as a studio unit. Some indications of this are that, in its most common form, it operates at the two most common domestic speeds (3.75 and 7.5 ips) and not the typical studio standard of 15 ips, and it has quarter-track stereo heads, rather than the more "professional" half-track stereo. It is possible to find a TD 20A with 7.5 and 15 ips speeds as well as a half-track stereo head, but these machines are less common.

Aside from the standard TD 20A model, Tandberg also offered the TD 20A-SE and TD 20A-SES (the latter of which apparently exists, but I have found essentially no info), which have improved fidelity, particularly in terms of their signal/noise ratios, and with the SES apparently having XLR sockets. The -SE, at least, also has 7.5 and 15 ips speeds as the standard options.

This machine is capable of remarkably excellent fidelity. The owner's manual states an ideal frequency response of 15—30,000 Hz at 7.5 ips! This is with the specified Maxell UD-XL or equivalent.

Anyway, I will not go into too great detail about the functionality, since there is a large amount of information on the internet already, as well as in the service and owner's manual, both of which are available at the bottom of the page. Also available at the bottom of the page is the data for both types of PROM chip responsible for control logic.


  • accepts 1/4" tape on maximum 10.5" NAB-hub reels
  • quarter-track stereo format
  • three heads: erase, record, and play (left to right, looking from the front, right-side-up)
  • 3.75 and 7.5 ips speeds, button-selectable
  • unique 4-motor design
  • separate recording enable switches for each channel
  • digital logic control of functions
  • connections on back:
    • two pairs of RCA jacks for L and R inputs - "Line 1" and "Line 2" (the second of which is in parallel with "Radio" input)
    • 5-pin DIN plug for "Radio" input and output
    • RCA jacks for L and R "Line Out"
  • connections on front:
    • 1/4" mono jacks for L and R mic input
    • 1/4" stereo jack for headphone output
    • 7-pin DIN plug for remote control


  • Two main PCBs, which are single-sided and use all through-hole components
  • 14 TTL logic chips, of the following types (in my unit):
    • 1....74LS00 (quadruple 2-input NAND gate)
    • 1....74LS02 (quadruple 2-input NOR gate)
    • 1....7406 (hex inverter, open collector, high-voltage output)
    • 3....74LS08 (quadruple 2-input AND gate)
    • 1....74LS14 (hex schmitt trigger inverter)
    • 3....74LS32 (quadruple 2-input OR gate)
    • 1....74LS138 (3-line to 8-line decoder)
    • 1....74148 (8-line to 3-line priority encoder)
    • 1....74LS279 (quadruple S-R latch)
    • 1....74S471 (256 x 8 bi-polar PROM)


I received this machine in fairly rough shape around 2006. It was purchased at a music store along with a large number of "logger" tapes recorded by the local radio station CKER. At first I thought it may have been used for logging, but these logger tapes were almost all recorded at 15/16 ips, so they wouldn't have been recorded on this machine. While it seems likely given its original condition, I'm not entirely sure whether or not it would have been in broadcast use at CKER.

Originally, it was incredibly dirty. The front panel had a coating of yellow residue and dust, and the heads and rest of the transport were covered in shed oxide, likely from playing tapes with "sticky shed"/binder hydrolysis. Most of the tapes in the CKER logger batch were affected by this. The left reel table was bent slightly, causing uneven rewinding. Many of the switches, particularly the front lever-type ones, were quite intermittent. Annoyingly, the very difficult-to-replace cast metal bottom piece that would cover the pinch roller was missing. (Luckily, I have since received a replacement part, sans lettering, from a very kind fellow from Norway, for which I am quite thankful!)

As well, even after repeated cleaning of the playback head with isopropyl alcohol on cotton swabs, the right channel of playback was still lacking in high frequencies. The original playback head is a ferrite design, which while it did not appear to have a significant "flat spot", it did have slight chips at the gap, effectively widening it, which would account for the loss of high frequencies.

This was around 2011, and at the time I decided I'd rather have a local tech work on the unit than do it myself. He was able to straighten the left reel table, and replaced the playback head with a significantly worn Studer head. Now, both channels sound about the same, but the wear of the head means that the tape contact is not excellent, and thus dropouts are heard even on excellent tape. I think this can be an interesting effect, but some day I may try changing the head again.

In early 2015, the 7.5 ips speed became intermittent and then stopped working entirely, so I finally had a look inside. Solving the speed issue was simple; the motor plugs just needed to be reseated on the logic/power PCB. I also discovered a problem that I hadn't been aware of, which is that a 10µF tantalum capacitor had failed, as evidenced by the below image:

Yes, it was originally green! After desoldering it and replacing it with an electrolytic capacitor, I discovered that it had a DC resistance (i.e. what would be measured by a standard ohmmeter) of only a few hundred ohms. When the machine was back together, it was clear what replacing this capacitor had done: engaging the transport (i.e. pressing play, etc.) no longer resulted in a burst of noise and hum on the output. I had suspected such a burst was abnormal, so it was good to find the cause and correct it so easily.

Update: In June 2016, in preparation for using the TD 20A in a live performance, I decided to replace all of the "Frako" brand electrolytic capacitors. These capacitors are known to age poorly in that they have a tendency of failing short, much more often than most other capacitor types, and can damage other parts when they do so. In another machine of mine, one failed Frako showed a DC resistance of less than 1 ohm! So, five capacitors were replaced, including one "Micro" brand that was mistaken for a Frako due to having a gold wrap and being mounted "face-down". All of the Frakos tested abnormally leaky. As well as the capacitors, all of the fuse holders were replaced, since the originals were extremely corroded. Here are the replaced parts:

For the replacement capacitors, four of the five are Nichicon VX series, which are general purpose axial units. The last one is a Nichicon VR, which is radial, since the original Frako was an axial mounted vertically. Here are the new parts installed:

Note that the Frako capacitors for each motor were not replaced, because these are not electrolytics, but rather some non-polarized type.

The Hardware

The TD 20A is a fairly well-built machine, in my opinion, though admittedly not as robust as the typical professional units of the time. For example, the tape tensioners are not terribly sturdy, and can become slightly bent (via the internal mounting arm) over time. There is also a plastic gear turned by a brushed DC motor, neither of which are ideal for longevity. In any case, a lot of the functional flaws and external features have already been discussed elsewhere, so I will stick mostly to internal features here.

In order to access the insides, four screws must be removed from the wooden cabinet; one on each side, and two on the bottom. The chassis can then be carefully slid out from the cabinet. The front panel can be removed by taking out four hex-head screws, which is necessary for various transport alignment tasks, as well as replacing the capstan or counter belts. A picture from the front with the cover removed:

A number of things can be noted. The tape path is fairly straightforward; there are two tensioners, one lifter, and five stationary guides, all with smooth surfaces (not rotating). As well, there is a flutter filter roller, and of course the capstan, pinch roller, and three heads (erase, record, and play). An optical sensor is used to ensure there is tape loaded; if not, it will only engage transport functions when the buttons are held down. The "fourth motor" can be seen to the lower right of the left reel table. This brushed DC motor is used to engage the pinch roller and disengage the brakes, and is essentially a substitute for solenoids, but with smoother and quieter motion.

Looking from the back, the two reel motors and capstan motor immediately stand out. All three are AC induction motors, with associated capacitors. The two main circuit boards are also visible; the upper is mostly logic and power circuitry, while the lower has most of the audio electronics. Note that the upper board is fiberglass, while the lower is phenolic paper, with both being single-sided with all through-hole components. Nice for servicing!

Some people have stated that the TD 20A contains many "unobtainium" logic chips. Looking inside though, there are 14 chips that are all fairly standard TTL chips, as listed in the Trivialities section. As of early 2015, all of these can still be purchased new or NOS for quite cheap; in fact, I already have replacements for most of these in my parts supply. They are also all socketed, so replacing them wouldn't even involve soldering.

Admittedly, there is one chip which, while still easily obtainable, might pose some difficulty if it ever needed replacing. This is the 74S471 (or TBP28L22, which has the same pinout), which is a 256 x 8 bi-polar PROM chip. If a new one was put in straight away, it wouldn't work, because it would have to be programmed. While I haven't done it since I haven't had to replace the chip, it is entirely possible. There are two versions of the chip: the "violet" for units with 3.75 / 7.5 ips speeds, and the "orange" for units with 7.5 / 15 ips speeds. Note that mine is the clearly the "violet" version, from the image below. In a previous version of this article, I stated that the service manual provided all necessary info for programming, but as it turns out, it is incomplete. Fortunately, an enthusiast from Norway, Knut Baardsen, has read out the data from both chip types, and has allowed me to host it here. See the links at the bottom of the page!

A final side-note about manufacturing dates... as far as I could find, the TD 20A was introduced around 1979. The logic chips (as shown earlier) appear to have date codes from roughly around this time, such as "7827" which I believe would mean the 27th week of 1978. However, the reel motors as well as the replaced Frako capacitors show manufacture dates from significantly earlier, such as "Nov 74" on the motor, and "02.74" on the smallest Frako (also as shown earlier). Sort of interesting.

I have made a short video of the machine, found below:

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

Last Updated: 7/23/2016