Summer 2025
This webpage covers my refurb of an Otari MX-5050 Bii 2 (the last "2" simply indicates 2 track). The Otari MX-5050 was sold over a long period of time: I believe the original model was introduced in 1975, with the "B" version in 1977, the "Bii" in 1982, and the "Biii" in 1990. Its one of the most popular and widely available semi-pro decks. It differs from studio recorders in a number of ways, for example lacking auto reel-size changing, lacking library wind, lacking locking reel spindles (until the last generation) and generally a bit less gentle tape handling (due to the use of mechanical vs electronic braking). It also has less sophisticated playback EQ adjustment: most studio decks have 3 settings to adjust EQ around 1K, 10K and 100Hz. This deck lacks the 100Hz low frequency adjustment. Nonetheless these are robust units made for long life and relatively easy service, hence many units are still in use today.
You can determine the year of manufacture for Otari decks by the 2nd and 3rd digit in the serial number. The MX-5050 in this project was made in 1984 - 41 years ago!
This upgrade consisted of normal maintenance: clean & de-magnetize the tape path, new pinch roller, complete alignment, but most of the time spent was replacing all of the aging electrolytic caps. In general I used quality Panasonic or Nichicon 105 degree caps (for reliability). For the audio section I used Nichicon Muse ES bi-polar caps in the audio path. As this is an older unit, the PCBs are single sided, making removal of old components fairly simple, but the boards are old and fragile, so care is required. That being said, the power supply caps on the control PCB are glued in place and require a bit of work to remove. I also replaced input/output XLRs with Neutrik gold units, and re-wired both input and outputs to have pin 2 hot, as Otari used pin 3 hot as the standard years ago. I don't plan to replace any of the op-amps, as in my experience its possible to cause more problems than solve. The same with tape transport bearings, and the heads, which all seemed fine in this unit.
This webpage is organized as follows:
1. Transport Section
2. Audio Section
3. XLR Re-work
4. New Pinch Roller
5. Meter Lamps
6. Side Panels & Feet
7. Repairs
8. Alignment
9. Completed Unit
Rear view of the recorder with the back panel removed. The large circuit board is the control PCB, responsible for most things non-audio in the unit.
You can also see the supply and take-up motors at the top, surrounded by the white felt straps that act as brakes. Power transformer at the top center, all inputs and outputs at the bottom.
For ease of service, that control PCB is on brackets to fold down for access to the machine's internals and the component side of the PCB.
Thats the capstan motor on the left.
Another view of the control PCB in the opened position.
You can see the component side of the PCB is labelled (as is typical) with component locations, part numbers, polarity, etc.
The capstan motor is on the center left.
Above: A closer view of the foil side of the control board. One unusual feature, rarely seen but very helfpul for maintenance, is the component markings printed on the rear foil side of the PCB making component identification easy.
Here's the component side of control PCB, removed from the machine. You need to disconnect about a dozen polarized plug-in connectors to remove the board. The right side is mostly the power supply (fuses, rectifiers, filter caps and regulators), other control functions are on the left. This shows the PCB after the electrolytic caps were replaced (but you couldn't tell from this photo). I counted 20 caps, a number of which are bi-polar.
MPB: Take new photo for below without all the dust on the PCB.
Next I tackled the capstan motor control PCB, shown here in the machine. This board is mounted above the capstan motor, and attached to a heatsink. Although there is one plug-in connector shown here, there are a number of soldered wires to at the other end with no easy disconnect....
I decided to leave those wires attached to the PCB to replace the dozen caps on the board. I made a cardboard shelf wedged under the motor to work on the board.
You also need to disconnect four TO-220 transistors (on the right) from the heatsink to get at the component side of the board.
This shows the board after the caps were replaced. The 3 blue trimpots on the lower left are for speed adjustment.
Here's the foil side of the capstan control board while I was working on it.
The speed adjustment trimpots are visible thru the holes on the lower right.
Next up was the counter PCB, shown here still in the machine.
The counter control PCB was a much simpler upgrade - just one cap to be replaced. This unit, however, had undocumented modifications requiring me to pull the board to work on it.
Here's the foil side after returning it to stock.
Like the control PCB in the rear, the audio section folds down to allow access for maintenance. That section can also be completely removed from the unit after unplugging a bunch of connectors.
Otari used a simple but ingenious cam system (circled here) to hold the audio section in a horizon position for checkout and/or repair work.
Closeup of the cam system. One slot is for the normal closed position of the audio section, and the other section (with the cam in it in this photo) holds the audio section in the open position.
Above: Like the control PCB in the transport section, the audio PCB also has component parts and polarities printed on the foil side. Much appreciated for working on the unit.
Although the open position allows access to the PCB, for my work it was much easier to simply remove the assembly from the main chassis.
Here's the audio assembly removed from the unit before my wholesale cap replacement.
And after replacing all of the electrolytic caps. As mentioned earlier, I used Nichicon Muse ES for all of the bi-polars (all the green caps), and mostly Panasonic FC for the polar caps. There are over 90 caps on this board, and because its 40+ years old and fragile, slow careful work is critical. I managed to accidentally lift just one pad, but connected that cap lead to the destination pad.
This small plug-board contains record equalization components, and simply unplugs after removing a single securing screw.
This is the VU meter driver PCB which is attached to the back of the meters, shown after the re-cap. This board supplies both the audio signal to the meters as well as the 8VAC for the meter lamps (more on that below).
This is the foil side of the meter PCB during the re-cap. Rather than remove a number of soldered connections to the board, I simply tilted the board down to replace the 6 caps.
The current standard for wiring XLRs is for pin 2 hot (non-inverting). Back when this recorder was made, there was no universally accepted standard, and pin 3 hot was widely used in Japan, including on Otari decks. There are a couple ways to address this problem, however custom cables or adapters are clumsy. The best solution is to simply swap the pin 2 and 3 wires on the XLRs.
This shows the stock input and output XLR wiring from the inside. This unit (and most, maybe all, Otaris) were supplied from the factory with pin 3 hot.
The quick and simple solution is to swap the pin 2 and 3 wires at the XLR, as I've done here. You'll notice the white and orange wires are reversed from the photo above.
I also decided to replace the line in and output XLRs with new gold Neutrik units from my parts bin, which happen to fit perfectly.
The rear of the unit showing the new Neutrik XLRs. I still have to replace the mic input XLRs at some point on the future.
The rubber pinch roller in this unit was rock hard (not surprising given its age), so I replaced it with a new unit from TerryRubberRollers.com
That new roller is shown here along with the plastic washer that sits atop the roller. There's a similar washer beneath the roller as well.
Note that Terry recommends cleaning the pinch roller with soapy water only, not IPA or any strong chemical to avoid premature aging.
The new roller and washer in place before the cap is installed.
To replace a dead meter lamp(s), you need to first remove the lower front panel (remove the 3 knobs first using a 1.5mm allen wrench). Then using a small screwdriver, simply pry the clear meter cover off from the bottom, as shown here. Re-install that cover by hooking the lip at the top and then pushing in at the bottom.
With the meter cover removed, this shows the original frosted incandescent lamps that illuminate the meters. Mine were both burned out in both meters.
To avoid anyone having to replace the bulbs again, I used soft-white LED replacements from Parts Express, here. These work fine with the Otari 8VAC supply, no other modifications needed.
This shows the new LEDs installed on both meters. Note the actual LEDs are mounted on a tiny flat PCB, which is aimed up at the meter faces.
The original side panels on my unit (made of particle board covered with thin plastic laminate) were chipped and damaged. I started by patching and sanding the panels to rough up the finish, then veneered them with cherry veneer, in process here.
The veneered panels with the cutouts for the handles.
To make the side panels thicker and more substantial looking, I added a backer of 1/2" plywood, visible to the right in this photo. You can also see that the original panels, toward the left, are particleboard.
After trimming to the correct size, I glued solid cherry stock around the perimeter, in process here.
Solid cherry trim glued on, front and back sides shown here.
Panels were then sanded and machined to get a round-over edge all around. Just stained in this photo, satin polyurethane spray to follow.
Completed side panels, front and back, after finishing and installation of the original lift handles.
While I had the table saw out, I made a new set of feet of oak. The stock feet allow the recorder to slide around a bit, and because of how the originals are made, there is no way to attach rubber feet. I made the replacements a bit longer in the front to add stability, as the unit is very front-heavy due to the transformer and all 3 motors being on the front of the unit.
The new feet mounted on the bottom panel, painted and with sorbothane feet added.
This tape recorder had a problem recording in channel 1. I traced the problem to the recording bias relay RL401. These are apparently a regular point of failure in these machines. The original relay is an Omron G2V-2, 24VDC. That relay has been out of production for decades, and although new-old-stock replacements are available, they are crazy high prices. I contacted Omron and they suggested the G5V-2 DC24 as a modern replacement. Here's the Mouser link.
This photo shows the original G2V-2 on the bottom and the replacement G5V-2 on the top.
Since one had already failed, I decided to replace this relay in both channels: RL301 and 401, as well as the identical RL501 (hiding near the back of the PCB to the right). This shows the new RL301 & 401 relays in place, and the unit is recording fine in both channels now.
When I began the alignment process, one VU meter was reading a bit low when playing an MRL tape with a 1K signal. There is no adjustment for this in the MX-5050 (only the standard meter zeroing with no signal), so I increased the feedback resistor in the meter amp (R710), shown here. Now both meters read 0dB correctly when the unit's output is +4dB.
After cleaning and de-magnetizing the tape path, I adjusted the tape speed on the capstan motor control PCB. Since I don't have a highly accurate frequency counter, I connected a CD player with a 1K tone to one oscilloscope channel (the known accurate frequency) and an MRL test tape from the Otari to the other scope channel. By monitoring the lissajous pattern, I could dial in the exact correct speed. This shows the 15 IPS adjustment in progress on the capstan control board.
Next was the brake adjustment using a spring scale which I tested for accuracy before using it on the tape deck and found surprisingly accurate.
The azimuth adjustment (both 15 IPS and 7.5 IPS) was done using a Sound Technology ST-1500A, as was frequency response adjustment. This requires a special alignment tape, available from MRL.
I greatly prefer alignment with the ST-1500A as it allows you to see the full frequency response, and optimize for overall flatness, as opposed to the traditional method of aligning at a couple frequencies only (e.g. 1K and 10K).
This is a closeup of the 15 IPS IEC playback frequency response for the unit after alignment with the ST-1500A. The vertical scale is 2dB per division, and the double vertical bars are 100 Hz, 1K, and 10K. As you can see, the response is bumpy below about 300 Hz, but remarkably flat above that point. Overall, well within the Otari spec of +/- 2dB.
Note: this shows both left and right channels, but they match so closely they are atop one another.
The parts that were removed and replaced.
The completed unit up and running.