Boss RSD-10

A sampler and a delay in one compact box, with a two-second memory that still surprises people.

Overview

The RSD-10 gets filed under "delay," but it really needs its own drawer. It carries both sampler and delay capabilities. The service documentation is candid about the split. The circuit description states that the RSD-10 handles both jobs, but the notes concentrate on the sampler side. They assume the reader already understands basic digital delay from the earlier Roland SDE-1000 and SDE-3000 and the Boss DE-2. So the delay half is treated as settled territory, and the sampler is the part worth explaining.

The memory is the headline. Sampling time runs to a maximum of 2000ms, and the delay time spans 2ms to 2000ms. Two full seconds of capture in a pedal-format enclosure made this thing strange and desirable in equal measure. The frequency response shows how that memory is spent. The direct path is quoted at 10Hz to 30KHz (+1 dB), but the sampler/delay path is 20Hz to 7KHz (+1/−3 dB) with a FINE × 0.25 note attached. That's an honest accounting of early digital memory. The sampler half is band-limited in a way the bypass path is not.

Physically it's a compact, metal-cased unit - 218mm wide, 169mm deep, and 44mm tall, or roughly 8-9/16" by 6-11/16" by 1-13/16". It weighs 1kg, about 2 lb 3 oz. The dimensions put it in the same footprint as the rest of Boss's compact line rather than the larger rack-oriented Roland delays. Power comes from a 9V DC supply - the BOSS PSA-100 in its 120, 220, or 240 variants - and the unit draws 100mA at 9V. Input level/impedance is quoted at −20 dBm/1MΩ and −10 dBm/47KΩ. Output is −20 dBm/2KΩ and −10 dBm/2KΩ, with an output load impedance of more than 10KΩ. Residual noise is specified at −95dBm (IHF-A) with the LEVEL switch at −20dBm, which is quiet for a box doing this much work.

The sampler architecture is where the RSD-10 gets interesting, and the documentation goes into real detail. Incoming audio hits the sample-and-hold stage continuously, but nothing downstream moves until a trigger pulse arrives. In MODE A, auto recording, the trigger is derived from the input signal itself. The unit listens and starts capturing when there's enough level. In MODE B, manual recording, the trigger comes from an external control input: a keyboard, a pad, or a pedal. Once triggered, the signal is sampled instantaneously, converted to binary PCM data by the A/D converter, and written into the RAM chips (IC11 through IC16). The main controller coordinates all of it, with timing locked to a master clock that is itself governed by the keyboard interface.

The keyboard interface is where the RSD-10 stops behaving like a delay and starts behaving like an instrument. The circuit generates a trigger pulse, a pitch control signal, and a volume control signal from a single keyboard audio input. Pitch tracking uses a phase-locked loop with switchable low-pass filter time constants. With only SW1 closed, the time constant is short and the VCO can chase the fast transients in a percussive or attack-heavy keyboard sound. Once SW2 closes as well, the time constant lengthens and the slower PLL reaction holds the VCO frequency steady on the sustained portion of the note. When the keyboard jack is left empty (pad or pedal triggering), both switches open and the PLL path disconnects. The VCO frequency is then set purely by the FINE/PITCH control, with no trigger or volume signal sent from the interface at all.

The counter start/stop logic explains a quirk. A trigger pulse drives Q4 and IC7a/b to produce a counter reset and a flip-flop start. That releases the counter stop signal and lets the main controller jump to the beginning of memory. When the memory wraps, a TEMP signal from the controller resets the flip-flops and the counter stops again. One trigger gives one full pass through memory. If a new trigger arrives while the counter is still running, only the counter reset is active, so memory starts over from the top. In delay mode, the counter stop line is permanently pulled low to defeat stopping entirely. In auto-record, the auto start trigger also flips the controller's WRITE pin high at the end of the pass, which protects the RAM from being overwritten. To release that write protection, the MODE switch has to be cycled briefly to B and back to A. The manual documents this rather than hiding it.

Factory alignment involves a handful of trimmers on the MT and SUB boards. RT-2 sets compressor distortion ratio, RT-3 sets expander distortion ratio, and RT-4 sets dynamic bias (1.5Vpp at TP-5). RT-1 sets A/D/A bias, adjusted for a symmetrical waveform at TP-4. None of that is unusual for the era, but a unit that has drifted needs someone who can read a scope and follow the procedure, not someone who just swaps parts.

The RSD-10 was designed to work with a specific set of accessories. A DP-2 pedal switch, an FS-1 foot switch, or a BP-1 pad controller can drive the external trigger input. A RAD-10 rack mount adapter and a BMR-5 Micro System Rack let it live in a rack. The printed-in-Japan marking and the BB-2 documentation code point to the same manufacturing lineage as the rest of the Boss compact line. Collectors note that the RSD-10 is not a common sight on the used market. The accessory list is part of why: the pad and footswitch options that make the sampler half useful are frequently missing from secondhand units, so a complete example is worth more than the sum of its parts.

eBay Listings

SHOP BOSS RSD-10 ON EBAY

As an eBay Partner, we earn from qualifying purchases. This helps support our independent vintage technology research.

Boss RSD-10

Overview

eBay Listings

Related Models