MOX internal architecture

Curiosity finally overcame inertia and I ordered the service manual for the Yamaha MOX6 and MOX8 workstations. (The Yamaha 24×7 part number is “S M MOX6/MOX8”.)

If you remember from my previous discussion about workstation internal architecture, the Motif XS synthesizer is Linux-based and has a 400MHz Toshiba TX4939 RISC CPU as its main processor. The TX4939 uses the MIPS instruction set and controls two SWP51L tone generator integrated circuits. Since the MOX is advertised as descendent of the Motif XS, I fully expected a MIPS architecture processor with only one SWP51L.

Check out the Yamaha MOX block diagram.

Surprise! The main processor in the MOX is the Yamaha SWX02 with an internal clock speed of 135.4725MHz. The SWX02 has an SH-2A CPU core and probably does not run Linux. The SWX02 is also used in the Yamaha PSR-S650 arranger workstation where it is clocked at the same rate. This processor seems to be Yamaha’s choice for cost-sensitive, mid-range products.

The MOX has one SWP51L tone generator IC clocked at 90.3168MHz. The SWP51L is fed by two 64MByte wave ROM ICs. The wave ROM components are Lapis Semiconductor MR26V51252R 512Mbit P2ROM devices in 32Mx16-bit configuration. One device provides a 16-bit high (H) channel and the other device provides a 16-bit low (L) channel into the SWP51L. The high and low wave ROMs communicate with the SWP51L over a 32-bit wave memory bus. The SWP51L has a separate 16MByte SDRAM on a dedicated interface to support digital signal processing (DSP). The DAC and ADC are also connected directly to the SWP51L.

The SWX02 functions primarily as a control processor. This is quite different from the PSR-S650 where the SWX02 performs tone generation as well as performing control duties. The SWX02 has its own wave memory interface and this interface is not used in the MOX. The S650 has a separate LCD controller IC. The MOX does not have a separate LCD controller and the LCD is connected to the SWX02 through its parallel general purpose I/O (GPIO) pins.

The MOX specifications describe the wave capacity as “355MB (when converted to 16-bit linear format)”. The physical wave ROM is 128MBytes total. Thus, Yamaha achieve overall wave compression of 2.78 to 1, or better.

The most interesting thing about the MOX is what it does not have. The MOX main logic board (DM) has unpopulated positions for:

  • A second SWP51L tone generator IC
  • Two additional wave ROM ICs (size unspecified) on the wave memory bus
  • An interface for a flash expansion module
  • A second WM8740 digital-to-analog converter (DAC)

Yep, Yamaha laid the ground for the MOXF. These positions are labeled “For future model” in the detailed circuit diagrams. One way to feel about that is cheated. A more rational way to view this situation is that Yamaha tries to lower cost through volume production (eventually) giving us more product for less money.

The MOX polyphony is 64 notes. The MOXF polyphony is 128 notes due, presumably, to a second SWP51L. A Motif/MOX note may use up to eight voice elements. Therefore, I infer that an SWP51L has a total tone generation capacity of 512 voice elements. Switching context to workstation arrangers for a momemnt, both the PSR-S950 and Tyros3 have 128 note polyphony. The S950 has one SWP51L and the Tyros3 has two SWP51B integrated circuits. I now believe that the S950 is a four element per voice synthesizer while the Tyros3 is known to be an eight element per voice synthesizer. (The S950 is voice compatible with the A2000, which is known to be four elements per voice.) Thus, I don’t think Super Articulation 2 (SA2) voices based on Articulation Element Modeling (AEM) technology are coming to the S950 or a new mid-range arranger workstation. Not without a second SWP51L, anyway. I’m guessing that AEM requires an eight element per voice engine.

It’s interesting to see how and where Yamaha shaved cost in order to produce a value-oriented mid-range product. It also provides geater justification for the higher cost in the upper end Motif and Tyros products.

See this article for an architectural overview of the Yamaha arranger product families.

Finally, Yamaha releases the source code for GPL’ed parts of the Motif XS, Motif XF, and S90 XS/S70 XS. See the Yamaha source code page. The MOX and MOXF are not mentioned on this page, giving further evidence that these products are not Linux-based.

A follow-up on the Yamaha SWP51

Here’s a follow-up on the Yamaha SWP51 tone generator.

Sound On Sound (SOS) mentioned that the SWP51 tone generator was first used in the Yamaha Motif XS family. The Motif XS has two SWP51 ICs with a master/slave relationship. Each SWP51 has 8MBytes of dedicated DSP SDRAM. The two SWP51s share wave ROM arranged in two banks (high and low) of 512Mbits each for a total physical capacity of 128MBytes. Each wave ROM device is a Spansion S29GL512N10TFI020 which is a 32M by 16-bit parallel NOR flash memory. (Yay, Spansion. I ate lunch in their cafeteria in Austin.)

SOS and others claim that the SWP51 performs sample compression. The Yamaha specifications state wave ROM capacity at “355MBytes when converted to 16-bit linear format,” meaning uncompressed size. The waveforms are compressed in order to fit into 128MBytes of physical memory.

There is only one thing that we can conclude for sure. The PSR-S750 and PSR-950 have twice the physical wave ROM space as the Motif XS and MOX (256MBytes vs. 128MBytes).

The Motif XS with two SWP51s has 128 voice polyphony while the MOX has 64 voice polyphony. Thus, the MOX most likely has only one SWP51. The PSRs have 128 voice polyphony. If the later version of the SWP (SWP51L) has the same number of tone generating and DSP elements as the first version, then the arranger keyboards are deploying the elements differently than the Motif family instruments. Without knowing the internals of the SWP51 and its variants, this is pure speculation.

By the way, the main processor in the Motif XS is the Toshiba TX4939 RISC CPU. The CPU is clocked at 400MHz. The TX4939 is a MIPS architecture processor with several integrated I/O controllers and interfaces. The TX4939 is capable of generating an audio sampling clock and supports PCM input/output. The processor runs Monta Vista Linux. Yamaha has an equity stake in Monta Vista and distributes the GPL’ed source code three years after initial product release. Yamaha have not released source for the arranger keyboards, so most probably, the arrangers use a different embedded operating system.

The MIPS instruction set and the Renesas SH-4 instruction sets are not compatible. If the arranger and workstation product lines share software-level functionality, it must be at the source code level.

Just in case you aren’t confused enough already. The Tyros 3 has two SWP51B ICs and two SH-series CPUs (Renesas SH7727 SH-3-DSP main, and SH7206 SH-2A sub). The sub CPU handles the SWP51Bs while the main handles the user interface, etc. The microphone input is routed to one of the SWP51B ICs. Playback/record audio is routed from the other SWP51B to a gate array that interfaces to the hard drive subsystem. Wave ROM consists of four 512MBit mask ROM devices arranged in two banks for 256MBytes of total physical ROM. Tyros 3, by the way, was the first Tyros with Super Articulation 2 (SA2) voices.