~ajxs

Yamaha DX7 Official Technical Analysis

2022.03.20
A chance discovery on eBay has surfaced some of Yamaha's original technical documentation for the DX7 synthesiser.
TL;DR: Previously unseen official technical documentation of the DX7 has surfaced, which is now available here.

In the closing remarks of my original DX7 article I wrote the only way that we could learn more about the DX7 from here was to disassemble the firmware, decapsulate the internal sound chips, and to obtain Yamaha's official internal documentation. In the intervening year since then the first two of those things have come to fruition. With this update, I am happy to announce that the final thing I listed has now been realised: Some of Yamaha's original technical documentation of the DX7 has surfaced!

The fantastic Grant of Music Technologies Group, was kind enough to forward to me an interesting eBay auction he found for a document titled Yamaha DX7 Technical Analysis1. Obviously this piqued my interest. A few weeks, and some exorbitant shipping costs later the document was in my hands. It was definitely worth the wait.

The 90-page booklet is certainly the real deal2. Laser-printed on high-quality glossy paper (slightly yellowed over the decades), it's clear that this is authentic Yamaha documentation. It's extremely well written, complete with many amazing diagrams and illustrations. As well as numerous interesting details of the DX7's engineering.

During the research for my original DX7 article I scoured the internet for any available technical material. I spent many nights sifting through Yamaha's voluminous patent literature, and poring over monochrome scans of service manuals, and schematics. In all my searching, I never encountered even a mention of this document anywhere. So rather than attempt to summarise, or re-print the document's information, I present it to you in its entirety.

I've scanned all of the individual pages3, and uploaded the compiled document to archive.org. It's now freely available here. I ask that readers please forgive the document quality. Archive.org perform their own re-compression of the uploaded PDF. Converting it from the original high-quality greyscale scan into a different format, giving the PDF document an unfortunate brown tinge. I also have no control over archive.org's automatic uploading of useless OCR text scraped from the PDF.
If you would like to host the original high-quality scan somewhere, please get in touch.

The descriptions of the individual internal circuits are detailed, and extremely informative. Among many other details, the documentation features an interesting classification of the data transferred between the CPU, and the EGS/OPS into three categories: EVENT, DATA, and TIME. This hints at a possible naming scheme used in the original firmware assembly to classify the data.

The technical details seem oddly specific in some areas, yet many of the most interesting technical details are conspicuously absent in others. For instance, the operation of the sub-CPU's communication with the main CPU is described in gruelling detail, including a detailed description of its internal data exchange format. However no details are provided of the internal implementation of the synth's sound chips.

After having disassembled the firmware, there is not much information here that is not already available. However for anyone interested in doing any serious firmware hacking, or reverse-engineering of their own this would be an amazing resource.

I was unfortunately not able to discover any details about the document's provenance. The seller was not from a technical background, and couldn't tell me more than that he had purchased it himself in an online auction some time ago. He was selling a wealth of other service manuals, so perhaps he had come into its possession by way of a technician liquidating their collection.

One particularly interesting detail is that the booklet is written in perfect English. Yamaha would not have gone to the trouble of having their technical material professionally translated without a good reason. The text accompanying the cartoons in the margin appear to be hand written. Does this hint that the booklet was originally written in English?
Also, who was the book's intended audience? Was it for the benefit of Yamaha's international engineering teams, or their international service centres? Yamaha would certainly not have published such a thing for the sake of the general public's curiosity. Could this have been documentation provided to the Canadian team responsible for the v1.9 "Special Edition" ROM? Who knows. If anyone has any extra details about the history of this book, I would love to know.

When I wrote my original technical analysis of the DX7, I really did not expect that any of the original technical documentation would ever surface. Needless to say, this has been a great surprise. Unfortunately I do not have high hopes for any more original technical literature surfacing. One can only wonder what kind of material Yamaha themselves would still have in their possession at this point. With the DX7 itself rapidly approaching four decades old, and the golden jubilee around the corner for many of Yamaha's original FM patents, I would imagine that as much information has become public as ever will. At any rate, I hope that making this document public provides some interesting reading for synth hackers for many years to come.


  1. I'd picked the title of my original blog post tentatively on the basis that I couldn't think of anything more appropriate. It's a very strange coincidence indeed that this piece of official technical literature bears the same name!
  2. I'm not sure how to interpret the numerical codes in the bottom margin of the front cover. The format of which matches that of the DX7's better known technical literature. It likely indicates the document was printed in October of 1984. The book itself is light on details of its authorship, instead it gets straight down to business.
  3. The book also included the synth's schematics in an A1 fold-out, which is not included in my scans. The schematics have been freely available online for some time as part of the DX7/9 service manual. I may at some point find a way to scan the full schematic in its entirety, as the existing version available online has been split into A4 size chunks.