| CBMSTEVE.CA - Personal pages of Steve J. Gray |
| Changes | Info | Prototypes | Remakes | PET Projects | CBM Projects | Modding | 3D Models | KiCad | Github | Ohio Scientific | Other | My Software | My Collections | Facebooks |
Welcome! This is a project to recreate a "Commodore V364" computer. The V364 was to be the top model in the TED series of computers, which includes the Plus/4, C16, C116, C232, and C264. It reached pre-production stage but was cancelled when Jack Tramiel left Commodore. Only a handful of these computers were made and few survived. The computer is very similar to the Plus/4 in terms of styling and function, but adds a numeric keyboard and speech capabilities based on the Commodore Magic Voice speech system.
For a long time I have been interested in this machine, as it would have been the first Commodore machine with built-in speech. Sadly, there is no known schematics for the machine, and since it was never released there are only a few machines of which to study. The V364 used a custom gate array to assist in the speech generation and sadly is not available. Luckily we have schematics for the Magic Voice cartridge and have fairly good knowledge of how it works. The goal of this project is to make a machine that looks similar to the V364 and has speech capabilities that are software compatible to the original but not necessarily with original hardware initially. I also will be starting with a mockup case based on existing Plus/4 cases and then eventually to a full 3D printed case. For the speech I have an idea to interface an existing audio playback system and to re-write the firmware to support it in a transparent manner.
I started this project a while back, before the CLCD project and am just now making a project page. The experience making the V364 keycaps was helpful in doing the CLCD keyboard. And the 3D modelling experience from CLCD will come in handy for the V364 case.
The recreation will initially use the Plus/4 PCB. There will be additional room inside the expanded case to install additional modules or pcb's to implement the speech. A long-range goal is to re-create the real V364 motherboard, but that will depend on how the speech is implemented. Although it is unlikely, it's possible that additional gate array chips may be found or reverse engineered.
STATUS: I have a donor Plus/4 motherboard to use.
The V364 keyboard is basically a Plus/4 keyboard with a numeric pad added to it. In fact, the numeric pad is wired in parallel with the matching keys on the main area. This allows the numeric pad to work with existing firmware without modifications. The down-side is that the additional keys can not be detected individually. I designed a keyboard PCB with Kicad. It has options for using the original diamond-shaped arrow keys, or to use an inverted-T cursor layout like modern keyboards. To use the diamond cursor it uses small button switches. The function keys use mx switches, and I have duplicated the function keys so that you can pick either original wide-spaced keys, or Plus/4 spaced keys. This means the pcb could also be used as a replacement for the original Plus/4 keyboard. Diamond and function keys will need to be 3D printed to adapt to the mx stem and/or button stem.
For keycaps I went with MaxKeyboard.com who have a Custom Color Printed Keycaps service that uses a dye-sublimation process. Each key can be a different colour, from a choice of over a dozen colours. You provide the vector artwork for each key using their keyboard templates. You can put basically anything you want on each key. To recreate the V364 keyboard I used a high-resolution scan of an actual Plus/4 keyboard and I matched the font and symbols as best I could, and created the graphics symbols all in vector format. I worked with MaxKeyboard and created a Keyboard Layout Editor ".JSON" file for them to evaluate. They said they would create the custom keycaps based on that design. I then took the scanned plus/4 keyboard and used it to match the font and sizing and graphics as vectors and used their templates to try to match the V364 layout on a 104 key pc keyboard for them to use with my custom layout.
STATUS: I had a few keyboard pcb's made and have populated one with standard MX switches and button switches for the cursor keys. Custom cherry mx keycaps have been designed and have arrived, but I have not created 3D function or diamond arrow keycaps yet.
I have hacked two Plus/4 cases together to form what I call a "Plus/5". It has the added numeric pad but it is not an exact match to the V364 case since that case has a different style of function keys and the case is a bit deeper than the Plus/4 case. This case will do temporarily for development.
I have been working on a true V364 case using Sketchup. I used a Plus/4 model by Commodore Cheetah as a start. The model is a Plus/4 on the outside, but is hollow on the inside. It was designed as a prop and not as a solid printable object. I have been adding internal structures to try to match the Plus/4 case as a first step. After that I will make the modificications to turn it into a V364 case. I am trying to make the model modular with components that can be stretched etc to more easily adapt it.
STATUS: Franken-case complete. V364 case in progress.
The original Plus/4 firmware can be used as-is. It was used in the prototype. Even the "+4" ROMS can be used. The speech system firmware was implemented as a separate ROM that is activated on boot, much like standard TED cartridges. This ROM firmware was already partially disassembled for use in the YAPE and MAME emulators. I am trying to do a more detailed disassembly with the goal of having it re-assemble with ACME and then adapted to whatever speech playback system I decide to implement. This is an ongoing and possibly changing implementation. I have a system in mind and believe I can implement it with software compatibilitiy with existing BASIC speech commands, but not necessarily with add-on speech files (Magic Voice loadable files).
STATUS: In progress.
The original Toshiba T6721A chip used in the V364 was only ever used in the C64 Magic Voice cartridge. You can still get the cartridge on auction sites. I have never seen the T6721A chip available separately. Even if we could get the chip we would still need to find or recreate the custom gate array chip that Commodore used in the V364. The long-term goal would be to do this. The short-term goal is to get an alternate solution working that might be easier and yet still be software compatible.
The speech playback works by passing in a text string or word# from the vocabulary. If you put in a string the firmware will parse it into individual words, lookup the words in the vocabulary, and convert them to a word#. In either case a word# is passed to the the appropriate routine which initiates an interrupt driven playback of the word. The speech system looks up the data for the word and feeds it to the T6721A which converts it to a waveform.
If you simplify this you have: word#, speech irq, wait for speech to finish, repeat.
The program that wants to play speech doesn't really care about the details or how it is done. So we can simulate speech using a different playback device if we patch the firmware to handle all the details of the process. If you think about the speech it's really just data that is processed into an audio waveform. So I have been looking for alternate speech modules that can be used instead. There are some simple and cheap MP3 playback modules that have simple parallel or serial interfaces. I am evaluating these to see if they can be used.
STATUS: In progress.
The following are resources used in this project:
Some pictures of the project development and results:
Created: Nov 27/2022. Last updated: Dec 13/2022, 11:00pm EST
Send comments or feedback to Steve Gray(sjgray@rogers.com)