Media Vision Memphis Multimedia Kit - An excercise in proprietary connector mappings

[Stinkerton18]

After having been alerted to a GoodWill auction for a Media Vision Memphis base unit, speakers, and power adapter in very good condition, I decided to get them, then go through the exercise of mapping out the pins to see if it's possible to build an adapter cable to connect most, if not all, the functions without having the matching Media Vision ISA sound card. Part of this post is a collection of notes and resources as I go through this in hopes that it would help others who may also be interested, or want to use the Bose speakers/base unit but have a broken/missing card.

Credit where credit is due, the below info started from this (now abandoned) post on VOGONS: https://www.vogons.org/viewtopic.php?t=63199
In that post, OP does link to some photos they took/posted on Amazon, so the preliminary info I have here is taken from cross referencing the ICs, their respective datasheets, and the visible/clear traces between them and the DB45 connector that goes to the ISA card.

Starting Assumptions​

From the start, we know the following:

• The "base" unit has a SCSI CD-ROM drive inside
• It is capable of audio CD playback without any connection or control from a PC/the ISA card
• From the thread above, audio from the CD drive passes through the speakers, amplified with an onboard amplifier
  • This implies the analog "CD Audio" output from the drive is not necessarily mapped to the ISA card/DB45 connector but instead muxed/mixed directly into the amplifier that feeds the speakers. Doing so would save on pins but also results in the "CD Audio" mixer in DOS/Windows having no effect either.
• The front headphone jack works as well, cutting out the speakers, and amplified at the same level as the speakers, which means the headphone jack is electrically connected in-line with the speakers and acts as a mechanical "cut-off" when a plug is inserted, typical wiring in such cases
• From the VOGONS post: Line In, Line Out, and Mic do not function in this "stand alone" mode, meaning they are likely mapped to pins for the ISA sound card to handle directly.
• There are MIDI Input and MIDI Output pins on the base unit, which of course will have to map to an MPU401 on the ISA card
• There has to be at least 2 pins mapped to "Sound output" the ISA sound card to play audio generated by programs/the OS through the speakers.
• The various connectors/SCSI bus/ICs share a Ground line or two, since this would greatly reduce the number of pins required
• EVERYTHING above has to map back to the ISA card via the single DB45 connector.

Collected Facts​

These are the facts that I've collected thus far

• The base unit has a Logic L53C80JC2 PLCC-44 SCSI Bus controller (datasheet found and attached)
  • This chip helps reduce the pins required to map back to the ISA card and allow the onboard play/pause/skip/previous/stop buttons to function for audio CD "sans PC" controls
• The base unit has a Zilog Z86E4012VSC PLCC-44 Micro controller (datasheet found and attached)
  • Further implies "onboard" logic and control of CD audio playback functions without a PC and without requiring a special SCSI CD-ROM drive
• The base unit has a Texas Instruments TI16L8-25 PLCC-30 PAL (datasheet found and attached)
  • First potential "gotcha" found on reverse engineering this. Chip will have to be removed to read/dump...

Current Actions​

The list of "What's next"

1. Create a Master Legend of connector pins and IC pins to use with the connection mapping table/matrix
2. Upon receipt, trace/map out all 45 pins from the DB45 connector to connectors/ICs using the Master Legend for to<-->from mapping names and designations
3. Remove the TI16L8-25 PAL and dump to a file
   1. Find someone willing to help with reading/understanding the dump taken OR
   2. Learn how to read/translate PAL code
4. Obtain a DB45 Male connector
5. Begin building a cable to break out MIDI/Line In and Out/Mic/Sound input from the ISA card
   1. Test/verify each as it's done to validate
6. Create Github repo to store "polished" results.
7. ???
8. PROFIT!!

 

[Stinkerton18]

First set of notes:
Looking here:

There's traces going from the SCSI controller directly to the Zilog, specifically pins 20, 21, 22, 24, 27 and 28 are visibly connected. Per the Logic datasheet, they're D7, D6, D5, D4, D1 and D0 respectively.

It's probably a safe assumption that Pins 25 (D3) and 24 (D2) are also routed to the Zilog microcontroller just through vias to the backside of the PCB.

 

[Stinkerton18]

Looking at the Logic SCSI chip, we can see (from those photos), Pins 17, 16, 15, 14, and 12 for sure are routed to the DB45, 13 likely is too but need to trace the via on the back of the PCB.

Those correlate to the CS, A2, A1, A0, and GND pins, which A2-0 are the Address lines, defined as "Inputs/Active high. These signals, in conjunction with CS, IOR, and IOW, address the 53C80 internal registers for CPU read/write operations", which is likely how the PC/OS would communicate (through the original ISA card) to the CD-ROM drive to read data from it.