Dragon32

DRAGON 32 CPU Issue 2

This project is a recreation of the SA2120 CPU Mainboard for the 1982 Dragon Data Dragon 32

Everything (excluding modifications) is taken from the SA2120 schematics. Component positioning is as close to the original too. The DRAM positions are as close as possible to the 1982 board but in preference use 4164s instead and set the memory configuration jumpers appropriately, this avoids transplanting a riser board to use the 16x16k configuration. The 16x16k should still work of course...

The two ROMS will need to be sourced from original boards or an adapter used to allow the use of modern, available EPROM or EEPROM components

The variable resistor (RV1) and motor relay (RL1) are adjusted for modern, available components. The cartridge port is based on a modern design but may still be hard to source

Important - Read This Bit!

September 2023: In the year since I started developing this board the availability of the MC6883P/74LS783N SAM chip has gone from "easy to find" to almost unobtainable. The only known and proven source is from DragonPlus Electronics. Please make sure you can find one before starting your build

There is work being done to create a replacement for the SAM chip but this is yet to become available

The rarer (but ironically easier to buy) 74LS785 should work as a substitute

Caveat

The scans on the original SA2120 schematics are not always clear and it has taken a fair amount of careful study to get everything correct but there is no guarantee at this time that the board design is as intended

The values of passive components are taken from observation of an original production board, in most cases those values do match the schematics, but not always

Revision 2.3 of this board is in a genuinely usable state but still with caveats

Working

In the current state (2.3) basic operation is tested working on CPU, RAM, Keyboard, Joystick and Video.

Working with caveat

Audio appears to be good but not fully tested (given the joysticks work and the board produces audio there isn't much to go wrong here)

Untested

Parallel port is untested in either mode beyond verifying signals to the port are present

Modifications

Joysticks

The two joystick ports provide a second fire button capability wired to consecutive keyboard rows, as per the Tandy CoCo. Unless a two button joystick is used this make no difference to the operation of the computer. It also requires the use of 6-pin DIN sockets, these are pin compatible with the 5-pin originals so a regular single button joystick can still be used.

The two extra buttons can also be disabled by omitting L3 and L4

Parallel Port

An extra jumper has been introduced to bypass the logic inverter on the STROBE line. When bypassed the parallel port should operate as a communication port for DRIVEWIRE. In effect this enables the printer port to behave as a TTL serial connection. Speed is limited by software and raw clock speed.

Memory Addressing

The board is configurable between 32k and 64k ram addressing using jumper 13/14. Position 13 is for original 32k ram configuration, position 14 is for an upgraded 64k ram. Note that this then requires removal of some additional passive components and assumes that suitable 64k x 1 ram chips are fitted in IC1 to IC8.

Jumper 9/10 has been added for consistency with some of the other early designs and for compatibility with the 64k upgrade. It is important to avoid populating C50-57 when using this option, along with C12, C58, C59 and R66.

See [World of Dragon: Dragon 32 - 64k Upgrade] (http://archive.worldofdragon.org/index.php?title=Dragon_32_-_64K_Upgrade)

Video

The Rev 2.1 board adds two (hidden) solder jumpers for adapting the board to NTSC output. By default these have a bridge putting the board in PAL mode so unless you want NTSC output there is nothing to be done.

To change the configuration the bridge on both jumpers needs to be cut and the option pads bridged with solder. This bypasses the additional logic for video sync. The jumpers are JP15/16 and JP17/18, these are positioned below IC15 and IC12 respectively. The default position is marked.

To complete the change an NTSC adapter board needs to be fitted. This piggybacks the 6847 position and adds in a suitable NTSC video modulator. The output of the NTSC board goes to PL8, next to the PAL colour crystal XL1.

In this configuration there is no need to fit a number of components:

• IC9
• IC10
• IC11
• IC14
• IC21
• IC22
• IC26
• IC27
• IC29
• IC30
• IC32
• C2
• C3
• C5
• C6
• C7
• C24
• C25
• C28
• C46
• C47
• C71
• R2
• R3
• R4
• R17
• R18
• R33
• R34
• R35
• R44
• XL1

Substitutions

Many of the original components are simply unobtainable or just very hard (and expensive) to locate. The transistors will need modern substitutes as the BC141, BC212 and 2N2369 are all obsolete but thankfully these are fairly run-of-the-mill transistors and shouldn't cause any particular issues

The video op-amp SFC2318D at position IC10 is especially hard to find and this one component can easily double the price of populating the board if an original is used... I've identified a Harris HA-5111 as a likely candidate, it is an audio op-amp rated for upto 40MHz so plenty of bandwitdh for a simple PAL/NTSC video signal. There are of course lots of other single channel op-amps out there in a dip-8 format but they tend to be expensive (as in really expensive) possibly cheaper than a real 2318D but not by much, some are stupidly expensive (nearly 100x that 2318D). The HA-5111 should be about £0.90

The CPU, PIAs and VDG are all fairly easy to obtain online and possibly even new (MC6821s are still made and there are lots of compatible alternatives). A and B rated components should all be usable but make sure the processor is an "09e". The Hitachi HD6309 is usable but may cause crashes in some software due to lazy coding

The 4000 series logic chips are particularly tricky - some have modern equivalents. The MC14053 and MC14050 can be replaced with modern CD4000 series chips. The MC14529 was replaced with a CD4000 series logic chip but even that is now obsolete with no valid replacement

The LM1889 is another hard to find chip but still reasonably cheap online when found

The SAM and the two ROMs are genuinely unavailable except from donor boards but the ROMs can be replaced with 27C64 eproms fitted with a suitable conversion shim board.

The Future

As these parts become harder to find it is inevitable that replacements and redesigns will be needed. Abandoning the original composite video output circuits and moving to a (more) modern RGB or VGA output instead removes a big chunk of the more troublesome hardware, notably the LM1889 and SFC2318 along with a number of other common logic chips. If a composite output is still required an ADR724J can be used to generate the necessary signal, all in one chip

A cheat replacement for the 6847 exists, using a raspberry pi pico but this could easily be reduced even further an FPGA (assuming such chips are readily available again)

The ROMs can be replaced with a single EPROM/EEPROM, even opening up the option of switchable ROM images given the size of a modern EPROM's addressable memory. I have a mini conversion board that fits into a single socket with a jumper to the E pin of the second socket. (see https://github.com/jimbro1000/Dragon32RomFix)

The one thing that isn't so quickly removed from the equation is that MC14529 - a dual 4-channel data selector. Onsemi and Harris used to made the CD4529 that replaced the original MC14529 and it is only fairly recently that the chip became obsolete so there is stock out there but the chip is somewhat specialised and as such really isnt easily substituted even with a modern SMD package.

The most likely candidate here is to use a GAL chip to simulate the switching behaviour of the MC14529 combined with discrete analogue switches (such as 74LVC1G3157). This will necessitate a new board design or a plugin board to match the pinout of the 4529 chip