A500_ACCEL_RAM_IDE-Rev-2

Improved design attempt for Amiga 500 in socket 68000 Accelerator, FastRAM and IDE Interface

Warning

This design has not been compliance tested and will not be. It may cause damage to your A500. I take no responsibility for this. I accept no responsibility for any damage to any equipment that results from the use of this design and its components. IT IS ENTIRELY AT YOUR OWN RISK!

Overview

The main purpose of this design is to improve on of the Rev 1. As the Rev 1 design was essentially a proof of concept, Rev 2 supports a larger CPLD (95144), SPI Flash, buffered IDE Interface, 2 Spare IO header pins and 2MB SRAM. However, most importantly support for the MC68SEC000 which can be aggressively clocked.

Appearance

Nice 3D model:

... and the populated PCB with only 1MB of SRAM (FastRAM) and 128MByte DOM:

My design goals for the Rev 2 design were to:

1. Use a larger CPLD to contain all the logic (and more) from Rev 1.
2. Support up to 40MHz with 0 Waitstates.
3. Support IDE.device (credit to MHeinrichs https://github.com/MHeinrichs) for a simple and quick IDE interface.
4. Have a SPI interface / SPI Flash for available to eventually support Flash based Kickstarts.

Here is the performance overview at 30MHz:

... and at 40MHz:

BOM

For those wanting to build their own hardware, here is the BOM;

Regarding the CPU, I purchased from here (I get nothing from this. The parts I received were new and legitimate so I wanted to share the source);

https://www.ebay.de/itm/MC68SEC000FU20-IC-MPU-M680X0-20MHZ-64QFP-NEW/311524970603?hash=item488855d06b:g:qe0AAOSwYIhWleGF

Known Issues And Pending Changes

While populated and debugging Rev 2 naturally issues where found and better ideas came to mind. Also contributions from the Amiga community (http://eab.abime.net/showthread.php?t=89165). The following corrections / improvements are pending the next design iteration:

1. Change all SMD capacitors and resistors to 0804 packages with hand-soldering footprints to improve soldering.
2. Correct SPI Flash footprint. Or considering to remove it as I see no immediate benefit.
3. Increase thermal relief around THD and SMD GND connections to improve soldering (as PCB is 4 layers).
4. Add 10uF filtering around the power components (+3.3 volt regulator, CPLD and 68SEC000) and add pull-ups/downs on JTAG.
5. Route PIN 20 of IDE interface to +5V to support DOMs without the need for a power cable.
6. IDE_IRQ to be routed to the CPLD (additional reason why IDE.device was used). Additionally add IDE_WAIT to the CPLD to have the option to support /DTACK Waitstates.
7. Add XTAL Clock buffer.
8. Change RAM Control signals to have /CS tied to GND and connect /OE and /WR to CPLD. RAM access can be faster.
9. Support a cascaded AutoConfig chain (implemented - still to be tested).