[Coco] MC6883 x 74LS785

Jayeson Lee-Steere cocolistemail at titaniumstudios.com
Sat Sep 10 20:41:29 EDT 2016


Daniel,

I can confirm the same behavior. I tried switching a 6883 into a PAL "A"
Coco 2 board (same as your Misedo) and it indeed only functions as 32k.
Cocochek reported 32K tested OK and Sailor Man crashed on startup. This was
with an 8x4164 RAM module and my SRAM module, neither of which should not
be affected by the different refresh size.

My guess is that the address decode lines (S0-S2) work differently, and
looking at that section of the datasheet (Page 5, Address Decode), they do
differ. Earlier Coco's use some extra logic to properly decode between ROM
and RAM in the upper 32K and IIRC from when I reversed that circuitry, that
logic is different on the Coco 2 board.

Bummer.

- Jayeson

On Sun, Sep 11, 2016 at 4:14 AM, Daniel Campos <daniel.campus at gmail.com>
wrote:

> The same 4164s that I'm using in the 2B now for testing with the 6883,
> were also tested on another CoCo2 with a 6883 and worked just fine. Tested
> and retested with the CoCoCheck utility.
>
> Daniel
>
>
> Em 10/09/2016 16:12, Bruce W. Calkins escreveu:
>
>> This may or may not relate to the issues:  The 74LS785 has a different
>> RAM refresh rate than the 74LS783 AKA 6883.  Also, not all RAM responds
>> well to the different refresh rates.
>> Bruce W.
>>
>>
>>
>> On 09/10/2016 12:17 PM, Daniel Campos wrote:
>>
>>> Thanks to everyone that privately sent me the 74LS785 datasheet. I have
>>> done some more tests and discovered that the CoCo2B board only recognizes
>>> 32Kb of RAM when using a MC6883. If I replace it with a 74LS785 it works
>>> just fine and recognizes the full 64K RAM.
>>>
>>> - Also tested replacing the 2x 4464 DRAMs for 8x 4164 DRAMs (the board
>>> has the places for both), but the result is the same.
>>> - For those who asked, yes, I tested all the 3 MC6883 used in the tests
>>> and they work just fine on another CoCo2 (not 2B) that I have here.
>>> - I have tried to replace the CB 1.3 + ECB 1.1 EPROM from another with
>>> the CB 1.2 + ECB 1.1, but with no effect.
>>>
>>> Daniel
>>>
>>>
>
> ---
> Este email foi escaneado pelo Avast antivĂ­rus.
> https://www.avast.com/antivirus
>
>
> --
> Coco mailing list
> Coco at maltedmedia.com
> https://pairlist5.pair.net/mailman/listinfo/coco
>


More information about the Coco mailing list