[Coco] The New Hires Interface

[John Kowalski]

Luis: I'm not even sure I understand how that sort of optimization is
even possible?  The comparator only returns one bit?  Or if taking two
reads per DAC update, a successive approximation routine won't be
feeding the same DAC values to both axis anyway because both axis will
take different paths in the decision logic.

In any case, we do a regular 6 bit reading first - using whatever
method.   So a better method if exists could be applied there.   but
not for the extra resolution steps.  The extra resolution works
completely differently and relies detecting propagation delays within
the DAC / comparator circuitry in order to sniff out extra
information.  Then we do some math and figure out what that extra
information meant.


Shain: Nah, I don't see any reason to change the pots.  Not unless
they're old and worn and noisy (signal wise).   The CoCo's joystick
hardware is actually very nice, and it doesn't really care too much
about the Ohm values of the pots because it uses them as a voltage
divider to get a reference between 0V and 5V.   Higher or lower Ohm
pots will divide the voltage the same.

We haven't actually tried this on previous CoCos.  I'd guess it might
work.  But the routine relies on the 6809 running at 1.79Mhz to read
the comparator as absolutely fast as possible after updating the DAC.
 At 0.89Mhz, it would have missed it's chance to read back any useful
results.   ...Unless the joystick electronics in CoCo 1s or 2s have
longer propagation delays...  then it might still be possible even at
0.89Mhz, but the program would have to be re-tuned to translate the
different timing & signals back together into meaningful data.



On 10/13/15, S Klammer <sklammer at gmail.com> wrote:
> Very nice indeed.  I hope to try this sometime over the coming weekend.
>
> Would there be any benefit to changing the pots within the joysticks?  Say
> to larger values (200k?) or a higher precision (1%?) version?
>
> Would the DAC on the previous CoCos support a similar result?
>
> Shain
> On Oct 13, 2015 6:08 PM, "Robert Gault" <robert.gault at att.net> wrote:
>
>> Luis Fernández wrote:
>>
>>> True 15 bits = 32000 points, my mistake
>>> Anyway, Moving from 64 to 320 points or more is a great
>>> accomplishmentCongratulations and forth
>>>
>>>
>> Just to be precise :) 15-bits maximum is
>> (2^15)-1=32767=$7FFF=%01111111 11111111
>> To carry this to a 16-bit word on the Coco, (2^16)-1=$FFFF
>>
>>
>> --
>> Coco mailing list
>> Coco at maltedmedia.com
>> https://pairlist5.pair.net/mailman/listinfo/coco
>>
>
> --
> Coco mailing list
> Coco at maltedmedia.com
> https://pairlist5.pair.net/mailman/listinfo/coco
>


-- 
John Kowalski / Sock Master
http://users.aei.ca/twilight/sock