[Coco] It's a small win, but a win nonetheless

[Dave Philipsen]

So are you running with just the comparator, no opamp before it?  Is 
your comparator input de-coupled from the cassette out signal going into it?


Dave


On 3/10/2017 8:26 PM, RETRO Innovations wrote:
> Finally, after fighting with the cassette port for a week:
>
> 55 55 55 55 55 55 55 55 55 55 55 55 55 55 55 55 55 55 55 55 55 55 55 
> 55 55 55 55 55 55 55 55 55 55 55 55 55 55 55 55 55 55 55 55 55 55 55 
> 55 55 55 55 55 55 55 55 55 55 55 55 55 55 55 55 55 55 55 55 55 55 55 
> 55 55 55 55 55 55 55 55 55 55 55 55 55 55 55 55 55 55 55 55 55 55 55 
> 55 55 55 55 55 55 55 55 55 55 55 55 55 55 55 55 55 55 55 55 55 55 55 
> 55 55 55 55 55 55 55 55 55 55 55 55 55 55 3C 00 0F 4A 20 20 20 20 20 
> 20 20 00 00 00 00 00 00 00 39 55 55 55 55 55 55 55 55 55 55 55 55 55 
> 55 55 55 55 55 55 55 55 55 55 55 55 55 55 55 55 55 55 55 55 55 55 55 
> 55 55 55 55 55 55 55 55 55 55 55 55 55 55 55 55 55 55 55 55 55 55 55 
> 55 55 55 55 55 55 55 55 55 55 55 55 55 55 55 55 55 55 55 55 55 55 55 
> 55 55 55 55 55 55 55 55 55 55 55 55 55 55 55 55 55 55 55 55 55 55 55 
> 55 55 55 55 55 55 55 55 55 55 55 55 55 55 55 55 55 55 55 55 55 55 55 
> 55 3C 01 02 00 00 03 55 55 3C FF 00 FF
>
> Notes (for posterity):
>
> This site was helpful:
>
> http://www.cs.unc.edu/~yakowenk/coco/text/tapeformat.html
>
> But, it is incorrect that bytes are stored "least" significant bit 
> first.  They are transmitted "most" significant bit first.
>
> I have a few lingering problems I need to fix, but I think I am over 
> the worst of it:
>
>  * The design runs on 5V currently, but will run on 3V, so the voltage
>    divider on the comparator will have to change.  No big deal
>  * The current design only works if my scope is connected to the
>    cassette data out pin :-)  Evidently, the scope probe is loading the
>    line, dragging it into the comparator center.  I'll have to thus try
>    some way to figure out the center of the signal without my probe.
>
> The above looks like I have a few straggler bytes to address, but the 
> bulk looks like it is there.
>
> Jim
>
>
>
>
>