[Coco] Coco Cassette interface help

[Gene Heskett]

On Tuesday 28 February 2017 17:26:05 Dave Philipsen wrote:

> > On Feb 28, 2017, at 1:25 PM, RETRO Innovations 
<go4retro at go4retro.com> wrote:
> >> On 2/28/2017 1:05 PM, Dave Philipsen wrote:
> >> All you want to do is amplify the signal that is coming from the
> >> CoCo cassette out. Use a capacitor to de-couple it from the
> >> comparator circuit. It won't matter if there's DC bias on the
> >> output of the op amp because you will effectively remove it with
> >> the cap.
> >
> > True, and I will give it a shot, but I'll continue to search for a
> > way to make the comparator idea work alone.  Amplifying the source
> > only to get it "over the bias" of the comparator just seems wrong to
> > me :-)
>
> If you think about it, in essence, that's exactly what the cassette
> deck is doing by amplifying the signal into the range of amplitude
> that is usable by an 8-ohm speaker which is slightly higher than line
> level.  The whole problem here is that the cassette out is line level
> and cassette in is speaker level. If you want to use the exact same
> comparator circuit used on the cassette input of the CoCo then you
> have to amplify your signal up to speaker level.  Alternatively, you
> can tweak the comparator to work with line level.
>
>
> Dave
>
And that would be the path I would take.

A capacitor to ac couple the input, and a cross input resistor of a quite 
high R, couple megohms at least, (which means the capacitor can be 
relatively small,) coupling the reference input dc wise to the signal 
input, and the input, assuming a usable signal to noise ratio, would be 
well recovered at any level from 10 millivolts up to the rail separation 
voltage peak to peak.  Such a circuit wouldn't even need a bias 
adjustment as long as it had say 10 seconds to stabilize its dc 
operating point after power up. The only item of caution would be that I 
would use an fet input op-amp, an LF357 maybe, and take precautions 
vis-a-vis static electricity so as not to needlessly blow the oxide in 
the LF357's input gates. A common 1n914  from each rail to the input so 
the voltage cannot go more than 1 vbe below ground, or 1 vbe above the 
rail would nicely bulletproof that.  The input capacitor of course 
should have femtoampere leakage at the working voltage of course if the 
signal level coming in is at true microphone level.  Thats lox clean 
pcb's and a higher quality capacitor you'd never get past the bean 
counters on a production line.  BTDT. Its a battle I was not able to win 
with a major scsi card maker 2 decades back.  So, everytime I ran into 
one of their cards, I pulled the si bus isolation diode and replaced it 
with a schotkey w/.o even testing it.  And that was the end of system 
crashes and trashed hard drives, totally and completely. Typically the 
next time I saw that box other than walking by it was when the psu's 5 
volt line was below 4.7 volts, causing it to auto-reset and reboot.

> > Still, I already was thinking of your idea, in the context of using
> > an LM386 to amp the signal brute force to get it to work.  So, I
> > agree it's a workable solution.  Thanks for the suggestion.
> >
> > I am also going to see if I can feed some known signals into the
> > comparator to see if I have messed up the design.
> >
> > Jim
> >
> > --
> > Coco mailing list
> > Coco at maltedmedia.com
> > https://pairlist5.pair.net/mailman/listinfo/coco


Cheers, Gene Heskett
-- 
"There are four boxes to be used in defense of liberty:
 soap, ballot, jury, and ammo. Please use in that order."
-Ed Howdershelt (Author)
Genes Web page <http://geneslinuxbox.net:6309/gene>