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

[Dave Philipsen]

On 3/12/2017 4:28 PM, Gene Heskett wrote:
>
>> To add some additional color, the sine wave coming out of the Coco1 is
>> a thing of beauty.  nice and rounded and smooth.
> Thats because the driver op-amp was too slow, and slew rate limited.
> Which is why it MUST be replaced with something considerably faster for
> drivewire to work at a usable speed.

The schematic for the CoCo1 doesn't show an op amp in the circuit for 
cassette data coming out of the computer.  There's a MC14050 acting as a 
resistor DAC. From there it's just a voltage divider and a capacitor to 
shunt off any hi freq. artifacts from the DAC switching.

At $A85C in the Color BASIC ROM you'll find a sine wave table:
    FCB $82,$92,$AA,$BA,$CA,$DA
    FCB $EA,$F2,$FA,$FA,$FA,$F2
    FCB $EA,$DA,$CA,$BA,$AA,$92
    FCB $7A,$6A,$52,$42,$32,$22
    FCB $12,$0A,$02,$02,$02,$0A
    FCB $12,$22,$32,$42,$52,$6A

which intentionally creates a sine wave.


>
>> I didn't pay a lot of attention to the Coco 2 waveform, but the Coco3
>> was jerky, sharp, and seemed unfiltered.  Cost reductions and
>> integration evident.
> Faster op-amps in the 2 & 3's.  The source of that waveform, assuming
> light speed parts is a 3 bit d-a, generating a 5 level approximation of
> a sine wave.  The speed of the parts rounded it off to something a cheap
> cassette recorder could handle w/o aliasing if it had a bias circuit,
> most didn't. Excellent white noise generators without the ac bias and
> erase circuitry. Also often responsible for doing a multipass erasure of
> your data tapes.

I don't have a CoCo 3 schematic in front of me right now but I'm pretty 
sure it's a 6-bit DAC just like the CoCo 1 which produces a 64 level 
output resolution.  The wave table above shows 8-bit values but the 
lower 2 bits are not sent to the DAC.

Dave



>> Jim
>
> Cheers, Gene Heskett