[Coco] Cassette data format

[Steve Bjork]

Somebody needs to read up on magnetic recording technology.  The 
cassette tape system is based on MOVING metal oxidized tape over a coil 
head receiver tuned to audio frequencies.  Basically, a coil moving in a 
magnetic field.

If the magnetic field was to shift from zero to full flux then you would 
get a ringing on the tape with that spike.  Also, the drop in the back 
to zero would also ring.  Also, any drop in quality of the metal 
oxidized on the tape would change the magnetic flux pickup by the 
recording head.  This would be pickup has a error in bit shifting.

But the real problem is the fact all circuits in the AUDIO cassette 
record is design for slow changing waveforms.  A square wave would not 
come out as a square wave on the other end.   This was the was the 
problem with the older computers that used a tape recorder to store the 
data.

The CoCo was one of the first "home" computer to use a sine wave to 
record the bits.  Because of this, they could jump to a higher bit rate 
than the older system.

Since the CoCo was not Tandy's first computer out of the gate, they 
could use what they learn to build a better computer.

Using the DAC to create a sine wave for recording to the cassette 
improved the overall reliability of the tape storage system.

I don't think you ever worked on a Apple II or TRS-80 Model I/III, 
because their "square" wave was a nightmare when it came to making 
production tapes.  It took less time to write the program than to get 
the cassette duplication house up to full speed.  The square wave format 
was that bad.

The CoCo tapes from the duplication houses work the first time.  I don't 
know how many times they asked for our "secret" in making the CoCo 
master to work so well.  "With the CoCo, they (Tandy) got the tape 
system right." is all I could tell them.

Steve


On 8/25/2013 8:33 AM, Mark J. Blair wrote:
> I wonder why they bothered to use a DAC for cassette output instead of just wiggling a digital output. Since the reader only looks at edge crossings, I'd think that sending a square wave to the cassette recorder should work just fine... assuming that the cassette recorder doesn't misbehave when it sees a square wave, i.e. ringing heavily on the edges.
>