[net.micro] Tape loading problems revisited...

YEKTA@mit-mc.arpa (03/09/83)

From:  Yekta Gursel <YEKTA@mit-mc.arpa>


	Here is some reasons and solutions for tape loading problems people
might be having with almost any computer:

(a) First of all; the signal that comes out of the computer is nothing like
a sine wave; usually they are square waves of frequencies 1200 Hz and 2400 
Hz.
(b) When these square waves are recorded and read back from the tape,
teh waveform that comes out of the tape recorder is nothing like a square
wave.  The reason for this is that a square wave has a lot of high frequency
harmonics, and tape recorder amplifiers have severe attenuation at high
frequencies.
(c)  This is where the "tone" control on the recorder becomes important.
Some cassette recorders have bad high frequency response.  ( For recorders
in the price range $35 to $100, the price is usually NOT correlated with the
high frequency response; a cheap recorder might have a better high frequency
response, believe it or not! )  What this does is to give back something
that resembles a sine wave.  Usually , the tape input ports on cheap
vcomputers are happy with a sine wave; because they have a pulse shaper at
the input.  SO, everything works fine.  If your recorder has a better
high frequency response; depending on the frequency "roll-off" curve,
the output will have a waveform that resembles a sine wave with ripples
superimposed on it.  Poor computers input port thinks these ripples are
data, and hell breaks loose.  The reason why the "tone" control helps
this is that when it is turned towards "bass" it makes the high frequency
response worse, thereby eliminating those annoying ripples.  Since the
ripple amplitude is proportional to the volume control also, usually
a reasonable combination of these controls works.  One might get a
tape recorder where there is no "reasonable" setting of those two
controls.  That is where the job starts becoming hopeless.  Remember
also that the "tape" itself also contributes to this.

(d)  There is a neat solution to the problems above:  If instead of outputting
square waves, your computer outputs sine waves; then almost all of the 
problems above will go away, unless you have a really uncooperative tape
recorder.  SInce almost no computer does this; one might try to make the 
waveform more like sinusoidal BEFORE it gets recorded on tape.  One (very
cheap ) way of doing this is to connect a capacitor ( about 1uf, but one
has to experiment, the value will depend on the computer. Do not get 
discouraged, this is VERY easy... )accross the tape output port.
This does not convert a square wave to an exact sine wave, but it comes very 
close.  It eliminates the high frequencies present very effectively.
This method  worked wonderfully in my AIM 65 for example.  Now, I do not
have to use the tone control at all, and the damned thing works at almost
any setting of the volume control as long as the signal amplitude is not
too small.  

	I hope this will help some of you enjoy your computer more,
whatever it might be...
					Best,
					Yekta

steve (03/14/83)

I just thought I would mention that the TRS Color Computer has a 6-bit D/A
inside and one of the things it does is synthesize sine-waves for recording
on tape - thus making the tape save/restore mechanism EXTREMELY reliable.
I guess on the fourth time around the Tandy engineers finally figured out
how to do it right.