PDA

View Full Version : storing analog data on vcr sound channel



Paul Guy
08-13-1993, 01:14 AM
In your letter:

>Does any one know a way to record analog data onto the audio track of an VHS
>tape? The data have to be saved in a form such that at a later time, I can
>download the data from the VHS tape into the computer.

I have made electronics devices to do this, it's not all that
complicated, but there are some things you need to consider. Most
important is the 'sampling rate'. The audio bandwidth of the sound
channel really doesn't extend more than 10-15 kHz, so if the signal has
any encoding, you won't come even close to the original transmission
channel bandwidth. Using pulse code, or pulse position (my favourite)
you can get about 1/10 the bandwidth (ie 1kHz) of the audio channel.
These types of encoding will give stable DC response. Pulse coding is
convenient if you want to run multi-channel data. The unit I have
constructed was designed to multiplex 16 analog channels, each sampling
at 50 Hz. It handled EMG signals that were processed by linear envelope
technique (absolute value, low-passed at 6 Hz).
If we say the limit for the audio channel is 1000 samples per second,
then the per channel sampling rate is just 1000/no_of_channels. The
channels here are the analog channels (not the audio channels). Using
the Nyquist criteria, the maximum frequency input for each channel is
then 500/no_of_channels. As you can see this is marginal for 1 channel
of raw EMG. For 1 channel of EMG you'd be better off by amplifying the
EMG signal and putting it straight into the audio VCR inputs. If you do
this, make sure that you have calibration signals on the same tape!
Another simple way of dealing with just one channel, is to get hold of
a few v/f and f/v (voltage to frequency, frequency to voltage)
integrated circuits and convert the signal to frequency and back.
This technique will give you stable DC response, and an output that is
independent of the audio VCR gains.
Of course the maximum/minimum frequencies must be arranged to suite the VCR
audio channels, and the signal to be encoded must be bandwidth limited
to suite the v/f circuits. Using the stereo feature on the VCR you can
get 2 channels!
Another trick is to use the audio channels to carry a modem signal.
It won't work with modems that are 'smart' or require a link with the
remote modem (it's not just separated in distance, but in time as well).
Modems that use protocol such as MNP,V42.bis, etc, are going to get
irritated when they can't get a return signal.
I have found that many of the VCR's do not preserve signal amplitudes
very well. If you record with a 0.1 volt input, chances are pretty good
you won't see that 0.1 volt on the output, even on the same VCR
playback. Your electronics has to be able to deal with that.
The 'HI-FI' models are much better for using analog data storage.
There is much less cross-talk, noise and junk. You should watch out for
artifacts at the tv horizontal frequency (15,750 Hz) and level
compression artifacts. The latter is not really a problem since the
encoded signals are usually fixed amplitude. The 'HI-FI' modes seem to
give better frequency response, pulse shapes are much better preserved
in this mode.
If you want more details about the circuitry please contact me
through e-mail.

-Paul


-----------------------------------------------------------------------------
Paul J Guy work phone:519-885-1211 ext 6371
paul@gaitlab1.waterloo.edu FAX:519-576-3090
pguy@healthy.waterloo.edu