Path: utzoo!utgpu!jarvis.csri.toronto.edu!mailrus!tut.cis.ohio-state.edu!cs.utexas.edu!usc!neuro.usc.edu!annala
From: annala@neuro.usc.edu (A J Annala)
Newsgroups: sci.electronics
Subject: Re: Digital recording on a VCR
Message-ID: <22821@usc.edu>
Date: 11 Feb 90 06:41:48 GMT
References: <21254@siemens.siemens.com> <13400@cit-vax.Caltech.Edu>
Sender: news@usc.edu
Distribution: usa
Organization: University of Southern California, Los Angeles, CA
Lines: 66

In article <13400@cit-vax.Caltech.Edu> toddpw@tybalt.caltech.edu (Todd P. Whitesel) writes:
>
>me and a friend have wondered how tough it would be to record a stream of
>digital pulses on a vcr tape and play them back (i.e. cheap HD backup, speed
>doesn't really matter). 

I have an eight channel sequentially multiplexed analog-to-digital converter
in my lab hooked up to a SONY PCM-F1 digital audio signal processor.  This
unit converts 16 bit ADC samples at 88.2 KHz and stores them on just about
any cheap commercially available low end consumer grade VCR.  It works just
as well in extended play mode as it does at higher speeds.  The SONY PCM-F1
generates a standard digital recording format called EIAJ PCM.  

The trouble comes when you try to hook this device up to an IBM PC/AT.  My
simple minded 16 bit wide synchronously clocked digital input board talks
to channel 6 of the second 8237A DMA chip just fine.  However, I am unable
to write any data to the hard disk (via turbo/microsoft c io library calls,
msdos file services, or direct calls to the hard disk read/write bios ints.
Does anyone have a clue why a DMA interface accepting 16 bit wide data at
88.1 KHz would interfere with access to my hard disk?  Does the controller
for the hard disk assume no dma cycles are present on the bus while it is
trying to write data to the disk?  Is there any solution that would allow
simultaneous DMA input and file oriented or raw hard disk output?

The only solution I have at the moment is to build a buffered digital
interface containing a FIFO chip (IDT7202 -- any know what this is --
where to buy a few of them -- where to get a spec sheet -- how large a
buffer memory is available) that will hold data while I mask (disable)
DMA input long enough to write to my hard disk.  This buffering scheme
was presented (but only briefly discussed) in Dr. Dobb's Journal Jan90
pp. 28-37, pp.94-95 with a hardware outline and supporting MSC5 code.
Does anyone make clocked digital input boards with large fifo buffers
for commercial purchase?

Thanks, AJ Annala, USC Neuroscience Program, HNB-534A, LA, CA 90089-2520
Telephone:  (213)743-3251 (office)  (213)743-8868 (lab)

-------------------------------------------------------------------------
For further information on constructing multi channel analog/digital to
VCR adaptors the following articles are quite helpful and detailed:

Bezanilla, Francisco.  A high capacity data recording device based on
a digital audio processor and a video cassette recorder.  Biophys. J.
47:437-442 (1985).

Lamb, T. D.  An inexpensive digital tape recorder suitable for
neurophysiological signals.  Journal of Neuroscience Methods 15:1-13
(1985).

Chiang, R. G. et al.  Recording electrophysiological data on video tape:
a superior and less costly alternative to conventional tape recorders.
Journal of Neuroscience Methods 15:15-20 (1985).

Blaettler, Markus et al.  Data recording and playback on video tape -- a
multi-channel analog interface for a digital audio processor system.  J.
Neuroscience Methods 23:139-148 (1988).

These systems are also available commercially from a number of sources.

Medical Systems Corp, One Plaza Road, Greenvale, NY 11548 (516)621-9190.

Vetter Instruments ... your neuroscience people get advertisements from
these people all of the time ... some of their adaptors are just simple
FM modulated carriers representing analog signals on modified VCR's ...
other adaptors are analog to digital --> multiplexed --> digital to PCM
--> recorded as bright spots on VCR.