Newsgroups: sci.electronics
Path: utzoo!utgpu!watserv1!maytag!focsys!jack
From: jack@focsys.uucp (Jack Houde)
Subject: Re: Re:Analog Signal on Floppy Drives?
Message-ID: <1990Jul25.142601.2905@focsys.uucp>
Reply-To: jack@focsys.UUCP (Jack Houde)
Organization: Focus Automation Systems, Waterloo, Ontario.
References: <13807@shlump.nac.dec.com> <13269@mcdphx.phx.mcd.mot.com>
Date: Wed, 25 Jul 90 14:26:01 GMT
Lines: 55

In article <13269@mcdphx.phx.mcd.mot.com> hbg6@citek.mcdphx.mot.com writes:
>In article <13807@shlump.nac.dec.com> sreekanth@rgb.dec.com (Jon Sreekanth) writes:
>>
>>In article <13183@mcdphx.phx.mcd.mot.com>, hbg6@citek.phx.mcd.mot.com writes...
>>>It's difficult to get any hard technical details from a salesman but as I 
>>>understand it, one frame takes up one track on the media. One field is on 
>>>side one and one field is on side two. As the disk rotates, it switches from
>>>one side to the other to construct the full frame.
>>

We looked into this option some time ago vis-a-vis putting video
information on a floppy diskette. Needless to say it went nowhere
real quick given a floppy's dynamics. A hard disk on the other hand
might hold promise. Most disks rotate at 3600 rpm which means that
it would rotate once every 60th of a second, exactly one field time!

The only problem is if the media can handle any sort of reasonable
bandwidth (ie 640 x 480, 256 x 256 or higher resolution).
 
>>About the inspection system you mentioned : I assume it  was storing digitized
>>(compressed) images on the hard disk, in which case it's just conventional
>>technology ? Or was it actually a hardware-modified drive  ?
>>

>Not being able to peek inside I have to go by what the sales guy said. He
>lead me to believe that they retained the drive sub-controller to run the
>heads and spin the platters but the analog lines from the heads were
>DIRECTLY connected to an external analog video board. I am inclined to

I seriously doubt this. You need to work out all sorts of
biasing issues here.

>believe it given their demonstration. The badge system works by swiping
>a card through a reader. The number of the card determines which picture
>is displayed. The time lapse from swipe to display on the CRT was about
>.1 seconds (judged by eye). It would seem to me that that's about enough

You can move a 752 x 484 image onto an SCSI disk in about 1 second (DOS)
which is 363968 bytes. .1 seconds would allow 36396 bytes which is
lower resolution (142 x 256 if we assume an interlaced image) so it
could still be digital.

>time to swing the heads out to the track but not long enough to read a big
>block of data and cram it through a flash converter. 
>

It takes 1/60th of a second to cram it thru the flash.

I would be interested in getting more information on these people
if its available.

Thanks.