[sci.electronics] fast A/D for video

clldomps@cs.ruu.nl (Louis van Dompselaar) (04/22/91)

Hi there...

Can anyone help me with my quest for a fast A/D for video
purposes. I now work with a A/D which only is 3 bit.
It uses the ordinary 8 comparators, combiner etc.

I want at least 5 bits (to get 32 grey shades) and might
even consider 3 x 5 bit for a lot of colours.

Louis
clldomps@praxis.cs.ruu.nl

elliott@optilink.UUCP (Paul Elliott x225) (04/24/91)

In article <1991Apr22.100053.1486@cs.ruu.nl>, clldomps@cs.ruu.nl (Louis van Dompselaar) writes:

> Can anyone help me with my quest for a fast A/D for video
> purposes. I now work with a A/D which only is 3 bit.
> It uses the ordinary 8 comparators, combiner etc.

> I want at least 5 bits (to get 32 grey shades) and might
> even consider 3 x 5 bit for a lot of colours.


Fujitsu makes 6 and 8-bit video A/D converters that convert at
20 M samples/sec (min, 30 M samples/sec typ), and they are pretty
cheap (around $10).

6-bit : MB 40576
8-bit : MB 40578

I've only seen these in the catalog, so don't know if they are really
available.


Brooktree makes lots of video conversion chips; they tend to be slower
than the above Fujitsu parts, but more highly integrated.  They have a
triple-ADC for color applications that looks interesting, and they will
gladly mail you lots of spec sheets and app notes.  They had a regular
media blitz going on a few months ago, but *now*, I can't find their
address in the stack of magazines on my desk.

There was a survey article on the subject in _EDN_ magazine (or it may
have been _Electronic Design_) a few months ago.  Perhaps a dozen
manufacturers were mentioned.

Fujitsu Microelectronics, Inc:
(408) 562-1000

-- 
Paul Elliott - DSC Optilink - {uunet, pyramid, tekbspa}!optilink!elliott
"...We spread out over all levels of all governments.  The threat of wars,
famine, and environmental disaster disappear." - Marilyn Davis, Ph.D.

tomb@hplsla.HP.COM (Tom Bruhns) (04/25/91)

clldomps@cs.ruu.nl (Louis van Dompselaar) writes:
>
>Can anyone help me with my quest for a fast A/D for video
>purposes. I now work with a A/D which only is 3 bit.
>It uses the ordinary 8 comparators, combiner etc.

If I read your address right, you ought to try Philips.  They have
some new chips for digital video processing, including A/D and D/A.
They sent me a sample of the A/D-D/A pair, and I don't think I had
even specifically asked for a sample.  Don't recall the exact part
number--sorry.  It's an 8 bit 40MHz part, 24 pin DIP as I recall,
for the A/D

jkubicky@nntp-server.caltech.edu (Joseph J. Kubicky) (04/26/91)

tomb@hplsla.HP.COM (Tom Bruhns) writes:

clldomps@cs.ruu.nl (Louis van Dompselaar) writes:
>
>Can anyone help me with my quest for a fast A/D for video
>purposes. I now work with a A/D which only is 3 bit.
>It uses the ordinary 8 comparators, combiner etc.

I used a Sony chip last summer - CXA1096P, I think.  It went up to
about 8MegSamp/Sec (8-bit flash type).  It was much nicer to design
with than a similar (at least, similar specs) Brooktree ADC, as the
Sony had separate analog and digital supplies (as any good ADC should)
while the Brooktree device (don't remember part #) functioned off a
single supply and required a 'calibration' pulse every couple hundered
usec or so (it was set up to happen in conjuction with the h-retrace,
but just got in the way for my application).  I also remember seeing
an add from TRW around that time (end of last summer) that they
were coming out with a video-bandwidth 8-bit flash device, but a
few calls produced no specs or samples.

Jay Kubicky