Shop OBEX P1 Docs P2 Docs Learn Events
SPDT silicon video switch — Parallax Forums

SPDT silicon video switch

BeanBean Posts: 8,129
edited 2005-04-07 17:55 in General Discussion
I'm trying to find an electronic SPDT video switch that will work fast enough that I can switch within a video line. All the switches I find with my friend (google) are some kind of matrix switches.

What I want to do is modify my video project so that it will overwrite a section of an incoming video signal. For example allow you to put 1 line of text at the top of a video signal. Like if you had a camera in your model plane the module would put a line of text at the top showing GPS coordinates or something. I do NOT want this to·"overlay" the existing video signal. The line of text will be simple black and white and the incoming video signal will not be present for those couple video lines.

Any ideas ?
Bean.


▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
"SX-Video Display Module" Available Now.

www.sxvm.com

"A problem well defined, is a problem·half solved."
·

Comments

  • SteveWSteveW Posts: 246
    edited 2005-03-27 16:45
    Do you want to switch it at 75 ohm, or something higher impedance? Both exist, but high impedance is easier / cheaper.
    Maxim certainly have devices to suit (and a generous samples programme [noparse]:)[/noparse]

    Steve
  • SteveWSteveW Posts: 246
    edited 2005-03-27 16:56
    Ah, having had a rummage, I remember how I ended up doing this last time - opamps with enable pins.
    The late lamented OPA632 from TI last time I was building test gear - now unavailable, but I can't magine there's no alternative.

    Steve
  • SteveWSteveW Posts: 246
    edited 2005-03-27 17:33
    More rummaging for suitable devices -
    OPA690
    THS4215
    OPA3690
    OPA2690
    THS4504
    (cut&pasted from http://focus.ti.com ) all seem reasonable candidates, with varying numbers of amps, all capable of driving the rather unpleasant load that video presents.)

    Steve
  • Jon WilliamsJon Williams Posts: 6,491
    edited 2005-03-27 17:34
    Isn't what you're wanting to do is mix video signals (incoming + overwritten text)? Your text generation would have to be based on the sync from the incoming signal -- I think this is how other overlay units work.

    ▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
    Jon Williams
    Applications Engineer, Parallax
    Dallas, TX· USA
  • SteveWSteveW Posts: 246
    edited 2005-03-27 18:02
    It is indeed - but someone, somewhere, has to do the switching between passthrough and locally generated content.
    Bean's gone for the easy option of switching whole lines, but teh same applies for pixel switching (although you start to care more about glitch energy as you switch, or things have shiny borders - which you may, or may not, be able to claim as features [noparse]:)[/noparse]

    Steve
  • Beau SchwabeBeau Schwabe Posts: 6,568
    edited 2005-03-27 18:26
    Bean,

    I have used all of the gates in a 4066 (one 4066 per channel to avoid cross-talk) paralleled
    together for a video A/B switch before. As long as you monitor the incoming "main" video
    to gen lock the signal and make your video switch during a Horizontal sync of the main
    video, I would think you should not have too much problem.

    ▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
    Beau Schwabe - Mask Designer III

    National Semiconductor Corporation
    Latest Company News
    (Communication Interface Division)
    500 Pinnacle Court, Suite 525
    Mail Stop GA1
    Norcross,GA 30071
  • SteveWSteveW Posts: 246
    edited 2005-03-27 18:35
    Can you get low enough to make a compliant sync pulse? It's painfully near ground, where 4066s don't exactly conduct well into a 37.5 ohm load, unless you've got split power.
    (However, if you're running high impedance or do have a luxuriant power suply (or can run AC coupled - ewwww), sounds like a fine plan)

    Steve
  • BeanBean Posts: 8,129
    edited 2005-03-27 21:01
    I plan to use the onboard comparitor to detect the incoming sync pulses. Look for the Vertical sync, then look for the next incoming horizontal sync pulses and switch to my video output, cutting back just before for the next incoming sync pulse.
    My video generation is non-interlaced, I guess I'm going to have to do some special handling for incoming interlaced video ??? I'll have to see how it goes.

    Thanks for the ideas... I think the 4066 would be the easiest.
    Bean.

    ▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
    "SX-Video Display Module" Available Now.

    www.sxvm.com

    "A problem well defined, is a problem·half solved."
    ·
  • Paul BakerPaul Baker Posts: 6,351
    edited 2005-03-28 15:24
    You can either ignore the frame field, and just do the 2 vertical line width (the effect of doing non interlaced on an NTSC/PAL) or monitor the commerical vsync (assuming your souce is doing commercial vsync), if you need some references on how to detect the difference between the even and odd field let me know.

    ▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
  • Beau SchwabeBeau Schwabe Posts: 6,568
    edited 2005-04-07 17:55
    Bean,

    I found something that you might be interested in.....




    24-hr samples with eval board available
    http://www.national.com/pf/LM/LMH6574.html
    http://cache.national.com/ds/LM/LMH6574.pdf


    BLOG:
    The LMH™6574 is a high performance analog multiplexer optimized for professional grade video and other
    high fidelity high bandwidth analog applications. The output amplifier selects any one of four buffered input
    signals based on the state of the two address bits. The LMH6574 provides a 400 MHz bandwidth at 2 VPP output
    signal levels. Multimedia and high definition television (HDTV) applications can benefit from the LMH6574's
    0.1 dB bandwidth of 150 MHz and its 2200 V/µs slew rate. The LMH6574 supports composite video applications
    with its 0.02% and 0.05° differential gain and phase errors for NTSC and PAL video signals while driving a single,
    back terminated 75 load. An 80 mA linear output current is available for driving multiple video load applications.
    The LMH6574 gain is set by external feedback and gain set resistors for maximum flexibility. The LMH6574 is
    available in the 14 pin SOIC package.

    ▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
    Beau Schwabe - Mask Designer III

    National Semiconductor Corporation
    Latest Company News
    (Communication Interface Division)
    500 Pinnacle Court, Suite 525
    Mail Stop GA1
    Norcross,GA 30071
Sign In or Register to comment.