DIY A/V Switch


I knew this was coming out, but [mike] was kind enough to remind me. [Ben Heckendorn]’s latest How-To is up on engadget. He provides a method of building your own A/V switch with as many inputs as you want, cheap. Each input uses a $.50 bus switch – literally, you could build one with 20 inputs if you really wanted to. (add some shielding if you do) If you’ve got too many game consoles, this might be your ticket. You can probably use the same trick to build a component video switch if HD’s your thing.

20 thoughts on “DIY A/V Switch

  1. Very difficult to hard wire to those surface mount parts and not have the leads break. But you can buy small PCB’s you can solder them to that then allow you to solder the wires to the PCB.
    This is a great app for switching those cheap composite video cameras. Make your own switch box to do monitoring on the cheap for many cameras.

  2. Bah-dum-bum!

    Uh, yeah. Anyway, somebody commented over at the Engadget posting about this that if you use it for HD component video, bandwidth might be an issue. What’s the frequency response of those bus-switch chips look like?

    I have a couple of craptastic 3-way A/V switches I got from some online outfit for dirt cheap, and while they seem to work fine when used as advertised (for composite + audio), using the three ports for component video (with a second switch to move audio) generates garbage. I don’t have the equipment to test them properly, but it seems that the audio ports have terrible bandwidth — they might even be filtered, it just occurs to me.

    Anyway, I thought this would be a cool project (+20 points for incorporating a cheap IR/laser/whatever sensor for remote switching? Should be easy enough) but wouldn’t want to get started on it only to find myself in the familiar, unfortunate situation…

  3. This is a very cool project, but why are you referring to it as a “trick”? This is exactly the sort of things the bus chips are meant for. It’s like saying (in 2007) that running a web browser on your PC is a really cool hack and Mac’s are your thing, you can probably use the same trick there.

  4. Nice. But of course it wouldn’t be a BenHeck tutorial without a complete spaghetti bowl wiring setup in the end. I wouldn’t want to put together more than 3 A/V inputs using this technique.

    That said, I know that making a PCB was well outside the scope of this one, but I’d love to see how that would reduce the space and complexity for higher numbers of inputs.

  5. I just checked out the datasheet on these things… they’re not really meant to switch analogue signals at all. While this may work, you’re operating these devices in an unintended operation mode and results may vary from batch to batch. I realize that we’re all hackers here or perhaps only wannabe hackers living vicariously through people like benheck, but it’s just something to be aware of.

    If you really wanted to do this, why not use something like Maxim, National, Analog’s or really *anyone*’s A/V switch chips? Using a high-speed LV244-equivalent doesn’t really seem like such a hot idea, when you can get the same (and likely better) performance out of a cheaper chip designed for the purpose.

    As far as SMT goes — there are 74LV/HC/ACT244s available in DIP form, too… :-)

  6. looks, usefull, if the quality is good, hopefully ill be able to stop manually switching my monitor cable between my comp and 360, i’ve tried kvm switches, but nothing under $50 has given me an unghosted image…

  7. it’s a neat project (not a hack… more of a diy) and haveing built one of these before myself (complete with IR control) I can whole heartedly agree with akmixdown, why use a a more expensive digital chip to do a less expensive analog chips job?

    In terms of AV switches quality is the second most important factor…(directly behind ease of use, else you’d be switching em by hand :P )

  8. ya um guys, that chip is defantly not designed for analog video switching… Those pins are designed for pure digital 0 or 5v signals, and likewise have absolutley no quality control on them, they just need to be able to keep the signal within ttl spec… You got lucky that the signal didn’t get converted to a ttl signal by sending it through the chip.

    If you are going to go to the effort to make a big howto on a project like that, at least use a chip designed for the application, ok?

    What you did is sorta like using a lm317 variable voltage regulator as a high power op amp, sure it will sorta work but the signal is going to be crap.

  9. well “…”, have you had a look at the internals of the chip in the data sheet? the chip only contains so-called “transmission gates” (you might want to google that term), not that different from what a CD4066 does (except for the lower ‘on’ resistance). i had built an audio crossbar switch with CD4052s when i was a teenager and it worked quite well. but i wouldn’t attempt to switch composite video (6 MHz of bandwidth!) with such a chip that has undefined input and output impedances. and a wrong termination can be quite ugly at these frequencies (ghost pictures). Not even speaking of VGA or even DVI/HDMI, that’s mere RF voodoo. Analog Devices has some fine chips for such purposes – but those need a proper pcb layout.

  10. It would be neat to incorporate a HD44780 screen into this project and display a friendly name of the device that is active i.e PS3, DVD, XBOX etc

    Also surely IR control would be an easy addition to this.

  11. Does anyone know if in this project, when switching between inputs, does the current “on” input button need to be pressed to turn off the current input before pressing the on button for the next input? Or does it auto turn off the last input.

Leave a Reply

Please be kind and respectful to help make the comments section excellent. (Comment Policy)

This site uses Akismet to reduce spam. Learn how your comment data is processed.