• State Resolved
  • Locked Locked
  • Replies 5 replies
  • Subscribers 16 subscribers
  • Views 1448 views
  • Users 0 members are here
Support feedback
Options
Options
Related

This thread has been locked.

If you have a related question, please click the "Ask a related question" button in the top right corner. The newly created question will be automatically linked to this question.

CD4066B: Confusing datasheet inconsistencies

Eric Page
Prodigy 110 points
Part Number: CD4066B
Other Parts Discussed in Thread: CD4051B, MPC508

I'm designing an 8:1 standard-definition NTSC video multiplexer using 2x CD4066B.  Video sources are small cameras that operate on a single 3.3 to 5V rail.

At paragraph 3 on pg 1, the CD4066B datasheet says, "...the well of the n-channel device on each switch is tied to either the input (when the switch is on) or to VSS (when the switch is off)."

At paragraph 8.1 on pg 14, this idea is reinforced: "...the well of the n-channel device on each switch is tied to either the input (when the switch is on) or to VSS (when the switch is off). Thus when the control of the device is low, the output of the switch goes to VSS while when the control is high the output of the device goes to VDD."

Both of these statements seem to contradict the bottom line of Table 1 on pg 14, which states that the output is high impedance when the control pin is low.  The sample application circuit in Figure 16 on pg 13 suggests that Table 1 is correct, as it shows four outputs bused together, but the first two paragraphs above lead me to believe that any OFF output would pull the bus to Vss.

Which interpretation is correct?  Can I bus eight outputs together, or will any OFF output short the rest to Vss?

If my video signal has a DC offset such that the entire 1V pk-pk signal is above 0V (I don't have the cameras yet, but I suspect it will be), can I simplify my design by tying Vss to 0V, and do away with the negative rail, or does the CD4066B require the negative rail for some internal purpose?

Thank you.

  • 0
    TI__Intellectual 2900 points
    Hello Eric,

    Thanks for contacting us for the datasheet confusion. Before addressing all your questions, may I quickly check if CD4066 is the best fit your application?

    Here are your requirements and our questions.

    Configuration: 8:1 MUX (CD4066B is 4-channel 2:1, while our CD4051B is 8:1. Why would you prefer 4-channel 2:1? )
    Signal format: Standard NTSC. (What is the data rate and bandwidth requirement? CD4066's bandwidth is 40MHz)
    Signal Swing: 0~3.3V or 0~5V because the camera is supplied by 3.3V to 5V.

    Additional question: is it possible that when the MUX's supply is 0V, the cameras still send active NTSC signals to the input of MUX?
  • 0 in reply to Fan.Wang
    Prodigy 110 points

    The project began as a 4:1 mux, and the 4066 seemed like a good choice.  Later it grew to an 8-ch device, and I just added a second 4066.  I wasn't even aware that the 4051 existed.  It looks like the 4051 is a better choice for me -- thanks for the tip.

    To be honest I'm not sure about the data rate and bandwidth requirements.  It's just a simple switch for SD video.  As I recall, the highest frequency component in an NTSC signal is the color burst, at 3.58MHz; the rest is a horizontal sync at 15.732kHz and vertical sync at 59.94Hz.   Given that the 4051's -3dB frequency is 20MHz and it's specified on the front page of the datasheet for use in televisions, I'll bet it will work fine.

    Yes, I believe  the output of the cameras will all be above 0V, but I haven't got one to test.  A colleague in the project has a camera on the way to test, so I should know for sure fairly soon.

    The cameras and mux will be powered from the same supply, so I don't think there's a case where the NTSC signal could be present on the input of an unpowered mux.  What is the concern here, and should I take some step to protect against this happening?

    If it turns out that the camera output is 0-3.3V or 0-5V, I should be able to power the 4051 with Vdd=5V and Vss=Vee=0V, correct?  It would be nice not to have to generate a negative rail.

    Thanks for your reply.

    Eric

  • 0 in reply to Eric Page
    TI__Intellectual 2900 points
    Hi Eric,

    Thanks for the detailed information. TI have 12 8:1 commercial rating switch. Here is list. You can narrow down your selection by your camera voltage swing range. If it is just up to 3.3V, you just need to make sure in the selection, Vdd(min) is less than 3.3V, while all 12 options can be used for 5V power supply.


    www.ti.com/.../analog-switches-muxes-products.page


    I was disussing the camera and our switch power sequencing is because if camera powers up first while VDD of our switch is not powered up yet, there will be undesired conducting path from the mux input to VDD. Then, you will need an extra feature called "power off protection".

    Here are three 8:1 devices which can support power off protection as well.

    www.ti.com/.../analog-switches-muxes-products.page off protection
  • 0 in reply to Fan.Wang
    Prodigy 110 points
    Thank you for the additional info. Looking at the switch/mux product selection page, it appears that the CD74HC4051E will be the best choice for me. It has relatively high bandwidth, low on-resistance, it's available in a DIP package and it costs no more than the others.

    I would have chosen one of the products with off-protection as you suggested, but among the parts with that feature, only the MPC508 is available in a DIP package, and it has no bandwidth specification and high on-resistance. I need to stick with a DIP package because the project will be assembled by the end user (it will essentially be a kit). I'll just make sure that power is sequenced such that the mux is on before the cameras receive power.

    Thanks again for all your help!

    Eric

    P.S. Should I submit a documentation feedback form for the issues with the CD4066B datasheet, or will this discussion suffice?
  • 0 in reply to Eric Page
    TI__Intellectual 2900 points
    Thanks Eric for the feedback of CD4066B datasheet. The discussion in this post is suffice.