• TI thinks resolved

TS3A4751: Bandwidth for TS3A4751

Part Number: TS3A4751

Dear Sir/Madam,

The bandwidth indicated in datasheet for TS3A4751 is 123MHz.

But it didn't mention whether it is at -3dB ?

However there is BW graph in datasheet shown below that at -3dB, the freq is about 70MHz.

Why the info is not tally ?

Thanks.

  • The graph does not mention many test conditions.

    Anyway, typical charcteristics are not guaranteed; if you're asking which number to believe, the answer is "neither".

  • In reply to Clemens Ladisch:

    Lee,

    Most of the switch datasheets report in the electrical specifications table the -3dB point below the insertion loss of the switch which could be an additional -0.1-0.5 dB depending on the switch. So you could also think of it as the ~-3.1dB point bandwidth from input to output.

    What did you like about the TS3A4751 that you were going to use this switch and why is the bandwidth important? Maybe we can point you to a different device?

    Thank you,
    Adam
  • In reply to Adam Torma:

    Hi Adam,

    I am looking for a switch which can carry 25MHz signal.
    TS3A4751 has been chosen because it has 4 channel , SPST configuration & BW = 123MHz.
    But I still doubt whether this switch can work properly with 25MHz signal without affecting signal integrity.
    After further check on datasheet, the BW shown in the graph is very much difference from it stated in the table.
    I really don't which one should refer to, pls help to advice.
    Thanks.
  • In reply to lee peng kiat:

    What kind of signal, analog or digital? In the latter case, do you need a specific maximum delay?
  • In reply to Clemens Ladisch:

    digital signal & no specific max delay
  • In reply to lee peng kiat:

    For a unidirectional digital signal, you do not actually need an analog switch that tries to preserve the original voltage level; you can simply use a logic buffer that outputs the same digital state seen at its input.

    The logic equivalent of a unidirectional SPST switch with an active-high control signal is 74xx126. For VCC between 1.8 V and 3.3 V, use the LVC family, e.g., SN74LVC126A.
  • In reply to Clemens Ladisch:

    The signal is digital bidirectional & used 1.8V.
  • In reply to lee peng kiat:

    What is my concern is Bandwidth whether TS3A4751 can be used ?
    Thanks
  • In reply to lee peng kiat:

    Okay, a simple buffer does not work for a bidirectional signal.

    The 125 MHz bandwith specification is for the - 3 dB point, and was measured in a 50 Ω system. The graph obviously used some different input or output impedance (but keeps it a secret).

    Your own circuit certainly does not use 50 Ω. If you know your input/output impedances, you can calculate the bandwidth in your circuit:

  • In reply to Clemens Ladisch:

    Thanks your explanation.
    What are the advantages of using buffer instead of switch if the source is unidirectional signal ?
    Does it subject to losses due to impedance mismatch ?
    Does it have Ron which would affect signal integrity ?
    How about propagation delay, which one is better ?
    If couple of buffers are put in between, will it cause any ringing, overshoot or other noise ?