TS5A9411 Bandwidth - What am I missing?

Hi Paul,

Thank you for your question. There are many factors that affect the signal bandwidth, and it is difficult to represent them by a single equation. For example, the capacitance changes for different frequencies, and the board layout itself will have a limited bandwidth.

The datasheet bandwidth spec comes from actual measurements during the testing of the part.

Let me know if this helps!

-Ryan

Thank you for taking the time to answer.   I know that the frequency response of a switch is more complex than a simple equation; but  it has been my experience that the parameters I cited are the primary determinant of a switch's bandwidth. 

But I asked my question because I believe there is something inconsistent about the BW rating. 

Parts with poorer specs (higher Con and Ron) in the the same TS5A family have higher bandwidth ratings (see TS5A2053 @ 330MHz) .  And since posting my question, I noticed a graph in the TS5A9411 datasheet which shows a -3dB bandwidth of about 500MHz at Vs=3V. (-3dB from the low freq gain).  Which is much closer to my number than the one in the datasheet.  

For the TS5A2053 a similar graph shows the BW rating to be taken at  -1.5dB (from the low freq gain, -2dB overall).  The TS5A9411 BW spec appears to be at -0.1dB.

So now I wonder how is the bandwidth number determined?  Is there a consistent method?

Hi Paul,

Thank you for pointing out these graphs. We are looking into this, and I will let you know of any updates.

-Ryan

Hi Paul,

The 100 MHz typical spec in the table takes part variations into account. In other words, it is an appropriate and conservative measure of the typical (not max or min) -3dB bandwidth for a large volume of parts.

The graph you are referring to on the TS5A9411 datasheet is a specific example of measurements on a representative part, and its intent is to show the general shape of the bandwidth curve rather than to be used as a source of bandwidth spec data.

What is the end application for the switches you are looking at?

Regards,

-Ryan

What is the parameter of the part which limits the typical bandwidth of the TS5A9411 to 100MHz?  

The max Ron and typical Con do not indicate such a limited bandwidth.  

The low Con and Coff make this part very appealing for my application but not if there is some unspecified parameter which will undermine the bandwidth.

Hi Paul,

We have further discussed the estimated BW, 100 MHz spec, and bandwidth graph in the datasheet. We are looking into methods used to determine the 100 MHz. We are also looking at a possible re-characterization of the TS5A9411 bandwidth to verify the datasheet spec vs. the graph.

In the meantime, what are your specific application requirements? If the 100MHz bandwidth is too low, we can recommend a different part for now based on your feedback.

Thank you for pointing this out to us.

Regards,

Ryan

Thanks, Ryan.  I look forward to hearing the results - when might that be?  Other than the doubt created by the BW spec it really is perfect for my application.  I am also considering the TS5A2053 but its capacitance is more than 2x.  Also the '9411 is physically smaller and the pin-out is easier to route.

Of course I welcome suggestions for alternatives.   I am switching an analog signal with a bandwidth of at least 100MHz.  Since the entire signal chain must pass 100MHz I want to optimize each piece - cost is also a primary concern.

When I model the switches I just use the Con/Coff and Ron.  I only look at the bandwidth rating as validation.

Ryan,

Can you tell me, please, what besides Ron and Con can effect the bandwidth of a part?

Or is it that the Con and Ron specs are not necessarily valid at high frequencies?
If this is true how do Ron and Con vary with frequency?

I am trying to keep my signal path as flat as possible to 100MHz.  It would help me if I better understood how these switches behave at high frequencies.

I would appreciate any insight or if you could point me to some other reference material I would be very grateful.

Thanks,

Paul Bryson

Hi Paul,

I am looking into your questions of resistance and capacitance vs. frequency.

In the meantime, have you looked at the TS5A3157? It has similar specs to the TS5A2053 but with a size and cost advantage. It is also rated at 300MHz typical bandwidth.

-Ryan

Thanks, Ryan.  I had not found the TS5A3157.