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.

TLK2711-SP AC Coupling Caps

Other Parts Discussed in Thread: TLK2711-SP

Hello,

Three related (multi-section) questions:

1. How do you optimize the 1 nF to 10 nF range expressed in the datasheet for your design output frequency (given that you will size the caps to match the trace width)? As an example, say the output data rate is 1.28 Gbps and for another application it is 2.4 Gbps.

Do you  try to find a cap value that brings 1/2πFC at the data rate closest to zero ohm? Do you consider the resonance frequency between these caps and the trace inductance and other secondary effects?

 

2. If two assemblies are communicating via a pair of TLK2711-SP transceiver (a transmitter to a receiver 10 meters apart and each using a different power supply), do you use the AC capacitors on the hi-side and the low-side of both the transmitter and receiver ends?

If so, how closely should these 4 capacitors match? If only two caps needed, which side (at the transmitter or receiver) should they reside and how closely should they match? Finally, It is recommended to place the Caps closest to the output connector at the transmitter end - I am assuming this is done to remove any common-mode current due to the entire route before the signal is existed the board, is that correct? Where should the caps at the Receiver end reside, closest to the TLK2711 chip?

 

3. For reliability, if one cap fails short (or maybe open), do you recommend parallel series cap combinations or will that cause too much signal integrity degradation or loss?

 

Thanks

-Ali

  • Hi Ali,

    Since the “TX” signals are connected to an “RX” for data to pass, only one set of AC Coupling capacitors are needed and these are typically placed near the RX pins.  However, they could be placed near the TX pins instead but not both.  Placing two caps in series will reduce the overall capacitance and could cause bit errors.

    This value could be changed if desired.  The 1nF-10nF value is the recommended value for most applications, but depending on the data transition density and the receiver’s characteristics a different value could be appropriate.  If the value is too small, the capacitor could become fully charged or depleted during longer strings of consecutive 1’s or 0’s respectively.  This can cause bit errors because the data will not pass, and also because the common mode of the signals will change and causing the receiver to improperly sample the data.  Likewise, if the capacitor value is too large, it could cause bit errors by preventing the AC data to pass at higher data rate due to a slower charge constant.

    Best Regards!

    Luis Omar Morán Serna

    High Speed Interface Group

    SWAT Team