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.

THVD1420: Interface forum

Part Number: THVD1420

According to the following Q&A, it seems there's no problem to implement redundant failsafe resistors.

https://e2e.ti.com/support/interface-group/interface/f/interface-forum/1228617/thvd1420-fail-safe/4640161#4640161

But in the case of a multi-drop connection, I think it seems there's a risk that current will flowed in unenergized unit from the energized unit through them. I think it should be designed with consideration about the noise margin and system operation, wouldn't it be better to rely on the IC's fail-safe function in terms of safety? Or will it still have the same problem even with the IC's fail-safe features?

  • You should need only one set of fail-safe resistors on the bus. So if they are in the section that stays powered, there is no problem.

    The THVD1420's receivers inputs have a high impedance when disabled, so if your system can avoid external fail-safe resistors, it should do so.

  • Hello Oda-san,

    1. Its not that there is "no problem" but  as long as the additional set of passive failsafe (only add 1 set of resistors) and they don't violate the loading conditions of the bus they are fine.

    2. There should only be 1 pair of fail-safe resistors if used - one on A and one on B. The voltage that the fail-safe resistor is pulled up to needs to be active if the bus is active. i.e. don't connect the fail-safe resistor to devices that may or may not have power. In multi-drop configurations you'd place it at the driver as if the driver is inactive so is the bus and the idle differential created by the fail-safes apply to the receivers still . 

    Back-powering is only a problem if the system is designed poorly - this is a biasing resistor - so the voltage needs to be on at all times the bus is active or you aren't biasing anything and just burning power. The receivers are hi-z when unpowered and can handle the biasing voltage (bus pins only) 

    There is a trade-off - which is that these load the bus so the max number of nodes will decrease and power consumption will increase compared to systems without passive fail-safes with the use of additional fail-safes but at the benefit of increased noise margin. 

    If your system is noisy then it could be unsafe to not increase the noise margin. If your system doesn't have much noise on the bus possible its not needed   

    Best,

    Parker Dodson

  • >Clemens-san, Parker-san

    I'm so sorry for late response. And thank you for your reply.

    In cunnrent our project, I'm desining only slave node and master node is desinged by our customer. According to your answers, I think we should be added extra fail-safe resistors to master node if we want increase nois-margin. But it should be considered about load violation of the driver.