THVD1410: Question about the resistor value between THVD1410 A/B terminals and the connector connected to external cable.

Hi Kazuya-san,

So if there is a 20 ohm load between A and B its not a short. RS-485 standard considers a short in the following circumstances:

Direct Short From A to B

A or B directly shorted to GND

A or B directly shorted to voltage supply that is between -7V to 12V (the common mode range of RS-485) 

The reason I bring this up is because there is basically a 0% chance you will get 250mA during the "short" condition described. 

Typically the impedance from VCC to A and B to GND when the device is driving a high are ~7 ohms each. There are two diodes in the driver - so the max voltage outputted by the driver is going to be around 4.4V. If during the described short condition that puts the current at around 4.4V/((7+10)*2) - which is about 130mA - but the increased current from the driver could impact the output impedance of the device and you could see less than the appx. I made above.  

So even if we give a margin of 150mA  that brings the total power consumed to around 450mW - which is still pretty high, but less than 1/2 of what you were expecting. 

If they are worried about shorts to power supplies they may still need to consider the 250mA limit - but in most cases that isn't going to be hit

With all that said - for the resistor sizing for 1kV surge:

1. They need to be at least 10 ohms - they can be slightly larger (generally 10 to 20 ohms is what we generally say - but to minimize negative SI impact due to the series resistors 10 ohms is usually what we push). If they are less than 10 ohms they may not be able to attenuate peak transients during surge which could cause damage to IC. 

2. They need to be pulse proof or else they really aren't adding much benefit. 

3. While the resistor sizing may still require larger resistors if shorts are a real concern - they are used for surge protection so bulkier resistors is not unheard of. 

So in conclusion:

1. Most likely they don't need 1.25W - there are just very few cases where that power dissipation would be justified as most of the time it should be much less than that. There would be no risk if they put that size in their - besides it being bigger and more expensive. 

2. They cannot lower the resistance of the pulse proof resistor below 10 ohms - that will essentially render the resistor more or less useless because it won't give the 1kV protection. 

3. The 1kV is specifically meant for that diode + resistor combination - so if a different diode is used you could see slightly different results - so testing in your system is critical to understand the proper break points. 

Please let me know if you have any other questions!

Best,

Parker Dodson