SN65HVD21M-EP: D input short to A & B outputs when driver mode is enabled

Hi Robert,

This is an interesting question, thanks for bringing it to E2E!

The standard 8-pin RS-485 transceiver pinout is designed to limit the possible shorts between any bus or supply pin and the digital-level pins by placing these pins on opposate sides of the package. This allows for good physical separation between the analog and digital domains and eliminates the potential for adjacent pin shorting to cause such a short circuit condition. Therefore, a short of this nature would most likely occur only if the signal traces of the bus and digital pins run nearby one another or share some off-board connector.

If such a short circuit were to occur, the failure mode would result in loss of functionality - as the bus signal pin would likely oscillate due to it being tied to the driver input. Because the digital pins do not share the same protection features or voltage ratings as the bus pins, it is also likely that the device would be damaged in a real system where the digital pin is exposed to potentially hazardous external conditions.

Let me know if you have any more questions. 

Regards,
Eric Schott 

Hi Robert,

So in the case where the device is in drive mode and D is connected to A:

1. from a damage perspective - there isn't a ton of damage to be expected when the device is driving as the differential output pins aren't outputting less than GND or greater than VCC. If in receive mode this changes because Ground potential differences can cause common mode voltage shifts that wouldn't damage "A" or "B" but would any console side pin like "D". However bus protection circuitry usually isn't enough to protect console sided pins so transient signals that could exist on the bus may not 

2. From a functional perspective - there is a delay between input and output and an input could repeat indefinitely (D goes high, A goes high after delay, D goes high again.....) so functionally the device isn't very functional and this setup will cause potential damage  in RX mode or due to transients on the bus that exceed the ratings on the IC console side pins (D in this scenario)

So in the second case:

1. From a damage perspective it is similar risk to "A" but there are a few more issues - basically when looking at shorting "HI" to "LOW" pins - explained in point 2 below. 

2. From a functional perspective B outputs the opposite of what is input to "D" so B goes low when "D" goes high - a direct short can cause a short from the controller to a value pretty close to ground. The actual device shouldn't be damaged necessarily but the controller may. When "D" goes low "B" goes high and "B" is shorted to the low input on "D" which is ~GND - which essentially is counted as a short circuit. The in-rush of current may cause damage on the pin unless there are other components to sink the current.. 

So ultimately its still not a good idea to connect these devices where D is shorted to bus pins.

D to A is going to be less likely to cause damage as D and A should be similar in value and large currents are't as likely. But using the D pin to communicate effecitvely is pretty much impossible because of the feedback from A to D. 

D to B is going to be more likely to cause damage because of the shorting of High to Low (D and B are opposite polarity)  which could cause too much current inrush into the "D" pin or possible cause too much current to be sunk/sourced from the controller controlling the device. 

Both situations in receive mode are prone to damage due to ground potential differences adding  common mode voltage that the "D" pin isn't rated for.

Please let me know if you have any other questions or needed clarification and I will see what I can do !

Best,

Parker Dodson