Damaged 1 Prop pin (Tx), could DIRA~~ avoid it?

Hi all,
I have successfully damaged a prop pin which was connected to a servo motor during a power failure at the servo motor side. I thought that only the Rx pin on the prop needs protecting via a 10k resistor but it was my Tx pin that was damaged.
May I know if placing DIRA[TxPin#]~~ could avoid this problem or its better that I place a diode?
I made another board that has 1 LED per available prop pins & have a simple problem run the LED & check the prop pins. That where I realised that Tx pin was damaged.
I have successfully damaged a prop pin which was connected to a servo motor during a power failure at the servo motor side. I thought that only the Rx pin on the prop needs protecting via a 10k resistor but it was my Tx pin that was damaged.
May I know if placing DIRA[TxPin#]~~ could avoid this problem or its better that I place a diode?

I made another board that has 1 LED per available prop pins & have a simple problem run the LED & check the prop pins. That where I realised that Tx pin was damaged.

Comments
I'm working on a circuit now that uses 100 ohms in series and a 0.01uF cap to ground to filter motor noise from the servo.
The closest thing to a guarantee would be to use an opto isolator with a 7.5KV breakdown voltage.
Murphy's law always applies.
When you mentioned 100ohmns in series, do you mean Prop Tx-assigned pin -> 100Ω -> servo motor's Rx and same for Prop's Rx -> 100Ω -> servo motor's Tx and a 0.01uF at the servo's Vcc/Gnd?
Thank you Kwinn. I think you've help me see the problem. The main adapter's current to the servos could be too high. I'd better check again. Also, could you please recommend a brand/part# opto isolator?
The major complication with series resistors is that the resistance along with possible capacitance in the load (servo motor) may form an RC network that will slow down the signal transitions and limit the speed at which the I/O can function. In the case of servos, this is pretty slow (10s of us) and you'd need a lot of capacitance between the servo control line and ground (say 5uF) for a 4.7K resistor to affect anything significantly.
Thank you.