Too much hand waving, too little schematic. I think I know what's going on, but don't want to decipher the word soup to understand the circuit.

I attached a schematic and added additional detail. Do the updates need to be approved? I am not seeing my update.

Edit: Looks like I needed to click the submit button twice because it was complaining about "alt text" for the images. I almost lost my work... whoops.

Olin Lathrop‭ These relay coils are close to a motor. Do you think magnetic fields and di/dt in the motor windings could be inducing currents in the relay coils? Like a guitar pickup?

Olin Lathrop‭ Thank you for the excellent answer! I have had issues with ground bounce due to the motor currents causing issues with other signals. I've had to be careful where I connect grounds of signals fed to an MCU. That very well could be what's happening. I will try investigating that further. I will also try measuring the source driver current and see what the coils are pulling for current. According to the TXS2-9V datasheet the coil resistance is 1620 Ohms and operating current is 5.6 mA. Perhaps these coils are pulling more than expected.

Olin Lathrop‭ I hooked up three relay coils directly to the power supply set to 9V. The current read from the power supply was 16.6 mA (16.6 mA / 3 = 5.53 mA). The maximum current read from the DMM was 16.46 mA (16.46 mA / 3 = 5.49). I've switched the relay on and off several times using the output enable on the supply. The relay switches cleanly on the scope. There are no sudden spikes. I assume this probably isn't a fair comparison since the slew rate of the supply is probably slower than the source driver.