Strange driving of RS-485 transceiver

[dobsonr741]

I found a strange way of hooking up an RS-485 transceiver in a slave peripheral I was reverse engineering.
Has anyone seen this before? The TX signal drives the enable pin, not the TX input. Why?

[oPossum]

Collisions don't produce bus contention. Same as CAN bus.

[PCB.Wiz]

I found a strange way of hooking up an RS-485 transceiver in a slave peripheral I was reverse engineering.
Has anyone seen this before? The TX signal drives the enable pin, not the TX input. Why?

Yes, this becomes CAN-like with almost open drain, but that needs care with non fail-safe BUS drivers.

Below is another slight variant, where TX drives DI and Inverted TX drives Enable, that has a couple of advantages.

The PNP in this case has a slower turn off than turn on, so the 485 driver pulses the line the reverse way, then floats.
This means you do not need the hassle of managing the Enable timing. It drives and auto-reverses.

Note also the BUS is pre biased, as the floating state has lower noise margin. The pulse time is enough to kick it over the threshold Schmitt.
It would be best suited to the skewed threshold / failsafe RS485 parts.

There are some RS485 parts that now have this 'auto-direction monostable' inbuilt. (THVD1406, THVD1426)

[SiliconWizard]

Yeah, it just makes the idle level (high) of the TX out as a "recessive" state. Yes, kinda like CAN. While probably some care must be taken, this is a very simple trick.