I’m building a board that bridges the common types of DMX connector (various forms of XLR, RJ45, tails and product-specific terminals). There is a possibility that someone could accidentally have made a wiring mistake and introduce voltages outside of the DMX spec range which would be damaging to certain transceivers. I would like feedback on this clipping circuitry I have designed, tested and simulated.
I’ve worked with a number of DMX devices that have various cabling termination interfaces, such as XLR (which is further broken down into male and female, with 3 and 5 pin versions of each), RJ45, and bare tails. This can make interfacing between different devices a bit of a pain, often involving looking around the workshop for the right cable. To make my life a bit easier when doing testing on the workbench and on site, I want to be able to adapt between these various termination forms without having to lug around an assortment of cables. Simply bridging the positive, negative and shield pins of these terminations is simple enough, and is what I’ve done in EAGLE, however I have also allowed for 12-way CambdenBoss terminal connectors that we use in our products. The issue with this is that if someone is not paying attention to what they’re plugging in, they could be connection cables with ~24V present. This would damage certain types of transceiver that could be present on the line, such as the SN75176 (which we use in a couple of our products too).
The circuit I have tested (attached above) clips the input voltage between ~-8.7V and ~12.4V. I have tested this circuit both with the scope to confirm it operates as intended and with a couple of different DMX products. It seemed to work (initially R1 was too high, causing packets to be corrupted. I reduced it to 2.8K and everything seemed to come through without any packet loss/degradation). I’ve put ~31VDC into this circuit with a device using an SN75176 that has no input protection and it was still operational after I removed the over-voltage input.
Despite my testing, I’m not comfortable in saying that the circuit is suitable for its application (I’m not terribly familiar with the DMX standard nor what typical DMX circuitry comprises). For instance, I don’t know what kind of issues I might be introducing with diodes attached to a DMX line, nor what I could be doing to the DMX line with a 2.3K resistor attached.
In case it wasn’t clear, this is not for a commercial product for sale, it’s just a tool I will use in-house to make testing easier.
Here's a schematic i've whipped up in EAGLE showing what i'm proposing (obviously, no input protection is present).