AC coupled termination

[jdurand]

I'm updating some of our hardware and figured it's time to add more support.

A draft copy of DMX512-A that I have suggests using AC coupled termination (0.047uF in series with 120 ohm). That draft also has an RDM section and which only shows the normal 3 resistor termination/bias.

The final specification only says the termination should be 120 ohm impedance which would allow the capacitor but the RDM termination was removed from the specification.

Is it ok to use the capacitor coupled termination and use the standard bias resistors (about 560 ohm pull up and down) as with other RS-485 systems or does RDM really insist on the three resistor version with the specified values?

I also noticed a timing error in one of charts in the DMX spec., but that doesn't apply here. I did report the error, but it's a bit late since it's been published.

[prwatE120]

Stick to the termination details as outlined in the final, released versions of the DMX512-A and RDM standards. RDM requires the use of the line bias termination at the "controller" end of the wire, and the standard details this with topology and values.

Peter Willis

[jdurand]

Since I'm making user-programmable hardare, I can't assure they're actually using RDM or something else. Are there any other uses for the EF1/2 in DMX512-A that would require some other termination? I doubt it, but thought I should ask.

[prwatE120]

Not that I can think of - and certainly not documented in a manner that RDM is - i.e as an ANSI standard.

Peter

[jdurand]

Originally Posted by prwatE120

Stick to the termination details as outlined in the final, released versions of the DMX512-A and RDM standards. RDM requires the use of the line bias termination at the "controller" end of the wire, and the standard details this with topology and values.

Peter Willis

I have been carefully reading the RDM spec. and the DMX512-A spec. and they both say the termination is 120 ohm impedance, NOT resistance. Also, the RDM spec. says the differential voltage shall be at least the number specified. So, none of that precludes using the AC coupled termination which makes it easier to pass the over voltage tests, especially on a receiver. However, I see from other postings that people are trying to detect far end (DMX512-A) terminations by looking at the DC bias levels. Sorry, that won't work with AC coupled terminations.

As with many other things, it looks like if I put in AC coupled termination it will meet the spec. 100% but break other people's equipment which somehow makes it my fault (just like my 800 frame per second DMX data with low time slot counts). So, you guys win, I add large power resistors to pass the tests and keep your stuff working.