SWA termination OOPS

[RandomGrinch]

whereas I brought it into the context of things electrical (which is what the discussion is about)

Sticking with electrical (sort of), at the end of a SCSI bus, we use a 'Terminator'!

 

[JohnW2]

Sticking with electrical (sort of), at the end of a SCSI bus, we use a 'Terminator'!

... as did the 10BASE2 network which I originally had in this house (and the coax cable of which still exists in the dust below floorboards!). Somewhere I probably still have a box, if not drawer, full of 'terminators' used with that!

Kind Regards, John

 

[davelx]

Sticking with electrical (sort of), at the end of a SCSI bus, we use a 'Terminator'!

and also at the end of a DMX chain, though they are often built in the equipment and switch selected.

 

[SUNRAY]

and also at the end of a DMX chain,

And frequently not used.

 

[davelx]

Indeed, and most of the time you get away with it, it depends very much on the various lengths of cable that are in use, as you can sometimes get a reflection from the unterminated end which is when things start to misbehave.

 

[JohnW2]

Indeed, and most of the time you get away with it, it depends very much on the various lengths of cable that are in use, as you can sometimes get a reflection from the unterminated end which is when things start to misbehave.

If I recall correctly (from decades ago - so no promises!), one could rarely 'get away with it with 10BASE2 - at least, not in the sense of everything always working reliably. If the end of the coax run was not 'properly terminated', all sorts of things could go wrong, albeit sometimes only intermittently.

Kind Regards, John

 

[davelx]

If I recall correctly (from decades ago - so no promises!), one could rarely 'get away with it with 10BASE2 - at least, not in the sense of everything always working reliably. If the end of the coax run was not 'properly terminated', all sorts of things could go wrong, albeit sometimes only intermittently.

Kind Regards, John

You are correct! The thing was than some devices, esp. repeaters, had built in terminators that could be switched in and out and in some cases people would have both the internal terminator selected and put an external one on the t-piece. I cut my networking teeth on 10BASE5 (aka ThickWire - bee-sting taps) when Xerox,Intel+DEC jointly released Ethernet as a commercial LAN technology (which also needed terminators each end) and then 10BASE2 (ThinWire), way before the IEEE 802.3 committee got in on the act.

 

[JohnW2]

You are correct! The thing was than some devices, esp. repeaters, had built in terminators that could be switched in and out and in some cases people would have both the internal terminator selected and put an external one on the t-piece.

Indeed - and, again, if I recall correctly, I think that doing that could also cause problems, since the termination was then lower impedance than it should be, hence the risk of 'reflections' etc.

Kind Regards, John

 

[SUNRAY]

Indeed, and most of the time you get away with it, it depends very much on the various lengths of cable that are in use, as you can sometimes get a reflection from the unterminated end which is when things start to misbehave.

I keep a number of DMX terminators in various bits of kit but the reality is I very rarely use one in my small systems, primarily as it's rare for me to design a system but rather they build up until there's enough.

I encountered my first 'flickering lamp' recently but adding a terminator did little, it turned out to be a broken screen on the very last cable bt it only affected an earlier fitting. Replicating the fault at home was easy and by careful selection of cable lengths I was able to find the sweet spot where a LED lamp didn't respond to DMX. It was fascinating to see the effect of fine tuning the position of the 'sweet spot'.