Do you have any citation for the minimum length? I was wondering if it didn't matter for rx/tx'd swapped patch cable.

ANSI/TIA/EIA-568-B.1 does not state a minimum for Cat5/6. Max is 100m.

Makes sense since most router/switch cabs I've worked in use 18" patch cables.

Short patch cables are OK because you're just extending the distance of a longer cable.

You can get crosstalk issues between the pairs on very short cables.

As far as citation goes, if you want the official line you need membership on the IEEE site, which I dont have so i'm unable to source an authorative link. However if you do a Google for 'cat5 "minimum length" IEEE' or 'cat6 "minimum length" IEEE' you'll get a batch of references to the IEEE specification noting that the minimum reccommended length is 1.2m.

When referring to a 'patch' cable, this really doesn't apply, because the term 'patch' assumes that you're linking into a patch pannel, and as previous poster highlighted, this means you're simply extending an existing device to device link. Hence why you can go out and buy 50cm patch leads from Belkin or whoever.

After doing that bit of searching for citation i came accross a few interesting posts surrounding the reasons for this limit. It's not so much overloading the Rx with voltage (which i previously assumed), but on half duplex connections, such short cables don't allow enough time for the CSMA/CD (used for collision detection) protocol to detection errors and correct them. However on full duplex connections i don't belive this would ever be an issue.

Regards,
Justin

So it's the same reason for the maximum length - or rather the opposite - there's not enough time for two distant nodes to notice the collision before they start transmitting (excuse my clumsy phrasing).

Well I ordered 5m Cat6 anyhoo

only

Postage included?

No and you have to spend

That's the logic my girl friend uses. Oh gosh, so do I.

hehe, don't we all