ethernet hub

why does an ethernet hub have a power supply?

Reply to
bob
Loading thread data ...

skrev i en meddelelse news: snipped-for-privacy@i40g2000cwc.googlegroups.com...

To supply power to the electronic circuits in the hub.

formatting link
Some hubs have some kind of traffic monitoring e.g. lanview.

/Rune

Reply to
Rune Christensen

For the same reason as every other piece of electronics. A hub regenerates signals and does various bits of sanity checking. It's not just the signal cables soldered together!

Reply to
John Rowe

just thinking, theoerietically. There are passive hubs and active hubs. Passive hubs don't boost the signal. i'd think that perhaps they don't always need a power supply. 'cos i've seen usb hubs that don't use a power supply.

Reply to
q_q_anonymous

skrev i en meddelelse news: snipped-for-privacy@g10g2000cwb.googlegroups.com...

They get power from the usb connector

/Rune

Reply to
Rune Christensen

FWIW

Token Ring interfaces (the PC end) supplied power on a seperate pin. (Unless your kit was not standards complient).

It is often quite practical to suck power from the signals on an RS232 interface.

I am pretty sure that Centronics also has a power pin, or maybe it is (was:) just a 5V reference.

Normal ethernet (not the new Power over Ethernet) interfaces don't have much spare. I am sure though that you could run a digital watch off of it.

Reply to
anybody43

Not with Ethernet there aren't. With Arcnet, yes.

USB provides a certain amount of power. Ethernet doesn't unless it's set up to support "Power Over Ethernet", which is not a required feature.

Reply to
J. Clarke

Passive hubs are really ugly in a technical sense, lotsa transistors steering currents around, and they only work on 10baseT. Poke around the RadioShack WWWebsite for a schematic.

Reply to
William P.N. Smith

YES! thanks, I even see the schematic, and the transistors. Looks like no power supply. Infact, it's just called a Y adaptor!!

formatting link
schematic mentioned on that epanorama link
formatting link
somebody wrote on that epanorama link

" Someone used to make what looked like a passive 3-port RJ-45 device, but I haven't seen them in a while. Looks like Radio Shack used to sell a 10BaseT passive hub (278-785), but all I can find on Google are people complaining that it doesn't work "

I have seen (powered) 10 base T (repeater) hubs. Anybody reading this that doesn't know. They should buy 100BaseT or

10/100 rather than 10BaseT. They look the same (at least on the outside) 'cos sockets are the same - RJ45 . But 100BaseT - being 100MBps is of course faster.
Reply to
q_q_anonymous

actually, i'm not sure. The title says something about AC and 250V

Does it use a power adaptor?

thanks

Reply to
q_q_anonymous

I think not. This gizmo seems to OR the Tx+ and Tx- of any of the three attached devices and send it to the Rx+ and Rx- of the other devices, recreating what you'd see on a coax Ethernet. I would expect the limiting factor to be that voltage is reduced by twice the voltage between base and emitter is when the transistor is turned on, for each additional port. Pretty soon, receivers will be unable to detect when one of the devices is on.

Seems pretty clever, actually, within its limits.

Bert

Reply to
Albert Manfredi

I think it is wrong, though, unless some are MDI and some MDI-X.

I would have expected pins 1 and 3 to alternate around the loop, as also for pins 2 and 6, where each ports pins 1 and 3, or 2 and 6 are three diodes apart. Also, it likely only works for fairly short cables for the reason you mention, and also because the termination impedance will be wrong. Though 10baseT is fairly insensitive to termination impedance.

It seems to be an application node for the RJ-45 jacks.

Note, though, that you can probably buy a repeater or maybe even a switch for the price of three jacks and a box.

-- glen

Reply to
glen herrmannsfeldt

A hub is a multi-port repeater - a signal received on one port is repeated out all the other ports, with the voltage and clocking regenerated.

Since the signal voltages are returned to nominal values, power is needed.

No power is needed for a crossover cable, linking only two stations.

Older versions of Ethernet (10Base2/Thinnet, 10Base5/Thicknet) did not generally use hubs, simply attaching stations to a coax segment.

Reply to
Wrolf

That said, I see no reason a hub (or small switch for that matter - who the heck still actually uses hubs? I mean other than for monitoring, for which they are still invaluable ) desgined for edge use couldn't pull it's power from a POE switch that it is uplinked to...

Reply to
snertking

Huh?

Not on ethernet. At least that I have seen or heard of. Can't see how that's possible.

Arcnet maybe...But ethernet?

Reply to
snertking

there are 2 definitions of ethernet.

definition 1 There's ethernet as in Ethernet I and Ethernet II. That's the most correct definition, that you use. No hubs. Just thin or thick coax.

definition 2 wider definition of ethernet. Some refer to IEEE 802.3 as Ethernet, because it is the successor.

10 Base 5 I think. It just took over. The 802.3 spec itself doesn't call itself ethernet. Marketting probably did. (note, marketting can influene specs too. IEEE marketting put 802.3x(that's full duplex) in 802.3 even though 802.3x has no CSMA/CD. one of R seifert's posts explained that ) The only reason anybody might think that calling 802.3 ethernet might gain some technical legitimacy is in my opinion probably because Charles Spurgeon (author of the o'reilly book - ethernet the definitive guide) calls it ethernet. Coming to think of it. I bet he did it to sell his book. Plus, charles spurgeon isn't AFAIK a god or warlord , and he isn't inspired by "the spirit of Gd", and he doesn't even claim to be. So, I don't see much of an argument for technical legitimacy to definition 2 even in the technical sense. It's just a marketting thing. Sadly enough, hub has also become somewhat of a marketting term though, like ethernet, it has an original - technical I guess - definition (hub and spoke topology)!

-- qq

Reply to
q_q_anonymous

q_q snipped-for-privacy@yahoo.co.uk wrote: (big snip)

I believe that hub describes the topology, and that both switches and repeaters are a subset of hub. Since the early hubs were only repeaters, hub became synonymous with repeater (and it's shorter to write), but isn't strictly correct.

-- glen

Reply to
glen herrmannsfeldt
[snip]

Actually, that is incorrect. The term "Ethernet" appears 24 times in the substantive text of IEEE 802.3-2002. Specifically, 802.3 operation at

1000 Mb/s is called "Gigabit Ethernet."

Not marketing--the *market*, i.e., the people who make and buy Ethernet products. In addition, Bob Metcalf (the inventor of Ethernet, both the original technology and the term) declared to the world that, "I say what Ethernet is, and I say IEEE 802.3 is Ethernet." This statement got a huge round of applause at the meeting where we celebrated the 10th anniversary of IEEE 802.3.

Other than the group that sells IEEE products such as their books and standards, there is no "IEEE marketing" organization that decides, or even tries to influence, what material is included in IEEE standards.

And Bob Metcalf. And me.

Reply to
Rich Seifert

Because the PoE infrastructure would double the cost of the small switch. Of course, what you are talking about _is_ available:

formatting link
(or goto 3com.com and find IntelliJack in their product list) is a PoE switch that lives in a standard wall box, runs off PoE, and does "PoE forwarding" (on port 1). They start at $120 and go (way) up from there.

Reply to
William P.N. Smith

[this post may appear twice. if so, it's an exact duplicate]]

ok, those are some heavyweights!

That gives it legitimacy And if you say that gives it technical legitimacy, then I suppose it does. It's your call. You worked on the specs of Ethernet II (and I ?)and on 802.3 . And of course Bob Metcalfe is Bob Metcalfe!

But , i'd like to question that. I have some familiarity with your posts and your switch book

Switches are, according to your book, - the latest bridges. But you don't refuse to use marketting terms. So in that instance . You, given your credentials, that establish you as an elite electronics and computer techie to say the least, give your blessing to a marketting term, but still do not consider it a technical term. You use the term later in the book, to get concepts straight, but you're clear, it's only a marketting term.

If you do not give technical legitimacy to the term Switch. Or rather, you are clear that it's purely a markettign term. Doesn't the same apply to the term Ethernet?

Do you really want the term Ethernet, which has both technical meaning, and a marketting meaning, to officially go down the marketting meaning.

The term Ethernet becomes as loose as the term Switch. Ethernet is almost the new thing on Layers 1,2 . And Switch is the new thing on Layer 2. Whatever they want to sell. At least the term "switch" was messed up to begin with.

Aren't you yielding technical definitions to marketting unnecessarily? You agree that 802.3X was only part of 802.3 'cos of marketting. I guess the only similarity there is that it looks the same on the outside as the equivalent HDX 802.3 technology. One has no choice but to call 802.3X , 802.3X. But one needn't take the category seriously on the technical level. If you'd call 802.3X Ethernet too, That's yielding totally to the ridiculous marketting definition. Once you take one step on the marketting definition, you're on the slippery slope and have to go all the way. There are no multiple marketting definitions. Just 1 loose one.

I can understand how the people behind Ethernet. Yourself, and Bob Metcalfe especially, would want 802.3 , and 802.3X, and any latest technology to be called Ethernet. But your place in computer, and internet history, is assured. And your contributions to usenet, clarifying concepts, and clarifying much of the confusion brought on by marketting, as 'just marketting', has been of great assistance, and is clearly greatly appreciated.

I can understand you endorsing the term Ethernet as a marketting term. I find it painful to comprehend you giving that marketting term technical legitimacy.

I recall you wrote that you are involved in the marketting too. And I guess Bob Metcalfe as a visionary, is also keen on marketting. So, even as technical people. If you endorse the marketting term with your marketting hats on, or rather, without taking your marketting hats off. Then, surely it only endorses the marketting term.

i'm wondering if in reading my long winded laborious post , you're ROTFL or falling asleep at your desk, or both ;-)

But these are important issues. In the future, terminology may become such a mess, there will be a revolution, and people will be going through your words with a toothpick, to try to gather the original meanings, the techie meanings. But if techie inventors give technical legitimacy to marketting terms, then what hope is there to save us from the future catastrophe? You may be a legend now, but you could be a hero!

Reply to
q_q_anonymous

Cabling-Design.com Forums website is not affiliated with any of the manufacturers or service providers discussed here. All logos and trade names are the property of their respective owners.