Microvaxen bits etc.

From: Russ Blakeman <rhblake_at_bigfoot.com>
Date: Wed Oct 28 16:44:23 1998

For a small (up to 30 node) network I *prefer* the 10base 2 coaxial setup, since
there's no fooling with hubs. Tees, terminators, and ready made cabling of various
lengths are relatively inexpensive from places like CNA ( http://www.cnaweb.com/ )
and others and there's generally no chance of messing it up. UTP cat 5 cabling for
the 10baseT systems is very easy to work with as well but the cost of hubs
generally kicks that in the butt. While I have a crimper for just about any coaxial
(twinax, BNC, etc) as well as modular connector (RJ11 to RJ45) I prefer to get
ready made cabling where it will work. Obviously if you're running coax or cat 5
cables through conduit to wall plates and all you will have to crimp at least one
end on but generally most people with home or small business networks don't sweat
this at all and leave it exposed. As for the damage factor, every setup is differet
and the ways to protect it are generally the job of the installer, just like it is
when installing phone wiring. Cable ties, clips, etc are all available to neatly
bundle the finished setup and keep the cabling out of harm's way. If you do get a
bad end just repair it whether it's twinax, coax, cat 5, twisted pair, etc. My
biggest pain is in installing those :"IBM Data" type connectors on the end of 4
wire ethernet data cabling. It should be easy but dependant upon the gauge of the
internal conductors, you may have a bit of a time making the insulation
displacement contacts accept the wires without breaking them.

Summary: Use what works best for you and your project. If you want to be a biog
burly hairy guy and make every one of your own connections, then do it. Otherwise
get factory done ones, just as cheap in many cases.

Now if someone could make twinaxial connectors as easy to assemble as others we'd
be set on S/34-38 and AS/400 systems!

Sam Ismail wrote:

> On Wed, 28 Oct 1998, Tony Duell wrote:
>
> > The Coax is no harder to install than twisted pair. It's no more likely
> > to be stepped on and there are proper wallplates etc if you're worried
> > about it getting pulled. And I find the strain relief on the BNC to be a
> > lot better than that on an RJ45.
>
> I don't. Its far easier for a coax to get pulled out of the BNC connector
> than it is for a twisted-pair cable to be pulled out of an RJ-45.
>
> > Hmm.. Fitting a plug on a piece of twisted pair behind a machine rack is
> > not fun. 8 wires to get into the right channels, brown and orange colours
> > to tell apart, no thank you.
> >
> > The BNC was _designed_ to be trivial to fit, actually. The best way to do
> > it is to use a solder type plug and ignore the instructions :-). Then :
> >
> > Put the nut and rubber washer over the cable. Remove about 3/4" of the
> > outer covering. Fan out the braid, put the T-shapped ferrule over the
> > inner insulator and push it down inside the braid. Trim off the braid
> > around the outside of the ferrule, cut the inner insulation (but not the
> > inner wire) flush with the end of the ferrule (use a sharp knife for
> > this), pull off the inner insulation, tin the centre wire. Put on the
> > little disk insulator, cut the inner wire a couple of mm beyond that, put
> > the pin on, solder in place. Put the outer insulator and plug body on the
> > end of that, slide up the nut/washer and tighten it all up.
>
> So you're saying its easier to work with a soldering iron, solder, a pair
> of dykes, a knife and a 3-piece BNC connector behind a machine rack, as
> opposed to a one-piece RJ-45 that requires one tool to cut, strip and
> crimp? Wow. You ARE a glutton for punishment.
>
> > It took longer to type that than it does to do it.
>
> I can crimp a BNC connector in a few seconds with the right tool, and I
> don't bother with the soldering. Crimping seems to do the trick quite
> nicely.
>
> > A couple of tips. Get the right cable and plugs. There are several sizes
> > of BNC plug (not just 50 Ohm and 75 OHm - acutally different sizes of
> > ferrule, etc) to fit different cables. A lot of hassles start when you
> > use the wrong plug. Also get a couple of the special flat spanners for
> > tightening up the plugs. A lot easier than trying to use pliers.
>
> And don't use any cable other than RG-58A/U or RG-58C/U for thinnet. If
> you use anything else (i.e. straight RG-58) you're going to start running
> into trouble if you begin to add nodes or have long distances in between
> nodes.
>
> > Anyway, with 10 base T, there may be a dozen cable coming out from the
> > room where the bub is to other machines in the house. With 10 base 2,
> > there's normally only one coax cable.
>
> Like I said, as long as you are connecting computers that are within a few
> feet of each other then coax is a good choice.
>
> > > around corners, over door frames, etc. And 4-port hubs can be had for
> > > less than $50 these days. A worthwhile investment for having a more
> >
> > _4_ port hubs??? Surely you jest :-). Us workstation types have 4
> > machines per desk with ethernet ports :-).
>
> Ok, an 8- or even 16-port hub can be had for around $100 or less.
>
> > Since when have any _classic_ computers had 100Mbps network ports? Heck,
> > I'm looking out for any original 3Mbps stuff :-)
> >
> > Which reminds me. 10 base 2 is on-topic here (over 10 years old). I am
> > not so sure about 10 base T
> >
> > > Of course, this assumes your machine can use 100baseT. Old machines like
> > > your PERQs and PDP11's would be quite happy with 10baseT.
> > >
> >
> > I thought this was classiccmp....
>
> Well, ten years from now you'll regret using coax when you want to plug
> that Pentium II with a PCI 100baseT ethernet card onto your network :)
>
> Sellam Alternate e-mail: dastar_at_siconic.com
> -------------------------------------------------------------------------------
> Ever onward.
>
> Coming in 1999: Vintage Computer Festival 3.0
> See http://www.vintage.org/vcf for details!
> [Last web site update: 09/21/98]
Received on Wed Oct 28 1998 - 16:44:23 GMT

This archive was generated by hypermail 2.3.0 : Fri Oct 10 2014 - 23:31:30 BST