[ Date Index ] [ Thread Index ] [ <= Previous by date / thread ] [ Next by date / thread => ]
On Friday 12 August 2005 00:30, Keith Abraham wrote: > >Minor point, but I doubt having the PC plugged in makes much difference. > > > >"ground" on any modern building is almost certainly connected to a > >circuit breaker that trips should any current flow, it is probably only > >connected to earth via being wired to neutral behind the circuit breaker. > > > > > Read the electricity section of the building regulations. Every electrical > supply has to have a valid earth (ground). Just think of a bathroom shower > without a proper earth. Scary!! > Absolutly, A proper earth must be present, The current breakers to which (Simon) refers are ELCB are based on a differential trip, if the Current (live) != Current (neutral) with in 20mA or so it trips, All earths remain connected. It is typical to bond neutral to the earth at some point in the supply system, but never assume neutral is the same potential as earth, it typicaly floats around up to 10v above ground but its exact level is due to demand and in particular 3-phase inbalance of the grid can have a significant effect. > >In that sense they are very poor "earths", hence the networking > >"purists" (read rich people) recommending optical fibre (or wireless!) > >between buildings, rather than wire networking cables. > > > > > Probably optical fibre because of the lower impedance and greater > bandwidth. Ok we have electrical saftey earth and signal earth confused together. The problem between buildings is that there may indeed be a local difference in earth levels at each end. If we assume TTL signals for an example TTLs operate with 0 - 5v logic. BUT the switching levels are defined as Logic 0 less than 0.4V, logic 1 greater than 3.5V, so a tiny 0.4v earth difference and all the logic 0's are suddenly undefined. The only way to resolve local earth differences is either using a big fat earth cable to bond the two sites together, to have systems that are differential and can reject the common mode offsets or to use non electrical contact methods as metioned. > Equalising the potential is the important bit. Static usually has a high > potential > but a low current. Even this very low current can damage a chip but the > current > will only flow where the potentials are unqual. > A Texas engineer once told me that antistatic mats and wrist bands were > mainly > used to keep the girls who assembled the componets from moving around > and generating a fatal (to the chips) static potential. Anti static bands do work, any proper production facility will also have antistatic flooring tiles as well as workmats and wrist bands. Basicly they are very high impedance connections through to an electrical ground so any static (high potential but low energy) will dissipate very quickly. Its not current damaging the chip but voltage break down, in the same way as when you get a static shock what is happening is that electrons are punching through a couple of skin cells causing damage and pain, in a FET (field effect) transistor high voltages of static will punch through the gate distroying the device. > > Always unplug the computer from the mains. Its the only way you can be sure > you won't get a lethal shock. Well that is always good advise if you are in anyway unsure of electricty. I have seen instances where the earth was not connected in a plug, in this case the PC chassis will go to 110v (due to X2 capacitors). > I merely roll up my sleeves and rest one arm on a metal part of the > chassis. (All > motherboards have quite a large earthing pad(s) which will, at some > point, be > grounded to the chassis.) I've used this method for years and never lost > a chip. > > >We should of course test this with a Van de Graf generator, and an old > >PC, at the next LUG meeting, because I'm not 100% certain. My experience > >is that it is quite hard to damage most PC components through static > >(accidently!) but that may just be that the weather is too damp in Devon > >to get a really good static build up. I think that PC components have a reasonable degree of antistatic protection built in with various clamping and zennor diods built in to all the input connections. > > > > > You can build up a potential approaching 30kve (at very low amperage) by > rubbing your hand briskly against your trousers. No jokes please ;-) Apparently it has been seen at a petrol station that a static discharge caused in this kind of fashon ignighted the petrol vapor. This may just be another urban myth though. But if you want to see staic problems come and vist my factory on a dry day in the summer. -- Robin Cornelius --------------------------------------------------- robin@xxxxxxxxxxxxxxxxxxxxx http://www.cornelius.demon.co.uk http://sourceforge.net/projects/rt2400 GPG Key ID: 0x729A79A23B7EE764 http://www.biglumber.com/x/web?qs=0x729A79A23B7EE764
Attachment:
pgpuOFoM5qZAl.pgp
Description: PGP signature