usenet@carssdf.UUCP (UseNet Id.) (07/10/89)
Symtom: Lightning Storm nearby causes damage to one or both ends of the data line. Usually type '188 or '189 devices. I need some help to get through this summer. Could anyone please email (I will post summary) suggestions that have helped others, or explanations to some of the following mythical tails that follow. Plugging devices into orange outlets helps. :-) It helps even more if the ground (green) from these outlets is tied back to a central point and earth grounded there. (No connect to box/conduit) It would seem to me that this would only provide a superior return path for the induced charge in the data line to return to ground. Question: Are we dealing with a magnetically induced current, or a capacitive, static pickup of some kind. Using shielded data cables will ward off evil charges/inducements. Well, charges maby, but inducements, I don't think so. If so, what do you do with the shield, besides connecting to pin 1 at each end and probably just generating a ground loop current, and about half the equip. I work with doesn't even have pin 1 connected. Placing expensive ($50 each end) protection devices at each end will neutralize evil. DO THESE WORK? What are they? Ferite Beads, cap, diac, MOV? If they are necessary why not put the magic circuit on the Terminal or the CPU (Computone?) driver/receiver circuits. Socket all the '188 & '189 devices, like WYSE does on their WY-60 terminal, but not yet the WY-995 mux board. If in doubt, pop 'em out. At about a dollar a piece they are cheaper than some fuses, and ignore the whole issue. Turning off the system durring potential for lightning will help. At least it is more humane, the system will never see it coming. :-) OR- Potential for latch up and thermal distruction of chip will be reduced? (How should I know when I got my Clarkson degree in '71 LS logic was not invented yet, RTL was the word and the computer was optically isolated. We never saw it!) I realy like the new modular telephone connector system for snapping together a system, but am I just stretching out antenna waiting for disaster? Does anyone know where to get SHIELDED MODULAR CABLE? Any help would be appreciated, John Watson, Middlesex, New Jersey Please reply to ...!rutgers!carssdf!usenet
macy@fmsystm.UUCP (Macy Hallock) (07/14/89)
In article <184@carssdf.UUCP> you write: > >Symtom: Lightning Storm nearby causes damage to one or both ends of the >data line. Usually type '188 or '189 devices. > Ah! One of my favorite subjects: Lightning Damage. Strange and awesome stuff, lightning. I am a telephone engineer with about 20 years of experience with telecommunications and datacom equipment. I will share a few practical suggestions with you. If you want the textbook and/or technical explainations, go read the many books on the subject... then learn the hard way anyway! :-) >Plugging devices into orange outlets helps. :-) > It helps even more if the ground (green) from these outlets is tied back > to a central point and earth grounded there. (No connect to box/conduit) > > It would seem to me that this would only provide a superior return path > for the induced charge in the data line to return to ground. > Although it is an unsafe practice (not to mention contrary to UL and Nation Electrical Code) I have found situations where the only cure was no grounding at all! Not recommended, though. Yes, isolated (orange) outlets are a little helpful. Mostly they help with noise, not lightning induced transients. Use a good surge supressor on your power, too. My personal favorites: Transtector for best at any price. Trippe-lite for cost effective use. We put Trippe-Lite IB-20's on every telephone system we install. Note: Do not be decieved into thinking a UPS solves powerline transient problems. Get info from Oneac or Best (UPS mfr's) on this if you need more info. We use Tripp-Lite OMNI 450 or 1200's for cost effective UPS w/regulation & surge suppression. They are not the best on the market, but they work well for us. >Question: Are we dealing with a magnetically induced current, or a capacitive, > static pickup of some kind. > All of the above, plus more. Most damage is from EMF induced transient voltage spikes, but a nearby strike can do unbelievable damage. A direct strike... What do you think 200,000 volts can do? The idea is to make a more attractive path to ground for the induced transients than through the equipment. There's more, but I only have 20 minutes for this discussion... >Using shielded data cables will ward off evil charges/inducements. > Well, charges maby, but inducements, I don't think so. If so, what > do you do with the shield, besides connecting to pin 1 at each end > and probably just generating a ground loop current, and about half > the equip. I work with doesn't even have pin 1 connected. > First, do not connect pin one at each end. Since pin one is your frame ground, connecting the equipment together makes the cable a path for transients. We usually ground the main cpu/hub very carefully (an 8 AWG solid wire, run directly to the entrance water pipe, no bends or corners, as short and straight as possible and connected to a bus where all the shields connect.) All equipment located at the main cpu is directly grounded to this bus. Orange outlets are used and the feeding electric panel is confirmed to be double grounded: to the water pipe and a ground stake (at least 5' from the foundation and 10' long) per National Electrical Code. This is as close to ideal as possible without violating code. What about pin seven? Well... we have been known -not- to carry though pin seven either. I know this is not receommended, but the isolation and removal of ground loops often justifies this - if the data will still pass. (I said practical, not textbook) I still prefer braided shield (and not the 70% stuff, either) but foil shield is often a necessary evil. Twisted pairs are advantageous if used properly. >Placing expensive ($50 each end) protection devices at each end will neutralize > evil. DO THESE WORK? What are they? Ferite Beads, cap, diac, MOV? > If they are necessary why not put the magic circuit on the Terminal or > the CPU (Computone?) driver/receiver circuits. > Here's a useful circuit: put 50 ohm resistors (1/2 or 1/4 watt) in series with each RS-232 lead as it comes out of the system/terminal. Put a 24 volt MOV (or, butter yet, a Tranzorb) on the lead with the other side to ground on the opposite side of the resistor form the system/terminal you are trying to protect. The resistor isolates the equipment and helps make the MOV appear to be a more attractive path to ground. This method solves most problems for us. This can also be used for phone line protection, but use 150 volt MOV at the highest wattage you can find and use 1 watt resistors and have the phone company install a Three Element Gas Tube Protector on the incoming cable (Bell often uses two Two Element units - not the same) Be sure the phone company grounds the daylights out of the protector AND DOES NOT USE YOUR GROUND WIRE! >Socket all the '188 & '189 devices, like WYSE does on their WY-60 terminal, > but not yet the WY-995 mux board. If in doubt, pop 'em out. At about > a dollar a piece they are cheaper than some fuses, and ignore the whole > issue. > Agreed. I bought a bunch of 1488 and 1489's at the Dayton Hamvention this year for .25 ea. I'll use every one of them, too. >Turning off the system durring potential for lightning will help. At > least it is more humane, the system will never see it coming. :-) > OR- Potential for latch up and thermal distruction of chip will be > reduced? This seems to help...but its no excuse not to do the items mentioned above. > >I realy like the new modular telephone connector system for snapping together >a system, but am I just stretching out antenna waiting for disaster? Does >anyone know where to get SHIELDED MODULAR CABLE? > Yes, it is nice, until someone plugs a terminal into a real phone jack and ringing blows it out. This is why DEC went to those "weird" modular plugs that look like standard modular phone jacks but aren't. Belden makes a flat shielded cable suitable for eight pin plugs. Its expensive and hard to get. I use standard cord and be sure the outside two conductors are ground. (this will never pass FCC emission tests!) You do know about the polarity reversal on each pair in these cords, don't you? ORA Electronics and MOD-TAP and Neveda-Western and Leviton have all this stuff. I prefer ORA's prices, though. Well, hope this helps. I will try to personally reply to queries as time permits (be specific!). If I get enough requests, I will write and post a more detailed article to comp.dcom.modems and comp.dcom.telecom (If Patrick will have it!). I did try to keep this short...and it shows. Regards to all, Macy Hallock fmsystm!macy@NCoast.ORG F M Systems hal!ncoast!fmsystm!macy 150 Highland Dr. 216-723-3000 ext 251 voice, office Medina, OH 44256 216-722-3053 voice, home 19:00-23:00 EDT New path just being set up: uunet!aablue!fmsystm!macy (Thanks JB!) Personal opinions are my own. All disclaimers apply. Your milage may vary. Caveat Emptor. E Pluribus Unum.