Saturday, August 3, 2019

Some Steps Forward, Some Steps Back

I don't know how interested folks are going to be in my cleaning up after the lightning strike, but it's how I spent the day and how I'll spend some more.  Yes, I heard about the El Paso shooting, but remember my First Law:  First Reports are Always, Always Wrong.  It's not worth paying too much attention to. 

There were good steps and bad steps today.  On the good side, I tested all three CNC machines and they're all fine.  Since the key seemed to be long wires, I was concerned about the wires between the controller boxes and the motors conducting picked up voltages into the stepper motor controllers.  I didn't use a caliper to measure the accuracy of the moves, but I did move all four axes on both mills and both axes on the lathe.   There are distinctive sounds motors make when some forms of problems are present and I didn't have any of those.  They're probably fine. 

Yesterday, we thought Mrs. Graybeard's computer monitor had blown out.  She realized the comparison with the replacement monitor was using a plain VGA input while the blown one was DVI so she found a way to test the DVI output on her computer and that was what had blown.  When she went from the VGA output of her video card to the VGA input of her old 23" monitor, the monitor was fine, so she put the big monitor back and put the 17" monitor aside.   

Last night, my monitor, which had been perfect, started developing an oddball problem.  Latent failure showing up?  The problem was while typing, suddenly a large black square with a white padlock icon in it showed up in the middle of my screen, flashing on and off.  The monitor (Dell) has a group of controls on the lower right side of the screen that bring up a menu of adjustments; color temperature of the display, which input its using, stuff like that.  That display started flashing on the screen.  It's a problem that shows up in web searches but the usual suggestions didn't work for me.  This morning, it got worse, going to solid color screens in green, blue, red and gray.  As time was going by, it kept getting worse, so I took it off the system and replaced it with the spare 17" monitor.

It was time to look more into the ham shack computer/monitor and see what was up with that.  Recall that the computer wouldn't start and the monitor looked like it was dead.  Some quick voltage measurements and inspections in the box confirmed the computer's power supply looked dead.  We have some spare computers that we never tossed out and I put one of those in place on the radio bench.  It immediately came on and the monitor sprung to life.  Changing over to a new computer is going to entail reinstalling all of my ham radio software and it was bitchy to get everything working and to play nice together.  Most ham gear and software requires a serial port - just as ham manufacturers started putting serial ports into all their gear, the computer companies stopped using RS-232 serial ports and switched to USB.  My radio has a USB port, but the accessories use serial ports. 

The old computer had a dual serial port card in it, so I moved that card and the hard drive from the old system into its replacement.  The good computer now wouldn't start.  Since I only put two things into it, taking them out one at a time showed the serial port card kept the PC from starting.  Whut?  Then I took a better look at the card.  Uh oh.  Parts were blown off it.  So I looked at the other end of the serial cable link.  This serial port connected to my antenna rotator controller.  It had parts blown off it, too.


Computer card on the left - the top IC has major damage, and the components to its right are blown, along with traces on the PWB being blown.  A chip capacitor appears to be blown physically off the board.  There are three components visibly destroyed in the rotator interface on the right.  Top right, there's a voltage regulator that's blown into two physically separate chunks (no, I didn't move one for emphasis).  The top most IC, between the four black cylinders (electrolytic capacitors) has half its package blown off and the chip closest to the bottom edge of the card has its lid blown off, too.  It looks like the overvoltage might have come out of the rotator serial port, traveled into the PC, blew out the parts, shorted high voltage onto the power supplies, which blew up the power supply.  After our experience with the monitors, I tried starting up that computer without the serial port card in it - still no go.  The power supply is really dead.
 
The really good news here is that it looks like we've gotten over the hump and have identified everything that needs repair or replacement.  Some of it can be worked around and lived with, like the DVI output of a video card not working.  Some of it might require a replacement, like my 23" monitor or our water heater controller unit.  Some of it is "just sit down and work on it" stuff, like moving apps to my replacement computer in the shack.  It looks like the work for the next few days will be in the ham shack.  Between getting that computer running, figuring out some odd results I get on the antennas (which I need software on that old computer to test with) and trying to see why my amplifier won't turn on, it will take a while.


13 comments:

  1. Ouch! The cable up to the rotor let the lightning demon into the house. Hope the rotor is OK. What type is it?

    The PC that serial card was in is likely 99.99% toast. What else was plugged into the same circuit that PC was plugged into? I'd start looking at *everything* that was on that breaker.

    ReplyDelete
    Replies
    1. Sure seems like that's the path. As a side note, one of the first things noticed right after the strike was the house's AC breaker for that room and the room beside it had popped.

      With the replacement PC running, I was able to run my AIM4170 VNA and look at all three antennas. My HF beam (a Tennadyne T6 LPDA) is toast. Since it has a toroidal balun at the feed, chances are that's iron powder or copper vapor now. My other two antennas are fine - the top antenna on the mast (my 6m antenna) included.

      The rotor is a Yaesu G-800DXA.

      The rotor cable, BTW, comes down the center of the tower to almost ground level. That's supposed to shield it from induced strikes.

      I have two switches - the closest to the antennas selects the antenna, its output goes to another switch that selects the radio. Both of those are Alpha Delta 4 ways with surge arresters in them. Both are fine.

      Delete
  2. I have that same rotor. If the rotor control box got blasted, I'd recommend rolling your own with an Arduino. Code is out there, along with complete build info, and you can build a first-class rotor controller for under $50.

    If the rotor itself is toast, well.....$$

    Sounds like your balun and/or feedline are history. Since it's an LPDA antenna (and a really nice one!), at least you don't have traps to worry about.

    Never heard about running the cables inside the tower for extra "shielding", but I suppose it might offer some extra protection, as long as you don't take a really close hit like you apparently did.

    DX Engineering sells a rotor cable surge arrestor that puts some kind of protection device on each line (up to 8 conductors), but it's $170. I built something similar to protect the 4 control and power leads for the SGC autocoupler when I get it mounted.

    ReplyDelete
    Replies
    1. Long ago, no later than '85, I was teaching a Novice class and when we did the student introductions, one guy said he did lightning protection on the Cape. When it came time to do that session, I had arranged to hand the class over to him and took a seat.

      He talked about how launch towers are man-rated. The voltage drop across any distance people could normally reach (call it seven feet) had to be below a certain limit. That's done by really good bonding at joints between metal pieces and a handful of things. One of the things he said was to run as much as you can down the center of the tower.

      Delete
  3. That's interesting, and something I don't recall ever hearing before. I understand the "why" of it, but just never thought about it. Since it's accepted practice at The Cape, I'm gonna SWAG that it's one of those "Last 2%" things.

    When I was working for Boeing I had to take several courses on "Electrical Grounding and Bonding", and a couple of similar courses that had "For Spacecraft" attached to the title. Aircraft (and spacecraft) are pretty much built the same way, with everything bonded together very well to minimize the voltage drop between sections.

    Since it's outside the normal, a lot of the people I worked with had a hard time grasping the concept of a bazillion amps running through something, and causing a significant potential difference between points only a few feet (or a few inches) apart.

    Do you have a "Single Point Ground Panel" where everything comes in to the house? What's the soil conductivity like in Florida? Here in the foothills it's not too bad, but up in the mountains it's a whole 'nother game! Solid granite doesn't conduct very well......

    ReplyDelete
    Replies
    1. Everything I did at my last job was to DO-160 (EMI and Lightning), as a sub contractor to Boeing. My tower installation predates that job, though. The aircraft and the hundreds of feet of cabling disperses lightning (in the pulse dispersal sense - they get wider and lower amplitude). I recall regularly testing to 1500V at 1500A. We took it as a point of pride for our VHF comms to play through a lightning strike to the antenna without resetting or worse. Not much you can do about the loud static crash beside audio limiting.

      Everything in the shack is bonded to an aluminum panel, which is bonded in turn to a Steelcase desk. Most of the radios are bonded through a 1/2" braid straps. The station is grounded to an 8' ground rod about four or five feet from the desk; the connection from the desk to the rod is the same braid. The tower is about six feet from that ground rod (and farther from the shack). The tower has two 1/2", 8' ground rods, and a 4" diameter aluminum pipe sunk four feet into concrete.

      Our soil is deep, on the sandy side (except for the topsoil organics) and no big rocks. When we sink the 8' ground rods they ordinarily go down with only hitting one rock that they either break through or push aside. I don't remember where my reference is, but we're slightly short of swamp conductivity. Most of the year we get enough rain to keep good conductivity.

      Delete
  4. I'm sure you to use some fairly fancy front-end protection to keep things intact! The Physics behind what happens when an aircraft gets hit is quite interesting.

    You have far better soil conditions than I do. Our soil is a mix of what ran off the foothills from erosion (granite, quartz, feldspar, the usual suspects), and organic material from centuries (10,000 yrs since the last ice age?) of untouched tree, shrub, and grass growth. It's kind of a brownish red (think "Red Rocks Amphitheater"), and a mix of very coarse sand, smaller sand, rocks, pebbles, and oh yeah, some "black stuff" from decayed vegetable matter. Very rich in minerals and nutrients, but pretty poor conductivity. The depth of the water table varies considerably around here, and the soil drains remarkably fast. Couple that with 15%~25% humidity during the Summer, and things get really dry, further degrading soil conductivity.

    Sounds like your grounding system could handle what happened if you'd had some surge protection/diversion on the rotor control cable. Since it's just fairly low DC voltage on the cable, you could build up something with some BIG "GEMOV" devices rated at.....hmmm...what would I use? Wonder how much of a hit the rotor controller can tolerate without damage. Since I have that same rotor I'll study the schematic a bit and scribble down some notes. Maybe 50 Volt GEMOVS? I'm sure you can do a comparable job to the unit DX Engineering sells for less than the $170 they charge.

    Anyway.....You're running braid outdoors? Where it rains? Uhhhhhh....that's a NO-NO. Even if you put adhesive-lined heatshrink over the braid, it deteriorates in six months. It happened all the time in SoCal to people I know who used it outdoors, "protected" or not. I understand the reasoning behind using it, but we're not in the lab. I learned "Commercial Grade" grounding helping repeater owners and at DirecTV, where they were absolutely paranoid about lightning protection.

    NOBODY uses braid outdoors.

    Sorry to be strict, no insult intended, but please don't do it. If you need flexibility, get some 'appropriate gauge' welding cable rated for outdoor use. The stuff is super flexible.

    Are you using anything like a PolyPhaser on the coax before it enters the house? I know your coax switches have the Arc Plugs in them, but it's better to try and keep surges on the coax outside the house.

    ReplyDelete
    Replies
    1. No need to mention "no insult intended". In this case, at least, it was my memory failure. The connection to the outside isn't braid, it's heavy wire. Likewise the connection from the tower legs to the ground rods. I looked at it five or six hours ago (and forgot to reply here) so I don't remember exactly how big. ISTRC it's #8. Stranded.

      No spark gaps or PolyPhasers on the outside. The antennas are DC grounded.

      The small project that turned into "most of the afternoon" was to replace the control panel for the water heater. That did nothing, so there's something else going on in the water heater. The internet searches are being unusually unhelpful.

      Delete
  5. OK, it surprised me when you mentioned braid running outside!

    I'm going to use #4 solid to bond my new ground rods together. It was surprisingly inexpensive here. Aluminum 'electric fence wire' is dirt cheap here, but I wouldn't even consider using it. *Maybe* for an antenna, but never for grounding.

    Gas or electric water heater? The electric ones like we have are dead-simple. A thermostatic switch opens and closes the 240VAC to the heating element! Identical to the thermostats for our electric baseboard heaters.

    ReplyDelete
  6. Back in the 90's(?), there was a guy who wrote books on assembling and troubleshooting computers and choosing software, etc, and he had a recommendation for prepping for lightning strikes. A Dr Tom, perhaps? Anyway, he suggested tying a knot in the cords. Supposedly, when a bolt hit the knot, instead of following around it, it would blow it apart, and not reach the other end of the cord. He lived in FL, and claimed to have never lost something protected with a knot.

    I recently tossed my copies, as they were WIN95, or earlier vintage. Prepping to move, and space for paper books will be at a premium.

    ReplyDelete
    Replies
    1. That's not unique to that author and not that far out. I was told by a broadcast guy that guys he knew would coil a couple of turns of their cables before going into the transmitter shack. The idea is it's an inductor and the lightning won't go across it. It will vaporize it. The thing is, the lightning will probably just jump another way. Does it jump from the power cord to something else in the room?

      The breakdown voltage for air is around 50,000 Volts/inch, so think of the number of inches from cloud top at 35 or 50,000 feet to ground. Once the strike gets started, it's going through. If what it hits is low resistance, it goes through. If what gets hit is high resistance, it gets vaporized. Trees have their bark explode. Things like that.

      Delete
  7. Yes, it would jump around the knot as it blew it apart. It's seeking a path to ground, and it can go into the floor, or jump quite a distance (I've seen a foot or more) to something with a lower potential.

    It's so unpredictable once it gets inside a room, that your best bet is to try to keep it outside. The knot in the card would be a "last 0.5%" type of protection.

    Besides the stupefying level of voltage Sig explains, there's an astounding amount of current behind it. The waveforms for "test levels" I've seen are something like 18,000 Amps.

    That's enough energy to laugh at our puny efforts to contain it, but we do the best we can.....

    ReplyDelete
  8. Big storm came rolling in. I unplugged everything. I was in a temporary rental and there was zero grounding (my bad, I know). Lotsa lightning. Storm passes, blue skies, birds singing. Cool. I plug everything back in. A few minutes later, KA-BOOM! Toasted the water pump pressure switch and blew the filament off of the supports in the 100 W incandescent light bulb on the computer/ham desk. No other damage despite having an HF and a VHF antenna in the air begging to be hit. Lightning is weird.

    ReplyDelete