SoylentNews is people
SoylentNews
from the of-course-it-is-impossible-to-get-online-using-a-desktop dept.
US Laptop Shortage Could Derail Remote Learning:
As students and teachers prepare for a return to in-person learning for at least some of the time this fall, many of the nation's schools are facing shortages and delays for laptops and tablets needed for online learning, an Associated Press investigation revealed.
Lenovo, HP and Dell, the nation's largest computer companies, have informed school districts that they are short nearly five million laptops.
[...] Last month, at the request of President Donald Trump, the U.S. Department of Commerce imposed sanctions on 11 Chinese companies, including Lenovo, AP reported. School administrators have asked the Trump administration to devise a solution because remote learning without laptops is impossible.
Lenovo has informed school districts of the supply chain delays and the trade controls set by the Commerce Department, which would cause another slowdown. Lenovo declined to respond to an inquiry from AP.
Have any Soylentils run into this?
(Score: 2) by RS3 on Sunday August 30 2020, @06:44PM (12 children)
Ugh, lightning that bad for you? Do you have really good ground rods? You may need several, spaced 6 feet apart. You may need special ones depending on your soil type.
And maybe a whole-house surge suppressor. Is it just coming in on the 240V lines? Do you have copper voice / DSL / caTV that need good lightning protection?
Lastly, maybe just the power supplies got hit and the rest of the computer is fine?
(Score: 0) by Anonymous Coward on Sunday August 30 2020, @08:02PM (1 child)
Afaik he lives in south Florida, with its tropical weather and low elevation.
(Score: 2) by JoeMerchant on Sunday August 30 2020, @09:13PM
Lived. We're a mobile society, don'cha know? (I only work with people from Minneapolis, not crazy enough to live there myself.) North Florida since 2006 for us.
🌻🌻 [google.com]
(Score: 2) by HiThere on Sunday August 30 2020, @08:36PM (3 children)
Ground rods won't stop sudden surges. They work fine for less sensitive gear. Possibly the problem is that surge suppressors wear out, or that you shouldn't have more than one of them between you and the main, or you get "echo surges". (Well, that's what I was told. I'm no hardware expert.)
Javascript is what you use to allow unknown third parties to run software you have no idea about on your computer.
(Score: 2) by RS3 on Monday August 31 2020, @05:01AM (2 children)
Uh, I don't agree. They can certainly help a lot. I admit it all depends on where and how the surge is induced. But ground rods are meant to dissipate lightning and other big surges. I had buried copper phone lines and had several modems and phones badly damaged by surges until I finally installed 2 additional ground rods.
The electric company's pole-mounted transformer that feeds my house has a big spark gap that every now and then would make a huge loud BOOM-BZZZZTTT sound- likely someone hitting a utility pole somewhere causing line shorts or downing wires. When that would happen our phones would jingle and uh-oh, things would be fried.
Since adding the 2 ground rods maybe 10 years ago, that horrible sound has happened but nothing damaged, except 1 surge suppressor.
More surge suppressors should be okay, but you need really good ground wiring and ground rods to give the surge somewhere to go.
(Score: 2) by HiThere on Monday August 31 2020, @01:46PM (1 child)
Well, maybe I should have said spikes. And by sudden I'm talking about the kind of thing that wouldn't affect an incandescent light bulb, but will burn out typical printed capacitor. For slower changes they're a good answer, even if not full protection, but a transient spike will get through before the lightning arrester gets into action.
That said, I'm depending on experts for this opinion. My own knowledge doesn't really cover this. My one experience with a lightning strike on the power lines left fragments of silicon spread out throughout the chassis of the computer, and the need for a total replacement. I suspect a ground rod would have prevented that.
Javascript is what you use to allow unknown third parties to run software you have no idea about on your computer.
(Score: 4, Informative) by RS3 on Tuesday September 01 2020, @12:28AM
I'm somewhat expert, but I don't deal exclusively with transients, etc. So it's not a simple thing to understand. For one, it depends on where the spike occurs- like in Joe's case, it probably came in mostly on the coax, and it wants go to a ground, so it went through the modem/gateway, which of course exploded, and along the Ethernets, taking out computers.
But another and major factor is how much energy is in the spike. Also how tall (peak voltage) the spike is, but generally the more total energy in a spike, the higher the peak voltage.
The actual working components of a surge suppressor are 1) spark gap, and 2) MOV (Metal Oxide Varistor). https://en.wikipedia.org/wiki/Varistor#:~:text=surge%20protector%20circuitry-,Hazards,this%20is%20not%20a%20problem. [wikipedia.org]
Spark gaps aren't super useful by themselves for electronics- very high voltage occurs before they "spark over", and they're a bit slower than MOVs, so they're only used for certain things beyond the scope of this topic.
MOVs are great, but as in one of the wiki pics, they can be destroyed during the surge, IE: lots of surge gets through after they clamp then explode. I've opened power strips and standalone surge suppressors inside of which the MOVs were blown apart. IE: nobody knew they had absorbed a surge and blew up in the process. Maybe a little window would allow inspection.
And most of the common (cheap) surge suppressors use very small MOVs. But you can buy big MOVs to make a serious surge suppressor that will survive more surges, and overall clamp much more energy before they explode.
For my phone's landline (being a tech / EE) I made a fairly simple surge suppressor consisting of some spark gaps (old special landline ones, tied to heavy wire tied to 3 ground rods), series resistors (to limit surge current), and more spark gaps. I've had to replace the series resistors a few times, showing that they did their jobs. I also cleaned the spark gaps- they're very precision and show "arc bites", which can have raised edges which can short out the lines, and a simple cleaning restores them. But the DSL modem survived many big surges.
I haven't done one for AC mains yet, but fast-acting fuses before the MOV would also limit the total surge energy and its voltage spike.
And an isolation transformer (1:1) would really protect things for AC mains spikes, but not for the lightning strike on the cable TV coax.
Overall I agree with you, but my point is: every surge is different- in its duration time and its peak voltage. If you get a direct lightning strike on your house or power lines, all bets are off. There are no guarantees with this kind of thing.
If lightening hits high-tension power lines, the transformer that brings that voltage down to 240 will dampen / absorb / spread out much of the spike, giving surge suppressors a chance.
If the surge is due to high tension lines shorting (due to car hitting utility pole, for example) that might not be as high a peak voltage, but duration would be longer, so more total energy and MOVs explode and much of surge gets through.
Power supplies can be built better. In (much) older PC power supplies I remember seeing MOVs, but most don't have them any more. Also, simply using higher voltage diodes and main power transistor would give better surge tolerance.
The bottom line- it's not an exact science because surges don't follow the rules. :)
(Score: 5, Informative) by JoeMerchant on Sunday August 30 2020, @08:49PM (5 children)
This part of the world (North Central Florida) is one of the worst. I had a similar strike about 10 years ago, different house but similar layout: acre of old oak trees around the house. Both times it was the tallest oak struck, massive damage - 10'+ splinters sprayed across the yard, this time that oak happened to have roots only a few meters away from our buried coax cable. Lightning came in the house on the coax, no apparent damage until it hit the ethernet connector in the cable modem - fragged that into dozens of shards, visibly scorched several of the shielded Cat6 connectors.
Most everything that fried - other than a couple of GFCIs - was connected to wired ethernet - didn't seem to bother much via the AC mains. Both AC mains and Coax have grounding rods down to the water table through sandy soil, and the soil was pretty wet at the time - although the AC mains grounding rod enters the ground in a sheltered location so it doesn't hit moist soil until a couple of feet below the surface. Steel roof, if that matters.
Lucky on 2 desktops, they were NUCs and the SSDs & RAM survived, just had to "upgrade" the bare bones box and everything was 99.9% back to normal (one weird Linux driver issue with the new Core i10 system, relatively easily cured.) That 2014 laptop was a bizarro case, somehow it was working fine before the strike, plugged in to a GFCI circuit that tripped and coincidentally that GFIC burned out with the 2nd near miss strike 2 weeks later. Laptop was on WiFi not wired network, but after the strike it wouldn't boot. All self-diagnostics checked out, so I just burned on a fresh copy of Ubuntu and it's back in business - sort of the opposite of the old NUCs where the SSDs survived intact but the power supply (and who knows what else) fried.
New surge protectors with ethernet filters in place now, much more WiFi reliance than before, using a WiFi6 mesh system now - I'm not 100% impressed with the mesh router, but it's pretty good most of the time. Another near-miss strike hit the neighbors' yard two weeks later, we had zero damage, but they lost their cable box again - 2nd one fried in 2 weeks for them.
For Raspberry Pi fans, I had 5 online at the time of the strike, zero damage to any of them (however, none were on wired ethernet.)
Yeah, that seemed to be the case for the NUCs, power switches were inoperative, but the RAM and better still SSD contents were intact. Might have tried to repair, but would quickly have spent more time on them than replacements cost.
🌻🌻 [google.com]
(Score: 2) by RS3 on Monday August 31 2020, @05:22AM (4 children)
Okay, here's my solution for you. Are you ready? :) Besides lots of ground rods and well-connected surge suppressors, cut the coax cable far from the house. Build a small doghouse to house the cable modem (gateway) in the box. At the Ethernet (LAN) jack attach a fiber media converter, and run fiber to the house.
To power the cable modem and media converter- either solar panels and batteries, or a long flexible fiberglass shaft in a plastic (PVC) sheath and a motor-generator pair. It's only maybe 10-20 watts load so it's not too much. Or instead of the flexible shaft you could use pulleys and a long rope.
Okay, I admit I'm cracking myself up but it could work.
(Score: 2) by JoeMerchant on Monday August 31 2020, @12:05PM (3 children)
What you're suggesting might work, and then the lightning might decide to enter via some other route. We have friends who know someone who had lightning boil the water in their pipes, bursting them all over the house. I'm wondering if our metal roof might become a giant capacitor plate and couple energy into all kinds of other metal objects in the house even if they are air-gaped. It's kind of a classic adversarial problem: the common entry point used to be mains powerlines, but mitigations on the mains have finally reached a point where it's something else - not surprising that something else is the cable company TBH.
I've got 12 gauge landscape lighting around the yard, so that's a mesh of conductors ready to channel energy back into the garage - it seems that the AC-DC converter survived the hit, but the GFCI outlet that supplied it was destroyed - not once, but twice. I replaced it and that second strike took out the garage GFCI again, and instead of just tripping the in-house GFCI, it was also permanently damaged by the 2nd strike. Could be the landscape lighting, or - more likely I think - it could have been the robot lawnmower's perimeter wire. Luckily, I suppose, I did not get the mower repair parts until after the 2nd strike hit, so I didn't end up having to repair the mower twice. Just the thought of a ring of copper around the entire yard gives me shivers when I think about lightning striking _inside_ that loop. Not sure how the timeline went, but my final mower repair (after replacing the burnt-out loopwire sensor board) was in the loop wire itself - it just physically snapped, no burn marks or anything, just broke like someone cut it underground.
🌻🌻 [google.com]
(Score: 2) by RS3 on Tuesday September 01 2020, @12:49AM (2 children)
I ASSume your metal roof is well grounded??? Actually it should give you a lot of protection by being a capacitor plate- it would greatly dampen out the peak voltage. But if it's properly grounded at many points to good ground rods, you should be quite well protected.
And you have some really good lightening rods?????
The theory is that the lightening rod shorts out atmospheric "static" charges that can develop on and around houses and trees and "attract" lightening. If you've ever played with high voltage stuff (and who hasn't) you'll know about "corona discharge" and how sharp points seem to be a better conductor (inducer?) of charge into and out of atmosphere. So get some pointy rods up there if you haven't already. And they really need to be grounded- a lot! You may need 10 ground rods. Can one really have too many? The copper-clad ones are better than galvanized. One would get rid of any accumulating charge, but would not absorb much of a direct strike. I know you said you have at least 1 ground rod.
You probably know that depending on soil and grounding requirements, in some (many) installations standard ground rods are ineffective and we go to much larger diameter (and maybe length) rods. In the really serious situations a borehole is drilled, very long rod dropped in, and then conductive "stuff" poured in. It's all about conductive surface area in contact with earth (unconsolidated overburden. :) My dad used to talk about "watering" ground rods. Don't know for sure but I'd imagine in some locations an irrigation system might be called for.
As you probably know, if lightning strikes one of your trees, there will be a large voltage gradient in the ground for a large distance around the tree. So your landscape wires and mower perimeter (robot mower- very cool!!) wire are just big energy gatherers. And these energy spikes are unpredictable- they propagate like RF energy and it's very difficult to predict their paths. No surprise at all that your GFCIs are getting hammered.
I have to do some thinking about some solutions, like wireless direction system for the mower (maybe GPS or 2 WiFi antennas and some interesting triangulation algorithm). And if your landscape lights are LED, you could probably run them from a battery that could have solar charging, and easily swapped for a battery being charged in the house. Anything to get away from having wires coming into your house, other than AC mains. And as I mentioned above, you could get a big isolation transformer for more protection. Or not a huge one, but just isolate some circuits like computer, TV, etc. And real surge suppressors- big ones. I'll design you one- point is one with several big MOVs, fast-acting fuses, maybe some low-value series resistors to limit spikes. Think of a ladder using resistors for the sides and MOVs as the rungs. That way if the first MOV blows up, you'll have 2nd, 3rd, etc. Maybe such things are on the market- haven't searched.
If you know anything about RF and antennas, you'll know those conductive rods can have 0 volts at one end, and theoretical huge voltage at the other end, if they're in proper resonance / tuning. Your perimeter wire just happened to have a resonance peak where the wire popped- huge current at that point. Would've been cool to see some kind of waveform of the whole event, along the wire. Not willing to sacrifice my precious oscilloscopes to try it though!
(Score: 3, Interesting) by JoeMerchant on Tuesday September 01 2020, @01:28AM (1 child)
Apparently, that would be the trees. There's about a dozen left in the vicinity that should shield the house as well as the last one did.
Not really, if it were it might become the strike point instead of the trees - I'm pretty sure I don't want to draw lightning toward a 60 year old wooden structure. In that same storm that fragged my tree, a neighbor's pickup truck was struck, caught on fire on the inside - burned to a crisp.
I've wanted to engineer my own robot mower forever, I have the skills to do it a dozen different ways, I even have sufficient capital to build prototypes and maybe start small scale production if I can market it, what I lack is the time - damn day job with fat salary and benefits, making us all wage slaves aren't they? In the meantime, the perimeter wire solution is actually very practical from a safety and cost standpoint - even if I did engineer my own mower with some optical flow guidance, I'd probably end up using the perimeter wire for series production (liability concerns about runaway cutting equipment), and also to cut the cost and maintenance in half by dropping the video processing bits.
I said "zero damage" for the second strike, but that doesn't count the $70 worth of GFCIs that I had to replace after it.
Was kind of interesting learning to use the wire break tracer tool - it only had about 500' of effective break-finding range, but I've got about 3000' of perimeter wire, so after I decided that there was no break within 500' in either direction of the base station I basically moved 800' down one leg, cut the wire there, and traced from that point - 500' further: nothing, 40' back I found the break, so just ran a replacement wire 40' from the break to the cut that I made, which happened to be the 40' closest to the lightning strike tree also... still didn't find any burn marks on the insulation, kind of a letdown after the scorching on the ethernet sockets.
🌻🌻 [google.com]
(Score: 2) by RS3 on Tuesday September 01 2020, @07:41PM
I didn't think about the wire being that long but it makes sense. TDR? (Time-Domain Reflectometer)
I'm not so sure about your ungrounded metal roof philosophy... Have you studied, and consulted experts? Again, the theory is a charge develops around objects and attracts the lightning discharge. The pointed rods drain off that charge and the lightning hits somewhere else. People have put lightning rods in trees too.
Yeah, I got that. There's almost nothing predictable nor deterministic about surges, lightning strikes, and the resulting damages. We just do our best to mitigate and hope...
I'd have to look into getting rid of the copper surge collectors (various wires).
Regardless, your story inspires me to count my blessings. The old "I complained because my foot hurt, then I met a man without a foot."