One source I see is 3.30ohms/100 ft, round trip. That's 32 ohms, and if it takes 24 watts, I agree. But the DC-DC inverters I saw for $8.00 each take up to 72VDC in, and I doubt you'll need 24 watts. Put 72VDC in and you get less current and less drop.
You could double up the pairs; OR take one pair, use both sides and use a ground return.
He wants to keep cost down and not use a separate power line - he wants to run it on a single 22 ga pair with a tad greater than a one amp load at far end. That's why it gets complicated. And burying the cable requires gel-filled cable - and he already has the 2 pair 22 ga outdoor aerial drop cable.
Your newsreader must be sad or you would have been able to retrieve the 80 or so messages in this thread :-)
My Time Worthless cable modem uses 12 volts at 1 amp and my Netgear router uses 12 volts at 100 mA. It's 1.1 amps, but your resistance might be a problem.
Any long cable run takes a huge voltage drop hit at, say 50 ohms per
1000 feet, so you need about 12Ga cable to make the DC to DC converter work, and at that time youi might as well throw down 14 Ga extention cords at 700 feet and run 110vac :-)
Just like I don' recommend IW for outside, I don't recommend Aerial for buried, but it beats IW if that's all you have. The gophers like the air filled cable better than the IKYPIC anyway.
Not sure, I believe it one amp at 12 volts. Which would be in line with a wireless access point. Most I've seen are 12v @ 1 amp, so call it 12 watts required.
Right. Most of the access points I'm looking at operate between 5V &
12V, and between 380mA & 1.5A.
I'm now trying to juggle power consumption with feature list, signal strength, and receiver sensitivity. So I still haven't made a final decision on equipment.
However, I've spent the better part of the weekend trying to learn and understand all of this. (based on the excellent information I've received in this thread.) I've done calculations at 12V @ 1.5A and @
1.0A... and it seems like I'm going to be okay(?), if I've understood all of this correctly. Here's what I've come up with:
700 feet of wire = 1400 feet round-trip
22 gauge solid copper is rated16ohms/1000ft
At 1.5 amps: V = 1400ft * 1.5amps * 16ohms / 1000ft = 33.6 volts loss
At 1.0 amps V = 1400ft * 1.0amps * 16ohms / 1000ft = 22.4 volts loss
This means that if I put a 48v DC power supply at one end, I should get out either ~14Volts or ~25Volts at the other end. Both should be sufficient to power these devices. I think...
Regarding heat, here's what I calculated: 700 feet of wire = 1400 feet round-trip 22 gauge solid copper has 16ohms/1000ft 1400ft * 16 Ohms / 1000ft = 22.4 Ohms
With a 1.5 amp current: E = 1.5 * 22.4 = 33.6 Volts 33.6 Volts @ 1.5 amps = 50.4 Watts dissipated over the length of the wire 50.4 Watts / 700 Ft = .072 Watts per foot. Probably no fire danger, but could get toasty???
With a 1.0 amp current: E = 1.0 * 22.4 = 22.4 Volts 22.4 Volts @ 1.0 amps = 22.4 Watts dissipated over length of wire 22.4 Watts / 700 Ft = .032 Watts per foot Probably just fine?
I've also been looking at power supplies. Depending on the two devices I pick, it seems like I'd need something like one of thes 48V 1.8A ones, in a weatherproof enclosure.
formatting link
?product=16004+PS Am I completely off? What am I missing? Or do I have reason to climb out of the pit of despair... :)
There are filters normally designed to keep the DSL signal out of normal phones. As phones expect DC, they should work to get DC in and not bother the DSL signal. I believe they are just LC filters.
It might be restricted to 48V in. You get the most power out when the load impedance equals the source impedance, so 26.6 ohm load, about 53 ohms across 48 volts for about 0.9A at 24V, about 21W out.
But he can use both pairs for power, with the DSL line on one pair. DSL is designed to run on lines with DC or low frequency AC on them.
If the choice is 48VDC or 48V RMS AC, I would choose AC and put transformers on each end.
You'll need to calculate the additional power consumption of the linear voltage regulators. Now here is where the lessor voltage drop (may it be from using a larger gauge wire or higher input voltage) would be to your advantage - regulating 33 or 24 volts down to 12 volts, well...you've doubled you power consumption at the far end.
I don't see a problem using aerial drop cable stretched across the ground where no one is going to trip on it, nor do I see using 120 VAC outdoor rated 3-conductor wire with a GFI at the ped end.
Going with 120 VAC has several advantages: 1) You don't have to worry about voltage drop. 2) You can use the devices' supplied walwart power units. 3) Its cheap - 14 ga 3-conductor 120 VAC cable.
Going with anything else has disadvantages: 1) 24 or 48 volt converters at the head end. 2) Custom power supplies at the tail end. 3) Both of the above would cost more than running a 120 VAC line. 4) Don't use the earth as a common ground return - unpredictable conductivity.
If your really are concerned about an animal tripping over it and pulling it out along the ground - splice the 120 VAC line with several 120 VAC plugs and jacks as break-away points. Same for the aerial cable - make a couple of splices with Scotchlok connectors stuck in a 45º bend of grey PVC with the ends pointed down and silicone plug the ends.
Use a pair of Linksys WAP54G Access Points - one config'ed as an AP and other as a Client.
Why bother, you'll climb back in when your daughter reaches dating age.
Bottom line - you've asked good questions and gotten some good input from different directions. Granted, solar power and buried fiber would be the way to go ...but that's a luxury few can afford unless its a commercial venture with a client payback.
Now, if you look at the DC-DC supplies I pointed you to at MECI.com; you see they accept 37-72VDC, with a regulated output at [5, 12, 24, I think.]
So plan on getting 37 VDC into them; with an R of 25 ohms. Say 37 watts total:
37W/37V= 1 amp.
At 1 amp, E == IR = 1x25 = 25V drop. So as long as you source 37+25 = 62V in, you're OK. If you supply 72, you're better off. {I'm capping the supply at 72 so even with no load, the DC-DC inverters are OK. You could no-load feed in more, with a 70V zener crowbar at the top to insure here no overvoltage.)
I^2R == 1x1x25 = 25watts. Say it's double that somehow.
50 watts/700 ft = 0.071 W/ft = 6 mW/inch.
Measure a 1/2 watt resistor's length and make a comparison... Heat will not be an issue...
SURPLUS. You need 72 VDC 2+ amp supply. I'm sure you can find such with some digging... Weatherproof is an upside down bucket with the supply inside. Put mesh along the bottom to keep the mice out.
Make sure you put in a breaker / fuse to keep from starting a fire when a moose breaks the wire. With a DC supply you'll likely have a built in current limit but should fuse it anyway to be sure. With 120 VAC it's even more so as you're likely pulling from a source that's setup to supply 15 A or more. And if you do start a fire, how's your insurance?
I won't pretend to understand what you just said. I'm already on the edge of comprehension on all of this stuff. : (
I actually talked to a Qwest splicer over the weekend. I asked him about this very thing. He said that if Qwest found someone trying to run their own power over the same wire as is carrying the DSL signal, they'd shut off the connection. Apparently this opens up all kinds of potential problems. Then again, he also told me later that beyond the Network Interface Unit (or "NI"), which they provide, they don't really care what I do. So I'm not sure what the verdict is. But when I asked about equipment that would inject power into the DSL-carrying wire, I got the impression that it wasn't widely available, and could be costly. decaturtxcowboy implied the same in an earlier post.
glen herrmannsfeldt mentioned earlier that I might somehow be able to use LC filters to put power into the line. Very intriguing.
Why AC? This has me intrigued. From everything I've read (which obviously still just skims the surface), AC is typically more efficient, but usually comes in the form of high voltages and can bring more shock hazard, infringement of electrical code, etc. There's also the issue of dealing with low-voltage tranformers and heat issues, but then again, I guess I'm not avoiding any of that by going DC, either.
Wow, I've heard of others using X10 switches to remotely power cycle equipment. But they have a much more limited range. Never would have thought to use FRS radios. Nor would I ever be able to successfully execute this approach myself. : ) Definitely some out-of-the-box thinking. You should patent this, or write it up as an instructable.com or makezine.com article or something!
Ughhhh. That's enough to scare me out of this altogether! Fire is definitely a concern in the summer months. I thought I was safe from this by using lower amperages. But, like you suggest, I'll incorporate fuses just in case. Add one more thing to my learning curve.
Yep, found a few at 36-73 VDC. (Great site, by the way. Thanks for the pointer.) I noticed that these things are PCB mounts, which I assume means etching my own circuits, etc. Yet another world that I'd love to get into, but don't know much about yet. *sigh*
Yep, that makes sense.
Here's yet another variation I'm looking into: At the 120VAC and telco pedestal, I put a 58V @1.8A power supply
formatting link
V= 1400ft * 1.8amp * 16ohms / 1000ft = 40.3 volts loss 58v - 40.3v loss = ~17v usable at the peak, @ 1.8A.
At the peak, I use a small 25W power inverter: I'd have to disect it and do a bit of welding, but this should give me enough juice to power 2 devices just using their wallwarts (especially if their 5V each, barely if their 12V each). The routerboard 112 + miniPCI card says it requires max 10W. Most modems use less power than that.
What I'm not able to find is what kind of input amperage these inverters expect. But it seems to me that if I have 17V @ 1.8A, then that's enough for the inverters to supply up to ~31 watts (not including conversion losses and heat). Am I thinking correct?
A couple more thoughts:
I don't understand all of the theory yet, but I have this nagging suspicion that I'm half-crazed for even considering the above approach. Essentially, AC to DC conversion, then DC back to AC (inverter), then AC back to DC (wallwarts). I actually like the idea of the inverter and wallwarts being in the same enclosure as the modem and router out at the peak, because they'll all keep each other a bit warmer in the winter months. At the same time, I'm pretty sure this is not conventional practice. But is there anything absolutely *wrong* about it?
My calculations seem to just eek by. Not much margin for error. Even though the numbers add up (theoretical), I probably just need to try it out (applied). If I went ahead and ordered the converter and inverter (under total), and tried it this weekend in my front yard, do I run any risk of frying anything? It seems like the opposite is more likely; that I'll be underpowered. But could someone provide me with a sanity check?
Are we just talking about a higher power bill each month? If so, I can live with that.
700 feet of outdoor-rated 120VAC cable is expensive for my budget. I've also priced THHN in 1/2" gray PVC, but it's equally spendy. And
700 ft of trench, even just a couple of inches deep, would be 100% manual labor because of the rocky soil. (BTW, what's up with prices on wire? I swear Romex costs twice as much now as it did when I finished my basement 2 years ago.)
I'm thinking I can get converters and custom power supplies for under $100. Maybe even under $50 if my 25W inverter idea is do-able. I didn't know about your fourth point. There's an existing tower on the same peak, abandoned, that is grounded 2 or 3 places. Each wire that connects the tower to a ground rod is probably 1/2" thick! Which probably means pretty beefy ground rods too. But I have no idea if that stabilizes conductivity...
Clever. Whether I end up using 120V AC or reduced-voltage AC or DC, this sounds like a good idea.
Lol. That's a way off, still. : )
Thank you. I agree, I've gotten excellent input from many individuals on this thread (plus the countless sites I've visited to learn theory). Amazing how much information and knowledge is available through the Internet. (BTW, I've been taking notes and photos of this entire project. When I finally get it working, I'll do my part and share it online for the benefit of others.)
I looked more into this. It looks like most of these base unit injectors push 48 volts through the wire, but they're not meant for use on 700 feet. In fact, there seem to be quite a few threads where people are having trouble powering PoE equipment even at 300 ft. So, as you said, I'd probably need to replace the base unit with my own power supply. I hadn't considered using the units at the other end (the peak) to regulate the incoming current, though. Not a bad idea, and definitely simpler than trying to build my own PCB stuff. I guess I could further modify a unit by adding a barrel connector to power the modem (which wouldn't be PoE-compatible) I'll crunch more numbers here. Thanks for the suggestion.
Cabling-Design.com Forums website is not affiliated with any of the manufacturers or service providers discussed here.
All logos and trade names are the property of their respective owners.