Power + DSL signal in same 700ft cable length

Tell ya what...I got a 900 foot run of 24 ga six pair and will test it out with 48 v, 24 v and 12 v with different loads and see what I emperically come up with.

FRS Rules & Regs kinda puts the damper on that..bummer....

In the past, the elco line card had a fuse to prevent excessive foreign voltage form getting into their frame. But in actual practice, a short on the line will eventually time out the DTMF register (tone decoder) then place that obnoxious warble tone on the line...eventually, the lien card will drop into a "deny service" mode and remove the 48 volt talk battery. So not any real problems per se...

Yes and no...FCC Rules & Reg require anything connected to the PSTN (Public Switched Telephone Network) meet several requirements. In a nutshell - not inject a foreign voltage into the line, adequate protection form the 120 VAC mains (most devices use a walwart for that), and not inject excessively high audio levels. In other words, "do no harm to they side".

That's all part of the inside plant at the telco's Central Office. The line card has a transformer with two windings on the street side.

frame ---)||(----------- Tip to street side )||( audi )||(----- Earth ground path )|| )||(----- - 48 volts )||( ---)||(----------- Ring to street

Now the -48 volt side has a current sensor that detects when you go off-hook and draw loop current. The unbalanced -48 becomes a balanced pair with the transformer windings connected as shown. A balanced pair with twisted wire acts much the same way as a coax cable for noise immunity.

Just a simple inductor & capacitor low pass filter (or high reject) to keep the high frequency DSL signal out of the voice path.

Much to the disappointment for Thomas Edison who advocated direct current, alternating current won out for one simple reason...it passes through a transformer. Lets say you want to move 10 Megawatts across country. 120 MW = 120,000 volts @ ONE amp 120 MW = 120 volts @ One THOUSANDS amps Now think of a wire wire carry that kind of current? So they use transformers to raise the voltage at one end and to drop it down at the far end. Avionics use 24 volts AC at 400 cycles. The higher frequency means smaller transformers and 24 volts uses smaller gauge wire (thus lighter in weight).

Electric codes usually don't differentiate between DC and AC, anything under 50 volts (typically) is not code limited.

If I wanted to move power from point A to point B and avoid electrical code restrictions, I'd stick a 120 V to 48 v transformers at both ends.

(I wrote)

The shock hazard might be slightly higher for AC than DC, but not that much. I was guessing that 48V is the electrical code limit for low voltage wiring, but I didn't look it up.

AC because you can use transformers instead of DC-DC convertors.

Also, because it reduces the problems with electrolysis if the connections get wet.

-- glen

You should be able to solder them, but it won't be as convenient as ones with solder lugs.

-- glen

(snip)

Even smaller at higher frequencies. The usual PC power supply takes AC in and converts to 300VDC (voltage double on 120V, bridge rectifier on 240V.), then to about 20kHz AC with a simple oscillator and into a ferrite core transformer. There is also the regulator that controls the oscillator, but at 20kHz you can get a lot of power through a small transformer.

Some lower current power supplied, like cell phone chargers, run closer to 1MHz. That is why they are now so light.

-- glen

Actually, what I was asking is if there's anything absolutely *wrong* about me converting from AC to DC to AC to DC. (From 120VAC, to 58V DC, to ~18VAC with a inverter, then back to DC with wallwarts.) My calculations show that it would work. Besides seeming screwy, are there reasons that it's just a terrible idea?

Tonight I read a bit more about how tranformers work, and about "windings", and found several sites with illustrations similar to your ascii art above. Nice approximation!

In a post yesterday, I proposed converting AC to DC, then using an inverter back to AC so that I could then use the walwarts (thus converting back to DC). Here you mention just using low-voltage AC out to the peak. I spent about 2 hours today trying to find 120V to

48V transformers online, but only stumbled across one, which output 3 amps (too high). Do you have a source/hyperlink for such equipment? Or, if it were you, would you arrange a few transformers in serial to arrive at ~48 volts?

Also, you mentioned 120V to 48V transformers at *both* ends, but out at the peak, depending on my amperage, wouldn't I need something like a

*24V* to 120V transformer, due to voltage loss?

Geez, I need to learn more about tranformers. Back to my reading.

You're very kind. I'm getting anxious to experiment a bit. In fact, I went out and bought a multimeter today. But now I'm not sure if I should plunk money down on a couple of AC tranformers or on a AC-DC converter and a DC-AC inverter. : /

The 3 amp rating is a MAX, draw you can do, not a push that it does.

As to your other post, the wall warts want 120vac.

Compared to water: Voltage is pressure, current is flow, and resistance is well resistance. Smaller pipe is like smaller wire. So to get the a given amount of water (power/watts) to the end of the line, you can boost pressure or flow.

Here's what I'd do first then see how it looks bad or doesn't work.

At Pedestal: Powered by 120VA DSL Modem via its wall wart Wireless Router with directional antenna via wall wart (say a Linksys WRT54G to match the AP) Transformer putting out XX VAC at max of X.X amps (maybe)

Put XX VAC on your wire (or 120VAC)

At peak put in something like a Linksys WAP54G. It has 2 antennas. So you can put a directional antenna going to both your house and the pedestal. Set it to operate in wireless repeater mode.

Power it via an AC2DC power supply. You'll need the power requirements for the WAP54G, pick a supply with this output, then work backwards to match the AC input here to the voltage drop and AC output of the transformer at the pedestal. Then pick an appropriate transformer for the pedestal. If this turns out to be too much current on your wire at

48VAC then you might want to consider 120VAC and just use the wall wart that comes with the WAP54G at the peak. No mater what fuse the line from the pedestal to the peak (at the pedestal) to make sure you don't have too much current in the wire when the moose steps on it.

Doing it all this way gets you out of the data cabling business and just into worrying about power. Maybe submersible wire as would be used in a well. 18gage maybe?

OK folks, what did I miss?

No, there's nothing "wrong" with that, per se. If there's a compelling reason as in your case (you want to keep the feed around 50 volts and want to save money by using the existing spare pairs in your aerial drop cable), then go for it.

$17 48 volt transformer rated at 3 amps (max current you can run through it)

Also check out

Would like to put in a request for some classical electrical engineering or even physics. There seems to be a general lack of knowledge about Ohm's law rampant here.

carl

I dug a bit deeper. A Linksys WAP54G needs 5V @ 2.5A. Or at least that what the wall wart can supply. They don't have the input specs but another Linksys router I just got has a wart that supplies 12V @ 1A. Which is close in wattage. My wart says input of 120V @ 3A. Which has got to be an overload situation as 3A is 30 times nominal. And considering that these warts have to be made for a total cost of about a $1 or less I'm guessing it's a very simple circuit with a large inrush and not very efficient. And if the included draws too much current you could look at:

Watt, 85-264VAC input, 5V @ 3A output, .3 A input at full load.

So you run up a 16 AWG outdoor 3 wire cable. (The first one I found was $.30 a foot. I'm sure there's cheaper.) Try the Linksys supply and if it doesn't work put in a better unit. The unit above would draw .3 A max and have a max voltage drop of < 2 volts.

Put in some slow blow fuses or better yet mini-circuit breakers that self reset.

What else?

Put a slow blow fuse at the pedestal.

*A Linksys WAP54G needs 5V @ 2.5A ???*

All the ones I've used are *12 volts at 1 Amp*. You might be looking at the WAP55AG which is the 2.4/5.8 GHz dual-band model.

Linksys WRT54xx - 12 v @ .5 A Linksys WAP54xx - 12 v @ 1 A Linksys WAP55xx - 5 v @ 2.5 A

Above numbers are from the Linksys spec sheets. Don't have any wallwarts for the WAPs laying around, but the label on the WRT wallwart says: Input: 120 VAC 60 Hz 12 Watts Output: 12 VDC 500mA (Which would be 6 watts)

I called and asked. That's the numbers they gave me. ?!?

My wired RV042 router is 12v @ 1A which is nearly the same wattage and I'm guessing they're close.

Follow the link at

for Products Wireless Basic Networking

I did. It didn't show current so I called. I missed the voltage discrepancy. Stuff like this will drive you crazy when trying to plan. I hate having to buy something before I can figure out if it will work.

But I'll still bet that a 15W to 20W supply would work. Which means all the other numbers are the same. Even it if needed 30 Watts, the solution I proposed would not change.

The voltage requirements are shown in the PDF User Guides for each product. Some list current ratings, most don't.