Solar power battery system for WRT54GS

William P.N. Smith hath wroth:

We have a solution for the frog problem.

little green effigy on the left is "Kermit the Frog". We threatened to sacrifice Kermit if the frogs misbehaved.

All of my outdoor sites are near heavily traveled hiking trails or on mountain top destinations for climbers. Anything not bolted down or near the base of the tower gets disappeared. We also seem to have a class of hikers and climbers that fail to appreciate technology and go to great lengths to destroy anything not deemed "natural". The Coleman ice chest might work for protected or fenced installation, but not out in the open.

Grid power is nice. Dumb 4th hand story. I'm sure I've scramble the details somewhere.

Mountain top radio site wanted to expand onto an adjacent hilltop. PG&E wanted about $50,000 to install 4 phone poles and provide service. Estimated electrical work about $15,000 mostly travel time due to the remote location. Connection fee was about $2,000. I'm not sure of the dollar amounts but they're in the ballpark. USFS and CDF were not going to allow buried power or creative underground conduit power wiring. Solar power was possible but it appeared to be a continuous loss during winter. The size of the panels required were larger than the building roof or marginal southern exposure.

So, the sneaky owner installed a hydraulic system. 2000 ft of dual hydraulic hose between buildings was snuck in as part of the "signal wire" system. I don't know the hose types or sizes. A pump sends the hydraulic fluid to the remote site, where a generator converts it to sufficient electricity to run the remote site. Bernoulli friction losses were horrible but not so much that something would burn or burst. Overall efficiency was about 20% or less with 100 watts delivered from the generator. That was enough to keep the remote batteries up when solar was lacking. It only runs at night during overcast days.

It's really easy to overcharge a car battery. However, I've found that most automobile batteries, running in stationary service, simply sulfidate badly, short, and die from running the battery down below the 75% capacity mark. Some type of low battery indicator, alarm, or disconnect would be helpful to protect the battery.

OMG! Couldn't they do anything with a few amps at (say) 50 volts DC on those "signal" wires? 8*)

SMS hath wroth:

Yep, y're correct. 25 watts is what's needed to replace the power consumed by the WRT54G during 24 hours (88 watt-hrs) with about 3.0 hrs/day winter insolation. The 3.0 hours is the "equivalent" direct sunlight that a non-tracking flat plate collector would receive during the winter[1].

're correct that 25 watts is insufficient, but for a different reason. If the solar system only supplied enough energy to replace what was consumed by the WRT54G, then the system would take forever to replace any additional losses. The battery simply would never catch up and take forever to get to full charge.

That's where I found a screwup in my spreadsheet[2], in the calculation for days to recharge under full load. The calculation was inverted. Revised version at:

'm was a bit optimistic at 4.5hrs/day for mid winter. Changed to

3hrs/day.

Revised (and fixed) version specific to WRT54G at:

that even with a 40 A-Hr battery and a 50 watt solar array, it will still take about 3 days to fully charge the batteries under load in mid winter.

[1] It's really Kw-hr per square meter per day, based on 1000 Kw-hr/m^2/day. The number "3" would be the same as 3000 Kw-hr/m^2/day or 3 hrs of equivalent full sunlight per day. [2] Can someone check the spreadsheet for me? Every time I look at it, I find mistakes. I can't seem to check my own work.

William P.N. Smith hath wroth:

Worse. There were also two #6(?)AWG wires in the signal conduit bundle that were also snuck in as part of the "signaling" bundle. I think what happened was that the insurance company conducted a physical inspection after one of the other buildings had an electrical fire and made sure there was nothing risky in any of the other buildings. There may also have been some EMI generated by the charger current and picked up by 2000ft of 50 pair telco bundle. I don't have the details but was assured that it was impossible to supply power through the comm/signal conduit. I haven't actually seen this hydraulic abomination but have been assured by others that it does actually function.

Isn't the telco voltage ( 48V DC) kosher to run without all the issues associated with a high voltage power line?

BTW, you don't have to have PGE run your power. I've never seen construction where PGE didn't run the power, but in theory you could hire a contractor and PG&E inspects.

PG&E operates in a strange manner in that as far as I know, they are the only contractors that also get to sign off on their own work. That is, there are no futher inspections.

In regards to you other post, it is nit-picking, but if that box as a switcher on the input, it's not really a LDO. I'm pretty sure LDO just refers to linear circuits, though many switchers incorporate LDOs for their own internal use, or on the output to remove some switching ripple.

Jeff Liebermann wrote:

snipped-for-privacy@sushi.com hath wroth:

Yes. However, there's a total power limit on what can be considered signal wire. I don't have a copy of the NEC code handy to look it up.

PG&E has to install the poles and run the wires on the poles. The utility drop from the end pole to the building has to be underground and is always done by the building owner. That's part of the $15,000 in electrical work. The same work done on the ground would be about

1/2 of that.

Yep. Actually just about any utility district (water, road, telco, CATV) can sign off their own work. In theory, the county or state have oversight. In reality, they only act on complaints.

You're right. LDO is strictly an analog term. Especially since it's possible for a switcher to have a negative input to output voltage drop, where the input voltage to the switcher is less than the output voltage. Sorry.

[POSTED TO alt.internet.wireless - REPLY ON USENET PLEASE]

Powerup Bsp40-12 40 Watt Solar Panel for $242

GE 45w 12v Solar Panel for $257

Ip50f, 50w, 12v, Fiberglass Solar Module for $270

Sharp NE-80E1U 80w module for $320 shipped.

Any suggestions for mounting the Fiberglass panel so it can't be stolen, or otherwise damaged by horses?

number of them, but I'm not certain which one to choose

Would this be sufficient, given it's only listed as 4 amps?

FYI, this would be for use with a WRT54GS. I have a number of them (2.2 version, so they now have dd-wrt on them).

I'm now looking for appropriate enclosures, and debating antenna. The box that I need in the middle of a field to provide line of sight between two other locations should be fine with it's original antenna, but I'm not certain how well the original antenna will perform given they'll be inside whatever enclosure the WRT54GS is in.

snipped-for-privacy@gmail.com hath wroth:

and fiberglass NEMA enclosures. Put the antenna inside.

Any links to someplace whose website I can actually navigate? The main page had flash, and no way to get around it? I didn't see any reference to price or how to order from the main page.

Why not navigate in - the flash is horrible, but the site can be browsed w/ Firefox so...

Me neither, I suspect you need to find a dealer or phone them up. Mark McIntyre

I missed what size you decided to get. Most of the John Navas links were to bare panels. The Harbor Freight link includes a mounting arrangement, cables, and a couple of lights, with a poor power control box. The Costco link to the Coleman/ICP link includes mounting, a charge controller, and an AC inverter.

?Itemnumber=90599The Harbor Freight review

snipped-for-privacy@gmail.com hath wroth:

Yep. The main page uses Flash to navigate. Try:

I use the printed catalog.

This is the series that the local electrical supply house stocks:

's PVC, waterproof, and UV resistant. Note that there are other vendors of non-metallic electrical boxes. for portable use: One gotcha with internal antennas in PVC boxes. The active elements for the antenna should be kept away from the sides and the lid. That's because PVC is fairly transparent to 2.4GHz RF, but will detune the antenna slightly. You can sorta see the effect if you take a piece of small diameter PVC tubing and slide it over a high gain omni antenna. The resonant frequency goes down slightly. I built a biquad inside a PVC electrical box and ran into this problem. effect should be much less with fiberglass boxes, but I haven't tried those (yet).

Also, watch out for venting batteries. Leak proof batteries are NOT outgas proof. Yo can demonstrate that to yourself by simply putting a plastic bag around a common gel-cell and charging it. Not only will you eventually see some gas expansion, you'll also probably notice droplets of liquid on the surface. If you put the battery inside the box, put it in a seperate sealed section and vent that section. Otherwise, your expensive electronics will corrode.

Also, watch out for temperature problems. They do get rather warm in the sun. The lack of ventillation means that small sources of heat will tend to build up to rather high temperatures. The WRT54G burns about 4 watts. That's tolerable for a small box. Try to get some conduction cooling to the outside, spread the heat out with a metal plate, or give up and use a metal box with an outside antenna.

Assorted packaging attempts.

(type "solar" into the search box)

Jeff Liebermann hath wroth:

Some photos and ideas for creative outdoor packages:

're not solar powered but it does give an idea of what can be done.

Technically, you are using a DC/DC, just that it is built into the Linksys router.

Without a manual (what k> snipped-for-privacy@sushi.com wrote:

True.

You mean schematics? What company puts schematics online these days?

The Linksys is usually supplied with a 12V wall-wart, though IIRC they've been shipped with 5V wall-warts in the past. The WRTs are pretty forgiving of input voltage, which makes the ideal for battery power.

snipped-for-privacy@sushi.com top-posted:

That's because it can (and occasionally does) vary, depending on what Linksys can get good pricing on. And it doesn't matter, anything from

5-12V will work, and there's a bit of margin on either side as well.