My calculations were right - it happens that this is the 'home transponder' - once it is tuned, the stream contains the frequency of the other TPs.
A given TP can be only in low or high band, not both (duh!).
Strictly speaking, the RSSI reported by an 802.11 interface is in arbitrary units. There is no requirement that it be in \$dBm\$ or anything else; the only requirement is that stronger signals are bigger.
But let's just say that your RSSI is indicating received power in \$dBm\$. It will necessarily be much less than the AP's transmitter power for two reasons:
An intelligent AP will not transmit at full power unless necessary, to reduce interference with adjacent APs.
Most of the electrical energy radiated by the transmitter goes uselessly into space, where the receiver isn't.
Point #2 is essentially the inverse square law. If the transmitter emits 1W, than that means for any sphere centered on that antenna, then there is 1J of energy passing through that entire sphere each second. Neglecting things that might absorb or reflect that energy, this is true no matter how big the sphere is. But since the sphere has an increasing area as it gets bigger, the energy available in a given area is smaller. Dividing the transmitter power by the sphere area gives you the power available per unit of area at some distance \$r\$:
$$ \frac{P}{4 \pi r^2} $$
So say 1W transmitter, 10m away, the field strength (assuming an isotropic antenna) would be:
$$ \frac{1W}{4 \pi (10m)^2} \approx 796\mu W / m^2 $$
Given that your typical antenna in these systems is a lot smaller1 than \$1m^2\$, you'd expect to receive a lot less than even \$796\mu W\$, or \$-1dBm\$, even though \$1W\$, or \$30dBm\$ was transmitted.
Of course, this is just an approximation. Earth will reflect some of the power. Walls will absorb some of it. The efficiency of the power coupling between antennas depends on their relative orientation and polarization. No antenna is isotropic. But the basic truth still holds: most of the energy went off uselessly into space, simply because there wasn't an antenna around to receive it.
1: really what we care about here is the antenna aperture, which is related to the physical size of the antenna but is also influenced by other aspects of its design. For an example of an antenna with an aperture much larger than it's physical size, see the loopstick antenna. Still, that's a rather special case, and your typical wi-fi antenna is still going to have an aperture smaller than \$1m^2\$.
Best Answer
The output is around -40dBm into 75 ohms. Plus minus like 10dB. That's just looking at blank sky. I've never measured it when pointed at anything powerful (like a satellite or the sun.)
So, like 3mV RMS or so.
Peak to peak is pretty much pointless. It is noise. At any moment, the peak could be many times the RMS - and then a fraction of it.
You won't be happy with what you get from your oscilloscope. The signal is around 2.5GHz. I sure as heck don't have a scope that can handle that in my hobby room.
What you need is something to convert the RF to a dBm reading (which you can mathematically convert to microwatts or microvolts at 75 ohms or whatever.)
I use a MAX2015 as a signal strength detector, and an LTC2440 ADC to read it.
Both are overkill.
The MAX2015 has a dynamic range of 70dB. 20 to 40 would probably be adequate, but I didn't know that when I started. It is handy in that I pretty much don't have to even consider the input level. But, that is also a disadvantage. The higher dynamic range comes at the cost (I suspect) of higher noise from the chip itself.
The LTC2440 is a 24 bit ADC. I intended to make measurements down to less than 0.01 dB of difference, but the noise of the MAX2015 limits it to 0.01dB. For 0.01 dB, 15 to 16 bits would be enough.
My gadget is driven with servos. I use the high resolution of the signal strength to make pictures of ambient RF.
This is an image of a window in my work room, taken with my steerable satellite dish:
The shape is fairly clear, and you can see the cooler section of wall above it (hollow for a roll up shade.)
Below it is a cooler section where the radiator (hot water, actually cold when the image was made) can be see.
This image is probably of more interest. I had to recreate it from the recorded raw data, which why I didn't post it earlier:
That's the moon through a break in the trees and shrubs behind my house.
The moon itself peaks at an intensity of -40.4dBm. The shrubs along the bottom reach about -41dBm.
That's with my uncalibrated MAX2015 detector, and a dish that's about 70cm by 55 cm.
The moon should NOT appear that large. The image is about 20 degrees wide and 20 degrees high. The moon should be a dot. It appears so large because of the wide beam of a small dish.
Stars and planets will also appear very oversized. The solution to this is called deconvolution. I intend to implement that in my software as well.
Deconvolution is also the solution to the moon being somewhat football shaped (rugby, for the non-american parts of the world.) The dish is elliptical - it is higher than it is wide. That makes the beam flatter and that makes the moon football shaped. A round dish would deliver a round moon.
I did in fact use a modified satellite finder for my first experiments. It was way too noisy, but I could see that were details to be found and looked for a better way.
There are many alternatives to the ICs I used. Those were the ones I found that I could get hold of easily and make use of.
The detector is built as an add on to an Arduino.
The Arduino talks to the hardware, and communicates results back to a program on my PC.
The whole thing has a custom PCB, a hand built frame for the servos and the dish, and a bunch of software.
The controller also has provision to switch the LNB band and polarization.
The hardware works pretty well.
I really need get back to work on the software on the PC, though. I have a lot things I want to do with it, but it keeps getting pushed off.
I want to measure the cosmic background radiation. I want to make a map of the sky. I want to make pictures of common places and things with it. I want to carry out some experiments with polarization with it.
Damn. I want to quit work and just fiddle with that machine. :)