Graphical Wi-Fi Spectrum Analyzer for Laptop

Metageek (the source of the free inSSIDer app) have a couple that may fit your requirements. They aren't free and "low cost" is a very relative metric but you might take a look at

I use the first version of the Wi-Spy (Metageek) spectrum analyzer for sniffing 2.4GHz noise sources. The major problem is lousy sensitivity due to the small antenna. I had to add a reflector of sorts: which helps. Best to have a receiver that has a connector for an external antenna. I also have the dongle from Ubiquiti, which has the same sensitivity problem. (Note: Both are repurposed wireless keyboard/mouse dongles).

Are there any cheap S-band down converters? The wispy looks like quite a rip.

That's a nice package.

But why does there seem to be nothing available that provides a simple graphical FFT of the actual signal picked up by one's laptop. The antenna and receiver are already built in.

There are tons of these for audio input.

Ken Rockwell

Your audio input is actually sampling the "baseband" audio signal and delivering it into RAM as direct samples. It's trivially easy to do an FFT on the resulting data.

WiFi chipsets do not normally work this way. They're designed to track and demodulate the spread-spectrum signal within the chipset, and deliver only the recovered packet data through the PCI or USB interface. The bandwidth requirements at the interface level are far less that way.

The closest I'm aware of to a general-purpose solution for what you want, are the popular "DVB tuner" sticks. These can downconvert and deliver a slice of the RF spectrum several MHz wide, as I/Q samples delivered into RAM over USB 2.0, and you can run spectrum analysis on the resulting data.

Unfortunately I have yet to see any of these which can go up to 2.4 GHz... haven't seen one which goes above 1.something GHz. Since they're made for DTV reception, there hasn't been an economic incentive to build them capable of tuning up above the frequencies that DTV uses anywhere in the world, I suppose.

And, some of the newer WiFi modulations use a very broad spectrum spreading... 802.11N can use channels up to 40 MHz wide. Capturing this wide a slice of spectrum (even as 8-bit IQ samples) would present quite a burden for a laptop!

Now I get the picture. Thank you for explainig this.

Ken Rockwell

A 2400/500 MHz down converter in front might help.

The spread spectrum was originally developed for military purposes to avoid detectability by spreading out the signal so it would be buried in thermal noise. Of course, the signal presence is detectable close by a simple diode envelope detector, but at larger distance, you need to know the actual spreading sequence to detect the presence of a signal.

At ground based systems at room temperatures (300 K) the noise density is about -174 dBm/Hz.

A typical 2.4 GHz device will have an output power 1 .. 100 mW (0..+20 dBm) and if 40 MHz = +76 dBHz bandwidth is used, the power at transmitter output is -76 .. -56 dBm/Hz, so anything past 30 m (path loss 70 dB) would be buried in noise from a 1 mW source (assuming 0 dBi antenna gain).

I dunno. I'm still using MMDS components collected over the years. I'm down to about six assorted downconverters, all with serious problems, which will need repair. I don't see much MMDS stuff on eBay. Need one?

As a spectrum analyzer front end, I guess it's tolerable. The input range has to be moved from the original 2.5-2.7Ghz to 2.4-2.5GHz. That's fairly easy, but getting a flat frequency response is not. The IF output is 200-400 MHz. That works, if you already have a spectrum analyzer that covers the range. I've never measured the dynamic range, but I suspect it's not very good. Still, it's probably better than a wireless mouse dongle.

With Wi-Spy, you're paying for the software and development, not so much the hardware. Think of the dongle as a copy protection device. I reverse engineered the original when it was first announced in 2006. Ryan Woodings, the founder, was not thrilled, but didn't complain.

This might be what you want: It describes a method of using commodity wi-fi hardware to detect non-wi-fi signals using FFT techniques. Cognio is now owned by Fluke. Aruba Networks does what looks like the same thing, but inside their access points: There are probably others.

You may also want to look at the high priced wi-fi(WLAN) test equipment available to see what can be done.

When I used inSSIDer here to track a problem I found that from the dozen or so networks I can see, well over half were using the new wideband method.

When ebay was raining MMDS converters, I got a few. Since the service was dead, I assumed the supply wouldn't be around forever. Now that schematics have shown up on the interwebs, i suppose I can get around to hacking with them.

If California Amplifier made hammers, I'd buy one. Their cases are really rugged.

I use free inSSIDer to find a 'less occupy' space among the jungle Don't know if it's still free or not ?

perhaps you could plug a MDS downconverter into the input, you'd need a 12V power feed tee too