Peter and Jim, -------- Original Message -------- From: jim bell <jdb10987@yahoo.com> To: Peter Gutmann <pgut001@cs.auckland.ac.nz>, "wirelesswarrior@Safe-mail.net" <wirelesswarrior@Safe-mail.net> Cc: "cypherpunks@cpunks.org" <cypherpunks@cpunks.org> Subject: Re: Intercept receivers (was Re: Hackers Remotely Kill a Jeep onthe Highway) Date: Mon, 27 Jul 2015 15:31:17 +0000 (UTC) From: Peter Gutmann <pgut001@cs.auckland.ac.nz> wirelesswarrior@Safe-mail.net <wirelesswarrior@Safe-mail.net> writes:

...

For example, Ettus' USRPs, covering VHF to 6 GHz or so, starting under $1000, .>>that not long ago were in the $10,000s. The HackRF (which some have complained is little more than an IF strip) effectively covering down to below 10 Mhz is only $300 (though its performance, due to only 8-bit ADC, is not in the same league as the 16-bit USRPs). If some hardware hacker were to deliver a 14-16 bit ADC daughter board (there are afforadble chips offering up to 60M samples/sec) for the HackRF (it is provisioned to accept one) it could substantially improve its use.

The problem with the HackRF (and other USB-based devices) is the bandwidth of the USB connection, it's not that the HackRF hardware can't handle it, it's that you can't get that much data to the PC.

The HackRF contains a LPC4320 which has unused capacity for light DSP and other tasking. For any heavy DSP it could use a FPGA. It has an internal header for expansions like this (and/or a higher resolution ADC and its CPLD can be reflashed to accomodate such changes.

(I have a HackRF - I needed a cheap way to track down some odd RFI issues - and it's a pretty cool piece of hardware for the money, but, as I mentioned earlier, you do get what you're paying for. If I could make a wishlist, it'd be nice to have a rev.2 with some work done on the front- end...).

The HackRF, like all wideband, direct conversion, receivers, can suffer LNA saturation from nearby/very strong out-of-band signals (pagers and FM broadcasts). A cheap, simple, solution is a coax notch filter, in line with the antenna. I've been experimenting with these and they seem to work very well. From: jim bell <jdb10987@yahoo.com> Here's my idea for a probe for an RF device. The wavelength of a 2.45 GHz signal (WiFi) is about 12 cm. A quarter-wave antenna would be 3 cm. Or, that 3 cm could be in the form of a loop, at the end of a piece of.'hardline', or other coaxial cable. (hardline would be best, I think, because it is a small diameter. ) https://en.wikipedia.org/wiki/Coaxial_cable So, the diameter of the loop would be a bit more than 1 cm diameter, which is appropriate for finding tiny RF sources. One side of the loop would be connected to the center conductor of the coax, and the other side of the loop connected to the shield of the coax. Note: The coax should probably have many dissipative RF beads strung along its length, in order to prevent the cable itself from picking up RF, and delivering to the loop on the end, thereby to be detected. http://www.amidoncorp.com/small-ferrite-beads/ There are many different kinds of materials that such beads are made from; perhaps we can have an RF engineer chime in on the subject. He can also tell us if the loop is too-badly mismatched (RF impedence) to the coax to function well. Jim Bell Try it!