Hacking SDRs

Conundrum, Sat Aug 26 2017, 07:41AM



Interesting effects:

1) Modifying the system in this way actually reduces noise so much that it doubles effective performance with weak signals,
with the LF mod (see earlier post) by adding modified DLLs allows tuning down to 11 MHz.

2) Some blue LEDs are really good broadband RF noise sources.
The worst culprits are ones used on certain appliances, so much so that I replaced all the units on my laptop because they were causing intermittent problems due to HDD activity. It appears that under certain conditions the LEDs draw excess power from the chipset and cause random glitching!

3) Adding copper tape around the dongle makes for a good RF "virtual ground" because the case is slightly conductive.
Not sure why but must be a fluke with this particular DVB-T

4) I made a makeshift RF antenna using old label printer roller, some high value polystyrene capacitors from Stella and many metres of random wire from a defunct reed relay with the turns separated using Sellotape (tm)
This allows me to pick up phone NFC "tweets" from 6+ metres away even before adding ferrite rods or an inner spaced layer made from squares of plastic graphite zrzsg
see

Things to try:

5) I found my MW coil, made using the casing from a very old plasma tube.
Going to try some adjustments such as a resonant tuned antenna set for 60kHz (aka Rugby) with a lambda diode oscillator to bring it into the SDR's passband.
As these use under 250uA if set up properly they are ideal for low power circuits and far better than even a single stage opamp.
The tricky part is biasing, which I intend to overcome using a PNP transistor with output from tuned circuit fed directly into a high efficiency green LED which happens to emit IR well below the normal turn-on threshold.
The effect should be a subtle modulation of the valley point therefore adjustment in frequency by moving the NDR region.