I wonder how true that is, and if the difference was not simply band conditions. How long was the time lag between tests?
Still, I'm not fan of switching supplies.
I also wonder how he did the HF receive mod... I don't see any obvious mod signs.
4 things here about this.
1)No hardware mod will be shown if using a up converter.
2)SDR sharp is the only interface that has the rtlsdr.dll modded for HF reception.
3)HDSDR (as shown) has no modded driver for HF reception.up converters or direct sampling mods have to be used.
4)Band conditions has no bearing on the noise level generated by the dongle
Time lag 3 minutes same day which doesn't prove falseness. Its right there in the image HDSDR has a clock and calender.
rtlsdr.com is just a site that sort of announces ways of using and modding the dongle. If they feel it's legitimate they will make the article. For cynical types be assured anything is possible. If you have Google Chrome you can go to the guys website (translate it) which is in Japanese where the original documentation is.
The community behind it is on freenode ##rtlsdr go ask them about it. They work with rtlsdr daily. I posted here in RR to have people know the noise level can be reduced. I am sure the Japanese experimenter has tried capacitors with no luck so added the new diodes and removing the other one gave him less noise. Wouldn't we see a mod with capacitors instead of diodes?
From the looks of it takes a steady hand, etching skills, and good smt solder skill which is not easy because of the size of the components. I won't attempt it. Only thing I done to mine was remove the remote sensor diode out of two of my dongles and re soldered some connections that looked weak.