Page 1 of 1
Sensitivity specification for RSP2?
Posted: Thu Mar 16, 2017 4:57 pm
by KM4YRI
Are the sensitivity measurements for the SDRPlay RSP2 published anywhere? I don't see them in the datasheet. I'm just wondering how it compares to say, a radio like my FT-817 which publishes ~0.25 uV for HF bands up through UHF and VHF in SSB / CW mode. I've searched Google and this forum and haven't yet seen a single measurement for the RSP2. If anyone has come across any measurements, I'd be thankful for a link to them.
Re: Sensitivity specification for RSP2?
Posted: Fri Mar 17, 2017 10:24 am
by 13dka
The only related figure I know is the theoretical noise floor of -132dBm (but without specifying a corresponding bandwidth IIRC). The actual noise floor should be around that number tho (got corresponding levels at a very QRM-quiet spot), AFAIK it should be possible to translate "noise floor" to "sensitivity" in some way. -132 dBm is not bad at all (better than a lot of renown classic receivers) and I can confirm that the RSPs are very sensitive.
Whether or not that sensitivity can be utilized is a different story though, I think receiver sensitivity is quite meaningless except when using the receiver in ITU "rural/remote" environments, and at least on HF the maximum usable sensitivity of the RSPx depends a lot on the overall signal load coming from the antenna, IOW the rudimentary/missing preselection of (not only) the RSPs can limit the actual sensitivity a lot.
Re: Sensitivity specification for RSP2?
Posted: Mon Mar 20, 2017 11:29 am
by JMG
The RSP/SDRuno combo easily allows to determine the sensitivity of the receiver.
Sensitivity of a receiever best is specified by its noise figure NF.
The noise figure is the factor between the fundamental thermal noise power of a resistor and the total noise power of the receiver.
The resistor noise power is -174dBm/Hz.
One of the unique great features of SDRplay/SDRuno is the absolut numerical dBm power indication! It is the total, real, absolut power of the receiver within the IF channel of bandwidth B. Reducing the indicated value by 10logB will give the power within 1Hz.
Example:
Terminate the receiver input by 50 Ohm. Set a high gain. Place the frequency window at a spurious free position. Read the dBm value. Subtract 10logB. The difference to -174dBm will give you the NF.
Representative numbers:
Frequency 150 MHz
IF bandwidth B=10kHz, so 10logB=40dB
Indicated Power -131.5dBm
Noise figure NF = 174dBm - 131.5dBm - 40dB = 2.5dB
This measurment is quite accurate. The tolerance of the indictated dBm value has been measured. It is within +-0,5dB for most parts of the frequency spectrum.