dBm accuracy

Information and discussions on Spectrum Analysis
Post Reply
HNRAO
Posts: 8
Joined: Mon Nov 27, 2017 1:14 am

dBm accuracy

Post by HNRAO » Wed Apr 03, 2019 12:45 am

I've been using the Spectrum Analyzer software with my RSP1. I wanted to explore it's possibilities by measuring in dBm, the Temperature in kK of a known noise source. My noise source is calibrated, and I wanted to see if the spectrum analyzer would yield the correct value. So, I have calculated vs measured. With a known input to the SA (spectrum analyzer) of 6.80E+7 kK, I end up with a power value of 9.38E-13 watts. 10LOG(9.38E-13/.001) gives me a calculated value at 20 MHz of -90.3 dBm. Setting the SA in the average mode, 10 MHz bandwidth, and 1000 Hz/bin, I measure -89.0 dBm @ 20 MHz. When I convert that value back into T kK, it's not the correct value. What is the accuracy of the dBm measuring within the SA? Is this the best accuracy I will get from this? I think this is a fantastic piece of software, but maybe I'm looking for higher accuracy than this will provide.

Reason: No reason

glovisol
Posts: 662
Joined: Thu May 10, 2018 6:42 pm
Location: Piedmont, Italy

Re: dBm accuracy

Post by glovisol » Wed Apr 03, 2019 8:07 am

Before doing the measurement with the Spectrum Anayser software it should perhaps be advisable to do a noise measurement with your RSP1 and SDRplay and read the noise power on the SDRplay power meter for a given detection bandwith & RBW. The rated tolerance for the RSP/SDRplay is 1 dB and this could be a good starting point, because the accuracy of the RSP1 with the Spectrum Analyser software could not be better than the accuracy of the power meter in the SDRplay.

Forgive me if you have already taken the above into consideration.

Reason: No reason

SteveAndrew
Posts: 213
Joined: Thu Nov 16, 2017 11:06 pm
Location: Gold Coast, QLD, Australia

Re: dBm accuracy

Post by SteveAndrew » Thu Apr 04, 2019 4:04 am

HNRAO wrote:What is the accuracy of the dBm measuring within the SA? Is this the best accuracy I will get from this? I think this is a fantastic piece of software, but maybe I'm looking for higher accuracy than this will provide.
Hi HNRAO

There will a small difference in gain levels between radios due to a variety of factors. In addition so-called 'windowing loss' may/will affect the accuracy of the displayed magnitude of the signal. For this reason I have added a control marked dBm Trim on the Input panel. This control can be used to trim the signal peak so it displays whatever level you wish. If you have a signal source of known accuracy then you can use that to trim out any small discrepancy in levels. This value is saved at shutdown and loaded when the analyser is launched.

Steve

Reason: No reason

SteveAndrew
Posts: 213
Joined: Thu Nov 16, 2017 11:06 pm
Location: Gold Coast, QLD, Australia

Re: dBm accuracy

Post by SteveAndrew » Fri Apr 05, 2019 3:40 am

Hi

First, I suggest you run SDRplay's radio - SDRuno and see what levels you get. Do they match your calculated levels, or are they the same levels you are seeing on the analyser ?

There is some confusion with your figures. In your original post (this thread) you say that your measurement showed an error of 1.3dB. In your post in a second thread, you say you now get an error of 3dB. Can you clarify these figures for me. Can you tell me what reference level(s) you are using.

With regard to bandwidths, the following lists the bandwidths for each IF frequency. Bandwidth filtering is analogue and is carried out in the radio.

IF_2048kHz - Bandwidth is 1536kHz

IF_Zero - At all sampling rates bandwidth is 8MHz

IF_450 - Span 500khZ - bandwidth is 600kHz
- Span 200kHz - bandwidth is 300kHz
- Span less than 200kHz - bandwidth is 200kHz

Steve

Reason: No reason

HNRAO
Posts: 8
Joined: Mon Nov 27, 2017 1:14 am

Re: dBm accuracy

Post by HNRAO » Fri Apr 05, 2019 7:45 pm

Thanks Steve. I'll look at both of those items. I really appreciate the reply.

Reason: No reason

Post Reply