This page contains details of a noise measurement project. There are two parts - software and hardware.
The software comprises a program to measure noise using a PC sound card and calculate noise figures by means of a calibrated noise source. It also allows a noise source to be calibrated from another noise source.
There are three ways to use it. Use it to measure the noise figure of a receiver by putting the audio output of a linear receiver such as an SSB receiver with AGC and audio processing disabled into the PC and a noise source into the receiver. Alternatively either use such a receiver with the PC sound card as a measuring system for determining the noise figure and gain of a device under test, or build (hence the hardware bit) a simple DSB direct conversion receiver instead.
You will find the software here. If you use it, give me some feedback. Hardware is here. Documentation to follow. The noise source I used is the design by Bill Sabin W0IYH. The PCB and build information of the version that formed a Warrington & District Amateur Radio Society club project may be found here. These are the gerber files for the noise source - caveat emptor - check them carefully before you use them. I *think* this is the right set but am not sure.....
16/08/08: New version uploaded. On a dual processor system it had a tendancy to lock up. The source has been tweaked to stop this and a new executable built.
24/03/06 The latest version fixes a few 'hangup' bugs and allows the font size to be changed for displayed results to allow its use under Windows 98. The usb driver is moved out into the n2pk VNA DLL so that prople without the Windows DDK and Cypress code can modify and build the software.
Not RadCom's fault - I should have read more carefully.....
My thanks to Chris Bartram GW4DGU for pointing out that I was not clear about the local oscillator and the need for cleanliness. Indeed phase noise and reciprocal mixing are not the main concern here - it is AM that is the enemy. It just so happens that if it is of low phase noise then it must have low amplitude modulation (which is what we need). This would be true if the receiver were a heterodyne or a direct conversion receiver too.
Also thanks to Brian G0GSF for pointing out a stupid mistake on the first page. "RMS Power" - wrong. The measured power is averaged but please disregard the term "RMS". Thanks Brian