# How does one (easily) measure "power spectral density"?

Discussion in 'Ham Radio Discussions' started by VK3CHI, Nov 3, 2019.

1. ### VK3CHIHam MemberQRZ Page

Hi all,

Under new changes to Australian Amateur licence conditions, Advanced licencees are now permitted on 160, 80, 40, 20, 17, 15, & 12:

Any emission mode.
Where the necessary bandwidth exceeds 8 kHz, the maximum power spectral density from the transmitter must not exceed 1 watt per 100 kHz.

How does one (easily) measure the latter parameter?

As a potential test instrument, I have a spare SoftRock SDR which could be connected to my RF output via an appropriate attenuator pad. However I don't know what software would allow me to do the measurement (if any). What are the alternative test instruments available? (any cheapo eBay gizmos that might do the trick? or kitsets that I can solder together?)

Thanks!

Andrew
VK3CHI

It's just power divided by bandwidth. Here's an example. The total channel power in 10 MHz is -23.26 dBm. So the Power Spectral Density is -93.26 dBm/Hz (or -23.26 dBm - 10log(10,000,000).

For your new rules, the power density is 1 watt per 100 kHz or 0.00001 watts/Hz. So at 10 kHz of bandwidth, you can run 10000 * 0.00001 = 0.1 watts.

In other words, exceeding 8 kHz of bandwidth incurs a huge power penalty.

K8AI, AE8W, WI7H and 1 other person like this.

Good post above. To measure power spectral density you only need to measure power and occupied bandwidth. A calibrated Spectrum analyzer makes that easy but an accurate way to measure power and another way to measure signal bandwidth like your SDR is all it takes.

K0UO likes this.

Cheat sheet says don't run more than 80 watts with such a wideband signal. Typical rigs would have a hard time with that anyway.
Seems like a dumb way to implement legislation. What about the antenna and ERP? At 7kHz you can turn on an afterburner but at 8 you can't?
I'm not seeing the logic behind the ruling.

K0UO likes this.

I'm guessing the intention is to allow spread spectrum and other exotic wide band modes, but only at low power on HF.

6. ### VK4HATHam MemberQRZ Page

It allows us some very unique ways of experimenting using wideband modes for data and other digital applications.

7. ### VK4HATHam MemberQRZ Page

Vrms/rt Hz, the math is quite simple really. rt is the square root of the Hertz.

8. ### G0GSRHam MemberQRZ Page

But what does it say for bandwidths less than 8kHz which is the normal condition for LF/HF bands ?

Frank