Continue to Site

Welcome to EDAboard.com

Welcome to our site! EDAboard.com is an international Electronics Discussion Forum focused on EDA software, circuits, schematics, books, theory, papers, asic, pld, 8051, DSP, Network, RF, Analog Design, PCB, Service Manuals... and a whole lot more! To participate you need to register. Registration is free. Click here to register now.

Low-Cost RF power measurement using Shottky diodes

Status
Not open for further replies.

jabidof

Member level 5
Joined
Jul 25, 2003
Messages
80
Helped
0
Reputation
0
Reaction score
0
Trophy points
1,286
Activity points
675
I'd need to develop a power detector for level between -70dBm up to 0dBm and a frequency range from 5MHz to 20MHz.

Any idea? Full-wave rectifier but which diode? How to improve sensitivity?
 

Hi folks,

I need to measure RF power in <50 mW range. Does somebody have a simple circuit to share?

Many thanks in advance. :D

humber555
 

Diodes can be used as RF Power detectors (meters). With no amplifier the dynamic range of a Diode Detector is ~20dB.
Important for this kind of detector is the matching network.
Here is an application note from Agilent:
**broken link removed**
 

At these low frequencies gain is cheap. You can amplify your low level signal and then use the diode without any complex matching networks.

Is this a manual measurement or an automatic one? For manual, you can have two or more measuring devices with different gains.
 

Analog Devices has low cost log amplifier circuits that will cover that dynamic range. This is clearly the way to go.

Trying to do 70 dB dynamic range with diodes is impractical. You could go to the trouble having a cascaded diode array. Input has a 3 dB splitter with a diode on one arm, and 40 dB gain on the other arm followed by another diode. You could then sum the diode currents with appropriate gain to get a crude power to voltage out. It will be pretty non-linear. At the lowest levels, it is smart to use two diode for each station, 1 with DC bias and RF applied and the 2nd with just DC bias applied. The detected output is the DC difference between diode 1 and 2. If you use two diodes packaged together, this will take out diode junction voltage variations and eliminate maybe 90% of the frequency drift.

If you are really enthralled with trying to do this with diodes, you could consider chopping the RF input at, say a 10 KHz rate, and detecting/amplifying the 10 KHz ac tone out of the diode. This will allow you to go down to lower input RF power readings.
 

    jabidof

    Points: 2
    Helpful Answer Positive Rating
Thanks biff54,

I just read the AD8307 - Low Cost, DC - 500 MHz, 92 dB Logarithmic Amplifier datasheet. It is clearly the way to go :D. The final end will be to connect a pic microcontroller that embed an A/D converter + RS232 (or USB device). A cheap power meter...
 

Status
Not open for further replies.

Part and Inventory Search

Welcome to EDABoard.com

Sponsor

Back
Top