A quality RF power sampler can be built quite easily. This device can be used to sample RF from your transmitter, power amplifier etc in order to accurately measure output power, intermodulation etc.
In general a spectrum analyzer or power meter can accept up to + 20dBm (100 mW) maximum of RF power, however to stay on the safe side, we will limit this to 0 dBm (1 mW). So, if we want to measure power up to 1 kW, we have to attenuate by 60 dB !
I found a very good description by Don Jackson W5QN, download it as PDF here.
Don's sampler was designed for -50dB, however for -60dB I used following component values:
-
Ferrite tore material 43 (out of my junkbox ...), through which RG-400 teflon coax is passing
-
20 turns/windings on the core, evenly distributed
-
Rt = 2 Ohms (I put 5 pcs SMD resistors of 10R in parallel, providing ample capacity for power dissipation)
-
Rs = 47 Ohm
This provided -60dB sampling +/_ 0.5dB from 4 MHz up to 50 MHz , but is still useable up to 145 MHz taking into account the actual attenuation at that frequency.
Important remarks :
-
attenuation by 60 dB is quite a lot ! In order not to pick up stray RF with your measuring equipment, you must use a good quality (=with double shielding !) coax cable between the sampler and your measuring equipment.
-
the short length of feedline cable between IN and OUT ports exhibits some inductance, this results in SWR going up proportionally with frequency. In my construction I measured 1.2 : 1 on 145 MHz. The frequency response can be optimized by compensation of the inductance with a small capacitance (a couple of pF ...) between feedline and ground. I found that a capacitor made of a length of 3 to 4 cm RG-174 nicely compensated the inductance and was tested OK with QRO power. This compensation has no noticeable effect on the sampling attenuation.
-
I haven't seen big difference by exchanging 'IN' and 'OUT' ports on the sampler output, but to be consistent with frequency response vs output, always use same port as IN / OUT.
RF sampler (the compensation capacitor not yet in place)
Results
One picture says more than thousand words ... measured with FIELDFOX VNA and lab-grade calibrated termination 50 Ohm.
RF sampler output vs frequency (S21 measurement on VNA)
RF sampler SWR vs frequency - before compensation (up to 1.2 : 1 on
150 MHz)
RF sampler SWR vs frequency - after compensation ( better than 1.05 :
1 up to 150 MHz)
RF sampler output with 'OUT port' open (see W5QN comments)
Upper trace is output with termination 50 Ohm
RF sampler output with 'OUT port' shorted (see W5QN comments)
Lower trace is output with termination 50 Ohm