I put a detector like this one together when I needed to measure around 20dBm
at a couple of hundred MHz.
I stripped the resistors out of a BNC co-axial attenuator, wired an H.P. 5802-2800
Schottky diode between input and output pins, and a 10nF ceramic capacitor between
the output pin and the body.
It's a simple peak detector
The original was calibrated using an r.f. signal generator, a 50 ohm termination and a 10Mohm voltmeter, at, I think, 100MHz. It gave readings down to -20dBm, and results were within about 0.5dB from a 100kHz or so to 512MHz (the limit of the generator.)
A subsequent one was further calibrated to 40dBm at 30MHz with an h.f. transmitter
and a power meter.
40dBm is about the maximum in consideration of the max inverse voltage of the
diode.
It seems reasonable to assume that anyone who is interested in such a device will know both how to calibrate and use it, but, just in case, here's how it would be connected.