measuring amplitude of small high-frequency signal

[Tom]

Everyone,

I need to measure the amplitude of an AC signal of variable frequency
between 1 and 15MHz, with amplitudes in the range of a few hundred
microvolt (say 100..1000 uV). Available supply voltage is nominally
+-2.7..3.3V or +2.7V/-1.35V (battery powered circuit). The measurement
itself does not have to be taken often, a few Hz refresh rate will
suffice.

What I'm currently contemplating is to use a high-frequency opamp such
as the AD8061 (300 MHz GBW) to provide two stages of gain 20. The stage
outputs would be AC-coupled to get rid of the DC offset. This would
give me a signal of ca. 40..400 mV which I could rectify and filter.
For better linearity I would use an active rectifier circuit, probably
using Schottky diodes.

Another alternative I have considered is to use an analog multiplier to
square the signal, and then simply lowpassfilter it for the DC
component (from which the amplitude is easily calculated). However, I
haven't found analog multiplier chips compatible with my battery power
supply.

Does anyone of you have a better suggestion?

greetings,
Tom

 

[John Woodgate]

<[email protected]>) about
'measuring amplitude of small high-frequency signal', on Fri, 30 Sep
2005:

I need to measure the amplitude of an AC signal of variable frequency
between 1 and 15MHz, with amplitudes in the range of a few hundred
microvolt (say 100..1000 uV). Available supply voltage is nominally
+-2.7..3.3V or +2.7V/-1.35V (battery powered circuit). The measurement
itself does not have to be taken often, a few Hz refresh rate will
suffice.

In that amplitude range, an ordinary 1N4148 has a square-law response. A
Schottky would give more DC output. You can rectify the signal directly
(you won't get much DC, but enough) and then amplify the DC. The usual
technique is to use two diodes, one in shunt across the signal and one
in series with the load resistor, with the signal capacitively coupled
to their junction.

Ancient (1950s) RF millivoltmeters, such as the Rohde & Schwartz URV
used this technique. But DC amplification is far easier now.

 

[John Larkin]

Everyone,

I need to measure the amplitude of an AC signal of variable frequency
between 1 and 15MHz, with amplitudes in the range of a few hundred
microvolt (say 100..1000 uV). Available supply voltage is nominally
+-2.7..3.3V or +2.7V/-1.35V (battery powered circuit). The measurement
itself does not have to be taken often, a few Hz refresh rate will
suffice.

What I'm currently contemplating is to use a high-frequency opamp such
as the AD8061 (300 MHz GBW) to provide two stages of gain 20. The stage
outputs would be AC-coupled to get rid of the DC offset. This would
give me a signal of ca. 40..400 mV which I could rectify and filter.
For better linearity I would use an active rectifier circuit, probably
using Schottky diodes.

Another alternative I have considered is to use an analog multiplier to
square the signal, and then simply lowpassfilter it for the DC
component (from which the amplitude is easily calculated). However, I
haven't found analog multiplier chips compatible with my battery power
supply.

Does anyone of you have a better suggestion?

LTC and Analog Devices, and maybe Maxim, make RF detector chips.

John

 

[Jim Thompson]

Everyone,

I need to measure the amplitude of an AC signal of variable frequency
between 1 and 15MHz, with amplitudes in the range of a few hundred
microvolt (say 100..1000 uV). Available supply voltage is nominally
+-2.7..3.3V or +2.7V/-1.35V (battery powered circuit). The measurement
itself does not have to be taken often, a few Hz refresh rate will
suffice.

What I'm currently contemplating is to use a high-frequency opamp such
as the AD8061 (300 MHz GBW) to provide two stages of gain 20. The stage
outputs would be AC-coupled to get rid of the DC offset. This would
give me a signal of ca. 40..400 mV which I could rectify and filter.
For better linearity I would use an active rectifier circuit, probably
using Schottky diodes.

Another alternative I have considered is to use an analog multiplier to
square the signal, and then simply lowpassfilter it for the DC
component (from which the amplitude is easily calculated). However, I
haven't found analog multiplier chips compatible with my battery power
supply.

Does anyone of you have a better suggestion?

greetings,
Tom

15MHz is such a low frequency, why not an active-device peak
detector. You could even take some gain.

...Jim Thompson

 

[[email protected]]

Why not using a calibrated SW-receiver?
Thats the simplest solution.
Aperiodic measurement is difficult with tis small values.

rgds Rudy