Detect RF input power of a diode

[MobiNaz]

Hi all.

I want to test a diode and plot a curve against its input RF power and output voltage. For this purpose, I have attached the diode to an antenna. The antenna's output is matched to the diode input.

So the antenna receives power from a RF source, diode receives this power at its input. Now at the output of the diode I have attached a wire and am trying to see the output voltage by the oscilloscope. But I can see nothing. What am I doing wrong?

Can anybody please advise?

Thanks,
Nazifa Tahir

 

[BradtheRad]

1.

The diode requires a complete circuit loop in order to conduct current. Does your setup provide this? Your oscilloscope alone is insufficient. A 500 to 100k resistor can be connected between the diode and ground. Then connect your scope across this resistor to examine the waveform coming from the diode.

2.

The diode is typically connected to a coil which is connected to the antenna. There is a voltage swing on the coil. A tap from the coil is applied to the diode. Does your setup have this?

3.

The diode is used to rectify the incoming rf signal. The signal voltage must be sufficiently strong to drive the diode to conduct. Typically 0.6V if you use a silicon diode (but can be less with germanium or Schottky diodes). Can you be sure your antenna is getting a strong enough signal to provide this? Do you have a transmitter of your own or are you picking up local broadcasts?

4.

To see the rf waveform your oscilloscope must be able to operate at the rf frequency. Can you see any waveform coming from the antenna directly on the scope (without going through the diode)?

Or even if your scope cannot show that high a frequency, you can see the information present in the broadcast (assuming it's AM rf). Connect a small capacitance across the diode output and ground (in addition to the resistor mentioned above). This detection method is similar to a crystal radio set.

-----------------------------------------------------

Simple rf detector probe schematic:

http://www.qsl.net/n9zia/wireless/pics/rfprobe.gif

Several rf detector schematics:

**broken link removed**

Simple radio circuit:

http://www.aaroncake.net/circuits/cradio.gif

Website about crystal radios, also showing evaluations of several diodes (bottom of page):

Crystal Radio Circuits

 

[biff44]

diodes are very non-linear. You are probably generating some DC voltage, but it is microvolts and you do not see it.

You need to use an "RF" diode, probably a "zero bias" one to use with an antenna.

You need a capacitor at the output that can pass DC volts, but block RF energy from travelling down the wire to the oscilloscope, as a minimum.

Then, as abovel, there are ways to improve efficiency further with good impedance matcheing, etc.

 

[MobiNaz]

Thanks for the replying.

I am trying to detect a signal at 38.5 GHz. I am interested to see the DC component. I do not have the oscilloscope to see the 38.5GHz signal.

So the capacitor shunts the RF signal, that is provide low resistance to 38.5GHz. Around 4 uF would be enough? Do I still need a resistor? How do I choose its value?

I can not do anything about the diode I am afraid, biff44. At this frequency, you do not have much choice. But it have low barrier and at a short distance the signal theoretically can bias the diode.

Thanks in advance.

 

[BradtheRad]

Sorry, I have not experimented in frequencies that high. Are you using the type of diode that is suitable for 38 GHz speeds? There are several types. Such as fast recovery type which sounds like it should be right.

'Low barrier' refers to Schottky type which seems like it would be suitable.

A 4 uF cap would most likely absorb the entire signal. I think you need a tiny packaged cap in the area of a few pF. Post #2 has links to schematics showing typical values to detect audio (assuming that's what is modulating the rf).

A wavelength at 38 GHz is only 3/10 inch. This demands the shortest connecting leads possible. I have trouble picturing how components I can pick up with my fingers are able to operate at such a speed.

Have you tried testing your detector at lower rf frequencies?

 

[FvM]

1. Why don't you refer to a reference circuit from a microwave diode datasheet? It will also tell you about expectable output for a specific RF input.

2. If you're doing something different from known working diode circuits, please clarify with a drawing.

 

[MobiNaz]

Thanks for replying everyone. I am able to measure/detect the dc output of the diode by placing a shunt capacitor at 5.3 GHz. But for input values greater than -3 dBm the RMS voltage was low, does not coincide with simulated results.

Is it because of the capacitor not rated to support that much power, or is it because of diode saturation?

Thanks for the reply in advance.

 

[tony_lth]

For 40G, the shunt cap should use 0402 or 0201 package with SRF reaching to 40G. Murata has such cap.
I used some MMIC detector to detect power of 40G, and it can work good. The MMIC can give you a deector pin and a ref pin for compensating the temperature drift.