4
\$\begingroup\$

I have a doubt about how to connect the following device (a RF power meter, AD8318) to the external circuitry. Let's consider the device:

enter image description here

There is the input RF coaxial cable, the power supply terminals (VCC and GND) and the output port. Regarding this last one, you may see that there are two terminals: the square terminal (which is GND) and the circular terminal (which is the signal output pin).

I have built the following stage of this circuit and I have connected its input pins between the circular terminal and GND of power supply terminals.

So my question is: is the GND of power supply the same GND of output port? Obviously they are short - circuited (so a multimeter will see a short circuit), so at low frequencies I would use them indifferently. But since this device works quite high input frequencies (1 - 8000 MHz), I was asking if there may be some problems, since at those frequencies also a little short circuit behaves like a transmission line.

So, can I avoid to use the square output terminal and take the output signal between GND of power supply terminals and the circular output terminal?

\$\endgroup\$
4
  • 3
    \$\begingroup\$ You have a question, not a doubt. If someone told you how to connect the device, but you didn't trust that information, then you'd have a doubt. \$\endgroup\$
    – JYelton
    Jun 27, 2020 at 20:42
  • \$\begingroup\$ there's no high frequency on the ground. Why should there be? \$\endgroup\$ Jun 27, 2020 at 20:54
  • \$\begingroup\$ @JYelton "doubt" is commonly used to mean "question" in Indian English. \$\endgroup\$
    – alephzero
    Jun 28, 2020 at 10:45
  • \$\begingroup\$ That may be, but the site is a "Question & Answer" site, not a "Doubt & Answer." \$\endgroup\$
    – JYelton
    Jun 28, 2020 at 17:36

4 Answers 4

4
\$\begingroup\$

The output doesn't contain frequencies as high as the 8GHz the input is rated for.

The output is the power level, not a copy of the input.

Transmitters don't change output power that fast, so your output won't change that fast either.

That said, the datasheet shows the detector responding to a change from no signal to -10dB in less than 20 nanoseconds:

enter image description here

You'll have to deal with sharp transitions there, but nothing worse than you would encounter in dealing with moderately fast digital signals.

\$\endgroup\$
4
\$\begingroup\$

It's good practice to return the current as close as possible to the outward signal. This means using the square blob in the 'out' rectangle for the ground return. The board is layed out like this so it's easy to do, just put a 2-pin connector on the two blobs.

However, the ground between the 'out' ground and the PSU ground is very short and wide. The output bandwidth from that chip is only to a few 10s of MHz, so you can either ground with little difference.

\$\endgroup\$
3
\$\begingroup\$

can I avoid to use the square output terminal and take the output signal between GND of power supply terminals and the circular output terminal?

The GND of the power supply terminal is connected to the "circular output terminal" by a quite wide copper pour on the PCB layer 1. This is about as good a connection as you're going to get.

If you were to separate your connecting wires for the output signal to connect them further apart, you'd create a much larger wire loop than if you just connect to the two terminals of the output connector, possibly leading to increased magnetic interference with the output signal.

I recommend using the output connector as intended.

\$\endgroup\$
2
\$\begingroup\$

There will be a ground plane under those components, and maybe a power plane.

In extracting the amplitude information, just use the "round" signal and the "Square" ground just 2mm away.

\$\endgroup\$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge that you have read and understand our privacy policy and code of conduct.

Not the answer you're looking for? Browse other questions tagged or ask your own question.