# Measuring RF output impedance

Discussion in 'Electronic Design' started by Paul Burridge, May 1, 2004.

1. ### Paul BurridgeGuest

Hi,

I have a spare RF signal generator that has an unmarked output from a
type of socket I've never seen before. What's the simplest way of
establishing its output impedance? I've had a few ideas but no doubt
someone out there will know of something better, since I always seem
to end up making unnecessary work for myself.
Any suggestions?

p.

2. ### W3JDRGuest

Paul,
The most obvious method would be to measure the open-circuit RF output
voltage, and then measure the output voltage with a known load. The output
impedance is a simple calculation of the ratio of the voltages.
However, depending on the vintage and quality of the generator, the
impedance is likely to change with frequency, and possibly even with output
level. For source-impedance-sensitive measurements, it's generally a good
idea to put a small amount of fixed resistive attenuation at the output
(6-10 dB) in order to "fix" the impedance.

Joe
W3JDR

3. ### Jan PanteltjeGuest

Load it with 50 Ohms, and measure output voltage, remove load and measure
again.
If it is half, it is 50, else do the math.
JP

4. ### Ian White, G3SEKGuest

Would it be a car-type output socket (deep recessed centre pin)? Would
it be an old Advance generator with the rounded corners? If so, it's
almost certainly 75R.
If it has a resistive output attenuator with no DC blocking capacitor,
you can switch to maximum attenuation (minimum output) and measure Zout
with an ohm-meter.

5. ### Ralph MoweryGuest

The generators are more of a voltage source. They will deliver their rated
output when loaded to the proper impedance. Load it with 50 ohms and see if
it gives the rated output. If not try 70 ohms, or another value. One of
the reasons for using a 6 db pad is that it helps isolate the impedance of

7. ### Frank BemelmanGuest

If it is half, you've got a funny generator.

Why that?

10. ### Frank BemelmanGuest

I expect amplitudes to double or at least rise, after

11. ### Jan PanteltjeGuest

Yes - sentence construction could have been more clear,
it should have been inferred as 'If it is half *before you removed it*
now that would be wrong too, 'If it WAS half' OK.
No, actually it was correct, cause 'it' referred to the first measurement.
Did you collect many nits while picking ;-)?

12. ### John CrightonGuest

Hello Paul,
what is the brand name and model number of your RF signal generator.
Can you descibe the socket to us.
Regards,
John Crighton
Sydney

13. ### Frank BemelmanGuest

Well, I also wondered what it means to Paul, once he has
figured out that output impedance. Will it change his life
dramatically...

14. ### Fred BartoliGuest

Sure. He now have a nice opportunity to ask the same question once more ;-)

Thanks,
Fred.

BTW, got some news of our goldmine order ?
Not that I'm longing to have it, but they do seem to be waaaay slooooow....

15. ### Frank BemelmanGuest

"Fred Bartoli"
slooooow....

The last news was it got shipped on 21 april. I expect it 'real soon now'.

16. ### Paul BurridgeGuest

Thanks John (and others).

The model number ain't gonna mean a lot to anyone as this is a very
old piece of kit (1950s) that I keep mainly out of a sense of
reverence for the past. It's an old ex-RAF AVO. I have posted a
picture of the socket to a.b.s.e under this same thread title...
BTW, there's a 5p coin shown for scale, but since that won't mean much
to anyone outside Britain, the outer of the socket is approx. 1" in
diameter (which won't mean much to anyone in europe but it serves them
right for adopting the metric system.)

17. ### Paul BurridgeGuest

My main sig gen states "output EMF using 6dB pad" next to the socket.

18. ### Paul BurridgeGuest

Or double, presumably.
Well there you go; I knew there must be a more elegant solution to the
one I dreamed up which basically involved taking a spread of 10 carbon
resistors of from 10 - 1000 ohms and measuring the applied voltage
across each, then arriving at power transferred in each by V^2/R;
drawing a graph of the results and finding the point of maximum power
delivered.

20. ### Ralph MoweryGuest

My main sig gen states "output EMF using 6dB pad" next to the socket.
A pad is usually 3 resistors in a small container. They can be a T or a Pi
type. This is the way the resistors are configured in the pad. You may
also see them referred to as an attenuator. The number 6 db is how much the
pad reduces the signal in power. You can get them from about .5 db to 20
db. They have to be used for the impedance they are rated at. They are for
reducing the signal level and also to help isolate small differences in
impedance.
For the calibration to be accurate on your generator it sounds like the
scale is calibrated so you need the 6 db pad after it. A 5 db pad will
reduce the power by a factor of 4 or a voltage by a factor of 2.