RF amplifier for RF signal generator.

[firewalker]

Hello. I was given an RF signal generator (Leader 17A). It produces sine wave from 100 kHz to 100 MHz. This signal generator has not a 50 Ohms output. It uses a simple pot to the output to vary the amplitude.

Is there any easy to implement RF amplifier (low gain, 2~3 or even unity gain) to make the output 50 ohm? The power supply is 18.7 volts.

   

       

   

Alexander.

[PA0PBZ]

You could use an emitter follower with a 50 ohm resistor in series on the output, but considering the stability of the signal generator I doubt that you need a 50 ohm output.

[Mosaic]

Any 'amplifier' transistor that gives you a low impedance output drive in series with a 49+ ohm resistor should be ok.
IF VSWR is important then do a test with a non inductive resistor load of 50 ohms and 25 ohms, measuring the peak output with a good power meter at the device output. You can assess the true output  impedance from those figures and trim your output resistor to give a good 50 ohm match.

At 100Mhz peak frequency you should keep your output leads shorter then 1/10 wavelength to avoid transmission line effects. Note that transmission speed  depends on the permittivity of the substrate or the cable being used.

Note that most BNC connectors & cables etc tend to be around 52 Ohms unless they're high quality.

You can also include an AGC circuit to drive your amplifier feedback loop near the 50 ohm output resistor to correct for signal flatness across the spectrum. That's basically an AVAGO schottky HSMS detector pair into a dc smoothing cap pair for a DC level feeding a jellybean LM358 which then drives a PIN diode array or alters the bias of your amplifier to suit.

[PA0PBZ]

You can also include an AGC circuit to drive your amplifier feedback loop near the 50 ohm output resistor to correct for signal flatness across the spectrum. That's basically an AVAGO schottky HSMS detector pair into a dc smoothing cap pair for a DC level feeding a jellybean LM358 which then drives a PIN diode array or alters the bias of your amplifier to suit.

That's what I ment with "I doubt you need 50 ohm", the output of this thing will be all over the place for signal strength, drift and harmonics. You only need 50 ohm if you are planning to do accurate measurements, and just converting the output to 50 ohm will not be enough. You will need AGC and a low pass filter, and after you have that you will find out that it drifts a lot. The generator is nice as it is but there are too many problems to fix if you want something more accurate.

[cncjerry]

One other consideration wrt output impedance is that some signal generators, mostly audio, are made to operate into 600ohm impedance.  If you try to feed that into 50ohms you might not get anything out or maybe something highly distorted.  Try using your scope in 50ohm mode to look at your calibration signal and you'll see what I mean.

There are a number of buffer amps, all you really need is a wide band LNA.  You might be able to determine the output impedance of the generator and do a transform through the LNA to drive 50ohms to a higher level.

[rfbroadband]

regarding the use of an emitter follower...
The output impedance of an emitter follower is 1/gm, so the series resistor you need at the output is 50 -1/gm. Keep in mind that you introduce a voltage divider at the output which will lower your gain, thus with 50 in series you end up with ~-6dB instead of your desired 2-3dB.

If you use common emitter circuit (without degeneration) the gain is gm * RC. The intrinsic rce of the device is high, so you can neglect it if you keep RC low. So you could choose a small RC (50Ohm) and then adjust the gm by adjusting the current Ic.

If you use emitter degeneration your gain is (first order) Rc/Re and you can control your with a lot less dependence on the current.

Matching for 50Ohms is usually done for max. power transfer, keep in mind that the back on the envelope calculations described here are focused on voltage gain and not power gain!

Lastly, each circuit type has different noise and linearity trade offs. In your case I assume you don't care about noise, but more about linearity. If you run a source follower at low currents, linearity will be really bad.

Before you make a decision I would suggest you determine what the max. signal level is that you want to pass thru that circuit.

[Wolfgang]

Hi,

I used some classic RF transistors to make something like that:

https://electronicprojectsforfun.wordpress.com/rf-module-gallery/the-amplifier-module-gallery/medium-power-hf-amplifiers/

they are class A und have a relatively low gain. Also, they do not bun out when misused.