Measuring things is fun, but measuring MORE things is a lot of fun!  This article is about the electrical characterization of the receive preamplifier that is part of the UHF repeater recently donated to the Wilson County Amateur Radio Club.

The preamp compensates for the unavoidable losses in the duplex filter and enables the receiver to hear better.

This unit was manufactured by Lunar Industries of San Diego, CA.  I can't find any technical data on it, but that's unimportant for now.  Let's put it on the test bench and find out what we have.

Here is what it looks like:

Preamp, External View
Preamp, External View
(click for a large view)
Preamp, Internal View
Preamp, Internal View
(click for a large view)

The circuit looks simple, but don't be fooled.  At UHF frequencies, tiny things matter.  The widths of the traces, gaps between traces and the physical layout & dimensions of the components all affect the unit's performance.

You see that the +12V power input is in the form of a feedthrough capacitor going through the wall on the upper- left-hand side.  The wire passes through, and the capacitor itself is formed around the wire and connecting its other terminal to the chassis.  In this way, DC signals are allowed through the center pin, but AC signals are shunted to ground by the capacitor body which encompasses the center pin.  That's one way to get DC into an RF-tight container.  The ground pin is connected directly to the chassis.

The heart of the preamp is the 4-pin MMIC amplifier on the right-hand side with the tiny ferrite toroid around its input pin.  You can see the input path has a series LC circuit consisting of a toroidal inductor and a ceramic capacitor.  The string of resistors and capacitors are part of the DC biasing network to power the amplifier.

Let's take a couple of measurements.  First, we set up the spectrum analyzer and normalize its tracking signal generator to the test fixture.  That way, the analyzer can see the effects of the test cables and adapters without the device-under-test (DUT) in the circuit and calibrate those effects out of the final measurement.  Then we insert the amplifier as the DUT and sweep its input from 9 kHz to 1.0 GHz.  The spectrum analyzer measures its output as the frequency is swept and presents us with a graph showing the output power level at all frequencies.

In this case, I measured the preamp's frequency response with the power on and with it off.  The unpowered trace is shown in purple and the powered trace is yellow.

Frequency Gain, Powered vs. Unpowered
Frequency Gain, Powered vs. Unpowered

The green line represents 0 dB relative to the signal generator's power which is about -20 dBm.  Everything above the green line is amplification and everything below it is attenuation.

From this graph, we learn that the unit amplifies between 290 MHz and 796 MHz.  It is non-linear across its amplification range, but its useful range is assumed to be in the UHF frequencies around 400 - 500 MHz where we see that performs its best.  The non-linearity across frequencies is particularly unimportant in this application because we only care about one frequency -- that's the repeater's receive frequency.

Knowing that the repeater is currently set up to receive at 466.800 MHz, we see that the amp has about 22.4 dB of gain.  That is a huge advantage in receiving weak signals.  Let's say the receiver's sensitivity is -120 dBm.  As shown in the duplexer article, a signal arriving at the duplexer with -118 dBm power will experience a 2.95 dB loss (ignoring additional cable losses) and will be too week to receive.  But the preamp will bump that up to -98.6 dBm and it's now a signal we can easily hear.

There's a lot more to it than that, but the bottom line is that the preamp gives the repeater a strong advantage in handling weak signals.

OK, so that measurement was made with a -20 dBm test signal.  What happens when we put a weak signal in?  The Minimum power output of the tracking generator is -20 dBm, but we can run that through an external attenuator, then through the device-under-test and into the analyzer's input.  70 dB is the most I can attenuate and the resulting -90 dBm signal is so weak, we have to kick in the analyzer's calibrated preamp.  Even then, we're near operating near the analyzer's noise floor, given such a broad frequency range.

Still, we can see that the weak signal performance we need is in the same ballpark as the strong signal performance:

Weak Signal vs. Strong Signal Gain
Weak Signal vs. Strong Signal Gain
(click for a large view)

Here, the strong signal (-20 dBm) is shown in yellow and the weak signal (-90 dBm) is shown in blue.  We still have around 17 dB of gain down here.  Therefore, the linearity of the amplifier over a range of power is very good.

One more test.  So far, we've been sweeping sine waves into the preamp.  Let's put an actual FM modulated signal in.  Here we use an HP/Agilent 8648B signal generator to create a 466.800 MHz FM-modulated signal at -71 dBm.  We're modulating a 1 kHz tone with 7.5 kHz of deviation.  The spectrum analyzer and the signal generator are locked to the same calibrated 10 MHz OCXO master frequency standard:

FM Modulated Signal Gain
FM Modulated Signal Gain (click for a large view)

We see that the gain is as expected with total channel power of around -53.72 dBm.  Therefore the gain is 17.28 dB.

Through all of this, the preamp only consumes 38.5 mA of DC power.

There are many more tests to do; 3rd order dynamic range, noise figure, etc.  Those will have to come later . . .

Add comment


Security code
Refresh