GBPPR 0 - 1000 MHz Spectrum Analyzer

Select a picture for larger image.


[newspec-1]

The mess.  Prototype on the workbench.  It's built into a 19-inch rack mountable cabinet.  Individual modules are built into old California Amplifier MMDS downconverter cases.

[newspec-3]

IF amplifier and logarithmic amplifiers.  Video filters are on the left (I use relays to switch in the caps).  The 10k trimpot on the CA3140 is for offset adjust.  The AD603 is on the right, the AD8307 is in the middle.  Note the nylon hardware on the LM317 voltage regulator.  The tab is also the output so don't let it touch ground.

[newspec-4]

Close up picture of the IF amplifier and logarithmic amplifiers.  The Y-output to the oscilloscope is on the top, through the two ferrites.

[newspec-25]

Logarithmic amplifier bypass relay.  Switches the 10.7 MHz (filtered) IF to a front panel bulkhead RF connector.  This allows for demodulation via an AOR-8000 communications receiver.

[newspec-5]

3rd local oscillator and mixer.  The 60 MHz oscillator is on the left, feeding the mixer and post-mixer amplifier on the right side.

[newspec-6]

Close up of the 60 MHz crystal oscillator.  It may require some tweaking to oscillate on frequency, you're smart, I'm sure you'll figure it out.

[newspec-7]

Close up of the 3rd mixer and post-mixer diplexer and amplifiers.  The Mini-Circuits PIF-60 and PLP-30 are overkill, but where in my junk box so I added them.  Be sure to heatsink the transistors.  There is a 6 dB attenuation pad on the output of the PLP-30 but it's kinda hidden.

[newspec-8]

Resolution filters.  300 kHz filter on the bottom, 25 kHz crystal filter on the top.  The 25 kHz filter idea comes from N8QOH's spectrum analyzer.  Separate RF relays increase the input/output port isolation.  Use resistive attenuators to match the attenuation of each filter.

[newspec-17]

Second local oscillator (1170.7 MHz) and mixer.  70 MHz IF filters are lower right.

[newspec-18]

Close up picture of the second local oscillator and mixer circuits.

[newspec-12]

Sweep and sync board.  Outputs are via BNC connectors, the bottom one is the VCO sweep and the middle one is the X output to the oscilloscope.

[newspec-16]

First local oscillator and mixer.  VCO is a Z-Comm unit, I tore it apart to tweak it a bit.  The top sheilding was removed to lower the frequency about 50 MHz.  I have no further information on the VCO, it tunes from around 1100 to 2100 MHz.  RF input is via the lower right BNC connector.  1100 MHz IF output is via the top N connector and the LO tap is via the middle BNC connector.

[newspec-19]

Full spectrum display during testing, connected to a 50 ohm load.  Oscope is a Tektronix 465 in X/Y mode, 0.5 V per division on each channel gives us 70 dB of logarithmic range.  The far left signal is the "zero spur", the others are spurious images which I have no idea where they are coming from or how to remove them.

[newspec-20]

Full spectrum display during testing, connected to an external antenna.  AM/FM/TV broadcast are clearly shown.

[newspec-21]

Zoomed in on the FM broadcast band.  300 kHz resolution filter.

[newspec-22]

Connected to a homebrew 10 MHz marker generator.  0 - 300 MHz shown.

[newspec-23]

Connected to a homebrew 10 MHz marker generator.  Full spectrum shown.

[newspec-24]

-30 dBm, 150 MHz signal.