Phil,

I have compiled these data for Eddie's down converter

RF amplifier:
HP/ Agilent (now Avago) MGA- 86576
Gain:  ~ 21 dB
Noise Figure: 2.2 dB
1 dB compression: + 6 dBm

Mixer:
HP/ Agilent (now Avago) IAM-8
Conversion Gain:  ~ 15 dB
Noise Figure: 15 dB
1 dB compression: + 8 dBm

Filter:
Coilcraft P7LP507
500 MHz Low Pass Filter
Insertion Loss ~ 0.5 dB

IF amplifier:
Minicircuits MAV-11SM
Gain:  ~ 12 dB
Noise Figure: 3.6 dB
1 dB compression: + 17 dBm

The Fiber modulator:
Fiber Span AC231
RF Gain -0.8 dB
Output noise -135 dBm /Hz (Noise Figure 39 dB)
(from an email from Fiber Span rep. to Bill Sisk / Eddie Castro)
1dB compression point for this system is ~ +15 dBm. For a clean(i.e., very linear) signal, we recommend staying 10 dB or more below the 1dB compression point, putting the "peak linear power" at +5 dBm. The damage threshold for these devices is an input power of +20 dBm. Please do not exceed this maximum, even for very short durations.
Loss from dome to WAPP room: 2 - 3 dB ?

The total gain adds up to 48 dB for properly biased (as per data sheet) but both the Rf & IF amps are biased below the suggested voltages. There are also losses from Isolator, tiny RF cables in front of RF amp, low pass filter, output directional coupler. So, the final gain up to the input of the fiber modulator could be ~ 45 dB and at the Fiber Receiver ~ 42 dB.

Assuming an LNA + Mixer gain of ~ 35 dB and with + 8 dBm mixer compression point, any input - 27dBm will begin to saturate the mixer. Assuming 20 dB gain for the LNA and + 5dBm compression any input above  -15 dBm will begin to saturate the amplifier.

- Ganesh