One problem that comes up again and again in doing measurements is that we need the apparatus to be quieter than the thing we're measuring, ideally by at least a factor of two. Besides quiet, it should be wideband, have an accurately known gain that's flat with frequency, have a clean step response, and generally do its job while keeping itself out of the way. There's a wealth of detail in our app note AN-1 on photoreceiver testing.
Test gear is noisy, with many scopes coming in at 5-10 nV / √Hz and RF spectrum analyzers around 30 nV / √Hz. What's worse, noise contributions add in power: that 30 nV is over 1000 times noisier than its 50-Ω load resistor. The analyzer is general-purpose, and can't do everything well.
At EOI/HEO, we’ve run into such problems more than enough times. Thus we built ourselves an amplifier that’s good for just about anything of that sort: the LA22 Lab Amp from Hobbs ElectroOptics, which we use all the time.
![]()
The LA22 has an accurately calibrated gain of 100 ± 0.5%, and a flatband noise of 1.1 nV/√ Hz. If we put that in front of the analyzer, we effectively get an analyzer with 1.1 nV noise in 1 Hz with no calibration worries, plus very good pulse response for testing low-level circuitry. Check it out!