Quote:
or a marginal design that had to be 'salvaged' by their inclusion?
|
I wouldn't start drilling more via holes unless it definitely shows instability in a chosen setup. Normally, a VNA is used to measure the amplifier to collect s parameter data and this can be used to predict if the overall PCB module is unconditionally stable at its SMA connectors.
However, one crude method to look for instability is to simply fit a few inches of low loss coax cable at the output SMA of the PCB and leave it unterminated/open at the other end of the cable. Then do the same at the input SMA with another short piece of cable. Then sniff the PCB with a pickup probe and look on a spectrum analyser on a wide span to see if the board is oscillating up in the GHz region. Then try fitting a short at the end of each cable in turn to see if it still remains stable. Then repeat with longer cables of a foot or so. This isn't as good as a formal VNA test but it is a quick and dirty indicator of any potential instability as it explores several combinations of reflection coefficient at the input and output of the board and by playing with cable lengths and shorts/opens at the end of each cable an unstable amplifier can often be revealed eventually...
If you don't have a spectrum analyser then look at the DC current draw from the PCB and see if it changes once you fit the unterminated cables. Often this current will change quite a bit once the board starts hooting. But this isn't a reliable test method.