<azonenberg>
miek: I have incomplete support for an undisclosed VNA that isnt merged to master yet
<azonenberg>
was waiting on them to publicly release the alpha firmware version i was doing dev against
<miek>
ah ok, good to know. i'd be down to write a driver for my agilent vna once some of that gets merged
<azonenberg>
Yeah. for the moment i'm assuming you are doing cal on the front panel interface of the instrument, using the vendor firmware
<azonenberg>
since that varies so much from device to device
<azonenberg>
then basically extending the existing scope APIs to set the sweep range, number of points, and acquire s-param mag/angle waveforms
<azonenberg>
as of right now the internal representation for waveforms is dB mag / degree angle, which is optimized for display vs computation
<azonenberg>
my thought being that s-param waveforms tend to be small
<azonenberg>
so the cost of converting to e.g. complex real/imaginary is minimal
<azonenberg>
vs time domain scope waveforms can easily have tens to hundreds of megapoints, vna sweeps are usually a few thousand at most
<miek>
yeah, that makes sense. i was wondering about whether to use mag/phase vs complex internally, i guess it depends on what's going to be most commonly used by filters
<miek>
i wasn't sure whether it should be derived from Oscilloscope or just be its own Instrument class. i feel like there are enough big differences that it doesn't quite fit into the scope APIs
<miek>
one thing i found was that the concept of a channel in most bench VNAs is a bit different to what i was used to. each channel can have its own stimulus settings and calibration, and then it can contain multiple traces for the various s-param measurements
<azonenberg>
So right now what we have is each s-param is a channel with a stream for mag and a stream for angle
<azonenberg>
but we have to figure out where things ilke power settings fit in
<azonenberg>
this is still a WIP
<azonenberg>
Filters generally convert to real/complex internally but users usually want to see mag/angle
<azonenberg>
as far as deriving vs not, thats a tough question, it may be worth making a new common base class? i dont know yet
massi has quit [Remote host closed the connection]
<azonenberg>
(as usual, tagging active contributors that I hope will attend and am scheduling around, but anyone else in the channel is free to call in if you're available and interested)
<azonenberg>
also @aleksorsist if you're interested in seeing what the dev team is up to, you're also welcome to hop in