No crud.
The only time I get screens like yours is when I enter into calibration wizard ...
(Attachment Link)
And after applying calibration ...
(Attachment Link)
What do you see when you connect SHORT and OPEN calibration standards, after performing the calibration? DOTs or ARCs?
Why measuring the calibration standards on a VNA results in arcs, not dots on a Smith Chart as most people expect:
https://www.kirkbymicrowave.co.uk/Support/FAQ/Why-do-I-not-see-a-dot-on-the-Smith-Chart-for-the-opens-and-shorts/
I see DOTS.
Maybe RF experts can jump in and shine some light about this issue. We should see ARCs or DOTs? I experimented with HP VNAs and having no money to buy the calibration standards, I tried to make a SOLT set, and manipulated the parameters that you enter into the VNA to get DOTs, but then I read about the ARCs posted by Kirkby... Then I got a decent N-connector SOLT kit and calibrated the VNA and I saw the ARCs, consistent with the report...
So... DOTs or ARCs?
Here we go:
The calibration process is just telling the instrument that the measurement file it has saved internally must match what you have currently attached to the port.
You measure the three standards, and it 'scales' the actual measurements so that they 'fit' the three 'reference measurement files' (or model) as closely as possible (solving three equations with three unknowns)
In VNA's (and in the SVA), the reference measurements (or a model of one, C0, C1, etc) will be precisely measured or calculated from real standards, and so physics determines that the open and the short cannot be perfect - therefore they show arc's.
In you guys case, you do not have the reference measurement files for any of the listed cal kits - and you do not have a model either. The best you can do is adjust the delays in the User cal kit entries. And the SVA User cal kit doesn't have a reference file so it works - it just assumes perfect standards. Therefore you see dots.
Important thing to understand with VNA's, is that the calibration is just a way of
relating aligning what you see on the screen to reality. Nowadays with the ability to use data files instead of models to define them, means that the standards do not need to be perfect as they did in the past. Essentially you can use anything as a standard nowadays, as long as you have an accurate reference measurement of it. In the past such an arbitrary standard would not fit the limited model that was used (the C0, C1 model). Model fitting introduced errors. Therefore they needed to use very accurately machined standards. Which is why they cost so much.
On the SVA to use a user cal kit properly - when you have the reference measurements or the model - then you need to do it on a PC. Ignore the on device cal, and just process the raw sweep data on the PC.
Otherwise you could do it on the device by working out the delays which provide the best fit - and you will get model fitting errors obviously. But if you do that over small frequency spans it will work. To do any of that you need measurements of your cal standards by someone with a VNA and a good cal kit, or you need a decent cal kit yourself.