LibreCal temperature setpoint proposal

[dna35]

I propose to increase the temperature setpoint to a user adjustable setting. Starting at 35C in 5C increments to at least 50C.
Reason:
1- The current setting of 35C is insufficient to buffer external heat influences from nearby equipment.
2- Calibration in the field may require higher temperatures.

[jankae]

I am not a fan of this idea.

1. The point of having a fixed temperature is to keep a stable behavior of the LibreCAL compared to when its factory coefficients were measured. How do you propose to achieve this if the user changes the temperature?
2. Neither the LibreCAL nor the LibreVNA is intended for field use (as the lack of a display may indicate). And most lab equipment has a fairly restrictive temperature range for which its specification is valid
3. Do not put your sensitive calibration instruments next to a heat source
4. Did you read the documentation? You can actually change the target temperature, there is a SCPI command for that. The setting is not persistent across reboots and actually using it is not recommended (due to potential drift of the calibration standards) but if you really want to, you are free to set it to whatever you want

By the way: you are opening discussions, not issues. Fine for general questions but if you have a bug or a feature request (which this is), then an issue makes it easier for me not to forget about it

[dna35]

I considered this issue and there are methods:
1- Use a table of various temperatures to chacterisze the LibreCal
2- Measure the s paramteres at leat at 50C and do a linear interpolation

#4 is a valid point concerning drift of the standards but why was the temperature set point included in the SCIP command?

[jankae]

1- Use a table of various temperatures to chacterisze the LibreCal

How would you achieve this for already produced and sold LibreCALs?

2- Measure the s paramteres at leat at 50C and do a linear interpolation

Temperature drift is not guaranteed to be linear. And for a calibration device, its coefficients should match the actual behavior as close as possible.

why was the temperature set point included in the SCIP command?

Why not? As mentioned in other discussions, I am not a big fan of software restricting the hardware and implementing such a command is easy enough. Yes, the manual specifically mentions that it is not recommended to change because of temperature drift. But if you really want to, you could calibrate with your LibreCAL, change the temperature to 50°C and then measure the LibreCAL again. This way you can create your own calibration coefficients that are valid at 50°C.

I do not consider this a common use case and measuring the coefficients at various temperatures would significantly increase the production time per device.

[dna35]

Yes, #1 & #2 would be done at the factory
Fair enough response though.

The LibreCal is an excellent product and show cases itself when trying to use two VNA's for comparing mesurements. Previously it was a major task to maintain calibration, so this brings a level of professionalism to the LibreVNA.

Thank you for developing this product and supporting the LibreVNA all these years.

[jankae]

Yeah, I like your suggestion but at this point it is impossible (for already existing LibreCALs) or time-consuming (for new LibreCALs) while providing (in my mind) not that much more value. At the end of the day, all decisions are compromises.

A quick hint: on Github, using #+number links to other issues/pull requests/discussions. Very useful but not what you wanted to do here. When referring to specific points in my reply you should use something else (e.g. "Point 1:" or a quote).

[dna35]

Thank you for that hint, I was wondering why it turned blue.