Calibration
|
Does this mean that everyone that builds one will have to ship it to Dave for verification? I wish I would have just built the prototype only. I’m thinking I should withdraw the Beta #1 version and Jove can go with the prototype to simplify and speed up the project. Verification didn’t seem necessary with the MK I or II we just accepted it as good enough. We also have to ask why no one else took the initiative to build one.
Larry |
|
Hi Larry,
toggle quoted message
Show quoted text
If a builder sticks to the exact same hardware that you & Jim specify, they could be somewhat sure the results will be the same. Sub out different parts and all bets are off. I don't see it as verification, Larry, but characterization. The prototypes of the Mk I & II internal noise gens were in fact characterized here, and then each unit during production by Dick with his Anritsu VNA. Isn't independent testing de rigueur for any new design of an instrument? If Jove is to recommend something, Jove should have numbers to back up whatever it's recommending. I /think/ the new calibrators will be just fine, far better than ±1 dB. But we won't know that with certainty until they're measured against a 5722 and put on the spectrum analyzer. I think it's a GOOD thing we have more that one design to characterize. This will give us a better real-world envelope of expected performance. -- Dave On 5/10/23 18:49, Larry Dodd wrote:
Does this mean that everyone that builds one will have to ship it to Dave for verification? I wish I would have just built the prototype only. I’m thinking I should withdraw the Beta #1 version and Jove can go with the prototype to simplify and speed up the project. Verification didn’t seem necessary with the MK I or II we just accepted it as good enough. We also have to ask why no one else took the initiative to build one. |
|
Dave, Characterizing the noise generator is one thing but characterizing the noise source, SDRuno, RSP1A and the calibrator together as a system is another. That's what Jim Sky's new calibrator software does. And that's a standard everyone has and can use.?Since the Beta is more expensive and perhaps more involved to build I think it best to withdraw the Beta version then there is only one choice of which to build. No chich means standardizing the process.? Seems the simplest and best solution to me. I'm sorry now that I built something a little different?from the prototype. What happened was discovery of all the noise from the Ardurino UNO R3 had to be solved. That's why?I tried the optoisolator and separate voltage regulators. Jim says changing to the cheaper Nano solved all that.?Eliminating?the Beta version confusion seems the next best step.? Larry
|
Larry Dodd
Dave Typinski