On 3/2/22 7:07 PM, montanaaardvark wrote:
On Wed, Mar 2, 2022 at 08:13 PM, WD0AKZ wrote:
#1 VNA w/Load Calibration Load: SWR= 1.001 LogMag= -66.3db X= 5mOhm Z= 50.045
#2 VNA w/Load Calibration Load: SWR= 1.000 LogMag= -87.0db X= 10.0mOhm Z=
49.998
What I¡¯m concerned about is the differences between the LogMag (ReturnLoss)
and Z (Impedance) on the Open and Short tests. I would also think that a -21db
difference in LogMag between #1 and #2 on the Load test is a bit excessive, or
is this just considered ¡°Close enough for Government Work¡± (old saying,
somewhat military speak)?
The way I read this you used the exact same cal standards on both NanoVNAs, right? Standard practice for calibration in the labs I've worked in was to clean both connectors with denatured alcohol and use a calibrated (7 inch*lb) torque wrench for tightening the sets. For every measurement. I'm going to assume you didn't do this and just put them together by hand.
The way I look at this is to turn the return loss values from dB into magnitude by doing 10^(dB/10). I get the values:
(-66.3 dB) = 234 *10^-9
(-87.0 dB) = 2.00 *10^-9
I don't have much of a feel for how much to expect out of the H4 since I just bought mine, but while that's a big relative difference (234 to 2), it's between two rather small numbers. How many bits difference is that at the ADC? I have no idea. Look at the VSWRs: 1.001:1 vs. 1.000:1. I can't think of anything I've ever done where a return loss difference like that mattered.
I agree - the important thing isn't the 20 dB difference. That's down in the noise floor (i.e. you're looking at the difference between a reflection coefficient of 0.0005 and 0.00004).? That is an impressively good load. Or, more probably, you're seeing the noise floor of the calibration.? The mate/demate uncertainty of an SMA is probably greater than that.
