开云体育

Hello,
I SOL calibrated my NanoVNA using a 30cm cable + F to F SMA coupler and the SOL end pieces. I removed the coupler and attached the 30cm cable to a ~5cm 50 ohm coax cable soldered to the input of a matching network which goes to a 900MHz Igneon PCB antenna. When using the tip of a knife to short the soldered-on end of the coax cable I need to use -205ps offset delay to bring the majority of the points to the left edge of the Smith Chart. Positive values make the plot worse. Is it ok to use a negative number with the NanoVNA Saver? That seems counter intuitive.

Christopher,

in air—strictly vacuum—205 ps is equivalent to about 60 mm, which is not so far off the length of the F-F adapter that you had removed.

The numbers, when all is going well, tend not to lie.

HTH, 73,

Robin, G8DQX

Hello Robin.
Thanks for the response. The F-F adapter 15mm long. It was replaced with about 50mm of transmission line in the form of a female SMA with 50 ohm coax cable. My question is about using a negative value for the offset delay. +200ps looks terrible on the Smith chart, worse than no delay, whereas -205 creates a fairly tight cluster of points across the swept frequency. I'm fairly new to NanoVNAs and the NanoVNA Saver Windows app but all the articles and comments I read on Offset Delays are using positive values which to me imply compensating for a longer cable than what was calibrated. Is my head in vacuum?
Chris

Normally you try to calibrate at the reference plane where you want to take the measurements, and the delay is set to 0.
The e-delay is a method of adjusting the position of the reference plane before making a measurement, and could be positive or negative - but the most common use is positive.
If you have an additional known length of cable/connector between your calibration point and the device you want to measure, you would enter a positive delay to represent that additional length.
If you calibrate and then remove a piece of cable, making the measurement plane closer than your calibration, then you can put in a negative delay to compensate for that length of cable that is not there anymore.
Note that e-delay does not do a perfect translation of the reference plane: it only compensates for time (i.e. phase and delay), not other impairments like loss in the cable. So calibration at the measurement point is preferred, especially for longer cables that could be lossy at the measurement frequencies.

Hi Stan and others,
Thanks for the reply. Based upon my photos from earlier in the thread does a negative offset delay seem reasonable. It doesn't to me, but my experience is limited.
Best,
Chris