This is a copy of N2PK's website. Paul has decided to delete his website (See N2PK-VNA on Groups.io message #5, Jan 2020) from ~Feb 2020.
Paul has made availalable a ZIP file of his entire web site, which has been uploaded as is, no changes made except for this preface. <vk3pe>
Contact Paul at https://groups.io/g/N2PK-VNA| This correlation activity was provided by me using my VNA and by
Chip Owens, NW0O, using an HP (now Agilent) 8753C VNA. The common Device Under Test (DUT) was a series RC load, using surface mount components soldered directly to an Amphenol SMA jack - the same type of jack as used for my homebrew calibration standards. The components are an 0805 SMD 24.9 ohm 1% resistor in series with two 0805 100 pf 5% NPO capacitors in parallel. This DUT and its mounting, in addition to its precision and stability, was selected to minimize stray series inductance, which you will see is still discernible & quantifiable at approximately 1.5 nH. I used three homebrew SMA calibration standards for my tests. Chip used the HP85033C 3.5 mm calibration kit, which has tracability to the US National Institute of Standards and Technology (NIST). There was a reference plane offset of 34.2 ps between my VNA calibration standards and Chip's, as determined by his measurement of my short at 1 GHz. This offset was used as needed to correct one or the other set to permit careful comparisons of data at the same reference plane. Two comparisons are made. The first uses the measured complex reflection coefficients, expressed as a magnitude and an angle at each test frequency. In this case, my reflection data was adjusted to the front of the SMA jack to correspond to Chip's reference plane. location. |
| |rho| | angle rho, deg |
| MHz | N2PK VNA | HP8753C VNA | Delta | N2PK VNA | HP8753C VNA | Delta |
| 1 | 0.9961 | 0.9970 | -0.0009 | -7.15 | -7.19 | +0.04 |
| 10 | 0.7641 | 0.7643 | -0.0002 | -60.82 | -60.80 | -0.02 |
| 20 | 0.5541 | 0.5537 | +0.0004 | -94.70 | -94.72 | +0.02 |
| 30 | 0.4575 | 0.4565 | +0.0010 | -114.85 | -115.07 | +0.22 |
| 40 | 0.4102 | 0.4107 | -0.0005 | -128.24 | -128.27 | +0.03 |
| 50 | 0.3846 | 0.3838 | +0.0008 | -137.72 | -137.90 | +0.18 |
| 60 | 0.3695 | 0.3687 | +0.0008 | -144.79 | -144.90 | +0.11 |
As you can see the correlation for both magnitude and angle of the reflection coefficient at all test frequencies is extremely good. For example, at 1 MHz the difference in |rho| is equivalent to a 0.0078 dB difference in return loss at a VSWR of about 500:1! The second comparison uses the measured complex impedances at each frequency, but expressed as a series RC, Rs and Cs, which should approximate the nominal component values of 24.9 ohms and 200 pF at low frequencies. In this case, the HP8753C reflection data was adjusted to the back of the SMA jack to correspond to my reference plane, so that the connector parameters are not part of the DUT. |
| Rs, ohms | Cs, pF |
| MHz | N2PK VNA | HP8753C VNA | Delta | N2PK VNA | HP8753C VNA | Delta |
| 1 | 25.20 | 19.25 | +5.95 | 198.6 | 199.3 | -0.7 |
| 10 | 24.98 | 24.97 | +0.01 | 199.2 | 199.1 | +0.1 |
| 20 | 24.96 | 24.97 | -0.01 | 199.9 | 200.1 | -0.2 |
| 30 | 24.97 | 24.98 | -0.01 | 201.1 | 202.1 | -1.0 |
| 40 | 24.98 | 24.95 | +0.03 | 202.9 | 202.9 | 0.0 |
| 50 | 24.98 | 25.00 | -0.02 | 205.0 | 206.2 | -1.2 |
| 60 | 24.97 | 25.00 | -0.03 | 207.8 | 208.6 | -0.8 |
The only really significant difference in Rs occurs at 1 MHz. Based on the components used in this DUT and the relative trends of both sets of data, I would submit that the larger source for the discrepancy is an error in the HP8753C reflection coefficient return loss, earlier noted to be only 0.0078 dB different which is extremely small. For Cs, the general trend is an increasing value for both sets of data with frequency. As noted above, this apparent increase in capacitance with frequency is actually due to inductances in all three components. Adding a single series inductance of 1.5 nH to my VNA data and accounting for its impedance results in no change in Rs, but now Cs' = 198.9 +0.2 /-0.3 pF. Applying the same procedure to the HP8753C data results in Ls = 1.6 nH and Cs' = 199.2 +0.5/-0.4 pF. With this refinement to the equivalent circuit, the two nominal capacitance values agree to 0.3 pF or 0.15% with a worst case deviation of about 0.5% from 1 to 60 MHz. It's also worth noting that 1.5 nH is an extremely small inductance for HF work, yet this VNA is able to accurately measure it. Back |
Please feel free to contact me with any comments or questions
at: See above.
73,
Paul Kiciak, N2PK
