#78: Building VNA Calibration Loads

[@WestCoastMole]

Thanks for the hints snd tips Larry. They are very helpful

[@PixelSchnitzel]

Thanks, Larry! I've had some success in building OSL sets, but this helps a lot. I'm going to build some new ones now and also modify/remake a couple existing sets. My female N set, in particular uses a commercial terminator as the load which has a totally different plane than the open & short, which explains why it starts sucking around the high-HF and low-VHF ranges. I've also been working on an NMO set that I think will be pretty good.

[@w0qe]

If your system was exactly 50 ohms from the VNA through any cables and adapters and including the connector used for the 50 ohm load the reference plane for the 50 ohm load would not be important as the VNA would always see 50 ohms. However it is much safer to make the 50 ohm load have the same reference plane as the open and short. Many commercial terminators are not that precise and have reasonable SWR's for the intended purpose BUT are not good enough for the calibration load. You can make a better terminator. Larry

[@peterschmelcher2754]

Nice video. It is easy to over think a problem. The load became my hangup. I was thinking of grinding and tapering the width of surface mount resistors so that the radial voltage drop would match the radial electric of the traveling wave. Thanks Peter

[@bob456fk6]

Very interesting, Larry. I've measured your loads and found them very good. No need to pay a high price for multiGHZ loads that are overkill for ham work. -- W5BIG

[@w0qe]

Hi Bob, And thank you for designing the original AIM4160, AIM4170, VNA2180, AIM4300 etc. and hope retirement is going well for you. As a community we need reasonably priced equipment to be able to make measurements to order to understand RF concepts. I hope my videos encourage others to do exactly this. 73, Larry, W0QE

[@willhelmx8388]

Very nice video Larry, good explanation on construction, i tend to mill/grind my connectors to an exact reference point on o/s/l wich normally works fine and as you mention it´s quite important that you start out with some good quality connectors. 73´s /Anders

[@AllElectronicsChannel]

Very good! Have you ever tried to measure aliexpress N/SMA connectors? I'm using them go play around with microwave circuits.

[@w0qe]

No I have not tried to measure them. The concern would be if they are all the same dimensionally. If not you would need VNA software that would allow you to specify this and would need to know these differences. Also how accurate at DC and RF is the 50 ohm termination. I would also be concerned if what you buy today would be the same as what you buy in 6 months. Larry, W0QE

[@AllElectronicsChannel]

@@w0qe Thank you! They are really cheap.. but sometimes even the threads feel chinese.. hahaha Almost all my gear I bought used.. If it was new I think I would never thread chinese connectors.

[@wa4aos]

Thanks Larry, I recently purchased a VNA but cringed at the prices for OSL's. Most of my work will be under 500 MHz and your approach will exceed my needs. 73, Glenn WA4AOS DSM Labs

[@w0qe]

Glenn, Glad that I could have helped. Larry

[@andrewbaerm.d.3984]

Hugo Gernsback, best known as the Father of Science Fiction for whom the Hugo Award was named, was my grandfather. RE was his publication. What most Amatuers don't know is that he was intrumental in promoting both commercial and amatuer radio in their early days.

[@ramjet4025]

W0QE, Great stuff. What SIZE of 100 Ohm chip capacitors fit an SMA or N-connector? How large a value can you go on SMA and N connector?

[@galileo_rs]

Amazing results especially at 3GHz when you check the specs for the SWR of the connectors/adapters themselves. Can you upload the measurement files somewhere like github etc so that we can play with them in SimSmith?

[@w0qe]

Dragan, I don't know how to get the underlying data from my 3GHz VNA. I can get the plots via GPIB easily but not the data. Let me do some looking around and see if I can come figure out how to get the data. Larry

[@galileo_rs]

I did a quick search and found a few utilities that can do that, from matlab files to standalone utilities like this: www.ke5fx.com/gpib/readme.htm It would be interesting to see full scans but files from AIM would also work for what I had in mind.

[@w0qe]

I sent you the files. Larry

[@oneman7039]

hey does the resistor package size affect the optimal configuration (two parralell 50 ohms or something else)?

[@w0qe]

Yes, the resistor package does affect what is optimal BUT a larger package has more inductance and less capacitance. These 2 differences are in the direction of cancelling each other so either an 0805 package or the slightly larger 1206 or the slightly smaller 0603 perform nearly identically. Larry, W0QE

[@tomthompson6911]

Good, Larry!

[@w0qe]

Thanks Tom!

[@carouselradio6643]

Can you share the SimSmith files that you used for this video? Thanks & 73

[@w0qe]

Files uploaded to www.dropbox.com/home/W0QE%20KZbin%20Files Larry, W0QE

[@andrewbaerm.d.3984]

Can you give me a source for the 100 oum resistors and the part number. Thanks

[@w0qe]

Andrew, I'm sorry but I don't have the actual part numbers I used and once they have been separated from the reels there are no identifying marks to identify them. I realistically have several million SMD resistors that I have accumulated over the years. However here are some guide lines that should help. All SMD resistors from reputable suppliers are all very much alike. For a given physical package size small differences in series inductance due to trim methods such as balanced edge sense, L-cut, and S-cut have some effects on very low value resistors but are not an issue for 100 ohm parts. Likewise termination styles such as wrap and flip have an effect on the shunt capacitance but again at 100 ohms are not an issue. I use the SMD part size that best fits the connector (1206 for N connectors and 0603 for SMA) and measure the resistance value of parts so that when paralleled they will be 50 ohms +/--.5% which is easy to do by sorting a few parts. You can of course buy parts that are 0.1% that are still pretty inexpensive if you don't have access to a good ohm meter. Also the typical 200ppm temperature coefficient if fine but parts can be bought with a 25ppm temp. coefficient if you intend to do calibrations over a wide temperature range. Does this help? Larry, W0QE

[@andrewbaerm.d.3984]

@@w0qe Larry, thanks. I do not do a lot of electronics work so I don't have a lot of parts lying around. But I do have an account with Digikey so I should be able to get the resistors. BTW, have you ever heard of a no longer extant magazine called Radio Electronics?

[@w0qe]

Andrew, Sure I remember the magazine and read several issues of it bit never had a subscription. Larry

[@WaschyNumber1]

🖖 👍