Calibration Loads

This contains the scottyspectrumanalyzer yahoo group backup
Post Reply
Sam Wetterlin
Posts: 646
Joined: Sat Apr 28, 2018 6:48 pm

Calibration Loads

Post by Sam Wetterlin » Sun Apr 05, 2009 2:34 am

I may be jumping ahead here a bit, but .... When it comes to using the VNA features of the MSA for reflection measurements, you will need to calibrate with an Open, Short and Load (and have the necessary OSL calibration software, which I don't think is far off). The Load is a connector of the appropriate type (N male, female; SMA male, female, etc.) with a 50 ohm connection to ground on the backside. You can easily pick them up on eBay for $200-300 apiece. Ouch. Or you can make them yourself. Homemade Loads on SMA connectors have tested very well, but are a bit fragile. Mechanical stress from pressure on the center pin can crack the resistor(s) on the connector backside.



Here's an alternative idea: Buy commercial 50-ohm terminations. They have good frequency response in the sub 1-GHz range, but their DC resistance is not always as close to 50 ohms as one would like for a calibration standard. If they are cheap enough, you can buy 4 and measure the DC resistance. Hopefully one is fairly close to 50 ohms. Throw away the other 3 and you have your calibration standard.



There are now available on eBay many 50-ohm terminations for both male SMA and male N. That's because the economy sucks. About $2 each for SMA; $7 each for N. I bought a bunch of SMA Huber + Suhner terminations and measured their DC resistance. None of that 4-point Kelvin hassle. Measure a precision 49.9 ohm resistor with whatever ohmmeter you have, and a 50-ohm resistance should measure 0.1 ohms more than that. My ohmmeter measures a 49.9 ohm resistor at 50.2 ohms, so my goal is a termination measuring 50.3 ohms. If the measurement shows the actual termination resistance is in the 49.9-50.1 ohm range, you have an excellent calibration load with return loss in the 60+ db range.



What about female loads? Well, attach the male loads to a bullet (threaded) adapter, and you will have a female load with return loss pretty much equal to that of the bullet. Not ideal, but darn good.



My collection of eBay male Huber + Suhner SMA terminations produced about 50% with resistance in the 49.8-50.2 range. Those are very good. And about 25% in the 49.9-50.1 range. Those are excellent.



The eBay Huber + Suhner SMA terminations also have another advantage: when you attach SMA connectors, you are not supposed to allow the body of the male connector to rotate during the final tightening. That causes wear on the mating surface. You want the "nut" part to rotate but the body to remain still. This is natural behavior with coax cable connectors, but many male SMA terminations are designed as one piece, so the mating surface has to rotate. Not so with the Huber + Suhner terminations.



I haven't yet tested most of the N terminations, because they haven't arrived yet. I should get a good sample, because I accidentally bought twice as many as I wanted. I have tested only one N Load from a prior purchase, and its DC resistance is as close to 50-ohms as I can measure. I'm expecting that the NARDA and MECA N-Terminations now available on eBay will be very productive. Their specs for frequency response are excellent, so if I select the ones with good DC resistance I should be in good shape.



Bottom line: check out eBay by searching for "termination". There are some good opportunities there. There are several listings of lots of 5 or 10, with many lots available.



Sam W.

Post Reply