AnsweredAssumed Answered

Ecal Module Calibration Interval

Question asked by G0HZU on Nov 2, 2015
Latest reply on Nov 11, 2015 by G0HZU

Like • Show 0 Likes0
Comment • 7

I've just upgraded to an Agilent 4 port VNA and it came with a 4 port Ecal module N4431B-60006 (13.5GHz)

The Ecal module shows a 1 year cal interval and this expires early next year. The VNA has a 2 year calibration interval with about 1 year left. I have no idea how much it costs to calibrate the Ecal module but it will almost certainly be too much for me to afford/justify on a yearly basis. I bought the VNA and Ecal kit together as an auction lot.

So does anyone know how well these Ecal units hold calibration over an extended timeframe? When they go for calibration do they get new calibration data stored in them or are they just checked for compliance inside a cal window?

Thanks in advance for any replies...

No one else has this question

Outcomes

7 Replies

jvall Nov 7, 2015 10:40 AM
Mark Correct
Correct Answer
Well, much may depend upon what sort of accuracy you require. If you don't really need the full accuracy of the published ECal spec, then you may be able to be less rigorous with the timing. Another factor is how often it is used (due to connector wear.)

Providing these two pieces of information would help in providing a good answer.
Like • Show 0 Likes0
Actions
- G0HZU @ jvall on Nov 8, 2015 2:40 PM
  Mark Correct
  Correct Answer
  I don't need the full accuracy a freshly calibrated Ecal can give but I'd want it to be able to maintain performance at least as good as my homemade SMA cal kit up to a few GHz.
  
  I don't intend to use the Ecal as much as my home made cal kit as most of the time I'll be making measurements below 3GHz and I already get good enough performance from my own kit up to 3GHz. I also don't have to worry about connector wear with my home made kit as I can just make another.
  
  The Ecal is great for doing 4 port calibrations because of the speed/convenience. It's already making me lazy :) I guess I'm wondering how far these things drift over time if they aren't calibrated for a few years. I'm especially interested in how stable they are across 6-13.5GHz as my home made SMA cal kit is not going to work well up here.
  
  It's hard to predict how much I will use it but maybe 300 connections per year per SMA would be a rough guess?
  
  Edited by: G0HZU on Nov 8, 2015 3:37 PM
  
  Edited by: G0HZU on Nov 8, 2015 3:39 PM
  Like • Show 0 Likes0
  Actions
  - jvall @ G0HZU on Nov 10, 2015 9:38 AM
    Mark Correct
    Correct Answer
    Thanks. I will try to have our Calibration standards expert reply.
    
    >It's already making me lazy
    Using an ECal is like using a garage door opener; once you use it, you *never* want to go back!
    
    Our expert is on business travel but I sent him an email so hopefully he will reply sometime soon.
    Like • Show 0 Likes0
    Actions
    - G0HZU @ jvall on Nov 10, 2015 3:40 PM
      Mark Correct
      Correct Answer
      My cal kit is something I made many years ago and I use it with my own custom user cal file. Maybe it would be useful to show that my cal kit gives fairly good performance despite being home made.
      
      To compare the Ecal against my cheapy homebrew cal kit I tried doing the 20dB attenuator test found here
      
      http://www.kirkbymicrowave.co.uk/
      
      You can see in the image below that my homebrew cal kit is OK up to a few GHz but the plots for S21 and delta phase (wrt the Ecal kit) go very ripply above this. I did a full 2 port cal ( 2port reflection and thru) in both cases and then just put the attenuator inline.
      
      I have been using the homebrew cal kit for many years to help me model various n port components. I've compared it numerous times against 85033E cal kits and it gives very close agreement when measuring SMD components in a test fixture
      
      It has been very successful but it's also very tedious to set up the calibration with my existing cal kit. Especially for 3 or 4 port components.
      
      [URL=http://s1227.photobucket.com/user/G0HZU/media/20dB.gif.html][IMG]http://i1227.photobucket.com/albums/ee439/G0HZU/20dB.gif[/IMG][/URL]
      
      My concern with the Ecal kit is that it may drift out of spec enough to make it of limited use. Presumably there's a way to correct this and store the correction data inside the Ecal unit but I guess only Keysight can do this.
      
      Edited by: G0HZU on Nov 10, 2015 4:16 PM
      
      Edited by: G0HZU on Nov 10, 2015 4:38 PM
      Like • Show 0 Likes0
      Actions
      - G0HZU @ G0HZU on Nov 10, 2015 3:34 PM
        Mark Correct
        Correct Answer
        See also below an old comparison plot of my homebrew cal kit against an 85033E. This is an S11 plot of a combline BPF at about 1100MHz.
        
        This was actually a comparison of my VNA + homebrew SMA cal kit against a VNA at work using the 85033E so it isn't a very well controlled test but the results were still very good. The two plots almost overlay each other perfectly.
        
        [IMG]http://i1227.photobucket.com/albums/ee439/G0HZU/comb1100.gif[/IMG]
        
        Edited by: G0HZU on Nov 10, 2015 4:32 PM
        
        Edited by: G0HZU on Nov 10, 2015 4:33 PM
        Edited by: G0HZU on Nov 10, 2015 4:33 PM
        Like • Show 0 Likes0
        Actions
kenwong Nov 11, 2015 1:55 AM
Mark Correct
Correct Answer
The ECal's calibration accuracy specification is based on one year's expected usage. That is 2 calibrations per day for one year of working days., 52x5x2 connections and disconnections. Connector wear is the dominant cause of ECal's 1 year re-calibration requirement in order to maintain the specifided calibraton accuracy. For expert users that took care of the conectors, the calibration can last much longer, more than 2 years.
Like • Show 0 Likes0
Actions
- G0HZU @ kenwong on Nov 11, 2015 3:46 PM
  Mark Correct
  Correct Answer
  Hi Ken
  Thanks for the helpful reply. I now think I'm going to spend some time improving the user cal file for my current homebrew cal kit and maybe also make a fresh cal kit as a stored reference that will only get used as a cross checking device.
  
  This way I can limit the usage of the Ecal unit to special occasions once I've finished charactersing my various test fixtures with the new VNA and Ecal module. I'm also hoping the stored reference kit can be used as a (flawed) calibration signature to show any change/drift in the Ecal unit. I also have access to about 6 85033E 3.5mm cal kits at work and also a few Ecal modules to do cross checks at work.
  
  My previous VNA only covered up to 3GHz and I found that the homebrew cal kit was fine for my needs. I guess I'll have to get the Ecal unit calibrated at some point in the future but hopefully not for several years. I guess the experiment to see how well it holds calibration over this timeframe begins now... :)
  
  Edited by: G0HZU on Nov 11, 2015 4:43 PM
  
  Edited by: G0HZU on Nov 11, 2015 4:45 PM
  Like • Show 0 Likes0
  Actions

Ecal Module Calibration Interval

Outcomes

Related Content

Recommended Content