5uH Aerospace LISN: How dumb would I be to "throw one together"?

• #50 Reply
  Posted by TimNJ on 21 Aug, 2020 22:35
• Today, I measured the magnitude of the impedance (s11) and plotted in Excel, using NanoVNA V2.2. I ran two sweeps, two calibrations...100KHz-200MHz, and 200MHz-400MHz. (Originally just swept 100KHz-200MHz, forgetting that the DO-160 upper frequency is 400MHz, so I just ran a separate sweep and appended the data.)
  
  I tried to get the reference plane as close to the circuit as possible by soldering an SMA directly between  DUT_line and earth. Perhaps more accurately would be to measure the impedance at the actual power port on the enclosure. However, didn't drill the holes out yet, and also the "true" impedance of the network of the circuit gets skewed by even a little inductance...as I've been told. (Thanks Jay_Diddy_B.)
  
  Anyway, the impedance seems very well behaved when measured @ -10dBm. Definitely within the limits. I am not sure how one would characterize it near its true operational power.
  
  Please see the attached plot and measurement setup. Since this was my first ever VNA measurement, I'm hoping someone can give me some feedback. If it looks okay, I'll share the files.
  
  I've also seen people measure insertion loss of LISNs...but I do not understand its relevancy, nor do I see any sort of official specification.
  
  Thanks!

  ---

  

  
• #51 Reply
  Posted by Jay_Diddy_B on 22 Aug, 2020 00:41
• Hi,
  
  I think you can take advantage of the fact that you have two identical LISNs.
  
  Connect them like this:
  
  
  
  
  The two LISNs are connected back to back. Bias current can be passed through the inductors using a bench power supply.
  
  You can measure the impedance at the mid point using a blocking capacitor.
  
  The impedance that you measure is the two LISNs in parallel. This is easy to correct for, multiply the result by two.
  
  
  I have not tried this, but the theory is good.
  
  I might try it on some low current LISNs which are easier to saturate the cores.
  
  Regards,
  Jay_Diddy_B
  
  
  

  ---

  
• #52 Reply
  Posted by Jay_Diddy_B on 22 Aug, 2020 01:54
• Hi,
  
  I tried this in the lab with a 50uH LISN. The inductor used have a inductance versus current like this:
  
  
  
  
  If I test the LISNs connected back-to-back, in parallel, I get this at zero bias current:
  
  
  
  
  If increase the bias current to 9A, I get:
  
  
  
  
  This should be the same as two 10uH // 50 LISNs connected in parallel.
  
  This is because the inductance of each inductor has changed from 10uH to 2uH and there are five in series.
  
  At lower currents, less than 5A, the bias current has very little effect.
  
  Regards,
  Jay_Diddy_B
  

  ---