Dismiss Notice
Join Physics Forums Today!
The friendliest, high quality science and math community on the planet! Everyone who loves science is here!

Composite panel

  1. Mar 13, 2008 #1
    In testing the loss through a fiberglass panel I experienced instances where the received power level increased. I have checked and double checked my setup and changed the distances between the horn and panel, I still am seeing odd results. Has anyone had any experiences such as this. I am assuming that the rise is caused by reflections. Any thoughts?
     
  2. jcsd
  3. Mar 13, 2008 #2

    berkeman

    User Avatar

    Staff: Mentor

    Do you have the dielectric constant for the fiberglass panel? Would the reflections match what you would get for the reflection coefficient for that thickness of dielectric in free space at your frequency?
     
  4. Mar 13, 2008 #3
    I don't recall what the dielectric constant is. I am testing at 16 GHz. However, yes the reflections do match what the reflection coefficient would be. signal is increasing at some angles by 0.1- 0.2 db. loss in the panel should be no more than 0.3 db, which is what the results indicate. I just can't explain the increase.
     
  5. Mar 13, 2008 #4

    Danger

    User Avatar
    Gold Member

    This has really caught my attention and made me very curious... because I don't know what the hell you're talking about. The term 'horn' makes me think of microwaves, for some obscure reason. Can one of you give me a (very brief) 'square-one' summary of what the situation is so that I can make sense of the conversation?
     
  6. Mar 13, 2008 #5
    I am testing a composite panel in an anechoic chamber. I have a transmit antenna at one end and a receive antenna(horn) mounted to a positioner at the other end. The chamber is about 20 feet long. I record a reference power level. After recording the reference I place a flat composite panel on the positioner in front of the receive antenna. I then rotate the panel through a series of angles and record the loss at different angles. However, at some angles I am actually receiving an increase in signal of about 0.1-0.2 db. This result seems odd to me. I have tried to change the distance between the horn and panel and got similar results. I have a theory that maybe the increase is caused by reflections that are in phase with the received signal and are adding to the received level, but have no idea how to get rid of them.
     
    Last edited: Mar 13, 2008
  7. Mar 13, 2008 #6

    Danger

    User Avatar
    Gold Member

    Okay. Thanks; that makes more sense now. Are you dealing with EM then, or ultrasonics? Sounds suspiciously like you're planning to 'stealth' your car to avoid those nasty speed traps. :biggrin:
     
  8. Mar 13, 2008 #7
    I am dealing with EM the panel will ultimatley be a radome. That is if I can get some data that makes more sense. I am a student soon to be engineer someday and have looked in the IEEE archives and other handbooks but have found nothing on the topic which makes me believe that I am making a measurement error
     
  9. Mar 13, 2008 #8

    Danger

    User Avatar
    Gold Member

    Since I don't know anything about the subject, I'm going to toss out one of my 'ignorance is bliss' comments. Is there a possibility of contaminants in the fibreglass that might alter its properties?
     
  10. Mar 13, 2008 #9
    No, the fabrication process is strictly controlled. I don't think that it's a problem with the panel. I think that the problem is in the test. I just haven't been able to isolate it yet. My next idea is to simply test the panel outside the chamber using a different setup. I'm thinking of just a transmission test with a network analyzer. This is really the only other idea I have right now.
     
  11. Mar 13, 2008 #10

    Danger

    User Avatar
    Gold Member

    I just checked back and noticed that you had added this part in an edit. Have you tried changing the distance between the emitter and the panel rather than just between the panel and the receiver? Wouldn't that eliminate constructive interference?
     
  12. Mar 13, 2008 #11
    Yes, I agree. I have moved the panel towards the transmitter. The standard test is moving by 1/4 wavelength and then averaging the results. I still come up with the same results of an increase in received power. I think that it is obviously a testing error. Otherwise I have created a panel that increases the signal level of an antenna and I'm going to take everyone out to dinner.
     
  13. Mar 14, 2008 #12

    Danger

    User Avatar
    Gold Member

    :rofl:
    Okay, I'm going to take one more shot with an idea which is probably utterly useless. (One great thing about being clueless is that I can say any stupid thing that crosses my mind without fear of embarrassment. :biggrin:) Is there any possibility that your panel is acting as a 'beam splitter' (prism)? If so, could the two 'daughter' beams then interact within the collector horn? (That would seem to violate thermodynamics, but what the hell...)
     
    Last edited: Mar 14, 2008
  14. Mar 14, 2008 #13

    berkeman

    User Avatar

    Staff: Mentor

    Yeah, I don't think you should be able to get an increase over the free space value, at least not without some *other* multipath contribution. Look for other non-anechoic paths between your transmitter and your receiver -- how good is your anechoic chamber?
     
  15. Mar 14, 2008 #14
    I don't know when the chamber was last examined, but I do know that it has a flat field of about 5 feet. I am well inside of that. It is completely enclosed and lined with absorber. I am going to give it another whack tomorrow morning and try some different setups. If all else fails I'll have to ask some old salty radar engineers that I know. Usually they have plenty of suggestions.
     
  16. Mar 14, 2008 #15

    berkeman

    User Avatar

    Staff: Mentor

    I'm not so worried about a flat field volume, I'm thinking more of other stuff in the back corners of the chamber that might be candidates for re-reflecting energy that the panel is directing off-axis when you tilt it, where the re-reflection somehow makes it back into the RX horn. Seems a bit unlikely, though.

    Have you tried making the panel small, so that it only reflects the energy that is aimed directly at the RX horn? That would help to cut down on the possibility of other reflected energy (from farther out on the panel area) being reflected a couple times in the chamber and coming back into the RX horn via multi-path. Just a thought.
     
  17. Mar 14, 2008 #16

    dlgoff

    User Avatar
    Science Advisor
    Gold Member

    I'm wondering if the composite might not be homogeneous. Could it be that there are strips or large threads in the fiberglass cloth embeded in the panel? Causing some sort of lensing (fresnel diffraction) which is capturing radiation that would not have been received (i.e. outside of the receivers horn)?
     
  18. Mar 16, 2008 #17
    Problem solved. I changed the test fixture and made it as small as possible and added more absorber to try to get rid of any reflections that might be occuring. I got the results that were expected and was able to repeat the results. Thanks for all the help and suggestions.
     
  19. Mar 16, 2008 #18

    Danger

    User Avatar
    Gold Member

    Glad to hear that it worked out well. Don't be a stranger just because your problem is solved; we like people to stay in touch. :smile:
     
  20. Mar 16, 2008 #19

    RonL

    User Avatar
    Gold Member

    Kinda sounds like something Jerry Lee Lewis might have been working on when he decided to sing "Whole Lotta shaking going on".

    :biggrin:
     
  21. Mar 29, 2008 #20
    I believe you are correct in that the radome "gain" is caused by reflections adding contructively to the received pattern. These reflections can be coming from the transmitted wave being bounced off the chamber walls or radome/antenna fixtures. Another path for the unwanted signal is coming from the transmitter, bouncing off the receive antenna or fixture, then bouncing off the radome backside, then finally into the antenna. The 1/4 wave shift method attempts to cancel out the latter reflection path by shifting the reflection 1/2 wavelength by the time it enters the antenna. If you only shift the antenna 1/4 wavelength, then you will only be cancelling out reflections from the antenna. If you shift the radome, then it should cancel reflections from both antenna and fixture. If the reflections are coming from the walls or fixtures directly into the antenna I don't think the 1/4 method will make any difference. In this case you might check the integrity of your chamber absorber and fixture absorber. You did not mention the type of transmit antenna. If the transmit antenna is broad beam, then you're more likely to have reflections from the ceiling, floor, and sides of the chamber. You may help this if you obtain and use a narrow beam dish transmit antenna pointed directly at the receive antenna. If your open air patterns don't look like the actual antenna pattern, you've got something wrong with your basic chamber and test set-up. Hope this helps.
     
Know someone interested in this topic? Share this thread via Reddit, Google+, Twitter, or Facebook

Have something to add?



Similar Discussions: Composite panel
  1. Photovoltaic panels (Replies: 2)

  2. Panel Meters (Replies: 1)

  3. Solar panel (Replies: 8)

  4. Solar panels. (Replies: 1)

Loading...