Negative real impedance? (HFSS design)

[lechko]

Hi everyone, im new to this group so i hope you can help me...
I've been working with HFSS for a while now, designing RF coils for MRI. The structure is pretty simple - a single turn coil with a coax feed. the feed is blocked by a copper block on its port end, and the inner coax connects to the coil... (coil radius=7cm) the problem is that i keep getting negative real part for my impedance in low freq. (around 100MHz).
Happend to any of you? what's that supposed to mean in HFSS?

thanks in advance
Alex

 

[Higgler]

Negative real part, that's an error. Imaginary part is + or -, so maybe the data output is confusing you.
Impedance on electrically small antennas will be very high and very real.

Do you work for GE. Alot of their MRI's are at 63 Mhz.

kch

 

[biff44]

Been a coon's age since I used HFSS, But I do remember it being very fussy about some basic stuff, like making sure all the pieces of copper actually connect (not a thin unconnected gap anywhere), and being pretty fussy about how you draw your boundaries. You are going from a coaxial 50 ohms to free space wave impedance of 377 ohms. If there is anything wrong in how you tell it to do that, it might interpret the wierdness as gain (negative resistance). In the real world, of course, without some source of power (transistor, adjacent coil pumping in energy, etc) you can not have a negative resistance.

I know it is not much, but good luck.

 

[lechko]

Thanks for the info, that was helpful. Actually im connecting the coax to the coil just by cutting it and putting the end so that it touches the single turn coil... should this make any problems? By the way, i'm putting a lumped element in the end of the feed just by cutting a gap in it and putting a geometry with lumped element boundries... you said gaps are a problem in HFSS, is it this sort of gaps, and if so, do u have another idea of how to implement the contact and the lumped element?


Thank you all very much!

 

[biff44]

No, a deliberately placed gap is fine. I was talking about, say, going from 2 different width lines. Since you are "drawing" these lines in a CAD drafting program to input them, if for some reason the two lines do not overlap, and there is a microscopic gap, then all hell breaks loose. Maybe they have done something in the newer versions of the program to spot these, but we used to have to really zoom in at any transistion and make sure there was metal to metal overlap in the drawing.

 

[lechko]

thanks, that was helpful. i do have a transition from a 0.5mm inner coax to an infintisimal width single turn coil, i'll try to make the transition softer. one more thing... i made the coaxial line deliberately shorter than 1/4 wavelength since it's thin and the high modes are suppressed, i've heard that even if the modes are fine, HFSS tends to make problems just over the sheer length of the feeding coax... any truth to that?

Thanks again,
Alex