The Designer's Guide Community
Forum
Welcome, Guest. Please Login or Register. Please follow the Forum guidelines.
Nov 24th, 2020, 2:00am
Pages: 1
Send Topic Print
Inductor Characterization (Read 15228 times)
aaron_do
Senior Fellow
******
Offline



Posts: 1398

Inductor Characterization
Jul 26th, 2011, 12:15am
 
Hi all,


when I want to characterize an inductor, I need to route a ground ring using my thick top metal around the structure to ensure a short return path (with enough clearance so as not to de-Q the inductor). My question is, why do I need to complete the ring? Is it possible to have a break in the ring so as to avoid the de-Q-ing?

Second question. When I calibrate the setup up to the probe tips, doesn't that mean that the return path provided through the instrumentation is already de-embedded? In other words, why do I need to have a return path on the chip, why can't I use the return path on the probe station or through the instrumentation?


thanks,
Aaron
Back to top
 
 

there is no energy in matter other than that received from the environment - Nikola Tesla
View Profile   IP Logged
harpoon
Senior Member
****
Offline



Posts: 112
Hampshire
Re: Inductor Characterization
Reply #1 - Jul 28th, 2011, 1:34pm
 
can you provide some drawings to illustrate this ?
Back to top
 
 
View Profile   IP Logged
aaron_do
Senior Fellow
******
Offline



Posts: 1398

Re: Inductor Characterization
Reply #2 - Jul 28th, 2011, 11:20pm
 
Hi,


thanks for the interest. I've attached an illustration below. I have been advised that this is a good test structure for characterizing the inductor within.

My questions are

1) See the picture - Is it possible to cut the return path as shown? I am worried that the loop formed by the return path might de-Q the inductor. However, after thinking about it more, the path to ground will exist anyway, and so the only way to avoid de-Q-ing the inductor would be to space the return path further away, OR

2) is it possible to remove the return path entirely (see the picture)? Unfortunately, I do not understand the measurement equipment well enough to answer this. Does a return path exist through the instrumentation, and is it de-embedded from the measurement after calibration? If this on-chip return path is necessary, does that mean I need to include it in my calibration structures in order to de-embed it from the measurement?

hope I make more sense now. Let me know if there's anything I haven't explained sufficiently.


thanks,
Aaron
Back to top
 

IND_Return_path.jpg

there is no energy in matter other than that received from the environment - Nikola Tesla
View Profile   IP Logged
Frank Wiedmann
Community Fellow
*****
Offline



Posts: 671
Munich, Germany
Re: Inductor Characterization
Reply #3 - Jul 29th, 2011, 5:52am
 
If you have a good electromagnetic simulator, you can use the following approach: Simulate the entire test structure, including rings and everything, and compare the results with your measurements. If you have good agreement, you can be confident that your simulation setup is valid. Then you can simulate the inductor alone with this setup and use the resulting model in your design. This approach avoids some common problems with deembedding and is recommended by Dr. Volker Mühlhaus (http://muehlhaus.com/).
Back to top
 
 
View Profile WWW   IP Logged
aaron_do
Senior Fellow
******
Offline



Posts: 1398

Re: Inductor Characterization
Reply #4 - Jul 29th, 2011, 8:39am
 
Hi Frank,


thanks for the input. In the end, of course, I will make sure I simulate the whole test structure an try and correlate it with measurements. At the moment, however, I'm really just trying to understand how the whole thing actually works.

I'm mainly trying to understand how the return path affects my measurements. Another thing I suspect is maybe the ring is meant for shielding. However, i've been assured that it is necessary as a return path. I would appreciate any input.


regards,
Aaron
Back to top
 
 

there is no energy in matter other than that received from the environment - Nikola Tesla
View Profile   IP Logged
harpoon
Senior Member
****
Offline



Posts: 112
Hampshire
Re: Inductor Characterization
Reply #5 - Jul 29th, 2011, 1:42pm
 
Back to top
 
 
View Profile   IP Logged
RFICDUDE
Community Fellow
*****
Offline



Posts: 323

Re: Inductor Characterization
Reply #6 - Jul 30th, 2011, 5:33am
 
Is the device you want to characterize just the inductor?

If so, then the goal is to obtain enough measurements to de-embed any impact from the probe pad structure.

Completing the ring doesn't seem like a good idea to me. The inductor will be de-Qed by any magnetic coupling to the ring. Completing the ring will prevent some of the current in the ring from traveling through the return path of the probes (through the "G" pads). i.e. I think you want the loop to be completed by the probe grounds.

You also need to consider how the substrate is tied to ground. You want a realistic scenario representing how the substrate will be tied off to ground to capture return path losses in the substrate. This is part of the final device, so you want this information as part of the characterization.

I can't tell from your post if you are trying to capture a realistic substrate return too.
Back to top
 
 
View Profile   IP Logged
aaron_do
Senior Fellow
******
Offline



Posts: 1398

Re: Inductor Characterization
Reply #7 - Jul 30th, 2011, 5:28pm
 
Hi all,


thanks for the replies.

Harpoon, I will try and read that article. It looks very useful.

RFICDUDE, My thinking is that if I don't complete the ring, then any magnetic coupling to the return path will cause even more de-Q-ing since there would still be a loop through the return path outside the chip, and it would be much longer. Also, I've been told that the on-chip return path cannot be removed. This suggests to me that there is no return path through the instrumentation (i.e. the negative terminals of the two ports are isolated from one another). I'm not sure if that's true...

For the substrate return path, I think it won't have any impact at my frequency of interest (1-2 GHz), but I understand where you're coming from.


thanks,
Aaron
Back to top
 
 

there is no energy in matter other than that received from the environment - Nikola Tesla
View Profile   IP Logged
Frank Wiedmann
Community Fellow
*****
Offline



Posts: 671
Munich, Germany
Re: Inductor Characterization
Reply #8 - Aug 1st, 2011, 8:24am
 
I think the idea behind the ground ring is to have a good connection between the ground pads on both sides and also to have a well-defined field configuration around the inductor. However, the ring will influence the quality factor and even the effective inductance. So you should either use the measurement results only to verify your simulation setup with an electromagnetic simulator as I proposed earlier, or make the environment in your test structure as similar as possible to the environment in your design.

In order to have a well-defined environment, a patterned ground shield (see http://www-smirc.stanford.edu/papers/JSSC98MAY-cpyue.pdf) is also often used. In this case, you might be able to make the ground connection in your test structure via the shield instead of a ring, which might be more similar to the configuration in the design.

In our group, we have always used inductor models generated by electromagnetic simulation, using the substrate data provided by the foundry (without fabricating or measuring any test structures). The circuits designed with these inductors (all of which had a patterned ground shield that was included in the electromagnetic simulation) have always performed as expected.
Back to top
 
 
View Profile WWW   IP Logged
aaron_do
Senior Fellow
******
Offline



Posts: 1398

Re: Inductor Characterization
Reply #9 - Aug 1st, 2011, 6:27pm
 
Hi Frank,


thanks for the suggestions. For you point about having a well-defined field configuration, are you basically saying that the ground ring helps to define boundaries for the field?

Quote:
However, the ring will influence the quality factor and even the effective inductance.


Just from a academic point of view, I'm wondering does the ring directly add to the measured inductance (see the attachment)? From my understanding, this return path inductance must be de-embedded from the measurement by using open, short, and thru test structures with the same return path.


thanks,
Aaron
Back to top
 

ReturnPathInductance.JPG

there is no energy in matter other than that received from the environment - Nikola Tesla
View Profile   IP Logged
Frank Wiedmann
Community Fellow
*****
Offline



Posts: 671
Munich, Germany
Re: Inductor Characterization
Reply #10 - Aug 2nd, 2011, 4:19am
 
aaron_do wrote on Aug 1st, 2011, 6:27pm:
For your point about having a well-defined field configuration, are you basically saying that the ground ring helps to define boundaries for the field?

Yes, and the same applies for the patterned ground shield.

aaron_do wrote on Aug 1st, 2011, 6:27pm:
Just from a academic point of view, I'm wondering does the ring directly add to the measured inductance (see the attachment)? From my understanding, this return path inductance must be de-embedded from the measurement by using open, short, and thru test structures with the same return path.

With permission from Dr. Volker Mühlhaus (http://www.muehlhaus.com/), I have attached the relevant pages from a presentation he gave us a year ago. On page 10, you can see the effect of the ring on the inductance. According to Dr. Mühlhaus, this effect cannot easily be removed by deembedding (see page 9).
Back to top
 
View Profile WWW   IP Logged
aaron_do
Senior Fellow
******
Offline



Posts: 1398

Re: Inductor Characterization
Reply #11 - Aug 2nd, 2011, 7:43am
 
Hi Frank,


thanks for the reply. I'm afraid you have misunderstood my question. I understand that unwanted coupling to the return path will change the value of the measured inductance, as shown in your slides. Based on EM simulations, I know how far the return path should be in order not to significantly degrade my measurement.

From the figures in the slides you posted, the return path for the 25 nH inductor consists of a long path at the top, and a short path at the bottom. In this case, the short path will dominate the inductance of the return path, and it will be extremely small, both due to its short length, and the fact that it is coupled to the feeding lines.

In my measurement however, my DUT inductance is very small. I am worried that even such short return path inductance will overwhelm my measurement. So my question is, will my measured inductance actually be the sum of the DUT inductance and the return path inductance. Assume there is no coupling between the DUT and the return path.


regards and thanks for the help,
Aaron
Back to top
 
 

there is no energy in matter other than that received from the environment - Nikola Tesla
View Profile   IP Logged
Frank Wiedmann
Community Fellow
*****
Offline



Posts: 671
Munich, Germany
Re: Inductor Characterization
Reply #12 - Aug 2nd, 2011, 8:58am
 
I must admit that I have no experience with this as I have never designed any test structures. Why don't you just simulate your inductor inside the test structure and in the environment in your circuit and see how much of a difference it makes? As I have mentioned several times, I would not use the measured data directly for design anyway but only to verify the setup of the electromagnetic simulation.
Back to top
 
 
View Profile WWW   IP Logged
aaron_do
Senior Fellow
******
Offline



Posts: 1398

Re: Inductor Characterization
Reply #13 - Aug 2nd, 2011, 6:47pm
 
Hi Frank,


actually I have run EM simulations on a simple structure to see if the return path inductance directly adds to the DUT inductance. According to my simulations, it didn't, but that may be due to assumptions made by the simulator, such as an infinite ground plane underneath the substrate. I had this discussion with somebody who insisted that the return path's inductance does not add to the DUT inductance, but from my understanding of the measurement, it should. I'm wondering if that person also fell into the trap of seeing no effect because the structures he has used are large compared to the return path. Thanks for your time anyway,


Aaron
Back to top
 
 

there is no energy in matter other than that received from the environment - Nikola Tesla
View Profile   IP Logged
Pages: 1
Send Topic Print
Copyright 2002-2020 Designer’s Guide Consulting, Inc. Designer’s Guide® is a registered trademark of Designer’s Guide Consulting, Inc. All rights reserved. Send comments or questions to editor@designers-guide.org. Consider submitting a paper or model.