| Author |
Message |
fanshuo
Joined: 21 Jun 2007
Posts: 381
Helped: 5
Location: Netherlands
|
 20 Apr 2009 8:29 inductor model |
|
|
|
|
Hi,
I found in my model file that the variation of inductor at different models could be up to +-15%. And the parasitic resistance and capacitance are correlated(when Rs is max,Cox is min and vice versa)
I can not find any theoretical explanation on the big variation of L and the correlation.
Are these numbers are trustworthy?
|
|
| Back to top |
|
 |
abhisheksbox
Joined: 24 May 2007
Posts: 41
Helped: 8
|
20 Apr 2009 14:17 inductor model |
|
|
|
|
| Modelling of an inductor can be done in various ways, wht type of model can be used is depends upon the design methodology and the end objective of the circuit design.
|
|
| Back to top |
|
|
fanshuo
Joined: 21 Jun 2007
Posts: 381
Helped: 5
Location: Netherlands
|
20 Apr 2009 14:18 inductor model |
|
|
|
|
I am not going to model the inductor
I just need to use the model of inductor which provided by the foundry
|
|
| Back to top |
|
|
Google
AdSense
|
20 Apr 2009 14:18 Ads |
|
|
|
|
|
|
| Back to top |
|
|
mylovec07
Joined: 13 Apr 2009
Posts: 1
|
20 Apr 2009 15:38 inductor model |
|
|
|
|
| hi
|
|
| Back to top |
|
|
abhisheksbox
Joined: 24 May 2007
Posts: 41
Helped: 8
|
20 Apr 2009 15:56 inductor model |
|
|
|
|
| Then use the technology file details provided by the foundry
|
|
| Back to top |
|
|
fanshuo
Joined: 21 Jun 2007
Posts: 381
Helped: 5
Location: Netherlands
|
20 Apr 2009 15:57 inductor model |
|
|
|
|
As I said,
I can not explain the big variation of the inductance value. What could be the possible reason?
|
|
| Back to top |
|
|
padpan
Joined: 02 Aug 2006
Posts: 18
Helped: 3
|
21 Apr 2009 12:43 Re: inductor model |
|
|
|
|
It could be attributed to a variety of reasons. For example, what is the process node you are working on? Or which are the metal thickness and width variations for this process?
Another reason for this big variation could be the poor inductor modeling offered by the specific PDK. Generally, foundry inductor models may not be as trustworthy due to the characterization process applied and issues related to de-embedding that are not efficiently addressed. Needless to say that those models are topology-agnostic since they do not account for mutual coupling phenomena.
--padpan
|
|
| Back to top |
|
|
fanshuo
Joined: 21 Jun 2007
Posts: 381
Helped: 5
Location: Netherlands
|
21 Apr 2009 13:00 inductor model |
|
|
|
|
Thanks!
I am working on 0.13um
top metal thickness is 2um
I check the graphs provided by the foundry
it seems the model is good at typical situation
But no data found on corners from measurement
|
|
| Back to top |
|
|
padpan
Joined: 02 Aug 2006
Posts: 18
Helped: 3
|
22 Apr 2009 10:08 Re: inductor model |
|
|
|
|
What's the metal thickness and width variation? Those parameters are expected to dominate L variations...
--padpan
|
|
| Back to top |
|
|
fanshuo
Joined: 21 Jun 2007
Posts: 381
Helped: 5
Location: Netherlands
|
22 Apr 2009 10:29 inductor model |
|
|
|
|
well
what I found from the process manual is that its metal length can varies for 15%,which cause the inductance to vary
According to my opinion, the metal should have a accurate control.It should not vary 15%
|
|
| Back to top |
|
|
padpan
Joined: 02 Aug 2006
Posts: 18
Helped: 3
|
22 Apr 2009 12:21 Re: inductor model |
|
|
|
|
they probably refer to the overall spiral length variation. This may come up when the inductor dimensions are not precisely taped out during fabrication but still this is a huge variation... What's the Q variation by the way? If L is 15%, Q will be by far higher...
--padpan
|
|
| Back to top |
|
|
fanshuo
Joined: 21 Jun 2007
Posts: 381
Helped: 5
Location: Netherlands
|
22 Apr 2009 12:32 inductor model |
|
|
|
|
of course Q would be more.
But I cannot find it in the process manual.
I have no idea why the variation is this big
Also according to the manual, the series parasitic resistance increases when parasitic capacitance decreases.
I do not know why those two parameters are correlated.
|
|
| Back to top |
|
|
padpan
Joined: 02 Aug 2006
Posts: 18
Helped: 3
|
22 Apr 2009 14:39 Re: inductor model |
|
|
|
|
I suppose it refers to the width. If the conductor gets wider its capacitance increases and resistance decreases and vice versa...
--padpan
|
|
| Back to top |
|
|
fanshuo
Joined: 21 Jun 2007
Posts: 381
Helped: 5
Location: Netherlands
|
22 Apr 2009 16:15 inductor model |
|
|
|
|
to padpan,
ok I see.
Thanks.
But still the width does not effect the L value.
Do you have any idea of why the length varies a lot?
|
|
| Back to top |
|
|
padpan
Joined: 02 Aug 2006
Posts: 18
Helped: 3
|
22 Apr 2009 18:34 Re: inductor model |
|
|
|
|
It's huge variation. Frankly, I have no idea. Perhaps the foundry wants to extend the variation range to avoid customer complaints in case of inconsistency of the PDK model against measured silicon...
--padpan
|
|
| Back to top |
|
|
