The Designer's Guide Community Forum

The Designer's Guide Community Forum
https://designers-guide.org/forum/YaBB.pl
Design >> RF Design >> max limit for number of fingers
https://designers-guide.org/forum/YaBB.pl?num=1345131891

Message started by summi on Aug 16^th, 2012, 8:44am

Title: max limit for number of fingers
Post by summi on Aug 16^th, 2012, 8:44am

dear forum,
To increase fmax, lower gate resistance can achived by using multiple fingers, but is there any max limit, like if i use very big is there any effect which increase resisistance.

I know with higher number of fingers, layout become complex and may be some higher paracitic's which can be tuned out.

BR,
Summi.

Title: Re: max limit for number of fingers
Post by aaron_do on Aug 16^th, 2012, 8:42pm

Hi,

there is definitely an optimum based on the layout, however, it might be quite high. Why don't you try parasitic extraction? Even if layout were not a constraint, there would be some number of fingers after which the improvement would be negligible.

regards,
Aaron

Title: Re: max limit for number of fingers
Post by summi on Aug 17^th, 2012, 2:26am

Dear Aaron,
thanks for the reply. yes i have simulated and not noise figure went up after some time insteed of going down. So i would like to know why it is going up, in theory (or up to my knoeledge) it should keep on decreasing...tell me whats the flaw in my arguement.

Br,
Summi.

Title: Re: max limit for number of fingers
Post by raja.cedt on Aug 19^th, 2012, 8:05am

hello,
may be due to many fingers, connection resistance between those fingers may dominate. I am not sure, some one can help better.

Thanks,
Raj.

Title: Re: max limit for number of fingers
Post by RFICDUDE on Aug 19^th, 2012, 4:56pm

The exact reasons are not very clear, but it appears that as you continue to decrease finger length the interconnect capacitance and perhaps the via and/or interconnect metal resistance starts to dominate.

Minimum NF depends on minimizing thermal noise at the input and input referred noise from the output. So, a good clue to why the NF is degrading with further reductions in finger length is to monitor the most significant noise contributors as NF starts to increase. If it is output noise sources that cause the increased NF then a reduction in the maximum gain maybe the reason; otherwise, if the input routing resistance dominates then you will know it is the routing parasitics that dominate the degradation.

Have you looked at changes in the noise summary versus changes in gate length?