The Designer's Guide Community Forum

The Designer's Guide Community Forum
https://designers-guide.org/forum/YaBB.pl
Design >> RF Design >> Minimum lenght of RF device
https://designers-guide.org/forum/YaBB.pl?num=1282504948

Message started by unevb on Aug 22^nd, 2010, 12:22pm

Title: Minimum lenght of RF device
Post by unevb on Aug 22^nd, 2010, 12:22pm

Hi,

i) Is it possible for the k-factor of a amplifier to be less than 1 at a frequency at which Gt < 0dB ? I am seeing this result in simulation. Also, since gain is <0dB , is it possible for the system to be unstable?I expect the k-factor of a tuned circuit to have a shape like a parabola, with the minima being at the tuned frequecy. I have attached the plots.

ii) Why is it that the RF model of the transistor has a minimum length constraint greater than the minimum length allowed by the technology. I am using a foundry pdk for a 90nm process, but it allows a min length of 240nm on the rf models. Does this have something to do with the optimal layout of the RF transistor for performance?

Title: Re: Minimum lenght of RF device
Post by unevb on Aug 22^nd, 2010, 12:24pm

Attaching the plot

Title: Re: Minimum lenght of RF device
Post by vp1953 on Aug 23^rd, 2010, 5:02pm

It is not unusual for the RF devices to have a min geometry greater than the minimum required for the process. They usually have guard rings around them so that the environment around the RF transistor is similar where it is used; it is related to being able to characterize the device consistently across all process corners and temperature.

Title: Re: Minimum lenght of RF device
Post by aaron_do on Aug 24^th, 2010, 6:10am

Hi,

for question 1, from my understanding, the k-factor is used to represent unconditional stability. For a given port impedance, it does not necessarily mean that the circuit is unstable, it just means that there is some value of the port impedance which causes the circuit to be unstable. On the other hand, GT is the transducer power gain at the specific port impedance which is defined in the testbench. So perhaps the circuit is unstable at a port impedance where GT is more than 1.

Let me say however, that i'm not sure whether or not a circuit can be unstable (or not) when GT<1. It doesn't sound impossible though.

For question 2, I haven't seen that many technologies, but for the ones I have seen, there is not such a huge discrepancy between the minimum device length and that for RF transistors. Perhaps you are not looking at the right transistor.

cheers,
Aaron

Title: Re: Minimum lenght of RF device
Post by purplewolf on Aug 24^th, 2010, 8:06am

unevb wrote on Aug 22^nd, 2010, 12:22pm:

You should care about the k-factor upto to the freq to which device has power gain..when device has no gain it is acting just as a "passive" element..and passive elements have no stability issues.

Title: Re: Minimum lenght of RF device
Post by purplewolf on Aug 24^th, 2010, 3:26pm

W/L is proportional to gm..keep L min for maximum gm...
yes .. probably it has something to do with optimum layout or so..