Need help with designing a tappered CML buffers

[husseinadel]

CML buffers

hello

i am designing a tappered CML buffers,,,,,, and i have a question ,,,,,i want to set the output common mode to a certain value so as the swing not from vdd to ,, vdd-I*R,,,but somewhat lower than vdd,so, how can i set it,,is that possible? or not????,,,,i think that as the switching of current completely from one branch to the other while swinging makes the output reach vdd.....so,,can anyone share me in thinking?????/ thanks

 

[tsb_nph]

Re: CML buffers

hello

i am designing a tappered CML buffers,,,,,, and i have a question ,,,,,i want to set the output common mode to a certain value so as the swing not from vdd to ,, vdd-I*R,,,but somewhat lower than vdd,so, how can i set it,,is that possible? or not????,,,,i think that as the switching of current completely from one branch to the other while swinging makes the output reach vdd.....so,,can anyone share me in thinking?????/ thanks

Hi,
If i understand your question correctly, then yes, it is possible.
In an usual arrangement, the two load resistors are not connected to Vdd.
Instead, connect those two nodes together and then add an extra common mode resistor (Rcm) form that node to Vdd. Now the swing will start from Vdd - Itail*Rcm instead of from Vdd as there is a constant Itail*Rcm voltage drop across the common mode resistor.

I have attached a simple schematic to illustrate this.

Bharath

 

[husseinadel]

Re: CML buffers

hello

i am designing a tappered CML buffers,,,,,, and i have a question ,,,,,i want to set the output common mode to a certain value so as the swing not from vdd to ,, vdd-I*R,,,but somewhat lower than vdd,so, how can i set it,,is that possible? or not????,,,,i think that as the switching of current completely from one branch to the other while swinging makes the output reach vdd.....so,,can anyone share me in thinking?????/ thanks

Hi,
If i understand your question correctly, then yes, it is possible.
In an usual arrangement, the two load resistors are not connected to Vdd.
Instead, connect those two nodes together and then add an extra common mode resistor (Rcm) form that node to Vdd. Now the swing will start from Vdd - Itail*Rcm instead of from Vdd as there is a constant Itail*Rcm voltage drop across the common mode resistor.

I have attached a simple schematic to illustrate this.

Bharath

thanks so much,, but i want this mainly so that the swing will not be clipped
but this may make the transistors enters triode,,,so,,,is the usual design of cml swings up to vdd,, or not?

 

[krashkealoha]

Re: CML buffers

What tsb_nph shows works, and you don't have to switch up to Vdd. You may have to reduce your voltage swing, then increase your width in order to eliminate the clipping but still maintain the gain of the buffer. Else, use source followers to precede the cml buffer in order to increase your headroom for the cml buffer.