The Designer's Guide Community Forum

The Designer's Guide Community Forum
https://designers-guide.org/forum/YaBB.pl
Design >> High-Speed I/O Design >> LVDS tx at 1.8v supply
https://designers-guide.org/forum/YaBB.pl?num=1207313519

Message started by justdoit on Apr 4^th, 2008, 5:51am

Title: LVDS tx at 1.8v supply
Post by justdoit on Apr 4^th, 2008, 5:51am

Hi ,
I am designing a LVDS tx at 1.8v supply voltage .Since at lower voltages the headroom available for the upper current source is less , i used normal current mirror at the top and cascode current mirror at the bottom . The problem now is i am cing my rise time to be faster than the fall time and these two values differ almost by 30 to 40% of thier absolute values .For example if my trise is 60ps then my tfall is around 100ps . Can anyone suggest some good solution to avoid this issue ?????????

Title: Re: LVDS tx at 1.8v supply
Post by ywguo on Apr 6^th, 2008, 6:39am

Hi,

It is helpful to paste the schematic here for geting more constructive advices.

Suppose a common differential structure for your LVDS tx, pls take a look at your circuit from the following several aspects.
1. Is it turn on the sink current branch as fast as turn off it?
2. There is one more parasitic capacitor at the bottom because it is a cascode current mirror. Is the parasitic capacitor charged/discharged each time the sink current branch turn on/off?
3. Is the source current equal to the sink current?

Hope it is helpful.

Yawei

Title: Re: LVDS tx at 1.8v supply
Post by kanan on Apr 13^th, 2008, 9:43am

Hello Hari,
Are you using a make-before-break circuitry in switching on and off the switches? Doing that will keep the currents in the mirrors steady and you would then need worry less about the charging and discharging the cap at the current mirror nodes.

Putting in a cascode at both ends would be advisable I guess. Are you getting the same pull-up and pull-down current?

--Kanan

Title: Re: LVDS tx at 1.8v supply
Post by SATurn on May 30^th, 2008, 12:28pm

You may also increase the headroom for PMOS current sources by increasing the input voltage swing ! That's a very effective way to design low volage LVDS drivers.

SATurn

Title: Re: LVDS tx at 1.8v supply
Post by loose-electron on Jun 23^rd, 2008, 1:01pm

Get rid of the Cascodes!!!

The only thing a cascode gets you is a more ideal current source at the expense of design area and a loss of voltage headroom.

For a LVDS data signal -- When you look at the current variance with/without cascodes, you will see that you really dont need them.

Been there, done that, at 0.8u and now all the way down to 32nm...

Jerry