The Designer's Guide Community Forum

The Designer's Guide Community Forum
https://designers-guide.org/forum/YaBB.pl
Design >> Analog Design >> question on Elmore delay
https://designers-guide.org/forum/YaBB.pl?num=1463743085

Message started by Debdut on May 20^th, 2016, 4:18am

Title: question on Elmore delay
Post by Debdut on May 20^th, 2016, 4:18am

I understand that Elmore delay is a model for interconnect networks, the networks being represented by cascaded RC network.

I am trying to create a delay line with RC networks and calculating the R-C values using Elmore delay model. But after simulation I am not getting the correct delay at the nodes. For smaller networks I have seen that Elmore delay is different than delay calculated using Laplace transform. But for large network the Laplace transform is becoming very complicated.

Can there be a simpler way predict R-C values that produces delay which matches with simulation?

Title: Re: question on Elmore delay
Post by raja.cedt on May 20^th, 2016, 1:45pm

How you are finding delay with Laplace transforms? By using time constants?

Title: Re: question on Elmore delay
Post by Debdut on May 20^th, 2016, 10:37pm

I have calculated the transfer function and did inverse laplace for step input. Then finding the time taken for the output to reach 50 percent of input.

Title: Re: question on Elmore delay
Post by raja.cedt on May 21^st, 2016, 1:24am

HI-
Yes that is called time constant based calculation. Elmore delay is based on dominate time constant, means in the transfer function ignore all s power terms except constant and s-coefficient. Hence this pretty accurate if systems is dominted by single time constant. For a given network if one R and one C domintes this works generally. Please post which circuit is creating problem.

Best Regards,
Raj.

Title: Re: question on Elmore delay
Post by Debdut on May 21^st, 2016, 11:10pm

"Elmore delay is based on dominate time constant" - thanks for the info.

The circuit has three low pass RC networks just connected in cascade. It does not have a dominant pole. All the RC values are comparable. So what can I do to generate correct delays ... I mean how do I calculate?

Title: Re: question on Elmore delay
Post by Debdut on May 22^nd, 2016, 1:55pm

Thank you again, raja.cedt. Is there any way to rate your comments. It saved me.

I added a large time constant circuit and I am getting the precise delays that I want. Thank you.

Title: Re: question on Elmore delay
Post by raja.cedt on May 23^rd, 2016, 2:37am

Which book you are using for Elmore delay. If you are interested better look for 8th chapter in the following book or just read open circuit time constants method.

The Design of CMOS Radio-Frequency Integrated Circuits

Good Luck,
Raj.