The Designer's Guide Community Forum

The Designer's Guide Community Forum
https://designers-guide.org/forum/YaBB.pl
Design >> Analog Design >> delay line variation due to PVT
https://designers-guide.org/forum/YaBB.pl?num=1466127421

Message started by Debdut on Jun 16^th, 2016, 6:37pm

Title: delay line variation due to PVT
Post by Debdut on Jun 16^th, 2016, 6:37pm

I have done a circuit using cascaded RC networks (like a Elmore delay model) and at each point it has 1 ps delay. The circuit has, say, 100 RC networks, giving a total delay of 100 ps.
Now, due to process variations and mismatch in the RC values, the delay may change.
So what do I do? Suppose the actual overall delay is 105 ps, then should I consider that each RC network has a delay of 1.05 ps?
I was thinking like breaking the circuit into parts and measuring the delay of each part of the circuit. Like, if I break the circuit into 4 parts, and, say, the delays are 25 ps, 27 ps, 25 ps, 28 ps, Then, should I take the mean average or rms average of these values? If I take rms average, then into how many parts should the circuit broken?

Title: Re: delay line variation due to PVT
Post by raja.cedt on Jun 16^th, 2016, 11:53pm

I-
I didn't understand 2nd half of the Question. To answer your PVT independent delay, it is simply impossible without having any reference and -ve feedback. If you want very accurate consider DLL, otherwise you should accept whatever variation you are getting. To reduce the delay variation what you have implement with single RC or use an inverter rather than 100 segments.

Hope this helps,
Raj.

Title: Re: delay line variation due to PVT
Post by Debdut on Jun 17^th, 2016, 3:22am

Hello Raj,
I have to use the several RC networks. I want to find out how should I represent the delay of each RC network considering the overall delay of the entire network.
So, if the overall delay is 105 ps should I consider each element has a delay of 1.05 ps?
Or, should I break the complete RC network into equal/unequal parts and calculate the delay of each network using frequency distribution method?

Title: Re: delay line variation due to PVT
Post by raja.cedt on Jun 17^th, 2016, 4:38am

Dear Debdut-
You haven't given any information why you HAVE TO USE SEVERAL RC NETWORKS. No can help for this kind of Question. Always explain the situation properly (unless it is a top secret) as clear as you can , hence you get better response from the people.

From your Question, what I can say is use as much as you can use.

Best Regards,
Raj.

Title: Re: delay line variation due to PVT
Post by Debdut on Jun 20^th, 2016, 10:31am

I need to delay a waveform by certain ps. Maximum delay can be 100 ps. Minimum I am considering 1 ps as below that the RC values are becoming very low. So I need to find out the average delay generated by each RC network.

Title: Re: delay line variation due to PVT
Post by loose-electron on Jul 6^th, 2016, 10:30pm

digital signal delay? analog signal?