An Innovated Skew Locked Loop to Improve Read Timing Margin for External Memory Interface Applications in FPGA Devices

IP.com Prior Art Database Disclosure
IP.com Disclosure Number: IPCOM000139945D
Publication Date: 31-Aug-2006
More Like This Download

Publishing Venue

The IP.com Prior Art Database

Abstract

Skew is an important factor that limited the performance of the system, especially in the high speed application, for example memory interface. Double data rate memory controller can be implemented in FPGA devices. As new DDR memory spec offered higher performance with 800Mbs to 1Gbs data rate, the timing budget left to FPGA becomes much smaller. DDR memory interface has data strobe (DQS) transmitted along with a group of data (DQ) for data capture in the receiver side. DQS is edge-aligned with data during read and center-aligned with data during write. All the skew and jitter among all the DQ in the group and DQS will be treated as uncertainty and will be subtracted from valid sampling window. So it is important to reduce skew among all DQ and DQS for both read and write.

Language

English (United States)

Country

United States

Document File

6 pages / 34.4 KB

This text was extracted from a Microsoft Word document.
At least one non-text object (such as an image or picture) has been suppressed.
This is the abbreviated version, containing approximately 48% of the total text.

An Innovated Skew Locked

Loop

to Improve Read Timing Margin for External Memory Interface Applications in FPGA Devices

Invention

Skew is an important factor that limited the performance of the system, especially in the high speed application, for example memory interface. Double data rate memory controller can be implemented in FPGA devices. As new DDR memory spec offered higher performance with 800Mbs to 1Gbs data rate, the timing budget left to FPGA becomes much smaller. DDR memory interface has data strobe (DQS) transmitted along with a group of data (DQ) for data capture in the receiver side. DQS is edge-aligned with data during read and center-aligned with data during write. All the skew and jitter among all the DQ in the group and DQS will be treated as uncertainty and will be subtracted from valid sampling window. So it is important to reduce skew among all DQ and DQS for both read and write.

This invention provides an innovated scheme to continuously find the skew between DQ and DQS paths and adjust the delay chain setting accordingly to compensate the skew between two paths.

Background

Minimize the skew among all DQ and DQS path can help to improve data valide window. There are a couple of methods can be used to reduce the skew. Compensate the skew by programming the same delay chain setting for all devices or programming on-chip non-volatile elements for each individual device. This method needs to accurately measure the skew information in order to set the proper delays, which is not a trivial job. Also the process, voltage and temperature variation results in variations of the skew, which can not be compensate by fixed delay settings.  

This invention continuously finds the skew between DQ and DQS paths and adjust the delay chain setting accordingly to compensate the skew between two paths.

During a read operation, FPGA will capture data using data strobe. Figure 1 shows the read path in FPGA. Data (DQ) goes through input buffer and gets to register data input. DQS goes through input buffer and a phase shift delay chain and gets to register clock. The phase shift delay chain is used to shift DQS signal to the middle of the sampling window. There are skew adjustment delay chains in both DQ path and DQS path. The skew adjustment delay chain is a variable delay chain which delay setting can be programmed or coming from Skew Locked Loop (SLL). Bypassing the phase shift delay chain we want 0 skew between DQ path and DQS path.

Figure1 DDR Memory Interface Read Path in FPGA

Implementation

To continuously find the skew between DQ and DQS paths without disrupting the normal operation, a mimic DQ path and a mimic DQS path is needed. Figure 2 shows the block diagram of skew locked loop.

The SLL consisted with DQ mimic path, DQS mimic path, Phase Comparator, Counter Control Logic and two counters (DQ Counter...

First page image
You are not signed in. If you have an IP.com account, your download price may be lower or waived. Click here if you want to sign-in now.
Loading PayPal...
The full document comprises 6 pages and is available as a PDF document as well as a ZIP archive. The cost is $40.00 USD (depending on your billing address, sales tax may apply); payment may be made directly using your credit card or your PayPal account.

If you've already purchased this document, and wish to download it now you may enter the download access code you received in your original email receipt.