AN 917: Reset Design Techniques for Intel® Hyperflex™ Architecture FPGAs

Download PDF
ID 683539
Date 1/05/2021
Public
Document Table of Contents
Document Table of Contents x
1. AN 917: Reset Design Techniques for Intel® Hyperflex™ Architecture FPGAs
1. AN 917: Reset Design Techniques for Intel® Hyperflex™ Architecture FPGAs x
1.1. General Reset Recommendations 1.2. Reset Design Strategies 1.3. Analyzing Resets with Design Assistant 1.4. Implementing Reset Circuitry 1.5. Document Revision History for AN 917: Reset Design Techniques for Intel® Hyperflex™ Architecture FPGAs
1.1. General Reset Recommendations x
1.1.1. Limit Reset Usage 1.1.2. Hyper-Registers Require Synchronous Reset
1.2. Reset Design Strategies x
1.2.1. Asynchronous Reset Design Strategies 1.2.2. Synchronous Reset Design Strategies
1.2.1. Asynchronous Reset Design Strategies x
1.2.1.1. Recovery and Removal Checks
1.3. Analyzing Resets with Design Assistant x
1.3.1. Example: Asynchronous Reset Design Assistant Analysis 1.3.2. Example: Synchronous Reset Design Assistant Analysis
1.3.1. Example: Asynchronous Reset Design Assistant Analysis x
1.3.1.1. Adding SDC Constraints to Asynchronous Reset Circuitry
1.4. Implementing Reset Circuitry x
1.4.1. Reset Coding Techniques 1.4.2. Avoiding Common Reset Coding Issues 1.4.3. Generating Synchronous Reset Trees 1.4.4. Reset Removal on Multi Clock Domain Designs 1.4.5. Using the Reset Release Intel® FPGA IP 1.4.6. Adding Clock Cycles to the Reset Sequence
1.4.1. Reset Coding Techniques x
1.4.1.1. Converting to Synchronous Resets
1.4.2. Avoiding Common Reset Coding Issues x
1.4.2.1. Coding Synchronous Reset with Follower Registers 1.4.2.2. Coding Reset-Related Optimizations
1.4.3. Generating Synchronous Reset Trees x
1.4.3.1. Reset Distribution through Hyper-Pipelining and Hyper-Retiming 1.4.3.2. Adding Pipeline Registers and Automatic Register Duplication
1.4.6. Adding Clock Cycles to the Reset Sequence x
1.4.6.1. C-Cycle Equivalence 1.4.6.2. Determining the Reset Sequence Requirement
1. AN 917: Reset Design Techniques for Intel® Hyperflex™ Architecture FPGAs

1.2.1.1. Recovery and Removal Checks

During the de-assertion of a reset, the control to the output of a flip-flop transfers from the reset line to the clock signal, like a regular D flip-flop. To avoid the register entering metastable state, you must ensure that the reset is not de-asserted in certain time frames of the active clock edge.

In Figure 3, the Removal Time, Trem, refers to the minimum time, after the active clock edge, that the reset must be stable before being de-asserted. The reset Removal Check ensures that the de-asserted reset signal is not captured by the same clock edge that launches the reset.

Figure 3. Recovery and Removal Check Timing Diagram

Reset Recovery Time, Trec, is the minimum time between the de-assertion of a reset and the clock signal being high again. The reset Recovery Check ensures that the reset signal is stable for a minimum time after de-assertion, before the next active clock edge.

The Intel® Quartus® Prime Timing Analyzer computes recovery and removal slacks separately from set-up and hold slacks. You can find the summary and individual reports in the Timing Analyzer reports.

Figure 4.  Compilation Report Window - Timing Analyzer Reports
Level Two Title