ABCs of signal integrity for embedded developers – Part 1: Basic SI rules and methods
If this digital pulse is to be a successful transmission, when it travels from one end of a transmission media to another all the harmonics need to reach the receiving end without any time delay in order to re-create the same square pulse.
What if there is a time delay? The origin of propagation delay in a square pulse is when all the harmonics are delayed by a constant time (i.e. group delay). There are cases where the group delay is a function of frequency, due to which not all the harmonics are delayed by the same amount at the receiver side. This changes the shape of the signal in addition to any constant delay or scale change. This is because of the non-linear phase filtering properties of transmission lines, a topic that will be covered in an upcoming article.
Figure 3 shows how a digital signal looks once it loses its high frequency component due to the losses in the transmission media.
Figure 3: Loss of high frequency harmonics during transmission
The diagram shows that the rise and fall time of the signal has increased. The signal also exhibits ringing or unwanted oscillation of amplitude (to be covered in a future article). Due to an increase in the rise and fall time of the signal, the quality is compromised. In addition, timing violations will start creeping into the system. The signal will remain in a transition state for a longer duration and will ultimately cause latching issues in systems where the signal needs to be latched at a clock edge.
As shown in Figure 4, the required setup time, ts1, is no longer met due to the increased edge timing. The setup time that can be met is ts2; which is less than ts1. Due to this, the correct state of the signal might not be latched and the system might enter a metastable state.
Figure 4: Setup/Hold time violation with Rise/Fall time increase
What if the low frequency harmonics are lost instead? Since the low frequency harmonics have high amplitudes, the signal will only contain the low amplitude - high frequency harmonics (as shown in Figure 5). The resulting signal is no longer a useful input to the system. This loss in low frequency harmonics can be due to electromagnetic interference or even crosstalk.
Figure 5: Effect on square wave when the first harmonic is lost
In a future article, we’ll focus on the fundamentals of transmission lines as well as their properties and effects on high-frequency signals.
Akshat Garg is an Applications Engineer with Cypress semiconductor. He is handling applications and debugging issues related to all cypress memory products, capacitive sensing products, and automotive products. His keen interest is in system design. He can be reached at akhi@cypress.com.
Sachin Gupta is working as Product Marketing Engineer 2 in the PSD division of Cypress Semiconductor. He holds a B.Tech degree in Electronics and Communication from Guru Gobind Singh Indraprastha University, Delhi. He has several years of experience in applications engineering in SoC products. He can be reached at sgup@cypress.com.
Pritesh Mandaliya has a Master's degree in Electrical Engineering (Digital and Analog Mixed Signal Design) from San Jose State University. He is presently working with Cypress Semiconductor as a Senior Applications Engineer in Cypress' Memory Product Division. Responsibilities include creating and testing signal integrity and behavioral models of SRAMs, applications support for customers, maintaining documentation and application notes, and board-level-failure-analysis debugging. His interests are developing GUIs in Labview, VB, and VC++; and FPGA-based designs. Email: prit@cypress.com.
Loading comments... Write a comment