Beyond Design: DDR3/4 Fly-by vs. T-topology Routing

Reading time ( words)

JEDEC introduced fly-by topology in the DDR3 specification for the differential clock, address, command and control signals. The advantage of fly-by topology is that it supports higher-frequency operation, reduces the quantity and length of stubs and consequently improves signal integrity and timing on heavily loaded signals. Fly-by topology also reduces simultaneous switching noise (SSN) by deliberately causing flight-time skew, between the address group and the point-to-point topology signals, of the data groups. To account for this skew, the DDR3/4 controller supports write leveling. The controller must add the write leveling delays to each byte lane to maintain the strobe to clock requirement at the SDRAMs.

T-topology can be challenging to route, particularly double T-topology with four back-to-back SDRAMs, but it can be advantageous when using multi-die packages. The fly-by topology used in Figure 3 is much easier to route but does not work well with high-capacitance loads, such as LPDDR3 DDP (double die package) and QDP (quad die package) devices. IC fabricators basically arrange dies in parallel to increase package density which can also increase input capacitance by up to four times. Excessive ring-back is often present in the first few nodes of the daisy chain.

This is the reason why the T-topology was developed. However, if you are supporting only SDP (single die package) devices, then the fly-by is the most straightforward approach. It doesn't matter which topology you use, though—both fly-by and double T-topologies should work fine. If you are using a DDP device, then double-T topology works better than fly-by in terms of delivering a better system margin.

During a write cycle, using the fly-by topology, data strobe groups are launched at separate intervals to coincide with the clock arriving at memory components on the SODIMM or PCB, and must meet the timing parameter between the memory clock and DQS defined as tDQSS of ± 0.25 tCK. The PCB design process can be simplified using the leveling feature of the DDR3/4. The fly-by, daisy chain topology increases the complexity of the controller design to achieve leveling but fortunately, greatly improves performance and eases board layout for DDR3/4 designs.

To read this entire article, which appeared in the April 2016 issue of The PCB Design Magazine, click here.



Suggested Items

Why We Simulate

04/29/2021 | Bill Hargin, Z-zero
When Bill Hargin was cutting his teeth in high-speed PCB design some 25 years ago, speeds were slow, layer counts were low, dielectric constants and loss tangents were high, design margins were wide, copper roughness didn’t matter, and glass-weave styles didn’t matter. Dielectrics were called “FR-4” and their properties didn’t matter much. A fast PCI bus operated at just 66 MHz. Times have certainly changed.

Bridging the Simulation Tool Divide

04/12/2021 | I-Connect007 Editorial Team
Todd Westerhoff of Siemens EDA recently spoke with the I-Connect007 Editorial Team about the divide between users of high-powered enterprise simulation tools and those who need a more practical tool for everyday use, and how Siemens is working to bridge the gap. Todd also shared his views on why so many engineers do not use simulation, as well as advice for engineers just getting started with simulation tools.

Barry Olney’s High-Speed Simulation Primer

04/09/2021 | I-Connect007 Editorial Team
The I-Connect007 editorial team recently spoke with Barry Olney of iCD about simulation. Barry, a columnist for Design007 Magazine, explains why simulation tools can have such a steep learning curve, and why many design engineers are still not using simulation on complex high-speed designs.

Copyright © 2021 I-Connect007. All rights reserved.