Logic analyzers play a critical role in validation of today's sophisticated high-speed digital designs. With the ability to measure real-time activity of system components such as I/O, memory buses, communication buses, ASICs, FPGAs, and processors, logic analyzers provide key insight for effective functional validation. The logic analyzer can time correlate system measurements so designers can determine what type of problems exist between system blocks. This measurement versatility makes the logic analyzer an excellent choice to solve issues between multiple developers on the same project. New advances in logic analysis use models allow teams to better utilize their logic analysis investment.
While logic analyzers play a crucial role in helping the design team get critical insight into the system, the use of a logic analyzer for any given team member can be sporadic. Typically, the instrument gets used during a frenzied project development period after simulation and before release to manufacturing. A team experiencing functional issues with a prototype will turn to the logic analyzer to provide measurement insight. The team is generally under tremendous pressure to find and fix the problem so that the product can quickly move into production. Problems at this stage of development typically involve multiple disciplines. Consequently, multiple engineers will need access to measurements and time to analyze the results. For geographically dispersed teams, this collaboration will require multiple sites to share in the problem-solving task.
Two new logic analysis modes allow teams to better utilize their logic analysis investment. These new use models allow the logic analyzer to be better shared within the team, decreasing the required investment. The first mode is called "hosted" while the second is referred to as "offline analysis."
Hosted Mode
For decades engineers have sat in front of sophisticated test and measurement equipment, such as scopes, logic analyzers and protocol analyzers, to make measurements. Logic analyzer users interact with the instrument's front panel when taking measurements and to be efficient, needs captive use of the logic analyzer. The engineer must locate the equipment, and keep it for a fixed period of time. Tasks, including set up and analysis of acquired data, must be done on the front panel of the instrument. This means that even though the measurement itself may take only a few seconds, the user may tie up the logic analyzer for several hours because he or she needs it to set up the measurement and to view the results of the measurement. Another team member, wanting to use the logic analyzer, must wait for the equipment to free up. With the move to geographically dispersed teams, coupled with fewer prototypes and dramatic advances in off-the-shelf PC technology, a new use model has evolved. This use model is call "hosted."
Innovation in logic analysis architectural design has made it possible to decouple the acquisition hardware from the software user interface as shown in Figure 1. With this architecture, a copy of the logic analysis software application can be installed on a computer that the engineering team already has. This full featured logic analysis measurement software connects to logic analyzer acquisition HW using a high-speed link such as LAN. This enables an engineer to access and control a logic analyzer on the network without having to physically be directly in front of the logic analyzer. In fact, multiple team members can install copies of the software on their own computers and all now have sequential access to a single logic analysis measurement system. This architecture has some very distinctive benefits.
Figure 1: The "hosted" use model decouples the acquisition hardware from the software user interface allowing the logic analysis software application to be installed on the engineering team's computer for greater measurement convenience. |
Hosted logic analyzers gives multiple engineers full logic analysis measurement access to anyone connected to LAN—whether they are 1 meter or 1000 kilometers away from the logic analyzer they are using to take a measurement. This means that multiple engineers can sequentially access the same logic analyzer for measurements. Each ties up the logic analysis hardware for a few seconds while the measurement is being made. This frees up the logic analyzer for another engineers who needs to use it for a measurement.
Prototypes are typically expensive and limited, so this use model gives quick measurement access to entire teams from a single prototype. And team member can each make their needed measurements without leaving their desks. They can use their personalized setup, made on the application running on their personal computer. This mode can also be particularly useful when geographically dispersed teams need measurement access to a single target. The engineer making a measurement only needs use of the logic analysis hardware for the few seconds it takes to acquire the data and transfer this information to his computer.
What does this buy the engineering community? It allows test-equipment manufacturers to lower prices while giving engineering teams ever more processing. Versions of hosted equipment can be made without including a motherboard. For example a hosted logic analyzer like the Agilent 1690A sans motherboard can be controlled from a remote PC where all the control and processing is done. This architecture allows teams to leverage their existing PC investment without having to pay for the price of a PC inside of the test and measurement equipment
Offline Analysis Mode
A second use model for more effective team use of logic analyzers is the capability of using offline analysis. Offline analysis allows multiple team members, each running a copy of the logic analysis software on their own PCs, to setup new measurements and analyze previously acquired data. They can do this without the need to have the logic analysis acquisition hardware. Historically, both of these tasks have forced a team member to tie up the logic analysis resource. With offline mode, only a team member who needs to take new acquisitions needs direct control of the logic analyzer as shown in Figure 2.
Figure 2: Once an acquisition has been made via hosted mode, the data can be saved and analyzed in offline mode. |
Setups can be made and then downloaded, via a USB thumb drive or LAN for example, to the logic analyzer. Measurement setups can be saved on each person's personal computer. So, another engineer who walks up to the equipment and turns it off, or sets up a new measurement, doesn't erase the setup work performed by the previous user. Measurements already taken can be shared with multiple team members—even those who may be remote. Detailed post-processing of measurements can also be performed without tying up the logic analyzer.
Once an acquisition has been made via hosted mode, the data can be saved and analyzed in offline mode. The decoupled hardware and software architecture no longer limits the software processing power you can apply to an instrument task to the CPU inside of the logic analyzer. Using a hosted approach with offline analysis a team can increase processing power by move than 1000 times—all without any incremental investment.
The most recent logic analyzers can capture massive amounts of data. A single trace can be up to gigabytes in size. Post processing of this trace would traditionally be done using the motherboard inside of the logic analyzer. With offline analysis, engineering teams can more effectively leverage the investment they've already made in their own high-powered computers. Look at an example from an engineering team who recently converted to the hosted use model. An engineering team purchased a logic analyzer in 2002. Using this non-hosted logic analyzer, it takes 15 minutes to upload a 300M-byte data file from the logic analyzer's hard drive. Today, using offline mode it takes 10 seconds to bring up the same data file up on their PC for analysis.
Users now are no longer constrained to performing this post processing analysis with the traditional approach using the motherboard inside of the instrument. Why is this good? Because instrument manufacturers don't typically incorporate the very latest PC technology inside equipment. They must take the time and expense to qualify that any components can withstand the rigors of an industrial environment. Vendors typically embed an older generation of computer technology. This is done to make sure the compute platform is rugged and that the instrument vendor can get guaranteed supply over a number of years.
Some logic analysis application software has been architected with multithreading capabilities. This means that when run on a multiprocessor compute engine, such as a 4-way multiprocessor server, the tasks run by the software can be partitioned across the four processors. A fast server with large amounts of RAM and multiple processors can provide a highly responsive front-end interface to the logic analysis acquisition hardware.
Most engineering teams routinely upgrade their compute platforms frequently to provide sufficient horsepower for demanding applications. So, engineering teams can apply the last processing power even thought the test equipment may have been purchased years earlier. Additionally, hosted logic analyzer users can take advantage of mass storage on their own compute platforms instead of being constrained solely to the storage available in the logic analyzer itself. Upgrading or adding computer storage is much less expensive than upgrading the hard disk of a logic analyzer.
Logic analyzer architectures that support hosted and offline modes allow development teams to make more effective use of their logic analysis investment. These new use models leverage computing platforms in which design teams have already invested. In addition, they foster better team collaboration as test results can be shared and analyzed on a wide scale. The future of test and measurement equipment will include more products whose capabilities are closely tied to existing engineering team computer investments.
About the Author
Joel holds a degree in Electronic Engineering, an MBA from Regis University, and has completed coursework from Harvard Business School. He joined Agilent Technologies (formerly Hewlett-Packard) in 1989 as a product support engineer working with oscilloscopes and logic analyzers. During his 15 years with Agilent Technologies, his career has focused on bringing new and innovative system verification and validation solutions to research and development engineers in the electronics industry. Within Agilent's electronic design automation and test and measurement businesses, Joel has held several individual contributor and management positions. Instrumental in the creation of three start-up businesses within Agilent, he has recently submitted a patent application in the area of field-programmable gate arrays debug. His outside interests include digital photography and hiking in national parks.
