If you spend any time on a test bench, you probably use square waves to test components, subsystems, and systems. You can get square waves from function generators and arbitrary-waveform generators, but they can be expensive if all you need is a square wave. At my company, we often use square-wave generators (SQWGs) when that's all we need. But, we often need some signal conditioning between the signal source and DUT (device under test). In this two-part series, we'll present many of the circuits we use.
We often have the following issues with SQWGs and other generators:
The SQWG may have only one or two outputs, but we often need more test signals such as inverted signals of the output(s).
The output of the signal generator is not exactly TTL/CMOS compatible.
We may need to send the signal from single SQWG to several points in DUT or DUTs, where we won't want direct electrical connections between these points.
The output of the SQWG may not have enough drive capabilities -- higher current, voltage, or power may be needed.
We may need adjustable or fixed positive, negative, and/or bipolar square-wave signals.
We may need to protect the SQWG's output(s) because repairing a damaged laboratory generator may be expensive, time consuming, or simply impossible.
We may need many square-wave signals with different pulse durations and frequencies that can be synchronized with the signal from a single SQWG.
We may have one SQWG but need several externally enabled and disabled (strobed) square-wave signals, etc.
Many low-frequency loads require considerable power during switching and the SQWG may not be able to supply enough voltage, current, or power.
In all these and some other cases, we've developed modules that extend the output capabilities of the SQWG. We will call the modules EPMs (extension and protection modules).
David Patterson, known for his pioneering research that led to RAID, clusters and more, is part of a team at UC Berkeley that recently made its RISC-V processor architecture an open source hardware offering. We talk with Patterson and one of his colleagues behind the effort about the opportunities they see, what new kinds of designs they hope to enable and what it means for today’s commercial processor giants such as Intel, ARM and Imagination Technologies.