Choosing the right tools is essential in any development, testing or automation project. The wrong choice will cost you time and compromise the quality of your tests, while the right tools will guarantee you the right results in no time.
When building a test / automation system, the main tools you will need are measurement and control tools. To select the correct instrumentation, engineers must have knowledge and experience in:
- Technical requirements for measuring the device under test (DUT);
- Technical specifications (specific to each tool) that will have an impact on the application developed:
- The different categories of instruments available and the compromises made in terms of capacity, size, price, etc .;
- Differences between different instrument models in a particular instrument category.
Choosing the right instrumentation for an application is not an easy task, especially when there are so many tool options on the market. In this article, you will be able to discover the main categories of tools and learn the selection criteria after which to make the right choice for each application.
Analogic and RF Instruments
In the category of analog instruments and Radio Frequency (RF), the restrictions given by the laws of physics create a compromise between the accuracy of a measurement and the speed at which that measurement can be made.
A starting point in choosing analog and RF instruments is finding the answers to two questions:
- What is the direction of the signal? (input, output or both)
- What is the frequency of the measured signal? (DC, kHz, MHz or GHz)
Once you have answered these questions, you can choose the category of tools you need by consulting the table below.
DC
ANALOGIC HIGH SPEED
ANALOGIC LOW SPEED
RF
INPUT (Measurement)
Digital Multimeter (DMM)
Analogic Input, Data Acquisition (DAQ)
Osciloscop, Frequency Counter
RF Analizor (Spectrum Analyzer, Vector Signal Analyzer)
OUTPUT (Generate)
Power Supply
Analogic Output
Function/Arbitrary Waveform Generator (FGEN, AWG)
RF Signal Generator (Vector Signal Generator, CW Source)
INPUT & OUTPUT (Same Device)
DC Power Analyzer
Multifunctional DAQ
Osciloscop All-in-One
Vector Signal Transcever (VST)
INPUT & OUTPUT (Same Pin)
Source Measure Unit (SMU)
LCR Meter
Impedance Analyzer
Vector Network Analyzer (VNA)
Note that the table above contains the most commonly used tool categories and does not contain the full spectrum of tool categories.
After choosing the category we can turn our attention to the key specifications that we need to consider:
- Signal range: we need to make sure that the input range of the measuring instrument will be able to pick up the signal we want to measure;
- Insulation: represents the insulation of the instrument from the grounding point that has an impact on the noise immunity of the signal and safety;
- Impedance: affects the electrical charge and frequency performance of the entire measuring system;
- Bandwidth: we need to make sure that the measured signal can be picked up by the chosen instrument depending on the bandwidth (kHz, MHz or GHz);
- Sample rate: we need to make sure that the Analogue to Digital Converter (ADC) can measure fast enough to capture the signal;
- Resolution: is given by the resolution of the ADC (usually generated between 8-bit and 24-bit) and has a direct impact on the quality of the measurements;
- Accuracy: represents the maximum error depending on time and temperature;
- Sensitivity: the smallest change in measurement that the instrument can detect.
For a better view of the route that the signal makes, the simplified stages of the route are represented in the image below.
Insulation & Termination
Filtering & Coupling
Amplifier
ADC
———————- semnal direction ———————->
Digital Instrumentation
In the context of electronic functionality testing, digital instrumentation aims to communicate through digital communication protocols and to test the electronic features as well as the communication links of the protocols. An essential aspect in choosing digital instrumentation is digital communication in series or in parallel.
Serial communication standards have become increasingly used due to the limitations of parallel communication buses in terms of clock rate (1 GHz – 2 GHz). High-speed buses using serial communication send encrypted data that contains both data and clocking information via a secure differential signal. This eliminates the limitations of parallel communication. Sending data in high-speed serial communication format helps reduce the number of pins in the integrated circuits, resulting in reduced physical size. Another advantage of the high clock rate is the increased data transfer rate.
Parallel Bus
Serial Bus
- PCI 64-bit/33 MHz
- PCI 64-bit/66 MHz
- PCI 64-bit/100 MHz
- Front Panel Data Port
- EISA
- PCI 32-bit/33 MHz
- PCI 32-bit/66 MHz
- IDE (ATA PIO 0)
- ATA PIO 1
- ATA PIO 2
- ATA PIO 3
- ATA PIO 3 | ISA 16-bit/8.33 MHz
- U ltra-2 wide SCSI
- RapidIO Gen1.1
- GPIB
- SCSI | ISA 8-bit/4.77 MHz
- PCIe Gen1x16
- PCIe Gen2x16
- Serial RapidIO Gen2
- PCIe Gen3x16
- PCIe Gen1x8
- PCIe Gen2x8
- PCIe Gen3x8
- JESD204B
- PCIe Gen1x4
- Serial RapidIO Gen1.3
- PCIe Gen2x4
- DisplayPort
- PCIe Gen3x4
- HDMI 1.0 | DVI
- HDMI 1.3
- HDMI 2.0
- SD-SDI
- G igabit Ethernet
- SATA 1.0
- Serial FPDP | PCIe Gen1x1
- SATA 2.0 | 3G-SDI | JESD204A | 10 Gigabit Ethernet
- PCIe Gen2x1 | USB 3.0
- SATA 3.0
- PCIe Gen3x1
- USB 3.1
For a quick clarification of the appropriate type of digital instrumentation it is necessary to find the answers to two questions:
- What is your goal? (digital interface, custom digital interface or electrical and timing testing)
- How fast should communication be? (static or kbit / sec, Mbit / sec or Gbit / sec)
STATIC/LOW SPEED
SYNCHRONISATION & HIGH SPEED PARALLEL COMMUNICATION (100 Mbits/sec)
HIGH SPEED SERIAL COMMUNICATION (10 Gbits/sec)
INTERFACE (Standard)
Standars Interface Card – I2C, C
Synchronous Protocol interface – ARINC 429, CAN, GPIB, I2C, SPI
10 Gigabit Ethernet, Fibre Channel, PCI Express etc.
INTERFACE (Custom)
Digital I/O (GPIO)
Digital Waveform Generator/Analyzer, pattern Generator
FPGA, Serial Rapid I/O, JESD204b
INPUT & OUTPUT ON THE SAME DEVICE
Pin Electronics Digital, Per-Pin Paramentric Measurement Unit (PPMU)
BERT, Osciloscop
Digital communication can be implemented using two methods: software timing or hardware timing.
Software Timing
Applications that use timing software do not use a clock for input and output. Software controls I / O, and a programming language controls software timing. This programming language often runs on an operating system that can take milliseconds to execute a command in the software. In the case of software timing, the operating system timer is used to determine the execution rate of commands. Software timing is most often used in low-speed applications such as monitoring and control.
There are two types of software timing: deterministic and non-deterministic. By using a real-time operating system (RT OS) up to 1 µs can be achieved, but it must be borne in mind that the real-time operating system does not increase the communication speed but only makes it more deterministic. Non-real-time operating systems, such as Microsoft Windows, are non-deterministic. In non-deterministic systems, the execution time of a software command in hardware is inconsistent. Factors such as memory, processor speed, and the specifications of applications running on the operating system may affect runtime.
Hardware Timing
Hardware timing tools use the rising and falling edges of the clock to generate and acquire deterministic data. Hardware timing can generate and acquire digital data at Gbit / sec rates with high determinism. Some of the applications in which hardware timing is used are chip testing, protocol emulation and testing, digital video and audio testing, digital electronic testing, etc.
Communication Buses
Before moving on to defining the types of communication buses, we need to consider the stability, repeatability, and speed required for the application. This brings us to issues related to the work environment.
- Laboratory (accuracy, repeatability, low-level control, ease of setup and ability to automate tests)
- Production (speed, transfer rate, accuracy, optimization through programming and debugging interface)
Nowadays USB, PCI Express and Ethernet / LAN are the most popular communication options for instrument control. Some instrumentation manufacturers believe that these communication options provide the answer to all the communication needs of the instruments. We will continue by defining what bandwidth and latency are, and then we will go on to define the most commonly used communication buses.
When analyzing the technical advantages of different communication buses, bandwidth and latency are the most important features that we need to consider.
Bandwidth
Bandwidth measures the rate at which data is transmitted through the communication bus, usually MB / sec. A bus with higher bandwidth can transmit more data over time than a bus with lower bandwidth. Bandwidth can tell us whether the generated or purchased data can be transmitted on the communication bus and how much internal memory is needed for the instruments.
Latency
Latency measures the delay of data transmitted on the bus. If we were to compare a communication bus with a production line, the bandwidth would be the number of production lines and the speed of travel, and the latency would be the delays given by the waiting points on the production line. A low latency bus introduces fewer delays between data transmission and processing time.
GPIB
The GPIB communication bus (IEEE 488) was developed especially for instrument control applications. GPIB has been a robust and secure communication bus for 30 years and is still one of the most popular options for instrument control due to its low latency and acceptable bandwidth. It currently enjoys the widest built-in instrumentation, with over 10,000 instrument models using GPIB connectivity.
With a bandwidth of 1.8 MB / sec, it is the perfect bus for communication and control of stand-alone instruments. The latest revision (HS488) increased the bandwidth to 8 MB / sec. Data transfer is based on messages, often in the form of ASCII characters. Multiple GPIB instruments can be tied together over a total distance of 20 m, and the bandwidth is shared with all instruments on the bus. Although the bandwidth is relatively low, the latency of GPIB is significantly lower than that of USB and Ethernet buses. Tools that use GPIB are not self-detecting and do not self-configure when connecting to a system; although there is software for GPIB, and the wiring is sturdy and perfect to use in the most drastic conditions.
USB
USB is known for connecting to the peripherals of a Personal Computer (PC). The popularity of this bus has also reached the fields of testing and measurement. An increasing number of instrument manufacturers are adding the ability to control instruments via a USB connection. In many cases where a device has multiple USB ports, these ports end up being connected to a single control unit, the bandwidth is divided by all the ports.
The latency of the USB bus is reasonable, being higher than that of Ethernet but lower than that of PCI and PCI Express. The maximum cable length for USB devices is 5 m. Devices that use USB enjoy self-detection, which means that devices are recognized and configured as soon as they are connected to a PC. The disadvantages of the USB bus are safety and physical endurance, which are lower than the other buses analyzed here.
USB-enabled devices are ideal for applications that require portable measurements, data backup/acquisition, in-vehicle data acquisition, and more.
PCI
PCI and PCI Express have the best features in terms of bandwidth and latency of all the buses presented here. The PCI bandwidth is 132 MB / sec, which is shared with all instruments connected to the bus. The PCI latency is very low, reaching 700 ns, compared to Ethernet which has 1 ms. PCI communication is based on registries. Unlike the other buses described here, the PCI does not connect to other instruments via a cable. This is an internally used bus used for plug-in PC cards and in modular systems (example: PXI). In this case, remote measurements do not apply directly but there are extension capabilities up to 200 m via an MXI interface. Due to the fact that PCI cards are connected internally, the physical resistance can be considered directly correlated with the resistance of the system they belong to, for example, a PC.
In many cases, PCI cards may cost less because they use the capabilities of the PC to which they are connected: power, processor, display, and memory.
PCI Express
PCI Express is similar to PCI, being an evolved version of it. Much of the presentation in PCI also applies to PCI Express.
The main difference between PCI and PCI Express is that PCI Express has higher bandwidth and provides dedicated bandwidth for each instrument. Of all the buses analyzed here, PCI Express is the only one that has this capability. Data is transmitted over a point-to-point connection called a line at 250 MB / sec in a Gen 1 direction. There may be multiple lines for each PCI Express link, with bandwidth varying depending on the implementation of the connection and the instrument. An Ax1 link will have a bandwidth of 250 MB / sec, Ax4 will have 1 GB / sec and Ax16 will have 4 GB / sec. The PCI Express bus can be used in conjunction with software that was originally written for PCI (backward capability) cards. PCI Express can be expanded using external cabling.
PCI Express is ideal for high-performance applications, high-speed data acquisition and generation, and for integrating and synchronizing different types of tools.
Ethernet/LAN
Ethernet is a mature bus technology that has been used in multiple applications outside the spectrum of testing and measurement.
- 100BASE-T Ethernet has a theoretical bandwidth of 12.5 MB / sec.
- 1000BASE-T (Gigabit Ethernet) has a theoretical bandwidth of 125 MB / sec.
The bandwidth is divided by the network.
Communication on this bus is done through messages using packets that make the transmission difficult. For this reason, Ethernet has the highest latency of all the communication buses presented here.
However, Ethernet remains a great solution for creating distributed networked systems. It can be used at distances of up to 100 m without repeaters, and with the help of repeaters, there is no distance limit. As with GPIB, self-configuration does not exist. You need to assign an IP address and a subnet configuration for the tools used. Like USB and PCI, almost all modern PCs have an Ethernet / LAN port. This makes Ethernet ideal for distributed systems and remote monitoring. The physical Ethernet connection is more robust than USB but weaker than GPIB and PCI.
We offer complete development, implementation and maintenance services using the experience + qualification of dedicated hardware and software architects and engineers to guarantee the success of your projects. See the services offered by SelfLAB here and covered industries here.
References:
- National Instruments – Fundamentals of Building a Test System
