Monthly Archives: January 2019

What are Diode Array Detectors?

High Performance Liquid Chromatography or HPLC uses diode arrays for recording the absorption spectrum of samples when ultraviolet and visible light passes through them. This enables the user to gather qualitative information about the samples. Major applications of HPLC diode array detectors include agriculture, environment, and industries such as petrochemical, energy, chemistry, life sciences, and pharmaceuticals.

Diode array detectors of HPLC have the advantage of the ability to select the best wavelength for analysis. Therefore, when selecting a diode array for use as a detector in HPLC, one should consider features such as resolution, wavelength range, near infrared ranges, baseline stability, low noise, and peak integration. Some vendors also offer the technique of detecting using a configurable light path formed from fiber optics.

An HPLC has a tungsten lamp emitting light in the visible range. This light enters a deuterium lamp that adds the UV to the visible light, forming a polychromatic beam. As this beam passes through a flow-cell, the sample in the flow-cell absorbs certain wavelengths. The output light then enters a grating, which splits the polychromatic beam into its constituent wavelengths and these pass through a slit before falling on an array of photodiodes, which measure their intensities.

As the diode array detectors measure all wavelengths simultaneously, it is able to acquire the spectra as well as the multiple single wavelengths at the same time by the different diodes in the array. The diode array detector has high selectivity, as it can identify different substances by their spectra.

One of the major advantages of the diode array detector is the tungsten lamp offering light in the extended visible wavelength. Additionally, by controlling the temperature of the optical unit of the diode array detector, its signal quality improves dramatically. Moreover, the diode array detector does not require a reference diode.

While other types of detectors use a diode for reference, for a diode array detector, there is no direct measurement of a signal when there is no absorption. Rather, the HPLC uses a detector balance. This happens automatically as the user switches the instrument on or just before conducting a measurement. The user achieves a detector balance by setting the absorption values for all wavelengths to zero. According to the Lambert Beer’s law, this allows the measurement of all intensities during an experiment to be made relative to this zero absorption intensity.

To cater to baseline changes or drift during a measurement, the diode array detector uses a reference wavelength. The user has to select a specific wavelength as reference, and make sure there is no absorption in the wavelength during the entire chromatography measurement. The user then uses the relative changes of the reference intensities for correcting the proportional changes occurring in other wavelengths.

Five factors majorly affect the measurements done by HPLCs using diode array detectors. These are the slit width, the bandwidth, the response time, and the flow cells. The user has to adjust all of them to obtain the best response from the diode array detector when testing the absorption of a sample. For instance, the slit defines the amount of light the detector measures. The bandwidth defines the window for the data acquisition. The response time defines the time resolution, and the flow cell defines the flow range.

What are Linear Image Sensors?

Fairchild Imaging makes CMOS 1421, a linear image sensor. This is an imaging device with a wide dynamic range of 94 dB or 52000:1, with excellent linearity. The device is a linear sensor, meaning it has 2048 x 1 high-resolution imaging sensors. Fairchild has designed this linear sensor for medical and scientific line scan applications such as optical inspection or fluorescent imaging that require wide dynamic range, high sensitivity, and low noise operation.

With several acquisition modes, this photodiode pixel has an optical area measuring 7 x 10 µm with a pitch of 7 µm and a fill factor of 85%, making the operation of this sensor very flexible:

  • Read after Integration: This mode is ideal for applications with high quality signals
  • Buffered Read after Integration: is a high speed mode that integrates the next line while reading the current line
  • Read on Integration: This is a non-CDS mode, allowing the highest speed of operation
  • Multiple Read during Integration: This mode is for low-light applications, permitting oversampling during integration

Other than the above, a programmed mode, accessible through JTAG interface, meets a wide range of specialized imaging requirements. Readout cycles in this mode are controllable through external signals.

CMOS 1421 has several features such as very low dark current, very low readout noise, and non-destructive readout for fowler sampling. Along with anti-blooming drain and electronic shutter, the CMOS 1421 also features two independent gain settings for each pixel. The entire device is enclosed in an RoHS compliant CLCC and PLCC package of 22.35 x 6.35 x 2.85 mm dimensions. The device consumes 40 mW of power while operating from 3.3 VDC. Major applications of linear image sensors are in microscopy, photon counting, and fluorescent imaging.

CMOS 1421 has a pixel array consisting of a photodiode, a pixel amplifier, and a sample and hold circuit. Along with the above, each pixel has a noise suppression circuitry and a gain register. While the pixel-level gain affects the device sensitivity, it also has a bearing on the noise and conversion factor of the sensor.

Linear image sensors from Fairchild use thinned back-illuminated large area arrays. Fairchild offers custom capabilities such as extreme spectral band detection, low noise active reset CMOS architecture, and high-resolution X-ray imagery using these sensors.

These linear image sensors are ideal for visible, ultra violet to visible, and visible to near infrared spectrometers, and their enhancement makes them suitable for spectroscopy applications. The design of their pixels being tall and narrow helps light distribution from a spectrometer’s grating. If provided with UV sensitivity, these sensors do not need extra UV coating.

CMOS 1421 displays superior linearity, which is of extreme benefit to spectroscopy measurements. The device also includes an electronic shutter along with a built-in timing generator, which are useful in spectroscopy. The device is suitable for several applications involving scientific, industrial, and commercial activities.

New sensors based on CMOS match features with those of CCDs. Featuring simpler external circuit design, and simpler operation, CMOS 1421 linear image sensors are suitable for spectroscopy, displacement measurement, barcode scanning, and imaging.

What are 3-D Image Sensors?

3-D image sensors from Infineon are perfect for use in mobile consumer devices. These new REAL3 image sensors measure the time-of-flight of infrared signals, enabling sensing gestures the user makes in front of the screen. Infineon has designed the sensors with a perfect combination of power consumption, performance, functionality, cost, and size. The IRS238xC 3-D image sensors work in any kind of ambient light conditions and this makes them indispensable for reliable use in mobile applications.

The IRS238xC has high-performance pixel arrays that are highly sensitive to infrared light of 850 and 940 nm wavelength. This allows the device to perform unparalleled in any outdoor environment. Combined with this, Infineon has provided its patented SBI or suppression of background illumination circuitry in every pixel. The combination extends the nominal dynamic range of each pixel by nearly 20 times.

As the single-chip design has a high integration level, it allows the user to optimize the bill of material. Apart from this, it also reduces the design complexity and offers a small form factor. There are other advanced features as well, such as integrated high-performance ADCs, illumination control logic, a modulation unit with high flexibility, and circuitry for eye-safety that enables it to work as a laser-class-1 device. Interfacing is through a high-speed CSI-2 data interface.

The IRS238xC operates from an optimized in-built voltage supply unit, and it can self-boot as it has a full SPI master memory interface. Among the new features available on the sensor are, coded modulation and enhanced configuration flexibility. This allows the device to perform flexibly and robustly in multi-camera scenarios and similar use-cases.

The time of flight technology from Infineon works with stability, as high assembly yields prove, and this is a great boon for camera module manufacturers, as the IRS238xC not only simplifies calibration efforts, it also simplifies the camera module design. In short, IRS238xC combines the benefits of reliability, cost, size, functionality, and power consumption, making it indispensable for mobile 3-D sensing applications in all kinds of ambient light conditions.

For instance, the IRS238xC has the smallest module size giving 224 x 172 pixels, each of size 14 µm and with their own individual micro-lens. The suppression of background illumination or SBI provides each pixel with a 20-time gain in dynamic expansion against strong sunlight, but at minimum power consumption. The robust high-volume assembly of the device and its low calibration efforts offer an easy design and low system bill of materials for the designer.

IRS238xC offers time-of-flight technology for directly measuring the amplitude and depth of information in every pixel. It does this using a single modulated infrared light source that the chip emits to the whole scenery. The TOF imager captures the reflected light. The unit measures the phase difference between the emitted and the reflected light along with their amplitude values, thereby calculating the distance information and producing a grayscale picture of the entire scene all in one sweep.

Infineon provides algorithms that feature unique multiple benefits compared to other depth sensing technologies such as stereovision or structured light.

Making a PiPlateBot with the Raspberry Pi

Turtle robots of the yesteryears had always fascinated Robert Doerr and he decided to create one with the popular single board computer, the raspberry Pi or RBPi. He hid all components of the robot inside the plate case, and decided to call his robot PiPlateBot.

Robert Doerr is the owner of Robot Workshop, and a dedicated robot builder. He used the Bud Pi Plate case, as this was strikingly similar to the turtle-style robots computer science students used earlier.

The Pi Plate has a circular design. Twisting off the top allows easy access to the space within that can house additional components. That led Robert try to fit an RBPi inside, along with the other parts required to build a moving robot. According to Robert, the PiPlateBot is the only robot that runs on an RBPi and uses an off the shelf RBPi case. Robert claims he is using as many RBPi-type products as possible for the construction.

To get everything to fit, Robert had to cut two rectangular holes in the Pl Plate enclosure base. Then he glued servos to the bottom and clipped the RBPi and RoboPi boards on the top. On the boards, he placed a BZO power bank to work as a battery. To enable communicating wirelessly with the RBPi, Robert uses a USB Wi-Fi adaptor. This allows him to SSH directly into the RBPi.

The RoboPi is an impressive motor controller, and the most powerful one for the RBPi. Using an eight-core 32-bit Parallax Propeller RISC micro-controller at 100 MHz, it allows offloading hard real time IO from the Linux OS running on the RBPi, thereby giving timing with greater precision to projects. The RoboPi will work with any RBPi model.

Each core on the RISC controller works at 25 MIPS, with each instruction taking only four clock cycles to complete. The RoboPi has three ten-pin IO module expansion connectors and they provide 24 servo-compatible headers. Some of these connectors use jumper selectable power from the internal 5 VDC or the external servo power supply for powering the sensors.

The user can connect servos to screw terminals that provide external power. There are eight headers for setting up an eight channel 0-5 V analog to digital converter. The user has a choice of using MCP3008 for 10-bit AD conversion or MCP3208 for 12-bit AD conversion.

Therefore, while the RBPi does the high-level thinking, the Parallax Propeller chip on the RoboPi board handles all the IO controlling and the real-time tasks. As the RoboPi controller has both C and Python libraries, Robert plans to write a Logo Interpreter to make the PiPlateBot use Logo to emulate the early turtle robots.

As the PiPlateBot has only two servos controlling its two wheels, the robot actually wobbles when operated. Robert had to use furniture gliders to prevent this. He attached them to the front and rear of the PiPlateBot. A sonar sensor fitted on the PiPlateBot allows it to sense its surroundings.

Building a robot is the fun way of learning to use the RBPi, and a great way to learn programming on the SBC.

What are Laminated Bus Bars?

Rather than use one solid bar of copper, the industry prefers laminated bus bars. These are fabricated components with layers of engineered copper bars separated by flat dielectric materials, bound together into a unified construction. Laminated bus bars offer several advantages—improved reliability and reduced system costs. They are available in various sizes and shapes, some as big as a fingertip while others more than twenty feet in length. Several industries use multilayered bus bar solutions routinely and they include telecommunications, computers, industrial, military, transportation, alternative energy, power electronics, and many more.

Laminated bus bars are good for reducing the system costs, improving system reliability, increasing capacitance, lowering inductance, and eliminating wiring errors. Additionally, the physical structure of the bus bars also acts as structural members of a complete power distribution subsystem. Multilayer bus bars function as a structural integration that other wiring methods cannot match.

The decreased assembly time and the internal material handling costs for laminated bus bars bring down the overall manufacturing costs. Assembly operating procedures can be difficult to follow and assemblers often resort to guesswork, leading to wiring errors. Using laminated bus bars eliminates this totally, as installers have to terminate various conductors at specified locations. Not only does this reduce the parts count, it also reduces ordering, inventory costs, and material handling.

Fabricators can make laminated bus bars fit specific needs and customize them for maximizing efficiency. The use of laminated bus bars, therefore, helps the organization build quality into processes. With reductions in wiring errors, the organization has fewer reworks, and they can lower their quality and service costs.

Laminated bus bars offer increased capacitance and lower inductance, resulting in lower characteristic impedance. The benefit to the industry is greater noise cancellation and effective noise suppression. Manufacturers can control the capacitance by using dielectrics of various thicknesses and different relative K factor.

Multilayered bus bars can replace cable harnesses—this eliminates mistakes in wirings. Moreover, failure rate of bus bars is extremely low, while wiring harnesses fail very often. That makes repairing and or replacing wire harnesses an expensive process, while using bus bars in the system is adding an effective insurance.

According to physics, a conductor carrying current develops an electromagnetic field around the conductor. As laminated bus bars have thin parallel conductors with thin dielectric material separating them, the effect of inductance on electrical circuits is a minimum. With opposing potentials laminated together, the magnetic flux cancellation reaches a maximum. Semiconductor applications routinely use laminated bus bars to reduce the proximity effect. GaN and or SiC high frequency circuits also use laminated bus bars to reduce high electromagnetic interference.

Using wide and thin conductors and laminating them together to form bus bars actually decreases the space requirement, thereby allowing a better airflow in systems and improving system thermal characteristics. Moreover, the flexibility of these bus bars provides the industry with a wide variety of interconnecting methods. Assemblers commonly use tabs, embossments, and bushings for installing laminated bus bars. Manufacturers also offer pressed-in fittings that can integrate into the design. This makes laminated bus bars compatible with almost any type of interface.

What is a PCB Via and How is it Made?

Vias are actually holes drilled into PCB layers and electroplated with a thin layer of copper to provide the necessary electrical connectivity. Three most common types of plated through via are in use—plated through holes, blind holes, and buried holes—with plated through holes running through all the layers of the PCB. These are the simplest type of holes to make and the cheapest. However, they take up a huge amount of PCB space, reducing the space available for routing.

Blind vias connect the outermost circuit on the PCB with other circuits on one or more adjacent inner layers. As they do not traverse the entire thickness of the PCB, they increase the space utilization by leaving more space for routing.

Buried vias connect two or more circuit layers in a multi-layered PCB, but do not show up on any of the outer layers. These are the most expensive type of vias and take more time to implement, as the fabricator has to drill the hole in the individual circuit layer when bonding it. However, designers can stack several buried vias in-line or in a staggered manner to make a blind via. Therefore, buried vias offer the maximum space utilization when routing a PCB. Fabricators of high-density interconnect (HDI) boards usually make use of buried vias, most often using lasers for drilling them.

Drilling a Via

At positions for the vias, the fabricator drills holes through the PCB using a metal drill of small diameter. He or she then cleans the hole, de-smears it, and de-burrs it to prepare it for plating. Rather than removing copper as is normally the case with the etching process, the fabricator then adds a thin layer of copper to the newly formed hole through a process of electroplating, thereby connecting the two layers. For a two-layer board, the fabricator then etches circuit patterns on both sides. Via usually have capture pads on both layers.

The process of drilling a via hole using a laser is somewhat different. In general, fabricators use two types of lasers—CO2 and UV—with the latter able to make very small diameter via holes. UV laser-drilled via holes are about 20-35 µm in diameter. As the laser beam is able to ablate through the thin copper layer, capture pads with a central opening are not necessary. Most fabricators program a two-step process for drilling a hole with a laser beam.

In the first step, a wider-focused laser heats the top copper layer, driving the metal rapidly through the melt phase into the vapor phase prior to gas-dynamic effects expelling it from the surface. The laser repeats this for all via positions on the PCB layer.

In the next step, the program focuses the beam tightly and controls the depth the laser can burn. For blind vias, It allows the laser to burn through the intervening dielectric and stop when it has reached the bottom copper layer, before moving on to the neighboring via position.

The same process of electroplating as above deposits a thin layer of copper along the walls of the holes left behind by the laser beam, thereby connecting the two layers. The rest of the process for etching the circuit pattern on the two sides remains the same.

How useful are PCB Vias?

Designers use a plated through via as a conduit for transferring signals and power from one layer to another in a multi-layer printed circuit board (PCB). For the PCB fabricator, the plated through via are a cost-effective process for producing PCBs. Therefore, vias are one of the key drivers of the PCB manufacturing industry.

Use of Vias

Apart from simply connecting two or more copper layers, vias are useful for creating very dense boards for special IC packages, especially the fine-pitch components such as BGAs. BGAs with pitch lower than 0.5 mm usually do not leave much space for routing traces between neighboring pads. Designers resort to via-in-pads for breaking out such closely spaced BGA pins.

To prevent solder wicking into the via hole while soldering and leaving the joint bereft of solder, the fabricator has to fill or plug the via. Filling a via is usually with a mixture of epoxy and a conductive material, mostly copper, but the fabricator may also use other metals such as silver, gold, aluminum, tin, or a combination of them. Filling has an additional advantage of increasing the thermal conductivity of the via, useful when multiple filled vias have to remove heat from one layer to another. However, the process of filling a via is expensive.

Plugging a via is a less expensive way, especially when an increase in thermal conductivity does not serve additional value. The fabricator fills the via with solder mask of low-viscosity or a resin type material similar to the laminate. As this plugging protects the copper in the via, no other surface finish is necessary. For both, filled and plugged vias, it is important to use material with CTE matching the board material.

Depending on the application, fabricators may simply tent a via, covering it with solder mask, without filling it. They may have to leave a small hole at the top to allow the via to breathe, as air trapped inside will try to escape during soldering.

Trouble with Vias

The most common defect with vias is plating voids. The electro-deposition process for plating the via wall with a layer of copper can result in voids, gaps, or holes in the plating. The imperfection in the via may limit the amount of current it can transfer, and in worst case, may not transfer at all, if the plating is non-continuous. Usually, an electrical test by the fabricator is necessary to establish all vias are properly functioning.

Another defect is the mismatch of CTE between the copper and the dielectric material. As temperatures rise, the dielectric material may expand faster than the copper tube can, thereby parting the tube and breaking its electrical continuity. Therefore, it is very important for the fabricator to select a dielectric material with a CTE as close as possible to copper.

Vias placed in the flexing area of a flex PCB can separate from the prepreg causing a pad lift and an electrical discontinuity. It is important designers take care to not place any vias in the area where they plan the PCB will flex.

HiFiBerry & Raspberry Pi Put New Life into Old Loudspeakers

If you have some old stereo speakers stored away in your basement, chances are they connect through the old way—with wires—to an amplifier, and that is the reason they were banished to the basement. With HiFiBerry Amp+ and a single board computer, such as the Raspberry Pi (RBPi), you can resurrect your vintage speakers. Using the latest in open source technology, you can now use the renovated loudspeakers wherever you want, since they now operate wirelessly.

HiFiBerry offers their Amp+ as an amplifier for the RBPi. As it is a Class-D power amplifier, it is highly efficient as a stereo module, and you only need to connect the loudspeakers. This high-quality amplifier is ideal for setting up multi-room audio installations.

The amplifier is stable enough to drive 4-Ohm loudspeakers and those with higher impedance as well, pumping out 25 W of output power. However, the best part is the RBPi can fully control the amplifier. As the amplifier includes on-board digital to analog converters, you do not need external sound cards or DACs to provide the 44.1 KHz and 48 KHz sample rates. The board connects directly to the RBPi without needing additional cables, and this provides a full digital sound path for optimal audio performance.

The HiFiBerry Amp+ comes as a pre-fabricated kit, so it needs no soldering. It is a daughter board for the RBPi, and when the RBPi plugs into it, you need to connect only a single external power supply of 12-18 V to supply both the amplifier and the SBC, as the RBPi draws power from the Amp+. You can use the Amp+ with all RBPi models that have the 40-pin GPIO connector. The board sits on four small plastic spacers that come with the kit.

The specialty of the Amp+ kit is it converts the digital signal into audio with far greater clarity than the RBPi can, and delivers that to the speaker as a 25 W audio amplifier. On the reverse side of the board, the female connector is easily visible, so it is easy to plug in the GPIO pins of the RBPi.

On one side of the board are a jack for powering the board, and six wire-terminals. If for some reason you cannot use the jack to power the board, use the two wire-terminals on the left. The rest of the four wire-terminals are for connecting to a pair of stereo loudspeakers, using two audio cables per speaker.

As the board takes in 12-18 V supply and delivers power to the RBPi as well, it is important to not power the RBPi from its usual 5 V power supply. This reduces the number of wires to the assembly. As the Amp+ board is very small, it does not protrude beyond the RBPi. It is important to mount the board on the four plastic spacers to avoid breaking the GPIO pins.

The SD card for the RBPi can be of the 8 GB type and people have reported better performance with Transcend cards. However, you can use 16 GB cards as well.