Texas Instruments’ OPA333 & OPA2333

TI’s micro-power OPAx333 opamp goes where few opamps have gone before

No matter how advanced your digital circuitry is: sometimes, you need an OpAmp. Technology improvements allow for low-power rail to rail OpAmps such as the OPA333.

Running operational amplifiers from a single supply is a sure-fire way to cause problems while training. Generating positive and negative supply voltages gets tedious when working on a low-power design. Having to make do with supply voltages as low at 1.8 V complicates things further.

TI addresses that problem with the low-power operational amplifier at hand. The chip works with supply voltages as low as 1.8 V. Should you feel like splitting them, the limits become -0.9 V and 0.9 V. The largest allowed supply is a generous 5.5 V – a value well known to microcontroller circuit designers. Another interesting aspect is the advanced rail to rail capability: the output voltage can approach the supply voltages to a margin of 50mV. This makes the part ideally suited to sensor information conditioning.

Texas Instruments intends the part for precision applications. In addition to the low CMRR, the part has a temperature drift in the range of but 0.05uV/°C. Its quiescent current is around 17yA – compare this to older parts to find out that our new contender is much more efficient.

Treasure trove of knowledge!

Texas Instruments data sheet team outdid itself when describing the OPA333. The datasheet…

…doesn’t limit itself to the specifications. It provides valuable insight into circuit design and PCB layout for low power OpAmps. This makes the document valuable even if you work with other operational amplifiers. From a packaging point of view, Texas Instruments is flexible. The parts are available in SOT23, SOT and SOIC housings, the latter of which can be soldered by hand.

Should you need two operational amplifiers of the type, switch to the OPA2333. It combines two of the parts in a fashion similar to the one shown in the figure.

…the OPA2333 combines two low power OpAmps in one

It, furthermore, is somewhat price competitive – the two figures show you OEMSecrets best prices as of the writing of this story.

Nordic Semiconductor’s nRF52840

Nordic Semiconductor evaluation kits for cheap

Nordic semiconductor is the king of the wireless hill. Their chips support a variety of protocols used in the internet of things.

The recently announced nRF52840 Dongle uses the nRF52840 chip on a USB stick-like PCB. The $10 planar can plug into a PC. After that, it supports all Nordic semiconductor development tools.

…the PCB doubles as USB connector.

Keep in mind that debugging is not available on these evaluation boards. If you want to debug, a more expensive development kit such as the nRF52840 DK is needed. Due to the high price, distributor margins are wide. Take a look at the figure to see how OEMsecrets can save you money.

Prices fluctuate violently when emulation kits get expensive…

From technical point of view, the dongle is extremely versatile. It supports Bluetooth 5, but can also be used with other protocols:

• Bluetooth LE
• Bluetooth Mesh
• Thread
• Zigbee
• 802.15.4
• ANT/ANT+

In addition to these features, the product also has a set of GPIO pins available from the side of the PCB. The interface pinout is somewhat difficult to find, but is available for you right here.

Thanks to the presence of the supply pins, the PCB can be used in a standalone fashion. Program it from your PC, and solder it to your circuit to get started. Due to a price of less than 10 USD a pop, small series could even use the PCB like a process computer.

Nordic’s nRF-Dongle also provides distributors with a nice margin…

Laird DA1206C121R-10 Ferrite bead networks

Ferrite bead networks – it takes multiple beads to tango

Ferrite beads come in groups – if your circuit needs one, it usually doesn‘t take long until a regulator makes you add another. Laird now provides a resistor array-like solution for the ferrite bead problem.

For those of you new to the concept: a ferrite bean is a resistor which rejects high frequencies. Look at the impendance chart shown, which we took from the data sheet.

…the resistance of a ferrite bead increases with frequency

It takes but one look at to understand what a ferrite bead is good for: place it in series with a component to eliminate high-frequency radiation. Popular candidates would be switching regulators, but also a variety of other chips such as amplifiers which work by modulating carriers.

Laird’s part differs from the competition in that it combines four ferrite beads in a 1206 housing. This makes the part ideally suited for controlling outputs. Put it next to a chip which needs to be disciplined, and don’t worry about pick and place costs.

From a technical point of view, the part is a classic ferrite bead. Its maximum current current is 300 mA continuous, its resistance in the DC range is around 0.2 ohms.
The main issue is that its impendance starts to raise in the range of more than 10 MHz. This makes the part poorly suited to small amplifiers, whose carrier wave is in the range of 100 kHz or so.

Other than that, the part is a total buy. The figure shows the pricing, which is very much in line with four single ferrite beads. Furthermore, using dedicated ferrite beads requires more space on the PCB.

Keep repairability in mind when using this part. While the component is on the market since 2004, one never knows when it might be discontinued. Furthermore, if your equipment is used in areas with limited parts availability, saving maintainers the hassle of finding a ferrite bead array might be appreciated.

NXP PCF2127

Lord of the time

Adding real time clock information to microcontroller driven systems is problematic. As time goes by, small errors in computing routines add up. Given that users hate entering system times, a real-time element is valuable.

NXP – once Philips Semiconductors – is an old hand in the space of real time clocks, making the PCF2127 a classic. Our figure shows a basic application circuit taken from the application note.

The extra battery ensures timing continuity if system power is unavailable…

Source

Let us start out with the most important question: the part is pretty accurate. The small version PCF2127AT promises a three PPM accuracy from -15 to 60°C, while the PCF2127T achieves the same from -30 to 80°C.

A question of complication! RTC ICs work with some kind of serial interface. Given that NXP developed the I2C protocol, the chip makes for a great I2C real-time clock IC. Furthermore, the chip can also work as an SPI real-time clock – the figures show two operating modes.

The PCF2127 makes for a great SPI real time clock…

…but can also handle the I2C protocol perfectly

In addition to providing a simple incrementing time keeper, the chip also has a calendar system. It even takes care of leap year computations, thereby allowing you to use microcontroller space for other tasks.

Finally, a timestamp function emits a periodic signal. This is useful smart meters, where periodic wake up can be handled via an interrupt generated by the real-time clock IC.
Sadly, developers who still hand-prototype the circuits won’t be happy. The IC comes in SO16 and SO20 variants, both of which are priced similarly.

Given that NXP has a long tradition of designing sensors, the Dutch could not resist adding a temperature sensor and 512 bytes of static RAM. This allows you to store a small amount of information in a remanent fashion. Finally, the chip can also supervise the battery voltage – the circuit can inform the user if the backup battery runs low.

Accelerate machine learning with the Xilinx ALVEO A-U200-P64G-PQ-G

Accelerate machine learning with the Xilinx ALVEO A-U200-P64G-PQ-G

Bitcoin mining experienced a severe acceleration with the introduction of dedicated accelerator ICs, most of which achieved effiency ratings miles above normal GPUs. Xilinx now introduces a series of FPGA-based accelerators intended for data center and machine learning jobs.

From a principal point of view, both ALVEO U200 and ALVEO U250 are a set of FPGAs which connect to the workstation via a PCIe link. In addition to that, a MicroUSB and a network interface are added to the accelerator cards, allowing data center administrators effortless access to the various configuration interfaces.

In addition to that, a wide variety of SDKs are provided: they allow developers to create kernel elements optimized for the needs of the ALVEO’s internal architecture.

Memory access traditionally was a bottleneck for FPGA accelerator cards. XILINX solves this problem by providing local memory. A total of four slots take one DDR4 DIMM each, XILINX officially permits only the use of the MTA18ASF2G72PZ-2G3B1 16GB unit produced by Micron.

…local DDR4 memory reduces bus contention

Keep in mind that Xilinx offers a total of four versions of the card. A-U200-P64G-PQ-G and A-U250-A64G-PQ-G are a bit heavier, but come with active cooling. A-U200-P64G-PQ-G and A-U250-P64G-PQ-G are a bit lighter, but are passively cooled and thus require airflow from the server blade.

Cooling the passive Alveos can become a hassle…

Given that the Datacenter Accelerometers are just starting to show up on the market, distributor margins are extremely wide. As of this writing, prices – as shown in the figure – range from 9000€ to 11000€ – definitely do run an OEMSecrets search before committing!

Distributor margins are extremely wide…