Allegro’s ACS723 Sensor Provides a Low Resistance of Just 0.65 Milliohms

Measuring currents is one of the oldest tasks of metrology…


Measuring currents is one of the oldest tasks of metrology. The knee jerk reaction involves adding a resistor to the current path and then analysing the voltage drop. Hall effect sensors provide a variety of benefits ranging from simplicity to galvanic insulation.

Allegro’s ACS723 sensor comes in a handy SOIC8 package, and provides a low resistance of just 0.65 milliohms. System design is simple – simply look at figure 1, which shows how to get the Allegro chip into your design.

Instead of having to play around with a differential amplifier, simply supply between 4.5 and 5.5 V to the component. Then, change the current to be measured to the device using pins one, two, three and four. Allegro provides two pins each in order to simplify layout and reduce voltage drops on the PCB and the interconnect.

The actual current output is then dependent on the selected version of the component: as of this writing, Allegro provides a total of seven different versions handling current ranges from 5 to 40 Ampere. In all cases, however, an analog voltage is put out which can be used without further meddling.

A small trap…

The only important consideration is the minimum input impendance of the attached amplifier, which is set to 4.7 Kiloohms. Most modern ADC parts do not have a problem meeting this – in the worst case, a simple operational amplifyer can be added to provide voltage buffering.

Whilst traditional resistor-based sensors need careful placement, a Hall effect sensor does not care about voltage potentials. You can set the sensing loop close to supply or close to ground – no special considerations need to be taken when designing the measuring loop.

Furthermore, resistors heat up significantly as currents raise. Our Allegro chip impresses with relatively constant thermic behaviour. Another very nice aspect of the product involves the presence of a bandwidth selection. Allegro provides you with a configurable filter, limiting the bandwidth of the filter to either 20 or 80 kHz. That way, unwanted noise and transients can be blocked out without having to perform additional filtering of the output signal.

Does it pay?

Feeding our price comparison engine yields a price range from £2.79 to £1.56 depending on the current range. Given the severe simplification and PCB space savings, going Hall effect definitely makes sense…

Texas Instruments’ BQ40Z50 Alternative

Many moons ago, creating a battery for a mobile device was simple…


The same, or different?

Many moons ago, creating a battery for a mobile device was simple. You grabbed a few nickel cadmium cells, trickle charged them and you were sorted. Today, advanced battery chemistry and their explosive behaviour make matters more difficult. Fortunately, TI‘s BQ40Z50 family provides a convenient alternative – and communicates via Intel’s SMBus standard.

The Texas Instruments chip handles primary protection for a battery pack consisting of between one and four Lithium-Ion or Lithium-Polymer cells switched in series. Providing a field effect transistor for charging and discharging is the responsibility of the developer; this is not a chip which works without the presence of a significant amount of additional logic.
Fortunately, TI provides suggestions for a lot of it. For example, the CSD17308Q3 part – conveniently found via OEMsecrets – is ideally suited to handle the charging and discharging of the cells due to its extremely low internal resistance.

Actual cell health is then determined via a group of relatively complex proprietary algorithms, which collect input from, amongst other things, a set of shunt resistors and thermal monitors which are to be embedded into each cell. Another nifty feature involves the SMBus interface mentioned in the introduction. This provides a simple way for the host processor to find out more about the current status of the battery. In systems which are based on an Intel microprocessor, this information can be used to enable or disable the overclocking feature of the CPU.

These features are rounded off with an SHA1 based authentication module and a set of flash based memories which stores lifetime information about the batteries performance and behaviour. This is especially useful for rejecting cloned batteries, and can also be used to reject warranty claims due to improper handling of the cells at hand.


A tale of two versions

Logic dictates that fitting all of this functionality into a 32 pin case requires some kind of compromise. In the case of the BQ40Z50, developers need to live with two problems. The above-mentioned need for a significant amount of external circuitry and the configuration, which takes place entirely using software. TI provides a programming utility called bqSTUDIO, which is furthermore described in a dedicated data sheet available here.

Finally, keep in mind that two versions of the chip exist; the deprecated BQ40Z50, and its successor BQ40Z50-R1. A quick comparison of the datasheets provided by the OEMSecrets team yielded no significant differences on the semiconductor side – the differences are likely to be found on the firmware side of things.

Analog Devices MEMS accelerometers

Analog Devices MEMS accelerometers offer ultralow power, precision impact and shock detection perfect for portable IoT applications.


These accelerometers are built to be intrinsically stable over time and temperature with no calibration required.

ADXL356/357 Accelerometers

The ADXL356/357 MEMS accelerometers feature low noise, low power and offer an excellent solution for Wireless Condition Monitoring nodes.

ADXL372/375 Accelerometers

Ultralow power, 3-axis, ±200g MEMS accelerometer features deep, multimode output FIFO, several activity modes and comes in a small, thin package.

Get your parts sooner with free 1-day shipping when you spend $50 or more with

Learn all about Texas Instruments’ comprehensive Power Management porfolio

Simplify your DC/DC design with LMZM33603

The LMZM33603 36V, 3A power module offers stellar EMI performance in a compact QFN package.
Order yours today and enjoy the peace of mind that comes with reliable performance, small solution size, and reduced design cycle time.

Power your processor with the industry’s first 5+3- phase PMBus controller

Powering high-current processors requires flexible solutions, and the TPS53681 PMBus multiphase PWM buck controller delivers with dual output 6+2/5+3 phase configuration. This multiphase driverless controller is paired with TI smart power stages for accurate current sensing and excellent thermal performance.

The next generation of TVS surge diodes is here

With a 94% smaller footprint, lower leakage and flat-clamping technology, TI’s Flat-Clamp portfolio provides more robust protection than traditional TVS surge diodes.

TI’s new TVS3300 family of Flat-Clamps gives designers precise control over clamping voltage vs surge current, even across temperature. It’s time to re-evaluate your TVS protection. Learn more in the whitepaper.

NEW: Industry’s smallest 8A single-rail module

Powering your 4A to 8A digital load just got smaller, easier, and more reliable. At 7.5mm x 7.5mm, the TPSM84824 17V, 8A step-down power module provides the smallest footprint in its class. And with TurboTrans™ technology, it delivers fast transient response with the industry’s lowest overshoot.

TPSM84824 comes with minimal external components, and pin- and footprint-compatible 4A and 6A alternatives.

Introducing TI’s first PMBus power module

Presenting an easier way to power your high-current digital core: TPSM846C23 is a PMBus-compliant DC/DC power module with an integrated inductor and FETs that crams 35A of output current capability into 240mm2. Two of these modules can easily be paralleled to deliver up to 70A of output current to the point-of-load from a nominal 12V or 5V input voltage.

Save space in your automotive design with TI’s synchronous buck-boost DC-DC converter

The TPS5516x-Q1 family provides a cost-effective single chip solution for a regulated output voltage between 5V and 12V, with up to 1A load current depending on voltage requirements. The buck-boost overlap control ensures automatic transition between step-down and step-up mode with optimal efficiency.
These buck/boosts provide a smaller solution size and higher efficiency than a discrete boost-buck converter solution.

Highest efficiency and lowest standby power without the audible noise

Want the lowest standby power? Need to simplify your system design for your AC/DC applications? Get to know our newest power factor correction (PFC) controller, UCC28056. With zero cross detection scheme (without auxiliary winding) and advanced burst mode, UCC28056 will help you design your next offline system.

6A three-level buck battery charger enables faster charge times for portable electronics

TI’s BQ25910, the industry’s first integrated 6A three-level buck battery charger, is a game-changer for how we will charge our smartphones, tablets, electronic point of sale (EPOS) and other portable electronics in the future. New power conversion technology reduces battery charging time up to 50% and allows for up to 5% efficiency improvements compared to conventional architectures.

Get great EMI and thermal performance the easy way

Any engineers out there looking for high-efficiency DC/DC conversion without the headaches that typically come with EMI mitigation and thermal management? Everyone? Ok, good.

Introducing the LM73605/6 and LM76002/3 families of wide VINsynchronous buck converters. Designed with high integration, thoughtful pin-out, and best-in-class thermal coefficients, LM73605/6 and LM76002/3 give you amazing performance without all of the headaches.

The world’s smallest 36V, 1A DC/DC solution

With a 27mm2 package size and only two external components, the LMZM23600/ 01 power module can convert a nominal 24V input to 3.3V in a solution size that is 45% smaller than competition. With the simplicity of power modules paired with the magic of WEBENCH®, you can bring the world’s tiniest 36V, 1A buck design to market with blazing speed.

To find the best price for Texas Instruments search OEMsecrets

The Benefits of MLCC Capacitors Part II

As discussed last week, MLCC capacitors will provide a variety of benefits to designers willing to use them…

As discussed last week, MLCC capacitors will provide a variety of benefits to designers willing to use them. However, recent developments have shown stock levels to become critical. The trade press and analysts are assuming that the situation will remain in a similar state for the foreseeable future. Fortunately, designers can react to this with a variety of behaviours.

Let us start out with the obvious one…

Electrolytic capacitors have improved a lot during the last few years. The problems seen on Danaher’s Tektronix TDS5xx/7xx series of oscilloscopes or in Boeschert power supplies will happen only if you use extraordinarily low quality parts. If your application allows it, using a low ESR electrolytic capacitor is a nice way to stay clear of the entire MLCC conundrum.

Should the larger weight or cost of electrolytic capacitors be unacceptable, going with the flow is an attractive choice. Companies such as Vishay, who are specialized on the automotive market, usually tend to be less squeezed than their consumer-oriented competitors. For example, a GRT188R60J226ME13D currently has a lead time of 36 weeks, while the VJ0805G226MXYTW1BC is available more easily (20 weeks).

OEMsecrets search, would then, allows you to minimize the distributor margin impact, allowing you to mitigate the total cost impact on your BOM – in some cases, price adjustments have made “consumer” parts more expensive than their automotive counterparts.

Smaller is better

Reputable German sources claim that Murata switched their manufacturing equipment away from classic case sizes such as 0805 or 1206, and instead prefer smaller ones such as 0402. From a person looking at raw materials, this decision makes good sense; the smaller the capacitor, the smaller the amount of rare earths required for its fabrication.

If your entire manufacturing takes place using pick and place machines, no significant action is required. Simply open your favorite EDA program, and reduce the size of the component footprints. 0603 size chips can be soldered by hand with good magnification.

Stockpiling works

Finally, the big question remains; should you stockpile parts, or should you not? Let’s start out with the obvious; with a tool such as OEMsecrets, your trawling net is significantly wider than that of the normal developer who is trying to source components from one or two distributors by hand.

This, however, does not mean that it does not make sense to stockpile parts. If your customer is willing to pay you for it, acting as a commis can be highly profitable. Having
a stockpile of a few thousand components is helpful also in that you can send them off later using the broker such as Kruse. On the other hand, most stockpiling operations fail for two reasons. First of all, companies tend to overfill their stocks anticipating demand which never crops up. Problem number two is the often insufficient quality of the stock-holding facilities – if you store capacitors wrongly, they might still be usable, but will be refused by reputable third-party stock resellers.

The Benefits of MLCC Capacitors Part I

Casual observation would claim that the main job of a parts search engine like OEMsecrets involves…

Casual observation would claim that the main job of a parts search engine like OEMsecrets involves finding cheaper sources for expensive parts. Recently, relatively low-priced MLCC capacitors have taken the lead in terms of search interest – let us tell you what this means for your daily work as a designer…

First of all…

Electrolytic capacitors are not the most desirable of products; they are heavy, large, and tend to fail early. Just a quick look at the various test equipment repair forums returns stories about expensive pieces of gear which croaked because some electrolytic capacitor decided to belch out it’s electrolyte.

Some years ago, a designer at a third-class German electronics trade show introduced the idea of trying to go electrolyte capacitor free. This can be achieved in two ways; firstly, by replacing classics such as the LM2576 with a higher frequency part such as the ACT4065 – they are not only cheaper, but also permit the use of smaller coils and smaller filtering capacitors. Remaining filtering capacitors which are still required are then chosen from MLCC families, paralleling multiple ones if necessary.

The inductor used for a 52KhZ switching regulator is much much bigger…

…than it’s counterpart on the 210KhZ design.

At first glance, opting for MLCC is a miracle. Let us compare the GRM21BR61A106KE19L to an electrolytic capacitor from a well known brand – by going through hole or by switching to a lesser brand, the electrolytic capacitor would of course become a bit cheaper…

Part GRM21BR61A106KE19L RNU1A100MDS1
Type MLCC, 10uF 10V Elco, 10V 10uF
Dimensions 0805 4mmx5mm, Throughhole
Weight 9.5mg Not published, similar part weighs 480mg with leads
Price, Q1 24 €-cent 35 €-cent

Taking a look at the actual specifications of the capacitor reveals the situation to be less rosy. The figure, taken directly from Murata’s datasheet, shows that the capacitance of the component decreases significantly as the applied voltage raises.

Actually getting this information can be tricky; whilst first-rate vendors usually provide this data themselves, designers working with cheaper parts can often find themselves forced having all oscilloscope and using the good old RC discharge curve to find out more.

As if this was not enough…

Recent news stories also point to shortages in the MLCC capacitor supply chain. Most distributors now provide warnings about manufacturing delays, with some parts having lead times of up to 38 weeks.

Whilst this might not be a problem for a large vendor such as Apple or Samsung, a smaller company with small order volumes will likely find itself on the short end of the stick if push comes to shove.

In short, MLCC capacitors do provide benefits, but are not a silver bullet for all capacitor-related problems. Tune in soon for the second part of our story, which looks at the various ways designers can handle the stock shortages in the MLCC capacitor market using OEMsecrets tools and a bit of ingenuity.

Wireless Freedom for the Pi-ple

When the Raspberry Pi Foundation introduced the Raspberry Pi Three, many a developer was fascinated by the integrated Wi-Fi module.


When the Raspberry Pi Foundation introduced the Raspberry Pi Three, many a developer was fascinated by the integrated Wi-Fi module. Sadly, practical performance was not quite as good. Let’s find out if the next generation brings any improvement.

The most important thing first: the Raspberry Pi 3B+ can now connect not only to 2.4 GHz networks but also to ones operating in the 5 GHz band. This means that you can escape congested areas more easily, albeit at the price of shorter range.

Performing actual Wi-Fi performance tests is a bit difficult – let us stick to using Ping and iperf, yielding the results shown in the table. The Raspberry Pi 3B had lower ping latencies, but the new model achieves better signal strength and more network throughput. This can also be verified in practice – surfing the web is much more comfortable on the new model.

Measurement Old PI New Pi
iPerf [ 3] 0.0-120.1 sec 5.90 MBytes 412 Kbits/sec [ 3] 0.0-124.9 sec 20.6 MBytes 1.38 Mbits/sec
ping 100 packets transmitted, 88 received, 12% packet loss rtt min/avg/max/mdev = 1.872/11.935/87.561/10.758 ms 100 packets transmitted, 84 received, 16% packet loss rtt min/avg/max/mdev = 1.728/14.319/151.091/23.329 ms

The inclusion of gigabit ethernet raised quite a few eyebrows: after all, the actual SOC is not able to provide more than about 480 Mbits of total bandwidth to all of the peripheral devices. In practice, this limitation shows up doing iperf runs:

0.0-120.0 sec 4.44 GBytes 318 Mbits/sec

Amper Tradeshow Fair

Whilst the Nuremberg-based Embedded World celebrates its birthday, the Czech Amper tradeshow is in the 26th year of continuous operation.


While the Nuremberg-based Embedded World celebrates its birthday, the Czech Amper tradeshow is in the 26th year of continuous operation. Let OEMsecrets guide you through the most interesting aspects of the fair.

First of all, semiconductor manufacturers and makers of test equipment usually don’t show up in person at Amper. Instead, they are represented by their distributors – funnily, one company can have multiple booths presenting it at the same time. This probably dates back to time of the Soviet Union: then, development largely took place with components selected from catalogs. As the Soviet empire fell, distributors took their place (see e.g. Howard Moon’s classic Soviet SST).

Given that Amper traditionally targets electrical engineering, one can clearly see that most cable manufacturers add intelligence to their connectors, switches and other electromechanical gadgetry – being unwilling to leave the value added services pie to third party suppliers is a wise business decision.

To speak of third party suppliers, there were many. For example, a start-up called 4Dot is offering the monitoring of ball bearings as a service: you buy the hardware from them, slap it on to your bearing and pay them some money every month. Another company offers a Wi-Fi button based on SigFox, which furthermore contains a few sensors. Finally, BigClown provides developers with an Arduino-like set of modules which can be combined into a ready-made PCB prototype easily.

In the metrology department, distributors show a lot of interesting components. For example, Transmille’s distributor provided one of their transfer standards which is the most modern 8.5 digit multi meter currently on the market (discounting new Russian units which are unlikely to become available in the West).

Danaher elected to become Farnell’s madness victim in this year’s Amper. Their DMM6500 multi meter, still in beta stage, was demonstrated at the booth. The system is not bad, but suffers from a slightly low resolution touchscreen which furthermore feels wobbly to the touch. Another issue is that the pricing of the unit seems to be extremely geo-dependent, in one country, it cost about €800, while another country’s residents have to pay about €1400. Help yourself, of course, by combining OEMsecrets with a travel sniffing service like GoEuro or Opodo.

Finally, technologies like SigFox and LoRa are coming to age. Starnet is a network operator who covers almost the entire Czech Republic with its custom LoRa network: developers of smart devices can use this low-power network as an alternative to 3G and 4G connectivity. According to the vendor, device congestion is not an issue: in practical tests, up to 20,000 units were handled by a single gateway.

People who feel gouged by Messe Nuernberg will be happy to find Amper significantly more affordable. Booths are cheap, while Czech hotels have yet to figure out the trick of really raising prices during a trade fair. Food prices are also very fair and in line.

To cut a long story short: for people who create smart devices, the world is your oyster – fairs like Amper show that a good business model can be found by trying to create value and reduce pains. OEMsecrets, of course, is always at your side when it comes to saving money during component purchases – let us help you take your next idea to fruition.