Skip navigation
1 2 3 Previous Next

Test & Tools

232 posts

MIT researchers create new sensors that warn you when plants are running out of water using carbon-based ink. These sensors are placed on leaves and let researchers examine the stomata (Photo from Betsy Skrip)


Plants can be a wonderful addition to any home. They make the room look nicer, can give off good smells, and just be a pleasant experience. But, what inevitably happens is you forget to water it, and there goes your plant in the trash. What if your plants could alert you when they’re running out of water? Engineers at MIT may have found a way to do just that.


The team at MIT created sensors that are printed on plant leaves that tell you when it's running out of water. The sensors rely on the plants’ stoma, small pores on the surface of a leaf that lets the water evaporate. When this happens, water pressure in the plant drops allowing it to suck up water from the soil in a process known as transpiration. When exposed to light, the stomata open ─ and it closes in the dark, which scientists are now able to study in real time.


So how does printing on a leaf work? The team used an ink made of carbon nanotubes, which are tiny hollow tubes of carbon that conduct electricity. The ink is dissolved in an organic compound called sodium dodecyl sulfate that doesn’t cause damage to the stomata. The ink can be printed across a pore to make an electronic circuit. Once the pore is closed, the circuit is active and the current can be measured by connecting the circuit to a device called a multimeter. The circuit breaks and the current stops flowing when the pore opens, which lets researchers measure when a pore is opened or closed.


Scientists study the opening and closing of the stomata over a few days and found that they can tell when a plant is running out of water. Results show that stomata take seven minutes to open after it’s exposed to light, while it takes 53 minutes to close when it gets dark. But during dry conditions, these responses change. If the plant doesn’t have enough water, the stomata take roughly 25 minutes to open and 45 minutes to close.


Not only could this save your house plants, but it could alert farmers when their crops are in danger. While there are devices that warn farmers of an upcoming drought, like soil sensors and satellite imaging, they don’t detail which specific plant is drying out.


Right now, MIT researchers are working on a new way to place electronic circuits by placing a sticker on a leaf instead of using the carbon-based ink. They believe this new research could have big implications for farming and could save more corps and plants in the face of drought or water shortages. And maybe now your houseplants can survive longer than a week.



See more news at:

Ford Performance is looking to improve your mental performance by relying on tactics used by professional drivers, VR, and an EEG. Prototype of a racing helmet integrated with EEG Ford is currently working on (Photo via Ford)


When you’re a professional driver, concentration is a must when behind the wheel. Letting your focus break could lead to a major accident, which is why they use certain mental training techniques. One of these methods includes a new brain scanning helmet that measures how racecar drivers improve their performance with mental training. This is part of a new study that explores these same mental training techniques could help us all deal with the stress of everyday life. 


While many athletes have been using mental training techniques to improve their performance for a while, it’s just spreading to the mainstream. Since people are constantly looking for new ways to deal with stress this new study looks at how these techniques used by athletes might improve our own brain performance.


Dubbed The Psychology of Performance Study, it’s being developed by Ford Performance, a motorsport branch of the US car maker in collaboration with King’s College London and tech partner UNIT9. The test works by using an EEG (electroencephalogram) headset that monitors your brainwaves. It’s ideal for this experiment since it’s versatile and has the ability to be used outside of lab setting. This way brainwaves can be studied in real-life situations and, in this case, VR.


Participants, which will include professional drivers and members of the public, will have their performance and brain activity measured throughout the test. Some participants will have prepared using mental techniques while others will have no preparation at all. This will allow researchers to see how these two groups perform with and without the mental training.


The team hopes by using VR that they can study how fatigue affects your driving and whether or not mental training can help improve focus and alertness for longer periods of time. Dr Elias Mouchlianitis of the Institute of Psychiatry, Psychology & Neuroscience, King's College London says the benefit of using VR is having “the subject is completely absorbed in your experiment; there are fewer distractions and you can control everything about the world that surrounds them in very precise ways.”


Aside from the study, Ford is also working on a prototype EEG race helmet for their motorsport team to be used in live simulations. Designers will integrate the EEG headset and sensors into a race helmet to measure drivers’ brainwaves in real-life practice environments. The results of the study will be published later on this month.



See more news at:

Scientists from EPFL are currently testing the use of VR to reduce phantom pain in paraplegics and those who suffer from spinal cord injuries. Phantom pain can’t be treated with medication, but VR may be the solution. (Photo from EPFL)


VR technology is still on the rise, but it still hasn’t become an every day thing for the masses. Typical complaints: the gear is too expensive and the games are usually just short demos rather than anything sustainable. But VR may soon find new life in the medical field. Doctors are using the tech in different ways to help them tackle different issues. A team of scientists from the Ecole Polytechnique federale de Lausanne (EPFL) have found a way to use VR to help paraplegics deal with phantom pain.


People who become paraplegics due to a spinal cord injury often have to deal with phantom pain, which sadly can’t be treated with medicine. With the team’s latest breakthrough, VR could be a cause for some relief. They had people wear VR googles, which showed live feed from a camera filming a pair of dummy legs. The camera was set up to mimic a person’s point of view in relation to their own legs. This gave the illusion that the dummy legs actually belonged to them.


From there, scientists then tapped the dummy legs and the area above the subject’s spinal lesions. After about a minute, the subjects felt like it was their own legs being tapped. They reported that the sensation helped reduced the neuropathic pain. Why does this happen? Team leader Olaf Blanke explains that the tapping is “translated onto the legs because the visual stimulus dominates over the tactile one.”


After the successful results, Blanke and his team are now working on a digital therapy that automates visuo-tactile simulations for those who suffer from chronic pain conditions and spinal injury patients. Amputees often experience a similar condition with phantom limbs, but the EPFL didn’t say whether this new technique could work for them or other people who have other conditions.


Incorporating VR into the medical field is becoming a more common practice. The EPFL isn't the first team to use VR for medical purposes. A team of Duke University researchers has developed a VR system that helps paralyzed patients regain some movement. A team from Oxford University is using VR to help paranoia patients face their fears and a team in Europe is using VR as a means to fight depression.


It’s great to see that VR is being treated more than a video game gimmick. While it may not be popular with all gamers, at least others have found ways to take advantage of the technology and use it for something that could change a lot of lives. It’s better than having those headsets live in the closet.




See more news at:

A demonstration, of sorts. A new system from Origin Wireless uses Time Reversal Machine technology to have WiFi detect the slightest movement and breathing in the room. This home security system uses WiFi instead of cameras. (Video via Origin Wireless)


I could have used this concept in my indoor motion sensing project.


Home security is a must when it comes to protecting yourself and your family, both inside and outside. For indoor systems, people turn to motion detection, which needs cameras and sensors to properly work, but there are some drawbacks here. Notably, the hardware installation costs can be high along with paying a monthly fee. And despite how comfortable we are taking selfies all the time, not everyone wants a camera on them all the time. This is where Origin Wireless comes in, which uses WiFi signals to detect movement.


So how does it work? The system uses “Time Reversal Machine” technology which is comprised of smart algorithmic work that doesn’t put a strain on the processor. The setup normally includes two hubs: one router for the “Origin” transmitter with the other routers acting as “Bot” receivers. These devices work on 5GHz over 802.11a, 802.11n or 802.11ac signals. The signal also relies on channel state information (CSI) to avoid any interference.


Normally, having WiFi signals that bounce around is bad, but this new device takes advantage of the delays to show the activity of the room. If something moves, the multi-paths will change along with the delay. The software can scan for changes 50 times per second and can detect any motion with an accuracy of 1 to 2 cm with a little bit of machine learning.  All this boils down to the machine can detect the respiration rates of everyone inside the room. Breathing rates are displayed on the live chart.


While it does sound creepy, almost like someone watching you while you sleep, it does show how much we can do with WiFi. And the company thinks they can do more with the system beyond security monitoring. They believe the system can be useful for the elderly who may live alone. Just imagine if they fall, the software can detect the rapid chance in the room along with a long silence.


Right now the device is still being tested and demoed, so we don’t have many real-world situations. But if it works as advertised, it could chance the way we think of home security. It may even be a tad bit safer than a camera set up since those can easily be hacked.


Have a story tip? Message me at: cabe(at)element14(dot)com

After several hiccups, the Thirty Meter Telescope has been approved for construction, but native Hawaiians still aren’t happy and won’t stop fighting the cause. Mauna Kea is considered sacred grounds that’s already home to 13 telescopes. (Photo from Flickr)


The Thirty Meter Telescope is meant to be the world’s largest telescope, but its journey has been a difficult one. Since its initial approval in 2011, its planned construction atop of Hawaii’s Mauna Kea has been halted numerous times. Recently, the project received major approval from the state. The state’s land board granted the project construction approval in a 5-2 vote, but this doesn’t mean those opposed to the telescope will stop fighting.


Mauna Kea is considered sacred grounds and holds religious and cultural significance to Native Hawaiians. Many activists have put a lot of effort into halting construction over the years, including blocking construction crews from heading up the mountain in 2015 and the project’s website being hacked in the same year. The state’s Supreme Court even stepped in to nullify the project’s permit in December 2015 since it was granted without giving those opposed a chance to air their concerns.


To make matters worse, the Thirty Meter Telescope wouldn’t be the first on the grounds. There are already 13 telescopes built on it. Mauna Kea is a popular spot because it provides a clear view for most of the year with limited light and air pollution. This new one is supposed to give us a deeper view into the universe. With it being three times wider than the current largest visible-light telescope and with a high resolution, it’s supposed to be better than the Hubble Space Telescope.


The project’s new permits come with some stipulations. First, they have to commit to cultural and natural resources training, and they have to follow strict environmental regulations. They also have to hire local residents for jobs generated by the project “to the greatest extent possible.”


Despite these stipulations, natives are still not happy with the move and are already filing motions to put the permit on hold until an appeal can be heard by Hawaii’s Supreme Court. Protest leader Kahookahi Kanuha said “For the Hawaiian people, I have a message: This is our time to rise as a people. This is our time to take back all of the things that we know are ours. All the things that were illegally taken from us."


Not all Native Hawaiians are opposed to the construction. Many believe it will create great opportunities for kids and will greatly benefit the community. Also, the observatories on Mauna Kea are a big part of Hawaii’s economy brining in about $60 million in earnings and taxes in 2012. With this latest addition, they would receive a great boost in earnings. Even if the new telescope goes ahead as planned, it’s clear protestors aren’t going down without a fight.


Have a story tip? Message me at: cabe(at)element14(dot)com

Google Clips is a small camera that can be worn or placed in a room that will autonomously take photos. Innovative or creepy? (Photo from Google)


I am considering making a Raspberry Pi bodycam for a future product. Then I saw this product last week, now I have to step up my design a notch or two.


Sometimes it feels like the phrase “Live in the moment” is uttered by people more than “Hello.” Whenever you pull out your smartphone to snap a photo or record a short video clip, it can be the first thing out of people’s mouths. There’s nothing wrong with wanting to capture a moment, but if you’re focusing too much on taking pictures, it can be a distraction. Google may have a solution with Google Clips, a body cam that automatically snaps pictures.


Clips is a small, square camera that you can place on your body or just leave in a room, which is recommended by Google. It uses AI to automatically take “motion pictures,” a new picture format that includes brief movement around the frame – similar to Apple’s Live Photo. It doesn’t catch audio, and it doesn’t use any kind of network connection, so you don’t have to worry about accidentally broadcasting something. To start capturing, just twist the lens to turn it on, then set it and forget it.


The camera keeps an eye out for anything it finds interesting, like certain people, facial expressions like smiles, and other indications it should start recording. Over time, it will learn faces and will take pictures of those people rather than a lot of strangers. Plus, it can also recognize pets. Some of its other features include a 130-degree field of view, Gorilla Glass 3, USB C, Wi-Fi Direct, and Bluetooth connectivity. It also comes with 16GB of onboard storages and offers up to three hours of smart capturing per charge.  


While it’s interesting idea, you can’t avoid the fact that it sounds creepy. Not too many people are going to be fine with a camera capturing random moments without their input. And Google knows this. Google product manager Juston Payne addresses this by saying the design of the camera is meant to be obvious. There’s no question that the little device on the table is a camera. Also, everything on Clips happens locally. The only thing that is synched with Google Cloud are the photos you save onto Google Photos. It won’t take pictures of faces it doesn’t recognize and the clips are only stored on the camera itself. They’re also encrypted in case you lose it.


Google Clips retails for $249. With a steep price tag, it’ll be hard to convince most consumers this is something they need. Smartphones are equipped with cameras and most of them are pretty good. While there will be some people interested in Clips, it may have a hard time finding a mass audience.


I am wondering if I could beat that price with my Raspberry Pi version.



Have a story tip? Message me at: cabe(at)element14(dot)com

Nike is using NFC tags in new NBA jerseys that connect with Nike’s app to give fans exclusive content. Nike’s new way of connecting with NBA fans (Photo via Nike)


When the NBA season kicks off, players will be sporting new jerseys by Nike; Adidas has been making their jerseys since 2006 but didn’t renew their contract this time around. But aside from wearing the classic Nike logo, the company is taking jerseys to the next level with NikeConnect.


The new jerseys come with authentication tags powered by Near-Field Communication (NFC) that can be hooked up to your smartphone via the NikeConnect app. This gives fans access to exclusive content from their favorite team and players, such as videos, tickets, highlights, and Gifs directly from NBA. Think of it as a NBA social network for your favorite team right at your fingertips.


To get access, all you have to do is scan the tag, and you’re in. As a bonus, when you buy a specific player’s jersey, you’ll get a “boost” code that makes that player better in the NBA2K18 video game. And, as you would expect, NikeConnect also wants to sell you stuff. The app will give you the chance to purchase limited edition products, like shoes and other gear picked especially for you depending on whose jersey you’re wearing.


Connect will give Nike more information about consumers, who bought what player’s jersey, where are they from, and where did they scan from. This information could then be relayed to the player giving them the option to interact with fans who bought their jersey rather than sending out a message on social media that millions will see.


So what about price? You’d think with the new technology, Nike would hike up the costs, but the company says this was not the intent. The new jerseys run between $110 and $200, which is pretty average for your standard jersey. The most expensive ones will be the authentic models, which is made from the same material NBA players wear.


NikeConnect could be used in other ways beneficial to NBA, like fighting against counterfeit merchandise. The NFL is currently doing something similar where NFC tags are used to keep track of memorabilia. If this new technology proves to be successful, don’t be surprised if other sports pick up on it. Nike already has a deal with NFL and big soccer clubs, like Paris Saint-Germain, Chelsea, F.C. Barcelona, and Juventus.


NBA Connected jerseys are available to purchase online and from retailers now. Just don’t expect to see your favorite player sporting these jerseys; they’re strictly meant for fans, at least for the time being.


Have a story tip? Message me at: cabe(at)element14(dot)com



Google’s new Pixel Buds go beyond your average headphones and will translate into 40 different languages in real time. Pixel Buds are more than meets the eye. (Photo from Google)


Babel Fish a reality?


Many thought Apple reinvented ear-buds, or were just crazy, with Air Pods. But once again, Google has blown them out of the water with their new ear-buds for the upcoming Pixel 2. The company unveiled the updated phone along with the new headphones at the Pixel 2 event this week. While they showed off a lot of updated tech, it was their new wireless headphones that stole the show.


What makes Pixel Buds different from your average Bluetooth headphones? The ability to translate 40 different languages in real time. Once the headphones are paired with the smartphone, you tap on the right earpiece and issue a command to Google Assistant. Along with playing music, providing directions, and making a call, you can tell it to “help me speak Japanese” and start speaking in English. The phone’s speakers will play your translated words as you speak them, which you’ll hear via the ear-buds.


Adam Champy, Google’s product manager, described it as having a personal translator everywhere you go. During their demonstration, there wasn’t any lag time when it came to translating. Of course, we’ll have to wait and see how this holds up once placed in the real world with background noise, weak WiFi connections, and crosstalk.


Google uses a similar method for its Translate app. Once you activate the live mic, the app will listen to your sentence in English and translate what you just said in 40 different languages. Other companies have similar technology, like Skype’s Live Translation feature, which works with four languages in spoken audio and 50 over IM. However, these translations aren’t necessarily in real-time since there’s a lag between when the original message was sent and when the translation arrives. 


Google’s Pixel Buds sound truly amazing. Imagine the possibilities when traveling around the world. The headphones will allow you to hold a natural conversation. You won’t have to rely on translation websites, which are spotty at best or carry around language dictionaries. No more awkward moments filled with hand gestures and cringe-worthy mispronunciations. It may even replace popular language learning tools, like Rosetta Stone.


Pixel Buds come out in November and will cost $159.


Have a story tip? Message me at: cabe(at)element14(dot)com

Hi, this is probaly not the right forum, I'll be grateful for hints on where to go, in element 14 or some other group/forum:

As surplus from my old job I have an old HP5308A counter/timer that does'nt work and I'd like to fix it.
I have manual and schematic for the HP5308A part of the instrument, and
testing around it seems the problem is that the 10 MHz clock signal is dead.

Unfortunately, the clock is generated in the top section, HP5300B, the measuring unit,
and for that I have not been able to find an manual or schematic.

Any help appreciated,

This series of posts relates to the work of  jancumps and  peteroakes in their WIP Programmable Electronic Load  blog.


In this post I'm going to look at stability and the way we might get some measure of how good it is other than looking at the step response as

I did in the earlier blogs. I did a post in the WIP blog on stability, so I'm repeating myself a bit, but it seems sensible to go through it

again before adding a bit more about interpreting the features of the open-loop response and how they relate to the closed-loop response of

the final instrument. Note that I'm learning how to do this as I go along, so this isn't a tutorial and you can't depend on it in the same way

as you could with a textbook - keep a skeptical mind and question what I do and whether it makes sense to you. If you know I'm getting it

wrong, I'd appreciate if you said.


To look at the stability I'm going to open the loop so that it's no longer functioning with the overall negative feedback that corrects for

the error term; now it just behaves like an amplifier. Here's how it looks as a circuit in the simulator:




I still don't know how you're supposed to do this, but I've developed my own method which seems to work ok although it's a bit fiddly to set

up. Then I'll plot the frequency response for the signal magnitude and for the phase - the two of them together are usually known as a 'Bode

plot' and are the basis of one way to derive a measure of the stability of a system. The one thing that complicates things for me with this

circuit (and my method of doing this) is that the bias for the MOSFET comes from the loop action, so to get the frequency and phase response,

which the simulator will do with a very small sinewave, I need to get that bias set up, which I do by manually putting a DC offset on the



It doesn't matter where I open the loop. We're assuming that the circuit is linear for small signals and won't limit, so the overall response

is the same whichever order the individual elements around the loop go in. I've chosen to do it after the integrator. The normal input is held

constant and a voltage generator drives the point where I've broken the loop.


I'm going to start with the situation where there's no inductance in the output circuit and look at the two cases from the previous blogs -

one with the original C2=4.7nF/C3=100pF values and the other with my revised C2=1nF/C3=220pF values.


Here's the magnitude plot. The simulator gives this to us as a gain in dB. I've combined the responses for the two cases on the same graph so

we can contrast them and see how they differ. [Just for comparison, the red line is the open-loop response of one of the op-amps used in the

circuit. I drew that on by hand, using the curve in the datasheet for reference].




If you're not used to using dBs, or are unpractised in their use, this is a voltage plot - what's being measured is simply the ratio between

one voltage amplitude and another (the output amplitude divided by the input amplitude) - so a gain of 0dB is a ratio of 1 (ie the output is

the same amplitude as the input) and each 20dB is a factor of ten (so 20dB is a gain of 10, 40dB is a gain of 100, 60db is a gain of 1000, and

so on). Negative figures are attenuation, so -20dB is a gain of a tenth. [Strictly speaking, we shouldn't use the term dB for a voltage gain

without specifying the impedances involved, but everyone uses the term for simple ratios like this, so I will too.] Plotting both the

magnitude and the frequency as log scales became the norm because simple circuit elements like capacitors and inductors have responses that

are basically straight lines on such a graph and, before the days of computer simulation, that allowed engineers to graphically piece-together

a circuit response from the circuit elements.


Here now is the phase plot. Again I've composited the two curves together - this time I've made one of them slightly fainter than the other so

that we can distinguish them properly. Remember that this is still the situation with no lead inductance in the output circuit.




The critical phase here is 0 degrees. (In a text or lecture, this will be presented to you as 180 degrees - see the note below for an

explanation of how what I'm doing here differs from the normal treatment. Sorry if this is a bit confusing.)

The shapes of the curves are quite complicated. Each part of the curve relates to different components within the circuit. Each time you see

the curve change direction, that's another component coming into play. The components that affect this are the capacitors and inductors since

they each contribute to the frequency response. Some of those components are obvious - they're drawn as capacitor symbols on the schematic -

but others aren't - the MOSFET has a fair amount of intrinsic capacitance, the leads in the output circuit will have some inductance, and both

of the op-amps have an internal capacitor to tailor its frequency response and ensure it remains stable at higher frequencies. As far as I can

see there are five integrators and two differentiators in total. We can already see, by plotting two curves, the way that C2 and C3 influence

both the frequency curve AND the phase curve. I'm not going to show all the experiments I did here, because there is too much of it, instead

I'll just say that C2 is responsible for that initial slope down on the gain plot and C3 is what counters that and lifts it at the end. [Keep

firmly in mind that this response is for the open loop and that the overall response will be different for the loop once it is closed and the

feedback is operating.] C2 has been sized to be the 'dominant pole' in the system - it rolls off the response at a much lower frequency than

the other integrators. The whole of the plot from 1Hz up to 1kHz is determined by that capacitor - it gives the 20dB per decade roll-off of

the gain plot and the 90 degree phase change on the phase plot. That's deliberate because it leads to a well-controlled situation that isn't

dependent on a fragile balance between several very similar elements that would put you at the mercy of small changes due to component

tolerances. Having a dominent pole is the standard way of compensating a transistor amplifier to ensure high frequency stability when you have

feedback over several stages and is also what is done for an internally compensated op-amp (which is, after all, a multi-stage transistor

amplifier that will normally operate with feedback). So far, with C2, what's being done here doesn't really differ from what an amplifier

designer would do. What's different is the area between 1kHz and a few tens of kHz where the differentiator comes into play. It lifts the gain

(there must be another factor at work here because it doesn't manage to get the response level) and it changes the phase in such as way as to

(hopefully) improve the phase margin. Above that, there are several other factors coming into play, including the response of the op-amps and

the RC network made up of the gate capacitance of the MOSFET and the gate drive.

In terms of stability, both of these curves are very safe. Up to the points where the gain curves cross the 0dB line, the phase is more than

90 degrees away from the critical phase of 0 degrees. And at the point where the phase falls to 0, the gain is something like -45db. Those

would be regarded as very safe margins.


Now here are the curves for the 4.7nF/100pF case with lead inductances (each lead, so twice for the circuit) of 0uH, 10uH, 100uH, and 1mH, so

that we can see the effect of inductance in the output circuit.






The inductance pulls the phase response down and steadily lowers the frequency at which a phase of 0 will occur. We can see why lifting the

phase response with the differentiator was of value because it partially counters the change, though it can't stop it if we throw enough lead

inductance in there. The effect on the gain curve is somewhat less evident, with most of the change occuring below the 0dB level - above that,

it's still dominated by the integrator response. However it's uncomfortable the way the gain curve sits just below the 0dB level, particularly

in the 20uH case where it moves back up before dropping away. Since that's close to the 0 degree point for 20uH, the gain and phase margins

are poor and the response will ring badly. The step responses in part 3 do show the 20uH case as ringing badly. If I were the designer,

that point would concern me, because you could imagine the whole thing actually oscillating if component tolerances resulted in the gain

curve moving up a bit more and the user happened by chance to get the output leads the right length. Overall, the phase margins are quite good

but the gain margins are poor - the value of the differentiator in improving the phase is countered by the way it brings the gain curve up

towards the 0dB level.


Could this be done better [the design I mean, not my amateurish analysis]? To be honest, I don't have a clue. I don't even know how you should

design such a circuit from first principles. Perhaps someone would like to tell me in the comments, or at least give me some clues.




One of the roadtests currently on the go is for a TI educational converter board that includes a very good programme of experiments with

commentary and discussions. In the commentary that goes with the experiments, the critical phase is presented as being 180 degrees. Since I

don't know anything about all this theory and TI most certainly do (they wouldn't do very well at designing DC-DC converter chips if they

didn't), my first thought was that I was getting things horribly wrong. However, when I first did this on Jan's WIP blog, I had sat down and

considered it very carefully and convinced myself I was right, so why the difference? Here's the argument: I am looking at the phase all the

way round the loop; if a sinewave is applied at such a frequency that the loop phase is 0 degrees, that will reinforce the original and the

oscillation will grow; if the loop phase is anything else it will tend to counter the original change, though if it's close to zero that may

take multiple cycles to settle which we would see as ringing. So is TI wrong? No, of course not! What's happening is that we're both right -

basically, we're talking about different phases. My phase shift is once around the entire loop to the point where I started and it includes

the 180 degrees implicit in the term 'negative feedback'. They are talking about the forward path and assuming the the response of the

resistive feedback plays no part, which for an amplifier or a converter at low frequencies would be true. So I think that's why we differ,

though I am slightly unsure of this - I really ought to see if I can get an up-to-date, modern, control systems textbook to work from rather

than the one I'm using [1] and see if I can understand better the conventional ways of doing this. I don't think it matters in the results I'm

getting - I've got a cross check by interpretation of the step responses for the same conditions, so I'm not far out - but, if you learned to

do how to do this properly in a classromm, stick with what you know and don't copy me.


[1] Dorf, R. C.  Modern Control Systems (2nd edition, 1974)


Does anyone have opinions and recommendations for a more modern book on control system design? [Preferably something orientated towards

electronics and common enough that I could pick one up secondhand for a few pounds.]


Programmable Electronic Load: Dynamic Behaviour: Part 1 Overview

Programmable Electronic Load: Dynamic Behaviour: Part 2 The Servo Loop

Programmable Electronic Load: Dynamic Behaviour: Part 3 Effect of Output Inductance

Programmable Electronic Load: Dynamic Behaviour: Part 4 Effect of Output Voltage Change

This series of posts relates to the work of  jancumps and  peteroakes in their WIP Programmable Electronic Load  blog.


In the previous parts, I looked at the effect of changes to the control input. Now I want to look at

the effect of holding the control input constant and varying the voltage of the output. That change

of voltage will result in the circuit having to vary the resistance of the MOSFET, to compensate and

bring the current back to the desired value, and the way it deals with that change may be different

to the way it handles a change to the control input.


Here's the circuit, much the same as before.




Initially, I'm going to try inductance of 1mH in each

lead, which is high but will show clearly what the effects are. I'm going to step the voltage of VG2,

the voltage source for the output loop, by 2 volts up and down between 8V and 10V. That's quite

arbitrary - I just want to get a feel for how it behaves. I'm going to have that voltage slewing

rather than a step change because any real power supply would have capacitance associated with it.


Here it is with the original compensation/servo components (C2=4.7nF and C3=100p)




And here with my alternative values (C2=1nF and C3=220pF)




I find this fascinating. Which is better? My immediate reaction was that my alternative values, which

so obviously ring, are worse than Peter's original values. And yet the excursion from the desired

value is less at its extremes and it settles quicker than the original values which, to a first

glance, looks much more controlled and graceful.


One thing that is evident to me now is that tuning a servo loop for a real system (rather than a

simple textbook example) isn't at all straightforward. Where there is more than one variable that can

change it get even more complicated because we're then making decisions about what is important to us

and doing trade-offs.


I'm going to leave this up in the air a bit - there are probably component values that would reduce

the excursion of the first waveform a bit and get it back on course quicker, and it would be natural

to experiment with that for a while, but I want to move on to looking at the open-loop response and

how a traditional Bode plot approach to stability views this.


Programmable Electronic Load: Dynamic Behaviour: Part 1 Overview

Programmable Electronic Load: Dynamic Behaviour: Part 2 The Servo Loop

Programmable Electronic Load: Dynamic Behaviour: Part 3 Effect of Output Inductance

Inexpensive, precise, stable and rugged, the most practical
solution for RTD simulation: Bulk Metal Foil trimmed to the exact value of PT100 @ +125C, 147.952 ohms. 


iNotes-quick webinar:


RTD Simulators – Product sheet:


Resistor technical datasheet:

Platysen’s Seal is an analyzer that tracks your stats, swimming strength, and movements. Like other wearable tech, Seal works with a companion app (via Platysen)


Whether you’re a fitness enthusiast or a professional athlete, wearable tech proves to be a great tool. A bit more underused than I thought they would be by now. Keeping track of your stats lets you know what needs improvement and movement reminders let you know when it’s time for exercise. But if you’re a swimmer something as basic as FitBit may not be the best tool. Thanks to a new device by Platysen, swimmers can get detailed stats about their techniques and forms.


Seal swim analyzers are small, wearable rings that help swimmers measure their hand’s movement and force as they push through the water. The built-in sensors also keep track of hand symmetry and rhythm. The device works with a companion app that gives you more information and your stats, like whether or not one hand is weaker than the other. Seal synchs up to your smartphone wirelessly giving you the option to share your stats with your coach for further training.


Right now, there’s no release date or cost for Seal as it’s still in the prototype stage. They showed off the concept at IFA 2017. Platysen hopes to turn the idea into an actual product six months from now. Meanwhile, the company also has another swim related device: The Marlin.


Marlin is a swim meter that tracks your performance through voice feedback using a bone construction headset. Worn on your head, your pace is reported to you while swimming with no interruptions. Your swimming performance is captured via GPS while in open water. There’s even a pool mode that captures motion data via sensors. To see how you did, connect Marlin to your phone via Bluetooth. Unlike Seal, Marlin is available for purchase with a price tag of $150.


While both devices are sure to be great tools for swimmers, Platysen isn’t the only company targeting athletes. Also appearing at IFA 2017, Samsung revealed a new partnership with Speedo. The two are working together to create a new app for the Gear Fit Pro 2 and the Gear Sport. The new app will let swimmers track their laps, and the time it takes to finish them along with burned calories and distance traveled. It’ll work with Samsung’s S Health app to view your stats in one place.


If these devices prove to be successful and actually work, don’t be surprised if more companies designed wearable tech catered towards certain athletes. Maybe next will come a wristband for tennis players that keeps track of how hard you swing. You never know.


Have a story tip? Message me at: cabe(at)element14(dot)com

health sensor.jpg

From a pen device that can determine cancerous tissue to health screening apps, more and more people are opting to go digital when it comes to healthcare. This pen-like device can tell you whether or not tissue is cancerous (Photo from University of Texas at Austin)


With so many initiatives and programs dedicated to staying healthy, more and more people are doing their best to keep a healthy and fit lifestyle. Whether it’d be getting enough exercise or adjusting your diet, it seems leading a healthy life is now more important than ever. To help with these goals, several apps and wearable devices have released to help keep track of your workout stats and even give you reminders to stand up, like FitBit. The popular wearable is a great way to ensure you’re staying active, but it could also help those with diabetes.


FitBit is teaming up with Dexcom, a company that creates continuous glucose monitoring (CGM) devices, for the new Iconic smartwatch. How it works is the CGM devices are paired with a sensor that’s placed under the skin and measures a person’s glucose levels every few minutes. Currently, you need a transmitter attached to the sensor to see the readouts on a smartphone or Apple Watch. FitBit and Dexcom are hoping to make this an accessible feature for the Ionic watch by 2018. But what about greater health issues that could lead to fatal diseases? Thanks to some devices in the works, detecting health issues early will come at ease.


First, there’s the MacSpec Pen, which is designed to help determine whether tissue is cancerous in roughly ten seconds. The small, pen-like device works by issuing a drop of water onto the surface of the tissue. The water then pulls in molecules from the tissue, which is analyzed by a mass spectrometer – a machine that rapidly measures the mass of chemicals. From there, the systems reports the status of the tissue.


This is a big step forward for doctors working with tumors. The status of a tumor can’t be determined until after surgery. If it turns out some cancerous cells were left behind, another surgery is required to remove them. With the new device, doctors can easily determine if they need to remove more tissue or if they already took out all the cancerous cells.



With digital health on the rise, it’s no surprise Google wants to get in on it. Recently, they acquired Seattle startup Senosis Healthy, a company devoted to turning smartphones into health devices. The company has already created apps that let you run checkups that normally require a sensor. There’s HemaApp which checks your blood’s hemoglobin count with the phone camera and SpiroSmart which uses the microphone to measure your lung functions.


So far, it’s not clear how Google intends to use the company’s assets. The team will apparently stay in Seattle and act as the “backbone” of a digital health team. It’s clear Google wants to do something larger and probably want to build on Senosis Healthy’s well-established apps. They may even want similar apps built into Android devices.



With these breakthroughs, people can not only monitor their own healthy more efficiently, but doctors are able to improve their procedures. It’s good to know that a lot of on the cusp technology is going into devices that could save lives instead of making your phone slightly better.


That was ten seconds, here are your results. That’ll be $1698.


Have a story tip? Message me at: cabe(at)element14(dot)com

Unfortunately my trusty old Rigol DS1052e oscilloscope suffered from damage on the way to the calibration laboratory.

Damaged Scope

The control knob for the trigger level had been broken off and the calibration lab advised that they could not repair the device, so I asked for it to be sent back to me to see what I could do with it.


A search of the internet revealed some information on the encoders utilised within the scope as there had been a few posts on forums of failures of the encoders.


Getting hold of the right encoder however has been a little troublesome. Details I found, identified them as an Alps EC12 series encoder, with 12 pulses per revolution, no detent, a single pole switch underneath and a 15mm long flatted shaft.


Searching the common electronic component suppliers within the UK did not produce the right encoder. There were several Alps encoders to be found that were either 12 pulses without the switch or 18 and 24 pulses with the switch. I eventually found a Bourns encoder that offered the right number of pulses along with the switch and mechanical size constraints, but these were not going to be in stock until December.


As I need the oscilloscope for some work testing generator rotors in October, I decided to opt for a 12 pulse encoder but with detents and without the switch, available from Farnell, part number 206-5051, as an interim measure to get the scope back up and running. The switch element just allows the trigger level to be zeroed quickly, which in all honesty, I can't actually ever remember having used.


Scope DismantledThe scope was dismantled relatively easily. The on/off switch cap has to be pulled upwards to remove it, a small piece of card wrapped around it allows a pair of pliers to be utilised to pull it upwards. There are then four torx head screws that hold the back cover in place. Two at the bottom are easily found, but there are also two screws at the top hidden by the handle, that needs to be positioned correctly to allow access to them.


With the switch and screws removed, the rear case can be removed. The case has to be prised over the two screws that hold the mains socket in place. These cannot be removed before taking the rear case off as they are actually a nut and bolt assembly, through the inner chassis.


With the rear case off, the screw adapters either side of the D-socket are removed to allow the rear shield to be removed. This just pulls straight off as there is nothing else other than location tabs to hold it in place. The top shield can then be removed by unscrewing the top two torx screws holding it in place.


One of the screws holding the front cover in place is located behind the power supply board and so the supply board must also be removed. The connector to the main board is unplugged, and a torx screw secures an earth connection to the shield. With this removed, the two nuts and bolts securing the mains socket to the shield are undone. This then just leaves four screws holding the pub to the shield. With these removed, the power supply board can be move forward to gain access to a three pin plug at the rear of the board. After removing the power supply, all of the control knobs are removed and then the front cover can be removed to gain access to the control board containing the encoders. There are only five acres that hold the front cover in place and these are easily found with the supply board removed.


The control board itself needs to be disconnected from the main board and then the four screws securing it are removed. The whole board can then be removed while feeding the connecting ribbon cable through the slot in the inner case.


Control Board RemovedNew encoder fittedBoard SolderingInspection of the board revealed that some manual soldering has probably already been undertaken during assembly, seen as white deposits around the original solder connections. The damaged encoder was desoldered and removed from the board and the new one soldered in its place.


The oscilloscope was then part assembled, with the control board, front cover and power supply board all refitted, to allow the oscilloscope to be powered up and the operation of the new encoder tested. Remember to reconnect the small plug behind the power supply board and the earth lead to the chassis.


With the new encoder for the trigger level functioning well, the rest of the oscilloscope could be reassembled. Now all I need to do is get the oscilloscope off to the calibration laboratory and it will be fit for use again.



Encoder dismantledWith the oscilloscope rebuilt, I couldn't resist dismantling the encoder. Four little tabs need to be prised upwards and then all of the switch and the rotary encoder can be taken out of the metal frame.


The plastic shaft has been snapped and there does appear to be some bits missing from it as there is no extension left to go through the encoder mechanism and push down onto the switch. The switch itself, just consists of the two contacts formed into the plastic case with a metallic disc installed over the top to act as both the switch to short the contacts and the spring loaded mechanism when the plastic shaft is released.


A more detailed examination of the encoder showed that there was actually eight contacts for each of the two outside pins of the encoder. This definitely isn't a 12 pulse per revolution encoder, which probably explains why I struggled so much to find one to replace it with.


Encoder TestingEight contacts to me suggests a 24 pulses per revolution encoder and this was confirmed by reassembling the encoder mechanism and testing. As I don't have any digital testingcapability other than a home built logic probe from back in the days when I was an apprentice, it boiled down to testing the encoder with an ohmmeter, which took a few of revolutions to confirm that it was indeed a 24 pulse per revolution encoder.


A quick scan through the Farnell website and I found a Bourns compatible encoder, part number 266-3518, to match the original Alps encoder, with the added bonus of them being in stock.


Might have been a better idea to have stripped the oscilloscope first and removed the encoders to identify what they were. Ah well, I will probably strip down the oscilloscope again after I have completed the testing required, and replace all of the encoders, as the other ones were showing signs of wear.

Filter Blog

By date: By tag: