Capacitors are all about storing and releasing charge. The main difference with batteries is that this charge is stored directly as electrical energy, whereas a battery converts to / from chemical energy in some form.

In an ideal capacitor, charge and discharge follow an exponential curve. Charging takes place when connected to a fixed supply via a resistor, discharging is a matter of placing a resistor across the capacitor:

(image copied from www.technologyuk.net)

The “time constant” is the level when the discharge reaches 36.8% or the charge reaches 63.2% of the original voltage. It can be calculated using the formula: T (seconds) = R (ohm) x C (farad).

This property makes it easy to measure the value of a capacitor: charge it up to a known voltage, then discharge it through a known resistor and measure the time it takes for the voltage to drop to 36.8% of the original voltage.

This is the approach taken by this capacitance meter kit by Radio Hobby Store:

There is excellent documentation including very detailed assembly instructions, leading to this:

And sure enough, it works as expected – measuring a 10 µF cap in this case.

The only two drawbacks I found is that it doesn’t measure caps larger than 50 µF, and that there is no on-off switch. With a tool like this, you tend to want to use it from time to time and put it away after use. Without the switch, you have to disconnect the battery each time – a bit awkward and inconvenient.

This meter is based on a pre-flashed PIC controller. There’s one button to calibrate its zero value, and a convenient “auto-zero” mechanism, which keeps adjusting for exactly 0 pF when no capacitor is connected.

The amount of solar energy available indoor is very limited… a very small fraction of outside, I’m sure.

Still, this unit has been running in the house here for a few years now, and not near a window either:

It’s a attractive goal: gadgets which you buy (or build) once and then use essentially forever!

We don’t use the alarm clock mode of this thing, so the beeper never goes off, but it does listen for the DCF77 clock standard transmitter in Germany once a day to stay in sync. It’s also slightly glow-in-the-dark, so this clock remains readable at night.

The fact that this clock works so well tells me that, with proper care, we should be able to run simple nodes inside the house with a solar cell of perhaps a few square centimeters, just like this clock.

And, whether battery- or supercap-powered, that’s precisely what I’d like to get going…

As with the AC-mains connected ultra-low power supply, I suspect that reliable startup will be the hardest part. Such an energy source will have very little spare energy and charge up very slowly, so when the JeeNode comes out of power-up (or brownout) reset, it’ll have to be careful to not cut off the hand that is feeding it, so to speak.

The tiny rechargeable Lithium batteries I mentioned recently are another way to try and retain some charge overnight, just like the supercap mentioned last week.

First thing to do was to charge it up for a day, using a 1 kΩ resistor and a 3.0V supply:

I adjusted the radioBlip sketch, to switch back to 8 MHz (because the ATmega will be running well below 3.3V):

And I used these fuse settings:

• efuse 0×06 = BOD 1.8V
• hfuse 0xDE = OptiBoot (512b)
• lfuse 0xCE = fast 16 MHz resonator startup

This should allow a JeeNode to work all the way down to 1.8V (the RFM12B radio only officially supports down to 2.2V but usually still works a bit below that). I also used a JeeNode with no regulator, and added a 100 µF cap to handle the peak currents during packet transmission (100 µF is a bit excessive – less probably also works fine):

And sure enough, even with 2.75 V left in the battery, it starts up fine and starts sending out packets.

Unfortunately, I accidentally shorted out the battery while fiddling with the cables – so the charging process needs to be repeated for duration tests :(

Elektro:Camp is a convention on Smart Metering, Smart Home, Smart Grid and Smart Ideas, in a BarCamp style.

October 2011, I attended this really interesting get-together about smart metering, monitoring, home automation, and more. Very heavy on technology. It’s basically a few rooms filled with lots of human ingenuity for two days :)

The Elektro:Camp event is scheduled twice a year in various locations, and is coming up again very soon:

You can find out more and sign up on this website.

Lots of people signed up already, from all over Europe.

Due to a silly scheduling mistake, I can’t make it this time, unfortunately. But if you like the stuff on this weblog, then you’ll most likely also be delighted with everything being presented and discussed at Elektro:Camp!

The tiny solar cell chip presented a couple of days ago has been doing some indoor sun-bathing:

I’ve left it alone for some 3 days, just to give it a chance to charge up the 0.47 F supercap as far as possible. The voltage after all that time (partly sunny on most days) is now only just over 2.78 V, so this isn’t really going to work indoor, I’m afraid. Nor outside during the winter, probably – it’s just too weak.

The other solar cell I tried is also a very small one, rated at about 5V but only 1 or 2 mA, IIRC:

It’s currently in the shadow, but during these same 3 days it has had its share of sunlight (still indoor, and again behind double-glazing). Much better: about 4.75V on the second day, and unchanged since then.

This might actually do the trick. I’ll wait for another experiment to finish and will then hook up the JeeNode running my radioBlip sketch to see how it goes.

In January, I described the concept of a Component Tester, and how it can help understand what various types of components are doing.

In theory, you can just hook up a small transformer to generate the signal needed for the CT, i.e. a 50 Hz ±10V sine wave. Here is the circuit again, which is delightfully simple as you can see:

That secondary voltage is a bit high, though. Here’s what I get with two 6.3 VAC in series, i.e. 12.6 VAC:

Half that would be fine. But it’s not really a sine wave, as you can see from the many harmonics in an FFT:

And it shows – here’s what you get when placing a 1 µF electrolytic capacitor under test:

Yikes, what a mess! With a (clean) sine wave, that would have been a (clean) oval!

I don’t think it makes sense to build a CT this way. Not with this particular small transformer anyway…

Got a tip from Lennart Herlaar a long time ago about a tiny CPC1824 solar cell from Clare with 4V output:

It’s packaged as a SOIC-16 chip, so clearly the light collecting capabilities of this thing will be limited. But with all this ultra-low power stuff going on here at JeeLabs, I thought I’d give it a whirl anyway. It’s trivial to hook up:

In bright sunlight, you get over 4V with a 100 µA short-circuit current according to the datasheet.

I added a BAT34 Schottky diode in series (which has a low voltage drop) and placed it all on a little breadboard together with a 0.47 F supercap – the solar chip is mounted on a little SOIC breakout board:

The initial voltage was under half a volt, but rising (very) slowly and steadily while exposed to light.

Let’s just leave this thing exposed to light near a south-facing window for a week or so, eh?

Yesterday, I charged a 0.47F 5.5V supercap to 5.1V and kept charging it for 24 hours to reduce the leakage current.

Actually, I lowered it to 5.01V in the last hour – there’s a slight memory effect, so right after lowering the voltage actually rises when power is disconnected.

Next step is to measure the supercap’s self-discharge time from 5.00V to 1.84V (i.e. 36.8% of 5V) – that’ll give the time constant of the RC circuit (the real capacitance, in parallel with an imaginary internal current leakage resistor). Note that this is not the same as the ESR of a cap, which is about charge & discharge current losses.

Ok, let’s disconnect the power supply and track the voltage readings in high-impedance mode. It is 10:17 here, and the voltage has just dropped to 5.00V – with the power supply removed.

…

Time passes. Unfortunately, waiting for the voltage to drop to 1.84V (i.e. 36.8% of 5V) would take a bit long, so let’s throw some math at this and come up with a quicker way to measure leakage current:

• for T = R x C, we need to measure a drop to 36.8% (i.e. a factor 0.368) of the original voltage
• since the charge decay curve is exponential, we can estimate when 0.5 T will happen
• this turns out to be the square root of 0.368, i.e. a factor of 0.607
• so with a drop to 0.607 x 5V = 3.033V, we know 0.5 x T
• let’s repeat that trick one more time, to get at 0.25 x T
• my trusty on-screen calculator tells me that the square root of 0.607 is 0.779
• so if we wait for a drop to 0.779 x 5V = 3.894V, we’ll know 0.25 x T
• four times that duration, and we have T, the RC time constant we’re after

Good. That means I only need to wait for the supercap charge to drop by roughly 1V i.s.o. over 3V.

…

More time passes. It’s now 0:26 after midnight, and the voltage has dropped to 3.98V – i.e. not yet the 3.894V we need to reach, but hey, let’s call it a day anyway.

That’s over 14 hours total, i.e. over 50,000 seconds = 0.25 x T, so the calculation now becomes:

• 200000s = R x 0.47F
• R = 200000 / 0.47 ≈ 425 kΩ
• so at 5V, the internal discharge current is 5V / 425kΩ ≈ 12 µA

Hmmm…. that amount of leakage is three orders of magnitude higher than with a 47 µF electrolytic cap, but it might still be usable as power source for a JeeNode or JeeNode Micro. Here’s my reasoning:

• suppose the JN/JNµ draws 12 µA on average – a tough target, but it should be feasible
• then we’re effectively draining the supercap twice as fast as its self-discharge
• it looks like the supercap can hold a charge down to 1.8V for 56 hours on its own
• note that 1.8V is too low for RFM12B use, but the microcontroller would still work
• with the added load from the JN/JNµ, this halves to 28 hours, i.e. slightly over a day
• so the challenge will be to fully recharge the supercap to 5V at least once a day

A solar cell might just do it – assuming it’s large enough to overcome a dark and cloudy winter day. And the good news is that supercaps can charge up very fast, so a short period of bright light could be enough.

Update – There’s a lot more to supercaps than this…

As suggested by @jpc in a comment yesterday, I had a look at some documentation from Panasonic, in particular Part 2. And sure enough, they show that a supercap can be modeled as a whole set of capacitors in parallel, each with their own – often substantial – series resistance. It takes a while to “reach them” with charge, so to speak. Which explains why a long charge time increases the charge and voltage:

And which also explains why the supercap tends to drop quickly at first:

Having seen the discharge tail off much more than expected (i.e. flatten out and retain voltage), I can confirm that a supercap behaves considerably differently from a plain electrolytic capacitor.

The good news, is that for our intended purpose, this might actually work out quite well: a solar cell, keeping the supercap charged up fairly well most of the time, with just night-time JeeNode activity to drain the charge a bit, and occasional dark days, expecially in wintertime.

Update #2 – Three days have passed, and the voltage is still 3.23V, so T will be over 6 days, and the corresponding discharge rate even lower than estimated above. Bit of a puzzle – the discharge tails off considerably, apparently. Which is good news in fact, because that leaves more charge for a JeeNode to use. I’m ending this experiment for now: real-world testing with a JeeNode sending packets will be more useful.

To be continued…

Now that I have this super-high-impedance multimeter, it’s time to revisit the venerable supercap:

That’s a whopping 0.47 Farad, the size of a little coin cell, and as you can see, this unit is rated 5.5V (most supercaps are 2.7V, I suspect that this is actually made of two 1F 2.7V units placed in series).

The beauty of a supercap is that it’s like a little battery, but with fewer limitations – you can’t really overcharge it, for one, because it doesn’t turn electric energy into chemical energy. There is no conversion: put 5V on it, and it’ll draw current and gobble up electrons until it reaches 5V, then it’ll stop.

So for example for solar-powered ultra-low power nodes, this could be a pretty nice solution. Solar cell -> diode -> supercap -> JeeNode. Max charge rate while the sun shines, and then it simply stops once the supercap is full. Only thing is to not exceed that 5.5V maximum, for which supercaps are very sensitive.

But there’s a problem. Supercaps can have a substantial self-discharge rate. When I connected 5.3V to it, the voltage immediately jumped to 5.3V, but when I disconnected that cable, it also dropped back to around 4.7V in just a few seconds – a normal capacitor sure isn’t supposed to work that way!

As it turns out, supercaps tend to “learn” to keep charge better over time. The longer you expose them to a voltage, the lower their self-discharge rate becomes. The isolation barrier needs time to build up, apparently (I’ve had this supercap on the shelf for over a year). Which is great, because presumably these cells would be kept charged most of the time, with the node only depleting them slightly when sending out a packet. So ideally, all we really need is for the supercap to retain enough energy overnight.

It’s time to put these unique components to the test!

The first encouraging fact is that indeed, when fed 5.1V for a couple of minutes, the discharge no longer jumps as radically when disconnected. It now drops to 5.03V in a few seconds, but tends to hold its value after that. So it does indeed look like these supercaps can be “taught” to better retain their charge.

This test is going to take some time. First thing I’m going to do is to just keep the supercap charged to 5.1V (note that the power supply voltage calibration is pretty good – slightly less so for the low mA’s):

Let’s just leave it there to stabilize for about 24 hours. Stay tuned…

And while we’re at it, let’s compare a regular transistor (a.k.a. BJT) to yesterday’s MOSFET.

Again a smal test setup, but this time it also needs a 10 kΩ resistor between input signal and base:

The reason for that extra resistor, is that the base of an NPN BJT is essentially connected to ground via what looks like a forward-biased diode. So the voltage on the base doesn’t normally rise above 0.7V. Without current limiting resistor, the transistor would get damaged (and perhaps also the source circuit into it).

Compare this to yesterday’s screen shot and you’ll see that a BJT behaves like a MOSFET, sort of:

The main difference is that the switching point is much lower, around 0.7V – which happens to be just about the point where the base-to-emitter junction starts to conduct.

Here’s the same as X-Y graph (with again the X axis adjusted to 500 mV/div for full scale):

Compared to the MOSFET, the switch-over is steeper, i.e. more like a digital on-off switch. Note also that although the base-to-emitter voltage will be at 0.7V, the collector-to-emitter voltage is in fact below that, almost zero!

What might not be immediately apparent from the above plot, is that a transistor has a much more linear behavior (even if steeper, i.e. with more amplification). In that small range between about 0.65V and 0.75V, it’s in fact a great linear amplifier – which is what transistors were initially used for, and on a huge scale.

A simple way to describe them is that BJTs are current-driven, whereas MOSFETs are voltage-driven.

For a nice article about how to use BJT’s for signal amplification, see this page on the PCBheaven website.

The BJT was at the start of the semiconductor revolution, decades ago. The MOSFET added a new and very different component, perfect for switching enormous loads with amazingly little power loss.

For the dual-voltage supply of a few days ago, either a MOSFET or BJT will probably work. With the BJT, there will be a higher residual voltage – so a check is needed to make sure it switches properly with a feedback pin voltage of only 0.41V. The MOSFET has no such issues, it’s essentially a controllable resistor: no bipolar junctions or diode-like behavior in sight.

In a previous post, I mentioned using a MOSFET to short out a resistor. So how does that work?

Well, a MOSFET is like a voltage-controlled switch. To be more precise, an N-channel enhancement type MOSFET is like an infinite resistance when the gate-to-source voltage is zero, and turns into a very low resistance when the gate-to-source voltage is a few volts positive.

To examine this in more detail, I created a test setup like this:

By applying a linear ramp voltage on the gate, we can see what it does with varying voltages. When open, the output should be 5V, and when conducting, it should drop to almost 0V. Let’s examine this in real life:

The blue line is the input voltage on the gate (by definition a sloped straight line), and the yellow line is the voltage on the output (i.e. between drain and resistor). Let’s try and read this:

• the gate voltage takes 10 divisions to reach 3V, so that’s 0.3 V/div
• the MOSFET starts conducting at around 1.8V and is fully on at ≈ 2.4V
• at slightly over 2.1V, the drain-to-source resistance is about 1 kΩ

The red trace is the derivative of the output, so the output change is maximal at just over 2.2V.

There’s no linear behavior, in terms of gate-to-source voltage (the derivative is never constant, except in the fully-open and fully-closed regions), but what you can see is that the MOSFET will switch just fine with a logic signal (anything switching between under 1.8V and over 2.4V will work perfectly).

There are more ways to look at this. Here’s an X-Y plot, with the linear ramp on the horizontal axis:

Note that – if you think about it – in X-Y mode, it doesn’t really matter what sort of signal is placed on the gate as long as it has the same voltage range. Here’s a sine wave to illustrate this perhaps somewhat surprising property:

It’s a good exercise to try and understand exactly why the two above screenshots are the same.

Lastly, here is a zoomed-in measurement, to get more precise data using the scope’s cursor features:

As you can see, a 0.33V change on the gate is all it takes between the “almost-OFF” and “almost-ON” states.

I’ll leave it as exercise for the reader to plot the resistance of this particular MOSFET at different gate voltages. With a bit more setup, the scope’s math functions should in fact be able to display that plot on-screen.

So there you have it: a MOSFET switches on voltage, and a scope + function generator makes it easy to see that behavior. Note that even without these instruments, with nothing more than a potentiometer and a multimeter, you could in fact derive exactly the same information. It would merely be a bit more work.

For a sensor I’ve been fooling around with, I needed a supply which can switch between 5V and 1.4V, supplying up to about 200 mA.

There are several ways to do this, but I decided to use the MCP1825 adjustable voltage regulator:

The trick is to create an adjustable voltage divider, using a MOSFET to short out one of the resistors:

When off, the MOSFET does nothing, with R2 and R3 in series. When on, R3 is essentially shorted out.

The regulator varies its output voltage (top of R1) such that the level between R1 and R2 always stays at 0.41V:

So the task is to come up with suitable values for R1, R2, and R3. Let’s start with the 5V output and R1 = 10 kΩ:

• 5V = 0.41V x (10 kΩ + R2) / R2
• then 5 x R2 = 0.41 x (10,000 + R2) = 4,100 + 0.41 x R2
• and 5 x R2 – 0.41 x R2 = 4,100, i.o.w. 4.59 x R2 = 4,100
• that would make R2 = 4,100 / 4.59 = 893 Ω

Now for the 1.4V output level (where R2′ is R2 in series with R3):

• 1.4V = 0.41V x (10 kΩ + R2′) / R2′
• then 1.4 x R2′ = 0.41 x (10,000 + R2′) = 4,100 + 0.41 x R2′
• and 1.4 x R2′ – 0.41 x R2′ = 4,100, i.o.w. 0.99 x R2′ = 4,100
• that would make R2′ = 4,100 / 0.99 = 4141 Ω

But that’s not quite right, because R2 and R2′ have to be in the range 10 .. 200 kΩ. This is easy to fix by making R1 = 220 kΩ. Then the above values all increase by a factor 22 as well – bringing both R2 and R2′ nicely in range:

• for 5V: R2 = 19.6 kΩ
• for 1.4V: R2′ = 91.1 kΩ

IOW, two resistors of 19.6 kΩ and 71.5 kΩ in series would work, whereby the 71.5 kΩ resistor can be shorted out with the MOSFET to take it out of the loop.

These are not very convenient values, for resistors in the E12 series – let’s try and improve on that. After all, we can choose these values any way we like, as long as their relative values stays the same. With 15 kΩ and 54.7 kΩ, R1 would have to be 168 kΩ. That’s not so bad, we could use 15 kΩ, 56 kΩ, and 68 kΩ in series with 100 kΩ, resp.

Or, better still, perhaps: 19.6 kΩ ≈ 10 + 10 kΩ, and 71.5 kΩ ≈ 33 kΩ + 39 kΩ. With R1 kept at 220 kΩ. This needs 5 resistors in total to get the desired results. Now let’s try it out for real, eh?

Yippie – it works! Voltages with 200 mA load are 1.38 V and 4.89 V, respectively – close enough.

With 5 V input, the output is still 4.86 V @ 200 mA, proving that the MCP1825 is indeed a low-dropout regulator. The switching edges look clean on the oscilloscope, with rise and fall times of ≈ 30 µs (1 µF cap charge/discharge).

Onwards!

As an alternative to supercaps, I recently ordered a Lithium rechargeable battery from Digikey:

It’s not quite what you might think, though: its size is only 6.8 x 1.4 mm, with a tiny 3.4 mAh capacity :)

Got ten of them, as part of a larger order, and they came packaged as follows:

So far so good, but now the crazy part. These batteries were sent out in a separate 23x23x5 cm box:

With a warning label …

… and another warning label:

The max discharge current of these things is 1.5 mA according to the specs. I doubt they’ll even go up to 15 mA when shorted! By the way, does that dented corner qualify as “damaged” ? … I want my money back! :)

Get real – or better still, read Bruce Schneier‘s works!

PS. Please don’t get me wrong: fire risks are very real – it’s just that the above cells have virtually no energy…

As I’ve mentioned before, an oscilloscope is a pretty nifty piece of test equipment. It can also be very expensive.

The following comment in my series on oscilloscopes is still a good summary of what it’s all about, IMO:

Oscilloscopes are the “printf” of the electronics world. Without a “scope” you can only predict and deduce what’s happening in a circuit, not actually verify (let alone “see”) it. Here’s what an oscilloscope does: on the vertical axis, you see what happens, on the horizontal axis you see when it happens. It’s a voltmeter plus time-machine.

That doesn’t mean you can’t get anything done in Physical Computing without one. A simple multimeter is a lot cheaper and will get you a long way in figuring out the electrical behavior of a circuit – not to mention finding shorts and connection mistakes. So the first thing to get is a multimeter, not a scope. Always.

The trouble is that ATmega’s are so friggin’ darn fast. We can’t observe events on their time scale, and more importantly: many problems will zoom past us and get lost before we have a chance to see anything!

So I’m going to revise my advice about oscilloscopes somewhat: if you solder together kits and basic components, then yeah, a multimeter is plenty. But if you hook up non-trivial chips and need to debug the combination of hardware and software, then you really need all the help you can get. Be it a logic analyzer for digital signals, buses, and pulse-trains, or a scope to investigate the electric behavior of a fast circuit.

Note that a logic analyzer can be a lot cheaper than a scope. The reason being that they are electrically much simpler – they just need to collect a bunch of digital logic levels (rapidly), whereas a scope needs to collect much richer signals (ranging from millivolts to hundreds of volts, and with all sorts of signal processing to make sure you’re seeing the real thing and not some artifact of the instrument itself).

If you’ve been following this weblog a bit, you’ll have seen quite a few scope screen shots in some of the posts. One of the most important uses for my scope here at JeeLabs is to figure out power consumption while trying to optimize a JeeNode’s ultra-low power mode. Power consumption is an analog thing, so that’s where a scope comes in. And when you look at the amount of detail a modern scope can show, it’s clear that this level of insight really comes from such an instrument. See the recent Watchdog tweaking and Room Node analysis for some examples.

Does that mean you have to shell out a few thousand dollars to do something similar? Not at all.

First of all, visualization isn’t everything. A couple of years ago, I used one JeeNode to measure the power consumption of another JeeNode, see the Power consumption tracker post, and the software for it. Less insight perhaps, and no geeky screen shots, but plenty of info to try and optimize the power consumption by trial-and-error. Just tweak your sketch and measure, over and over again.

Second point I’d like to make, is that such power measurements are fairly slow, so any scope will do. Even a 10 MHz model will be able to accurately display changes from one microsecond to the next.

There are a couple of ways to get such a “low-end” scope (don’t let that term fool you, any oscilloscope can be extremely useful as long as things don’t change too fast):

• Look for a second-hand unit, lots of them can often be found on eBay.
• Consider getting a USB-connected scope such as the DSO-2090.
• For PC’s there is software to create a basic scope using the sound card.
• Check out the ultra-tiny Xmegalab, its under $50 (plus shipping).

These last two options are lower cost, but more limited since they don’t really include a full “front-end” to handle a wide range of input voltages. For circuits with only a few volts, they may still be sufficient.

Normal “sweeping” analog scopes are ok, but storage scopes (analog or digital) are considerably better because you can “capture” an event and keep it on the screen to investigate. Such a feature will cost more though.

Here’s an example of how a €100 second-hand Tek 475 (analog & non-storage) scope can be used to measure that same power consumption as in the Watchdog tweaking post – it’s the same waveform:

Two essential tricks were used: 1) the watchdog is firing at ≈ 60 Hz, so the scope trace fires constantly, and 2) it triggers on one pulse but displays the next one, using x10 horizontal magnification.

The above screen shows 2 mA and 200 µs per division. The vertical scale could have been zoomed in further, but for the horizontal scale I’m sort of at the limit unless I start using delayed sweeps. Here’s the whole unit:

No storage, no screen capture, no USB, so this was done by darkening the room and holding a camera in front of the scope. It took a couple of tries, but hey – it is possible to estimate power consumption this way!

What I’m trying to say is that you too can do this sort of work with an investment of €100 to €150.

If you intend to do more with electronics (and let me assure you: this sort of fooling around is geek heaven, and addictive!) – then consider holding off just a bit longer if need be, and save up for a Rigol or Owon scope. These “DSO’s” are mature, have tons of useful features, and they can store lots of detail (that’s the “S” in DSO).

Is this a case of “if you have a hammer then everything starts looking like a nail”? All I know is that my insight in ultra-low power consumption and optimization has increased significantly since getting an oscilloscope.

Came across this graphic a while ago – US energy cost and consumption over the years:

For comparison, in 2012, electricity here costs ≈ €0.21 (i.e. $0.28) per kWh, including taxes.

Our usage (i.e. Liesbeth’s and mine) was about 3000 kWh in 2011. That includes electric cooking, but note that heating and warm water is provided through natural gas.

That puts us in the late 1950′s w.r.t. US electricity consumption levels – yo, Elvis! :)

I’ve started to get involved in a local initiative (see this Dutch website if you’re interested – “duurzaamheid” is all the rage these days, it seems), with all sorts of simple and not-so-simple ways planned 1) to consume less, 2) to switch to renewable sources, and 3) to fall back to natural resources for the rest. It’s not an all or nothing game, more a way to plot a practical trajectory for improving things over the next couple of years.

Here’s the JeeLabs neighborhood:

Lots of space to catch some sunlight on all those rooftops – but careful with that chimney’s shadow!

Now that solar energy has become so cheap (Wp prices including inverter have dropped below €1.70), we’re finally getting together with a couple of neighbors here to actually make it happen. This year, and hopefully before the summer is gone again!

The aim is to try and get 4000 to 5000 Wp onto our roof (16..21 panels of 100×160 cm), which would cover for our entire yearly electricity needs, even without pushing for further savings. For the 52°N latitude of the Netherlands, panel + inverter efficiencies are estimated in the 80..85% range, nowadays.

That’s just half the story, gas consumption is the other biggie – but hey, ya gotta’ start somewhere, eh?

The Hameg HMO2024 scope just got a firmware upgrade – wow, it just keeps getting better and better.

Support for up to 6 calculated values (was 2), based on any of the input channels – now with optional statistics:

And one of the things I really missed dearly – the ability to see all decoded serial data in tabular form:

The top two traces show the SCL and SDA data in analog form, the next group is the color-coded serial data, and at the bottom is the list of packets. As you scroll through the table, the traces adjust to show the related information. Still shown at the bottom are the 6 auto-measured items I configured in the first screen.

Last big new feature is the capability to search through stored traces, again with a table to help navigation:

It’s all firmware, evidently, but I hadn’t expected the development to keep on moving the capabilities of this oscilloscope forward to such an extent. And these aren’t just gimmicks, such features can be extremely useful!

The Heading Board has been sold out for some time now. I’ve not been reordering it because it’s a bit quirky (needing the IRQ pin as well) and probably also not really all that sensitive.

To continue to offer a solution, I’ve decided to switch to the Modern Device 3-axis Compass Board instead:

As you can see, it has a port-compatible header footprint on one side. The other side is for use with a 5V system, such as an RBBB or Arduino. Which is why there is also a 3.3V regulator on there.

The board is slightly smaller than a standard JeePlug and does not have port headers on both sides to support daisy chaining, but apart from that it’s totally JeeNode-/port-compatible. You can simply put it on the end of the chain if you want to mechanically stack this along with other I2C-enabled plugs.

Be careful about the orientation: it not the same as other plugs and there’s no “dot” next to the “P” pin.

I’ve added a very basic implementation in JeeLib to access the HMC5883 chip on this board, with demo:

Sample output:

This new board has now been added to the JeeLabs shop.

When you have two nearly identical sine wave signals and you want to compare them, one technique is to plot one against the other, creating what is known as a Lissajous curve.

Lissajous curves make nice images, and even nicer videos because of the phase shifts.

So let’s take two signal generators and try it out, eh?

On the X-axis, I’m going to plot a 10 MHz sine wave from the new AWG, described on this weblog a few days ago. The frequency accuracy and stability of its output signal is within 1 or 2 ppm, according to the TG2511 specs.

On the Y-axis, let’s connect a second 10 MHz sine wave from a cheap DDS, also described on this weblog a few months back. This has a simple crystal so I’d expect 50 to 100 ppm frequency accuracy, i.e. within ≤ 1000 Hz.

When you connect these to an oscilloscope and put it in X-Y mode, you get pictures like these:

The more the two signals are in phase, the more the result will look like a straight line, slanted at 45° (from the bottom left to the top right). When exactly 180° out of phase, it will show a straight line from top left to bottom right. Everything in between create ovals, and when the signals are 90° out of phase (either lagging or leading), the result is a perfect circle.

So the big thing about Lissajous curves is that they let you compare the relative phase of two sine waves.

In practice, signals from different sources will tend to change phase over time, i.e. “drift” as one sine wave is slightly slower or faster than the other. This creates a way to precisely compare two frequency generators: measure how long it takes for the phase to go from 0° to 180° (or 360°, which is 0° again), and you get an idea how long it takes for one signal to catch up (or lag) one full sine wave over the other. Trouble with this approach is that sometimes these cycles are too fast to see, let alone time manually.

With an adjustable frequency source, there’s also another way: adjust a known frequency until the shape stays the same, and you’ll have “measured” the frequency of the other signal in terms of the adjusted one since they must now be equal. It’s very much like tuning a musical instrument by ear and adjusting for a “zero beat”.

That’s what I did, and I ended up with the following result for this test setup:

IOW, the cheap DDS is running 0.03% slow – i.e. about 300 ppm! And it’s not even very stable, because very soon the DDS starts drifting again: an indication that it’s not holding its frequency really accurately either. This is not really surprising for such a low-cost unit off eBay – it’s still a useful signal source: lots of useful experiments and measurements can be done with such a fairly decent 0.03 % accuracy level, after all.

Ok, this concludes my first exploration into signal-processing – enough signal theory for now!

Triggered by the recent signal generator checks, and those FM radio stations creeping into the signal yesterday, I wanted to do another test to see how and when this happens, using a series of scope FFT snapshots.

Here’s a 50 Ω coax cable of 2m length hooked up, with 50 Ω termination on both sides but no signal. The scale is 5 dBm per division, and I’ve zoomed into this very low range with a baseline of about -105 dB. The following 200 MHz wide FFT measurements were all done with the scope input set to max sensitivity, i.e. 1 mV/div:

Note the slight FM radio station RF signal pickup, even with a fully terminated coax cable!

Same thing, but disconnected on one end, i.e. only one 50 Ω terminator inside the scope:

I don’t know why there’s a peak at 25 MHz – the signal generator was completely powered down.

Here’s the spectrum with no coax at all, i.e. nothing connected to the scope, but its 50 Ω shunt still enabled:

When also adding an external 50 Ω terminator, the lower frequencies drop ever so slightly further:

And here’s what happens when the 2m 50 Ω coax cable is attached back on, without the 50 Ω termination:

As you can see, the coax cable now acts as antenna, picking up a few more signals at 38.45 MHz and 46.38 MHz. And FM reception shooting up to 20 dB above the noise floor. Even though it’s shielded!

The slight drop in noise across the screen from 0 to 200 MHz is probably nothing more or less than the scope’s bandwidth: a 200 MHz scope is specified as having a 3 dB drop at 200 MHz, which fits amazingly well with what all the above screen shots are showing.

These tests confirm the superb signal processing specs of the Hameg oscilloscope front end: a -105 dB noise floor @ 1 mV/div maximum sensitivity. For even lower noise levels (and a higher frequency range, as would be needed for 868 MHz and 2.4 GHz RF measurements), probably only a “real” spectrum analyzer will do better.

Whee… with a multi-meter probe wire attached as antenna, I can easily pick up all major AM radio stations:

Tomorrow, I’ll close off with one more post about signal processing: accurate frequency measurements.

Update – As with yesterday’s post, these FFT’s were produced with the Rectangle window function. As a bonus, here’s the frequency spectrum produced by the noise generator in my new AWG:

Down by 30 dB at 50 MHz, but a pretty good source of white noise at lower frequencies. The AWG can add an adjustable amount of this noise to the generated waveforms – can be useful to see how well a filter, demodulator, or other detector behaves, for example.

As shown in yesterday’s post, once you’re looking at signals in the high MHz range, it’s easy to make mistakes. Looking at that screen shot again, you can see a whole bunch of 90..100 MHz spikes:

These are in fact local FM radio stations – being picked up by my clunky scope probe hookup to the AWG. In other words it’s an antenna: radiated RF signals are accidentally being received by the probe and mixed in with the conducted 25 MHz signal. The way to get rid of these is to use “shielding” to keep those radio waves out.

Here’s the same signal, using a 50 Ω coax cable all the way between signal generator and oscilloscope:

No more weird stuff, just the 25 MHz multiples – indicating that there’s a slight distortion in the generated 25 MHz sine wave. This is normal for any signal generated by a direct digital synthesizer, because the waveform is created through through a digital-to-analog converter, fed at 125 million samples per second in this case. IOW, it’s an approximated sine wave (20 datapoints per sine wave @ 14-bit resolution).

So the big change is that the FM radio stations are gone, and that the signal’s noise level is now in fact a few dB cleaner than before – as you can see from the slightly lower tail towards 200 MHz.

I did have to use a small trick to make these graphs comparable: the second one has the scope’s 50 Ω internal terminator enabled, so that the path from signal source to signal destination is now done by the book: a 50 Ω source (in the AWG), feeding a 50 Ω coax cable, terminated by 50 Ω at the destination (in the scope). This does “attenuate” (i.e. reduce) the signal level by half, so I had to raise the baseline by 3 dB on the second FFT screen shot to make the height of the 25 MHz peak identical in both screens.

One other minor difference is that the second graph is smoothed over 256 samples i.s.o. 64 – cleaning up the resulting line slightly more.

So you see… it’s possible to do RF-type stuff without understanding all the details – which I certainly don’t, yet – and get decent results. The 25 MHz wave coming into the scope is very clean: the first harmonic is some 50 dB below the signal itself, which means that the first harmonic has 100,000 times less energy than the main signal.

Tomorrow, another post about this topic: cables, termination, and noise…

Update – As John Beale pointed out in the comments below, the FFT baseline is caused by the choice of FFT windowing function. Here’s the 50 Ω coax example again, using a Hanning window:

Much better for comparing relative dB differences between peaks.

(Tomorrow’s post will also use the default Rectangle window, sorry about that…)

Earlier this week, I described how a fixed frequency can be used to stabilize others.

Well… as part of my continuing drive to set up a more complete workbench here at JeeLabs, I’ve decided to get another piece of equipment which relies on this mechanism, called an “Arbitrary Waveform Generator” (AWG) or “function generator” or “signal generator” – three names for essentially the same instrument, as far as I can tell.

An AWG produces a repetitive electrical signal, such as a sine wave or a square wave. Very roughly speaking, you can think of sine waves as “pure analog” and square waves as “pure digital” frequencies.

The unit I picked is a fairly advanced one, the TG2511 from TTi (Thurlby Thandar Instruments), in the UK:

(Check out that box underneath – as a reference it now makes a lot more sense, eh?)

It produces sine waves and square waves up to 25 MHz, and has tons of other waveforms built in, including ramp, triangle, pulse, noise, and more. In fact, since it’s an AWG, you can load any waveform shape into it, and it’ll reproduce it at up to 125 Mega-samples per second and 14-bit resolution (it goes up to 6 MHz in this mode).

Two other major capabilities of such a unit are: the ability to “sweep” across a range of frequencies and being able to “modulate” the generated signal with another one in numerous ways: AM, FM, PM, PWM, and FSK.

As with the Hameg HMO2024 oscilloscope and the GW-Instek GPD-2303S power supply, this thing can be remotely controlled over USB. So it can be driven from a computer to perform complex and/or lengthy tests.

This model does more than I need, but there was a good “price burner” offer at Distrelec, so I decided to go for it. Function generators are not the most important instruments for an electronics lab, but they are extremely useful to learn about all sorts of analog electronics, and to illustrate various concepts and effects “for real”. Note that for lower frequencies, you can generate rough arbitrary waveforms with simply an ATmega and a few resistors.

Here’s the FFT spectrum of its 25 MHz sine wave – a few spikes at 25 MHz multiples, as expected, plus a bunch of 90..105 MHz spikes which also appear when the AWG output is off (more about those tomorrow):

Such an AWG is not limited to strictly analog uses, by the way. This unit should also be able to generate a serial bit-stream, like an RS232 message, for example. Such patterns can be loaded via USB on the front panel.

I intend to put this instrument to good use here at JeeLabs, not in the least to create good examples for future weblog posts and to illustrate relevant electronics concepts in that huge playground called Physical Computing.

A week ago, there was a post about various clock options and their accuracy.

These clocks generate a stable pulse or sinewave, basically. But what if you need a different frequency?

Suppose you get a very accurate 1 pulse-per-second (i.e. 1 Hz) signal from somewhere, but you want to keep track of time in microseconds? IOW, you need a 1 MHz clock, preferably just as accurate. One way to do this, is to use a “Voltage Controlled Oscillator” (VCO). It can be any frequency really – the idea is to divide its output down to 1 Hz and then compare it with your reference clock. If it’s either too slow or too fast, adjust the voltage used to set the precise frequency of the VCO, and bingo – within no time (heh, so to speak), your VCO will be “locked” onto the reference and generate its target frequency, at just about the same accuracy as the 1 pps reference.

My Rubidium clock came with a 63.8976 MHz VCO as part of the bargain:

With no control voltage it generates a sinewave-ish very high frequency signal from just a 3.3V power supply:

That frequency is not as awkward as it looks: 638976 = 3 * 13 * 16384, so you can get 100 Hz out of it with a few simple dividers, as well as any integral fraction of that (including 1 Hz). Another way of going about this is to divide the clock by a simple power of two, say 256 or 4096, and then pass the resulting square wave to an ATmega’s timer/counter input. I haven’t hooked up this VCO to the Rb clock yet, since there’s a bit more logic involved – look up “phase locked loop” (PLL) if you’re interested.

Another source of very stable clock signals is the GPS navigation system (see also this note). Their clocks are used to be made a little bit jittery for civilian use, but this averages out over time, so you can still lock onto it and get a very accurate long-term reference. Look up Allen variance to find out more about short- vs long-term stability – it’s fascinating stuff, but as with most things: once you get into the details it can become quite complex.

To summarize: with a VCO you can produce any frequency you like given some stable reference. So I’m happy with my 10 MHz @ 10 ppt atomic clock, for those rare cases when I’ll need it. And for its geek factor, of course…

Do all these extreme accuracies matter? Well, apart from TDMA, think of it this way: an 868 MHz RFM12B wireless radio with 1 ppm accuracy may be off by 868 Hz. That’s no big deal because the RFM12B’s receiver uses Automatic Frequency Control (AFC) to tune itself into the incoming signal, but with bandwidths in the kilohertz range, you can see that all of a sudden a couple of ppm isn’t so academic any more!

There are several scenarios where it’d be nice to detect the pulse of a blinking LED – especially low-power, because then we can sense it with a long-lasting battery-powered setup, such a JeeNode or JNµ.

Fortunately, that’s fairly easy to do. I used this test setup to try things out:

The left-hand side is a test pulse, generating 10 ms pulses once a second to simulate a typical indicator light. It’s simple enough with no further explanation needed.

The right-hand side of the above circuit is the actual pulse sensor we’re trying to implement. It’s a voltage divider with on the upper half a fixed resistor (well, a trimmer, but we only have to adjust it once) and the lower half is a Light Dependent Resistor (LDR) – like these two examples:

We want to generate one electrical pulse for each incoming light pulse, in such a way that it could trigger an ATmega’s digital input pin. With a clean pulse we could then set up a pin-change interrupt and keep the ATmega asleep most of the time.

The trouble is that LDR’s and voltage dividers are analog i.s.o digital. One way would be to constantly read out the signal as analog input. But this sort of polling and continuous ADC use eats up quite a bit of power – a digital signal would be a lot better as it’d allow us to use pin-change interrupts.

No worries. A digital signal is also a voltage, but it has to stay under a certain limit to be treated as digital “0″ and above another limit to act as digital “1″. Here are the specs from the ATmega328 datasheet:

With a JeeNode running at 3.3V, we get: “0″ ≤ 1V and “1″ ≥ 2V. Note that in theory voltages between 1 and 2V will have indeterminate results, but in practice the signal will work fine as long as it doesn’t stay forever within that gray zone.

The trick is to make that LDR sensor as sensitive as possible. The LDR which I used is a fairly standard one (same one as included with the Room Board) and rises to over 1.5 MΩ resistance when dark. Let’s assume 1 MΩ as extra margin, then we could use 470 kΩ as upper resistor of the above resistor divider, and the resulting signal would be about 2.2V when dark.

The way I maximized the dark-state resistance was to place it in a small black plastic cap, as shown in the above photograph. This is essential, as you’ll see.

Now the actual pulse detection: the resistance of an LDR drops (quite dramatically) in the presence of light, so the trick is to place it close enough to the blinking LED that we want to “read out”. I placed my blinking test LED a few millimeteres from the black cap (which is open at the end, of course);

Here’s a scope snapshot of the LED pulse (channel 1, yellow trace) and the detected signal (channel 2, blue trace):

You can see the LDR signal dropping when light is detected, and that the LDR actually needs a bit of time to react. For 10 ms pulses, it’s plenty fast enough, though.

This configuration is probably ok – the voltage swings from about 1.8V (a marginal “1″) down to 0.7V (a clean “0″). The whole setup really depends on first getting the dark resistance as high as possible (i.e. shielded from any stray light) and pulling it down enough during the LED blink (i.e. close enough to pick up a good LED signal).

When the LED is inserted inside the plastic tube, the signal becomes much stronger – but recovery is slower:

It all hinges on the pull-up resistor, really. Which is why the best way to create this sensor is to use an adjustable 1 MΩ trimpot, and tweak it. You won’t need an oscilloscope or even a multimeter to get optimal results:

• very important: shield the LDR from stray light as well as you can
• pick as high a resistance as possible which still gives a “1″ signal (between 100 kΩ and 1 MΩ)
• place the LDR + shield near enough to the LED to generate a “0″ pulse
• tweak and iterate the above steps until it works reliably under all conditions

For minimal power consumption, the pull-up resistor should be as large as possible. Example: with an optimal pull-up of 1 MΩ and the LDR’s dark resistance about 1 MΩ as well, the quiescent current draw will be (Ohm’s law: I = E/R) 3.3 V / 2 MΩ = 1.65 µA, an excellent value for ultra-low power nodes. During the LED light pulse, this will increase to at most twice that (i.e. if the LDR resistance were to drop completely to 0 Ω).

Note that a more sensitive sensor design will be needed if you want to actually measure the length of the pulse with a decent accuracy, but for simple counting purposes where incident light can be kept out, there is nothing simpler than this LDR + pull-up trimmer, probably.

Update – More info about LDR’s on the LadyAda site.

Someone drew my attention to a very small PIR sensor on eBay:

The size is great, of course – but the current consumption isn’t: I measured 1.9 mA idle current @ 5V.

The other inconvenience, in the context of JeeNodes, is that this sensor expects a 5..9V supply voltage.

Using my new accurately adjustable power supply, I was able to establish that it actually works all the way down to 3.6V – with current consumption down to 1.3 mA. But that’s still far from the 50 µA current consumption of the PIR sensor used in the Room Board, so this rules out ultra-low power battery nodes. The detection range is specified as 2 to 3 m, not stellar but probably enough for many uses.

Here are the two sides of this really tiny sensor (whoops, I accidentally cracked the lens):

The chip marked 7144-1 appears to be a 4.4V LDO regulator, with excellent < 5 µA idle current and 0.06 V drop-out voltage under light load, but that seems to point to a circuit which really expects to run at 4.4V internally.

I have no idea what this spec means on the eBay page:

It’s definitely not referring to the idle power consumption of this sensor. Too bad!

Should we try and design our own PIR sensor? I wonder what that would take – some way to stabilize on an average detector level, and then detecting changes in that value? Using an ultra-low power op-amp?

My existing lab power supply delivers 30V @ 3A, which is more than enough for normal use, but it uses linear fine + coarse potentiometers, which are in fact not optimal for really fine adjustments. I’ve been using it a lot, and I really have been wanting something more convenient for quite some time.

So I decided to get a second and more high-end unit, the GW-Instek GPD-2303S:

It even comes with a “calibration certificate”, FWIW:

There are many lab power supplies out there, and I intend to come up with a really good option for low end use in the context of the Thursday Toolkit series, but I’ve got enough future projects piled up here to justify this instrument for JeeLabs. Everything other than the ultra-low power experiments will benefit from this.

BTW, if you’re looking for a DIY design which is coming along very nicely, check out the EEVblog episode list, where Dave Jones has over half a dozen fascinating videos about how he is designing a really nice Arduino-compatible power supply, with all the bells and whistles you might be after: finely programmable voltage and current range, an LCD display, rotary switches for adjustment, etc. Here’s the first one in the series.

The GPD-2303S delivers 2x 30V @ 3A, i.e. up to 180W of controlled DC power. There’s a 3-channel unit with extra 2.5/3.3/5V output, even a 4-channel unit, but I’ve got enough supplies here now to cover such needs.

Note that lab power supplies are designed to “float” w.r.t. ground. The reason for this is described in my two weblog posts here and here. So you can hook them into your setup in any way you like. Even doing some totally crazy stuff like a adding a 50V DC component to AC mains would be possible…

Anyway. The nice thing about this supply (even though its shape is a bit deep for my workspace), is that it includes two independent supplies which can be used in series (double the voltage) or in parallel (double the current) to get 0..60V or 0..6A capability, and that both voltage and current can be controlled and measured very accurately (not quite down to the 1 mV/mA levels as they claim, but close).

This power supply is not a really high-end one, though (which would cost even more), since there is no remote sensing, for example. So small losses over the cabling are not compensated for. I’m not too worried, because with large currents I’m usually not really concerned about 10 mV error.

More important is that it’s a linear power supply with only 1..2 mV ripple, and that the current limit can accurately be set to very low levels. By setting it to 50 mA, say, you can avoid most damage when hooking up things the wrong way – as so often happens while messing around with circuits.

Also very nice is that this unit is programmable – meaning that you can control it fully via USB. That opens the door to all sorts of stress and limits testing, i.e. plotting the effects of a slow voltage ramp on a circuit, for example.

Sooo… with this new addition to JeeLabs, I hope to stay out of Mr. Murphy’s path a bit more!

Tracking time is as old as… well, time itself really. FWIW, I stopped wearing a wristwatch about a year ago. When traveling, I often don’t have a convenient way to tell the time with me. I like it that way because it pushes me to leave a bit earlier and enjoy the journey a bit more, instead of stressing out to reach some location on earth at some particular point in time. “Onthaasten” as the Dutch say (“un-hurry”).

That wristwatch was one of the most beautiful time-pieces I ever owned, and darn accurate – less than a minute off per year. More than accurate enough for day-to-day use without ever adjusting it (except for DST).

Still, time is everything. Cell phones make very efficient use of bandwidth (and energy) by using time-division multiple access (TDMA), i.e. taking up specific time slots to get a transmitter to talk to the receiver it wants to reach, without collisions. It’s a common technique in many advanced networks, not just with cell phones.

TDMA requires all parties to be aware of time. No wonder that cell-phone towers need the Rubidium clock I described in the past two days. If your timing is off, you end up jamming others, and to avoid that the system then needs to introduce wider gaps around each slot. More gaps = more time & energy wasted, as each unit has to wait longer in receive mode to be certain it picks up the entire packet, and more gaps = more unused bandwidth.

For good timing, you need to have every node in sync within a millisecond or so. Perfect timing means a receiver can turn on exactly when the transmitter starts, and switch off right after the end. But the need for exact time is bad news for the simplest ultra-low power nodes, which tend to use an RC-controlled watchdog timer for the sleep modes. On an ATmega, watchdog accuracy is only about 10% in the worst case:

Ok, time to get into some terminology…

• one percent (%) is 1 per 100, of course
• one part per million (ppm) is 0.0001 percent
• one part per billion (ppb) is 0.0000001 percent
• one part per trillion (ppt) is 0.0000000001 percent

It’s easy to make mistakes with so many zero’s, so let’s approach it from another angle: a year has about 31.5 million seconds, so let’s specify time accuracy in the amount of error over a year. And let’s not fuss over 50%, I’ll round things up or down a bit for convenience.

My trusty old Seiko Lasalle wristwatch:

• estimated 1 min/year, i.e. an astoundingly good 2 ppm

ATmega watchdog accuracy:

• worst case: 10 % = can be over a month per year off
• if supply is 3.3V ± 0.1V and temp is 25°C ± 10°C: 1 % = within 3 days per year

The 16 MHz resonator used in a JeeNode:

• 0.5 %, i.e. less than 2 days per year off

The crystal normally used in RBBB’s, Arduino’s, RTC’s etc:

• 50 ppm, i.e within half an hour per year

The more accurate crystal used in a JeeLink:

• 10 ppm, i.e. within 5 minutes per year

The Precision RTC Plug, which will be released later:

• 2 ppm, i.e accurate within 1 minute per year

The calibrated Temperature Compensated Crystal Oscillator (TCXO), mentioned recently:

• 0.1 ppm, i.e. within 3 seconds per year

And then that new Rubidium clock:

• 10 ppt, off by no more than 0.3 milliseconds per year

Cesium clocks can do even better, by the way:

• 0.01 ppt, no more than 0.3 microsecond per year

And how’s this Quantum Logic Clock, mentioned in one of the recent comments:

• less than 0.3 nanoseconds off per year!

The clocks mentioned so far all try to be spot on as best as they can. That’s what a clock is for, after all.

But that’s not all there is. Next week I’ll describe how a fixed reference can be used to stabilize any frequency.

After yesterday’s intro of my “get your own atomic clock”, which is really just doodling, here’s the next step:

The clock, and the PCB panel it came attached to, has been placed in an all-plastic enclosure along with a little 15V @ 1.7A switching power supply. This thing needs quite a bit of power and actually gets quite hot. Nevertheless, I expect that placing it inside this relatively small plastic enclosure will not be a problem because much of the heat seems to be generated simply to keep the Rubidium “physics package” inside at a certain fixed temperature. For that same reason, I suspect that the heat sink on which this clock is mounted is not so much meant to draw heat away, but to maintain a stable temperature and improve stability.

Speaking of stability… here are the specs of this unit from eBay:

To get an idea: 10 to the power -11 frequency stability is less than 0.3 milliseconds per year error!

This particular unit (they are not all identical, even when called “FE-5680A”) also needs a 5V logic supply.

I haven’t yet decided how to bring out various signals, so I’ll hook up the 50Ω BNC connector on the back first and wait with the rest. Also needed: a LED power-on light, LED indicators for the “output valid” and “1 pulse-per-second” signals (via a one-shot to extend the 1 µs second pulse), and a 7805 regulator. Here’s the front – so far:

I don’t intend to keep this energy-drain running at all times, but it’ll be there at the flick of a switch to generate a stable 10 MHz signal when needed. One of the things you can do with it is calibrate other clocks, and compare their accuracy + drift over time and temperature.

Geeky stuff. For a lot more info about precise time and frequency tracking, see the Time Nuts web site.

Tomorrow, I’ll describe some of the trade-offs w.r.t. time for JeeNodes and wireless sensors.

Triggered by a video on EEVblog of a Rubidium frequency standard and its teardown, I decided to get one myself. These things are available from eBay for around €40 these days, as recalls from cellphone towers, apparently:

It’s about the size of a hard drive, it’s completely closed, and there really is nothing to it – connect power, wait a few minutes for things to stabilize, and out comes a 10 MHz signal – a perfect black box:

There’s a capacitor in the output circuit, so the resulting signal is AC-coupled and hence centered around 0V and 1.5 V peak-to-peak. There’s also a 1 pulse-per-second (PPS) output signal, with a 1 µs pulse (at first I thought it didn’t work, but the pulse is really there).

The big deal about such a Rubidium-based atomic clock is its accuracy. More on this tomorrow…

I ran out of USB BUB interface boards in the shop the other day (all my fault, not paying attention), so these last few days a slightly different version is being sent out. This is an earlier, but nearly equivalent, design by Lennart Herlaar – and as it so happens, I had a bunch of them lying around – a perfect susbtitute until the BUB’s by Modern Device are back in stock next week.

Here’s the USB FTDI Board in close-up:

As with the BUB, you need to solder on the 6-pin FTDI connector for use with JeeNodes and RBBBs:

The “settings” for this board is to pass the 5V supply voltage from the USB connector (solder jumper, already installed), and to set the logic levels jumper to work with 3.3V, as required by the JeeNode (or 5V for RBBB use). It’s not that critical, really.

Sooo… slightly different unit and shape, but does the same job as before, i.e. connecting up to USB for power, as well as serial-over-USB uploading and debugging.

After yesterday’s introduction, we’re ready for some more insight…

To summarize: a straight line going through (0,0) represents a purely resistive effect. The slope of the line is related to actual resistance. With resistors, once you know the voltage, you know the current (and vice versa).

Here’s a diode, i.e. a component with very specific properties (this shows why it’s called a semiconductor!):

With negative voltages, it just blocks (horizontal line, infinite resistance). With positive voltages it’s essentially a short circuit (vertical line, almost zero resistance). Note the “knee”: a diode starts conductiong at about 0.7V.

Here’s a blue LED:

Very much like a diode (the “D” in LED stands for diode, after all). Except that the knee is higher, at around 3V.

Here are three zener diodes of 3.3V, 5.1V, and 9.1V, respectively:

Note first of all that these diodes were connected in reverse compared to the diode and LED shown earlier, so the graphs are rotated by 180° compared to those. A zener is a regular diode, in that it conducts normally at around 0.7V. The difference is that when it’s blocking, it will at some point “avalanche” and start conducting anyway. This very specific voltage is what makes zeners special. But note how that avalanche knee is round and inaccurate for low voltage types. Zeners for less than 6V or so are not very precise for regulating the voltage – but 9.1V is fine.

Neat, huh? Each type of component has its distinctive analog signature when viewed on a CT!

So far, you’d be forgiven to conclude that a Component Tester is simply a hardware function plotter. With the horizontal axis being the voltage applied, and the vertical axis being the current flowing through the component.

Ah, but wait… here’s a 1 µF capacitor, showing that capacitors are fundamentally different beasts:

This is where things start to go crazy. No current at maximum and minimum voltage? Lots of current at zero volts? Positive and negative current at that zero-volt position? What’s going on here?

The thing to keep in mind is that this is not simply a function of voltage vs current. We’re applying a sine wave – a voltage which very uniformly and smoothly varies between -10V and +10V. Think of a swinging pendulum, oscillating over and over again in a constant pattern.

Note also that the component is being driven through a 1 kΩ resistor, limiting the maximum current through it. So we’re looking at the capacitor while it’s in fact part of a circuit – i.e. a 1 kΩ resistor in series with our 1 µF cap.

Let’s start at the right. The capacitor is fully charged to +10V, and our voltage is starting to decrease. When the voltage is +9V, the cap is still +10V, so it starts sending out charge in the form of current to try and regain the balance. So a positive current flows out when the voltage is at +9V. If that voltage stayed at +9V, it would soon stop, since the charge drops, and the capacitor reaches +9V equilibrium again. But as this happens, the voltage keeps on dropping. In fact, it drops faster and faster, so more and more current leaks out while catching up.

At 0V, the rate of descent (dare I say slope or derivative?) is maximal, as you can see when you look at a sine wave. So at that point, the capacitor is leaking charge as fast as it can – at the rate of 4 mA in this case.

The voltage doesn’t stop dropping, though. I keeps on dropping to -10V, although it’s slowing down again. So the current still flows out of the cap, but slower and slower. At -10V, the voltage is no longer dropping at all, and the charge will have caught up – no more current, i.e. 0 mA.

Now the roller coaster ride repeats the other way around. The capacitor has -10V charge (lack of charge, if you wish to look at it that way), and voltage is about to start rising again. This time, charge has to be fed into the cap to try and equalize voltages, and so the current is now negative.

And sure enough, the lower negative side of the circle goes through the same changes. Until we reach +10V again.

So what you’re looking at is not a function, but the path of a point in space, racing around a circular path (ok… oval, since you insist). That point in space leaves a trail on the screen, and that’s the resulting image.

Phew! Still there?

The reason this happens, is due to the fact that a capacitor has state (or memory, if you like). It will respond to an external voltage differently, depending on the amount of charge it currently holds. Applying +5V to an empty cap will generate a different current than applying +5V to a capacitor which is currently charged up to +10V, or whatever. Current will start to flow to balance things out, but this requires time.

Very loosely speaking, you could say that capacitors “live in the time domain”. Unlike resistors – which just resist the same way under any circumstance.

Here’s the trace of an inductor (the secondary coil of a small transformer in this case):

Hey, it looks like inductors also have state! And yes indeed, they do. Capacitors and inductors are very similar, electrically. They both “live in the time domain”, although through very different mechanisms.

The state of a capacitor is its current charge level, i.e. the “amount of electricity” inside it at any particular time.

The state of an inductor is the magnetic field level it has created. When you send an electric current through a coil, that coil becomes an electro-magnet, and starts generating a magnetic field around it. When the current stops, the magnetic field wants to keep going. But it can’t and it starts fading – while it does, an electric current is generated in the opposite same direction. This effect (plus a little resistance) is what causes the tilted shape shown above.

As you can see, it has the same weird effect: no current at maximum or minimum voltage, and either positive or negative current at zero volts.

The point of these little demos was to show how current and voltage stop being linearly inter-related with caps and inductors. Because they mess with time. The charge which came in today could come out tomorrow, for example.

With constant voltages, capacitors and inductors are boring. But when their time effects are pitted against voltages which change over time, then nifty things can happen. It’s probably fair to say that the discovery of DC (direct current) brought electricity to the world, whereas AC (alternating current) brought electronics to the world.

For measuring DC, you can get by with a voltmeter. For AC, you need a voltmeter-over-time, a.k.a. an oscilloscope.

I hope this gives you a feel for what’s going on in electronic circuits. The behaviors shown here are universal, i.e. caps will behave like this every time, no matter what else sits around them, and getting an intuition about how these components react to voltages is a fantastic way to figure out all sorts of more complex circuits.

There’s tons more to explore about signals and circuits: filters, phase effects, crazy stuff called “complex numbers” (values with a “real” and an “imaginary” part, go figure!), switching perspectives from the “time domain” to the “frequency domain”, and Fourier transforms. None of this matters, if all you want is to turn on a lamp or work with digital signals. But if you’ve ever wondered how electronic stuff really works: trust me… it’s fascinating.

Is anyone interested in any of this? I’d love to write a series about it one day, where intuition comes first, insight a close second, and where all the mathematics involved will become totally obvious (seriously!).

PS. Here’s my intuitive summary of what R’s, C’s, and L’s do (and what makes each of them unique):

• Resistors turn electrical energy into heat (no way back with a resistor)
• Capacitors turn voltage differences into electric charge (and back)
• Inductors turn electrical current flow into magnetism (and back)

Hameg scopes have often included a “Component Tester” (CT) and mine’s no exception. It’s a really nifty way to identify a component and understand its basic characteristics. It requires a sine wave signal and an oscilloscope:

Don’t fret too long about the above circuit (copied from this PDF, which I found via Google). It’s just to show that setting up something like this is very easy – but you do need an oscilloscope with X-Y capability.

The basic idea is to apply a sine wave of say 10 VAC @ 50 Hz to the part you want to identify, and to then display voltage over versus current through that component.

My scope has the equivalent of the above simple CT circuit built in:

Two pins, on which a 50 Hz voltage is applied which varies between +10V and -10V in the form of a pure sine wave. For some reason, the scope won’t let me take screen dumps to USB in this mode, so I’ll use camera shots.

Here’s what you see with nothing connected (note the full scale: ±10 V on X and ±10 mA on Y):

The horizontal axis shows the applied voltage, and as you can see, no current is flowing. Because air insulates!

Let’s short the two pins with a copper wire (I’ve reduced the image scale to reduce this weblog post’s length):

Of course: no matter what voltage we try to put between the pins, the wire will force it to 0V, and will simply pass -10 .. +10 mA of current. As you can see in the schematic, there’s a 1 kΩ resistor in series to limit the current.

These two images of open vs shorted set the stage. Now let me insert a couple of different components, so you can see how they behave when subjected to this 10 Vpp sine wave. First, let’s insert a 1 kΩ resistor:

Make sense? Now have a look at a 10 kΩ resistor:

So what we have so far, is resistance varying from zero ohm (shorted) to infinite ohm (open), with two values in between. It all ends up as a straight line, with the slope varying from horizontal to vertical. That’s it: resistance!

If you want an explanation: this is Ohm’s law, visualized. Voltage and current are proportional, i.e. V = I x R.

So much for the basic stuff. Tomorrow, I’ll show you a couple of considerably more interesting components.

Yesterday, I tried to get to grips with how a capacitive power supply works. Real samples are a bit messy, though.

So let me try something else this time, and simplify a bit:

The setup is very similar, but I’m leaving out the zener diode and the rest, and more importantly, I’m going to feed a real 50 Hz sine wave signal into this circuit, using my little sine wave generator.

Here again, because scopes need to measure with a common ground, I’m placing that common ground in between the resistor and the cap, and I’m using the scope’s internal “invert” feature to treat this as if the channel 1 probe were connected the other way around. Here’s what we get with this new setup:

The vertical scale of the resistor was adjusted to display the same amplitude for the resistor as for the capacitor.

• the yellow line is the voltage over the capacitor
• the blue line is the voltage over the resistor
• the red line is the sum of the yellow and blue lines

The scale of the red line is not quite accurate, but its shape is. So the red line is in essence the input signal.

So what’s going on here?

As you can see, all these signals are 50 Hz sine waves. That’s quite remarkable already. Obviously the red line is a 50 Hz sine wave, since that’s what we’ve been feeding in. But so is the voltage over the capacitor, the voltage over the resistor, and hence also the current through this circuit!

What you see, is a set of sine waves which differ only in phase and in amplitude:

• the voltage over the capacitor (yellow) lags the input signal (red): it’s forever trying to catch up
• the current through the capacitor, i.e. the voltage over the resistor (blue), is leading in phase

And something else, as we saw yesterday: the current through the capacitor is related directly to the slope of the capacitor’s voltage change (i.e. its derivative). When the yellow line is steepest, the blue line is at its highest.

Let’s throw one more calculation into the mix: power. Power is input voltage (red) times input current (blue):

Looks very sine wave’ish again! There is some amplitude variation, which can probably be attributed to signal asymmetry or amplitude differences between the different sine waves. The phase of this wave is different from the ones already shown, but note that it’s also twice the frequency, i.e. 100 Hz.

Let’s step back for a moment. With a purely resistive circuit (as used in a resistive transformer-less supply), the current and voltage would be in lock step, i.e. sine waves with exactly the same phase, and the power consumption would be maximal (high current at times of high voltage).

With this capacitive setup, currents get “moved around”. That means power consumption will be less than when voltage and current match up (since this gives the largest possible results).

I’ll include one more screenshot, this time using the same vertical scale for all signals (except power):

This puts things more in perspective: the voltage over the capacitor (yellow) is slightly lagging the input signal (red) now. And the input current (blue) is out of phase w.r.t. the input signal (red). So the power consumption (red x blue) is substantially lower than with a resistive circuit: when the current is maximal, the input voltage is only a fraction of its maximum range. IOW, we’re drawing current when it “costs” little.

This is why a capacitively coupled supply is cheaper: the electricity company charges us for real power (i.e. V x I). We’re being charged for what happens inside a pure resistor.

But we’re doing a lot more: we’re taking charge out of the AC mains line on one half of the cycle and pushing it back on the other half. It might seem as if that doesn’t use up energy, but it does: current through a wire causes resistive losses (in the form of heat), and “returning” that energy one half cycle later causes those losses again! So the electricity company sees its electricity turn into waste heat, and they can’t charge us for it.

Here’s a thought experiment: suppose you had a huge capacitor, and hooked it up directly to AC mains, next to the electricity meter. According to what we’ve seen, it’ll track the 230V cycles, with current exactly 90% out of phase with AC mains voltage. Huge currents would flow at the time of zero volts. You’d be charged relatively little for the large out of phase current that flows (not quite zero because a practical capacitor has some internal losses that look like a resistive component from the outside). But your garden would be nicely heated by that current in the resistance of wires between the electricity company and your meter…

You can read more about “real”, “reactive”, and “apparent” power on Wikipedia.

(with a tip of the hat to Martyn for helping me understand this stuff a little better)

(Thanks for your overwhelming understanding w.r.t. not making the Low-power supply available as kit!)

The concept of the transformer-less capacitive power supply still puzzles me – intuitively I still don’t get it:

(the above image was copied from this excellent site with a web-calculcator for it all)

So what exactly is going on here? Tomorrow, I’ll simplify that circuit in an attempt to really get it, but for now let me show you what I see on the oscilloscope. What I did was feed a ≈ 100 Vpp signal into the above circuit, which is essentially the same as in the Low-power Supply.

To interpret the graph, you need some info:

• the scope ground was connected between the capacitor and the resistor
• the yellow trace is the voltage over the resistor
• the blue trace is the voltage over the capacitor
• the yellow trace is inverted, i.e. negative voltages at the top, positive at the bottom
• zero is the middle of the screen for both signals

Here’s the scope capture:

Ignore the fact that these waves are hideously complex, ignore the red lines for now, and also note that watching voltage over a resistor is the same as watching current through that resistor. Voltage and current are always proportional in a resistor, that’s what defines a resistor (Ohm’s law: voltage = current x resistance).

So what are we seeing here?

Well… when the yellow line is high, the blue line rises sharply. When the yellow line is zero, the blue line is flat.

That makes a lot of sense: the resistor is charging the capacitor. And similarly for negative values, it’s discharging the cap (and then charging it negatively).

So ignoring the zener and the rest of the righthand side of the circuit, this is really all that’s going on: when the input voltage is highly positive, current flows in one direction, charging the cap and dropping to zero, and when the input voltage is highly negative, the whole process unfolds in the opposite direction.

One more piece of the puzzle: current is “charge per time unit” (in units: Amperes is Coulombs per second).

In other words, the capacitor accumulates the charge pushed into it, in either polarity. And while it does so, the resistor “takes the heat”, so to speak: it limits the current by creating a voltage drop over itself.

Please let this sink in, dear reader. It’s essential to get a solid intuitive grasp on what’s going on.

Note also that it really makes no difference at all how complex the input signal is. The resistor and capacitor work in tandem, sharing the task of dealing with that signal. In other words: the capacitor is always trying to catch up!

This is where it gets interesting.

You may or may not have given up in high school when it came to advanced maths / calculus – derivatives and integrals, in particular. If so, then get ready to finally get to grips with these incredible concepts.

Let me explain the red lines in the above image – they are generated by the built-in math functions of this scope. One red line is the integral of the value measured across the resistor, and the other red line is the derivative of the voltage across the capacitor. But here’s the big surprise:

• the integral of the voltage over the resistor is the same as the voltage over the capacitor!
• the derivative of the voltage over the capacitor is the same as the voltage over the resistor!

What’s the point? Well, this means that I didn’t have to measure both signals to see what’s going on. I could have omitted the blue trace, because it can be calculated from the yellow trace (and vice versa). Even though these traces have completely different shapes, they are in fact totally inter-locked and inter-related.

As I said before, the voltage over a resistor is proportional to the current through it. So the derivative of the voltage over the cap is the same as the current through it (the current flowing through the resistor and the capacitor is always the same, since they are connected in series).

The derivative is the rate of change, i.e. the slope of the graph. Integrating the current (i.e. the derivative) is like adding “all the little currents together over time. A capacitor is no more and no less than a “current integrator”.

And that’s exactly the same as saying that a capacitor accumulates charge. It’s like a tiny rechargeable battery, it takes the current pushed through it and it stores that current (as charge). As the charge accumulates, the voltage rises. Loosely speaking, this is the same as saying that it pushes back harder and harder against the incoming current. At some point it pushes back so hard, that no more current comes in. At that same point, there will be zero volts over the resistor, and the voltage over the cap stays constant. Check the graph to see where that happens.

One last observation is that the blue line is a lot smoother than the yellow line. That’s not surprising: when you integrate (accumulate) a jittery signal, things tend to smooth out. That’s why capacitors are also a fundamental component in filters, i.e. circuits which let some frequencies through more and others less. That schematic we’ve been looking at here is also known as an RC circuit – if you ignore the zener and the rest. One way to look at an RC filter is to see the capacitor as the sluggish part, and the resistor as taking up the slack. So with any input signal, the voltage over the cap is related to the low frequencies, while the resistor follows more the high frequencies.

Did this explain how a capacitive power supply works? Probably not. But first we need to get to grips with what a capacitor does, and hopefully this little experiment helped you get some intuition for what’s going on.

I’ll try to take this further tomorrow, by simplifying things a bit. Stay tuned!

Oh boy, I’ve been bitten by the collector’s bug…

Couldn’t resist an excellent offer on Marktplaats (the Dutch equivalent of eBay) for a fully operational analog scope built in the late 70′s. The Tektronix 475 is, ehm… slightly larger than the Hameg HMO2024:

Then again, they are quite comparable – both are specified as 200 MHz bandwidth. Check ‘em out:

The Tek will need to be re-calibrated, but as far as I can tell all the functions work and all the switches, knobs, and indicators are in good order. The previous owner said he had used it for a long time, but not intensively.

First thing I had to try was the analog clock, of course:

Apart from that? Oh, I don’t know. I’ll refurbish and re-calibrate it one day, there’s lots of info about this classic workhorse out on the web, and I really would like to learn how to diagnose and repair stuff. Even old stuff.

This is a single-beam unit, meaning it can’t show two signals at the same time. Newer and more advanced models are dual-beam, but this one has to either “alternate” between the two beam displays, or “chop” them up and draw bits of one and the other in an interleaved fashion. It’s also not a storage scope, so you’ve got to look very carefully if you want to see one-shot events – the display on the screen shows only as long as the phosphor glow lasts!

Fantastic engineering. Electronics, mechanical, operation, documentation, service – everything!

Would I have bought the Hameg if I’d had this one at the time? You bet – the “S” in DSO is a game changer, especially with microcontrollers and physical computing. But that huge Tek brick is more endearing :)

Heh, I’ve never before collected anything in my life – this is fun!

To avoid switching noise near the work bench, I wanted to drive my LED strips from a plain ol’ transformer -> bridge -> capacitor supply. It would have to supply about 2A @ 12V for my LEDs. Here’s the setup:

Ignore the diodes at the top, these were 1N4004 diodes which weren’t quite up to the task. The ones at the bottom right are Schottky diodes for a lower voltage drop, rated at 5A. The capacitor is 2200 µF, but ripple isn’t an issue.

Alas, the heat from this thing is excessive. I even went for an artsy swiss-cheese design:

Trouble is the transformer (and surprisingly not the diodes!) – I measured them both, with everything closed:

Diodes heated up to 75°C – but the transformer went all the way up to well over 90°C within the hour! AC mains consumption is 50W, so this silly thing is eating up half the power while trying to behave like a stove…

The transformer is a cheap 2x6V @ 2.5A unit I had lying around (draws 3.7W, even without any load). I’ll keep this as a high-ripple 12..20 VDC power brick for testing - but it definitely can’t be used continuously @ 2 Amps!

PS. If you’d like to meet up – I’ll be presenting JeeStuff at HackersNL coming Thursday evening in Utrecht.

A while back, Jeroen mentioned on the forum that he had an ancient GM5655 Philips oscilloscope, and since he lives in Houten we thought it’d be fun to put it side-by-side with my new Hameg scope I keep generating screen shots with in this weblog (ad nauseam for some, I suspect…). Here’s his vintage 1955 hardware (from a web image) – just imagine how attractive that must have looked to geeks back then:

It’s single-channel unit, with a sweep generator which can go “all the way” up to 30,000 Hz. The horizontal deflection can also be driven externally, making this scope X-Y capable. Vertical sensitivity is up to 60 mV/cm, horizontal up to 100 mV/cm. No positioning, no magnification, nothing. This is as basic as a scope can be.

Jeroen dropped by a few days ago and we had a go at it. Unfortunately, we couldn’t get it to work. The horizontal deflection appears to be broken (not just the sweep generator, since it also didn’t respond to external signals). Most probably one or more of the “tar capacitors” has failed. It’s always them capacitors with old equipment…

Neither of us felt confident enough to go about messing with this thing while powered on (vacuum tubes, and especially the CRT, run on high voltage). But the least we could do is open it up and marvel at how things were constructed back then. No semiconductors and no printed circuit boards – pretty amazing, if you think about it.

It’s a pretty scruffy unit, once you open it up!

Note the big tar caps and the “flying” construction, especially around the rotary switches:

Here’s the other side, with what looks like just the vertical input circuit:

Here’s the top view, with 6 vacuum tubes in total:

The bottom view, with one more tube and what looks like a bunch of electrolytic power supply caps to me:

All this was manually assembled, so each unit had to be tested and debugged, since it wouldn’t take much to hook up things incorrectly. The invention of the printed circuit board not only creates a mechanical platform to hold everything in place (especially for low-voltage semiconductors), it also acts as a self-documenting area during assembly, since all the spots are already marked with what goes where.

Amazing stuff – I’m tempted to get a (somewhat more modern) analog scope off eBay, just for the fun of it :)

To follow up on yesterday’s post, here’s a similar TCXO oscillator which is considerably more accurate:

It’s a 10.000 MHz unit which has been calibrated and claims a 0.1 ppm accuracy. That’s 10 MHz ± 1 Hz, i.e. 7 digits. Here’s the output signal, which appears to be AC coupled:

Not bad for a low-cost unit. I got it from the RFcandy shop. Current consumption is 0.94 mA @ 5V.

Once you start looking into this stuff, all sorts of things pop up on the web. There’s a group of amateurs who call themselves Time-Nuts (terrific name!) and who aim to keep track of time as accurately as they can – with pretty amazing results. In a comment yesterday, John Beale pointed me to a web page with info about the Rubidium clocks now available cheaply on eBay.

I found this page most entertaining, summarizing the history of accurate time-keeping in a few slides.

The RTC Plug uses a DS1340 chip to keep track of time. It has its own on-board coin cell and 32 KHz crystal, so that it can keep running when power to the JeeNode is switched off.

The 32,768 Hz crystal is rated at 20 ppm, with 3 ppm aging in the first year. A year has about 3600 * 24 * 365.25 = 31557600 seconds. With 20 ppm, the RTC Plug can be some 630 seconds off at the end of the year, i.e. close to one minute per month – worst case, that is. It’s fine for lots of uses, but evidently not all…

Here’s a new plug prototype, designed by Lennart Herlaar:

It’s based on the DS3231 chip, which includes a “TXCO”, i.e. a Temperature Compensated Xtal Oscillator. There are many ways to improve the stability of a crystal oscillator, but this one is pretty nifty because it’s low-power (unlike an oven-based technique, which puts the crystal in an controlled temperature environment). The idea is that the chip periodically measures the ambient temperature, and then adjusts the capacitive loading of the crystal according to a calibrated function of temperature. IOW, it knows how to compensate its on-board crystal.

The DS3231 is specified at 2 ppm for temperatures 0 .. 40°C, i.e. ten times more accurate than the above crystal. So this clock will stay on time with an error of at most one minute per year, i.e. about a second per week.

For something, eh… slightly better – you could consider a Rubidium Frequency Standard with 50 ppt accuracy, i.e. well under a millisecond per year error. But don’t expect to run it off a coin cell – it draws about 11 Watt :)

If there’s enough interest, this plug could be added to the shop…

There’s one dirty little secret about most digital storage oscilloscopes (DSOs) which puts them way behind the capabilities of their analog brethren: screen update rate.

For a scope to present a trace on screen, it has to do a lot of signal processing. After all, 1 GSa/s means its acquiring 1 billion data values per second. And although we’re not able to really see that with our eyes, we do see things that stay on the screen for a few milliseconds.

One trick is to turn on persistence. On analog scopes, that’s a pretty nifty feature whereby the image generated from the beam hitting the screen is made to stay visible for a while. Either via a phosphor coating which keeps on glowing for a while, or by constantly re-sending the same beam pattern to keep the visual display going.

Digital scopes can simulate this. All they need to do is leave the pixels on their LCD display as is, right?

Unfortunately, it’s a lot more complicated than that. If you just refresh the screen a few times a second, then you’re still going to miss a huge amount of detail. Let’s take a periodic 1 MHz signal: to display everything measured, the display needs to be updated 1,000,000 times per second, which means each “sweep” would only remain visible for 1 µs. That won’t do – our slow eyes wouldn’t see a darn thing, especially glitches which occur very infrequently. So what a DSO does, is simulate persistence by merging new sweeps into what’s being shown, and then take old sweeps out of the picture a bit later (try implementing that – efficiently – !).

To test this stuff, I used the following sketch (which is based on this pin flipping trick):

It generates pulses at ≈ 1 MHz, but every once in 10,000 pulses, it messes up the timing. Note that interrupts are turned off, so this thing is as stable as the clock it’s running on (a JeeNode with a 16 MHz resonator in this case).

One glitch every 10,000 pulses at ≈ 1 MHz means there are 100 glitches per second. That should be constantly visible on the screen, right? Well… not so on my scope. I enabled 10s persistence, i.e. sweeps will fade after 10s:

That’s two glitches (the actual display is a bit erratic, but the position of the glitch pulses is absolutely repeatable).

IOW, this scope is showing only one out of every 500 pulses on the screen, on average!

This is why DSO manufacturers report a “waveform/sec” value in their specs. In this case, we are getting 2,000 waveforms per second, even though a scope could theoretically sweep 500,000 times per second on this.

An analog scope would have no problem whatsoever with this. To show the same two pulses in each sweep, it could trigger a new sweep 500,000 times per second, or maybe even “only” 100,000 but that would still show 10 glitches per second, i.e. a nearly constant visual display of the glitches. Fifty times more often than my DSO, anyway.

That 2,000 wf/s figure is an optimistic one, BTW. With more channels enabled, or things like signal stats calculated or the math function applied, this value can drop substantially. Meaning you’ll see even less glitches.

What this tells me, is that the Hameg needs about 500 µs of processing time to get 1 channel of acquired data onto the screen in its most basic form.

Don’t get me wrong: technically that’s an astonishing achievement. It’s not just copying a few bytes and setting a few pixels. It’s (always!) performing the emulated-persistence trick – because with persistence turned off, I still see a glitch every few seconds, which means that the LCD display is showing the glitch for a substantial fraction of a second – much longer than the 1 µs (or even 500 µs) sweep rate would suggest. Besides, you can see that fading is also implemented, so it’s not just drawing pixels but actually fooling around with color intensities.

There are some scopes which get much higher waverform/sec rates, such as the Agilent 2000X (50,000 wf/s) and the Agilent 3000X (1,000,000 wf/s) series. But they are twice the price, and even have lower specs in other areas.

Does it matter? Not for me. Being aware of this is good, though – far more important than fixing it. Note also that IF – that’s a very big if – you know what you’re looking for, then there are usually other ways to find these things. In this case, triggering on a negative pulse width under 900 ns captures all those glitchy pulses, for example.

As so much in life, it’s all about trade-offs. Analog / digital, brands, and of course… your needs and budget.

Recently, the radioBlip sketch reached yet another milestone: it’s been running for 500 days on one LiPo charge!

Seems like a good time to start a new duration test, this time with a JeeNode Micro:

Since I don’t want to have to wait yet another 500 days, I decided to up the ante:

The JeeNode Micro includes holes for a CR2032 3V battery holder, so why not try that for a change, eh?

The sketch was modified slightly, to set the proper clock divider on startup:

I’ve set it up as node 18, and it’s been blipping happily:

Let’s see how many days that little JNµ can dance the Sputnik shuffle…

After having become reasonably familiar with the new Hameg HMO2024 oscilloscope, I wanted to find out just what its limits are. Came up with this screenshot:

There’s quite a bit of information in here.

First of all, these scopes can do double-rate sampling when you don’t activate all the channels. Channels 1 and 2 share some hardware, and so do 3 and 4, so by enabling only one of each block, the scope can actually double its sampling rate to 2 GSa/s i.s.o. 1 GSa/s and also “stack” its memory depth to 2 MB per channel, i.s.o. 1 MB. Or to put it another way: a 4-channel oscilloscope can be used as a 2-channel unit with twice the specs.

I switched the acquisition to display in “sample-and-hold” mode instead of the normal Sin X interpolation mode, so the steps shown above represent the real sampling that’s taking place. As you can see, there are 4 steps per division, with 2 ns/div, so that’s indeed 2 billion samples per second.

The top was zoomed out as far as I could while still showing the fastest supported 2 ns divisions in the detail window, which ends up being 50 µs/div. Running the scope timebase any slower will cause it to reduce its sample rate so it can still capture the required 12 horizontal divisions full of sampling data. In this case we see 12 x 50 µs = 600 µs of data on the screen, while the detail view is zoomed in to 2 GSa/s (25,000x).

That’s 1.2 million samples of data – a hefty pile of bits flying around in this thing while it’s running!

But there’s actually more, because of the stacked memory. When I pan across the 50 µs/div display, I can see that data was collected from 575 µs before the trigger point to 525 µs after the trigger point. Which is 1100 µs of collected data, IOW that’s 2.2 MB of acquired sample data – again, more than the specs in the brochure!

The vertical signal acquisition hardware is also very impressive. This is one of the few scopes I know of which will go down all the way to 1 mV per division. That’s real sampling, not some sort of digitally enhanced sensitivity, as you can see from the fact that the data steps really are 1 pixel in the vertical direction.

It might not seem important to go down that low, but it’s quite useful actually when measuring voltage drop over a shunt resistor. Which is what I’ve been doing quite a bit to display current usage of JeeNodes while in ultra low-power sleep. The other benefit is that you can still go down to 10 mV/div with the standard 1:10 probes that come with the scope (I’ve got a few 1:1/1:10 switchable no-brand probes for when I really need the extra sensitivity).

The red line is created by using the “Quick Math” function, subtracting the channel 3 signal from channel 1 in this case. The nice thing is that the math function supports 20x more magnification than the input channels. So with some trickery, this scope can display down to 50 µV/div: subtract a spare channel set to “GND” from the input signal, and magnify the resulting math output 20x. It’s a coarse (digital) way of magnifying the display, but still.

The actual data shown above is what the scope displays with no probe connected, BTW. This is the residual noise in the scope’s input circuitry and it really is impressively low-noise – just a quarter millivolt peak-to-peak.

At the other end of the range, the scope will go up to 10 V/div, i.e. 80V full scale. That’s ≈ 6 orders of magnitude of usable range on each channel. Wow… Hameg (and R&S) did some pretty serious engineering to achieve this.

PS. I’m ready to ditch my DSO-2090 USB scope – it’s unlikely that I’ll use it with this HMO2024 now at hand.

Update – same for the Logic Analyzer – a LAP-16032U, if you’re interested in it, let me know.

V-scoring is a low-cost way to cut up printed circuit boards. I use it in combination with routing to get al ‘em plugs made out of larger panels. It consists of slightly cutting into the PCB material from both sides – when done properly, the resulting boards can be snapped off quite easily by hand.

Recently someone gave me a really neat breakout panel for SOIC and TSSOP IC’s. These are useful to try out SMD chips on a breadboard. Except that the board I got was still panelized…

Can’t look a gift horse in the mouth, and since I don’t have a mini bandsaw or such, I had to find another way.

As it turns out, it’s quite easy to do the v-scoring by hand:

The basic idea is to lightly scratch all the lines about three times, and once there’s a decent groove, just take the hobby knife and cut two more times, pressing down fairly hard. Rinse and repeat for all the cuts on both sides, and things will come apart by pressing down firmly. Here’s what I ended up with after about 15 minutes:

The edges can be a bit rough, as usual with v-scoring, but those rough bits can easily be scratched off with the knife.

And this is what it’s all about:

Tiny chips, ready for breadboard use!

Have you noticed that solder fumes always tend to rise up into your nose? As the type of solder I’m using is particularly irritating, I decided to do something about it, ending up with a very low-tech solution…

Normal fume extractors (Dutch link) tend to suck the air in, filtering it through a carbon filter to actually remove the fumes from the air, but I find them a bit unwieldy. Yet another power cable to clutter my workbench.

Here’s my solution:

A small 12V brushless PC fan, which happens to still start up and rotate reliably at 3.6V, as provided by three Eneloop AA batteries. Stuck to a battery pack (with on/off switch) using hot glue, which also doubles as weighted base for it all. The fan is slanted slightly downwards, to better cover the area on the table.

The result is whisper quiet and generates a micro breeze which gently diverts the solder fumes away from my nose. This particular fan draws only 30 mA @ 3.6V, so it’ll last some 70 hours on a single battery pack. And since these are rechargeable batteries of which I keep a constant supply, that’s plenty for me.

Long live re-use!

To try and improve noise levels during measurements (and as general ESD precaution), I went “green”:

That’s a green ESD mat, covering almost the entire workbench. It’s hooked up to the radiator for grounding. Note that the mat only provides a “dissipative” connection to ground, the surface still has several MΩ’s of resistance. It’s just to get rid of static electricity and to offer a clean protective working surface. Got the mat from Farnell, BTW.

Here’s what I see when pushing a scope probe onto the mat:

A clear 50 Hz pattern of a couple of volts (the amplitude increases when the probe is pushed more firmly onto the mat). This is with a standard 10 MΩ high impedance probe. The big puzzle is: where does this come from?

My explanation for now, is that the scope ground is “floating” a bit, due to the different devices hooked up to the house wiring. Note that the mat is tied to the radiator, not to the ground of AC mains. Since there isn’t any current flowing through the heating system (I hope!), I’m inclined to trust it more as being the “real” ground potential.

It’s no doubt completely harmless. Measuring AC current between these two ground levels with my most sensitive multimeter (a Voltcraft VC940), I see a very occasional blip of up to 0.20 µA flowing. Well under a microwatt.

To give an idea how crazy things can get, here’s the mat with nothing connected – turning it into one big antenna:

Who knows, maybe one day we could harvest even this sort of energy, eh?

A long time ago, I got this DDS-60 kit, which is a small circuit based on an AD9851 DDS chip:

It has everything on board to generate a sine wave from 0..30 MHz, and my intention was to hook it up to a JeeNode (as part of a long term plan of mine to set up a more extensive wirelessly-controlled electronics lab):

Never got it to work at the time, but now with the new scope, there really is no excuse anymore. First check, as indicated in their build instructions, is to verify that the crystal oscillator is feeding a 30 MHz clock into the chip:

Looking good. Very impressive rise and fall times, BTW.

When driven at 30 MHz, the AD9851 output frequency is settable in steps of 0.006984919 Hz. In other words, a multiplier of 1000 will generate a sine wave of ≈ 7 Hz.

Here’s the output when programmed to generate 10 MHz (multiplier 1431655765, i.e. 9999999.9977 Hz):

Whoa… it’s 10 MHz, but a far cry from a sine wave. Ah, but that’s not really surprising: this thing uses DDS to synthesize a sine wave, as recently described on this weblog. With 30 MHz sample rate, i.e. 3 samples per wave, it’s not really possible to create a decent 10 MHz sine wave (not even a symmetrical shape in fact, as you can see).

But the AD9851 has a trick up its sleeve: it includes a “6x” multiplier option, which causes it to internally generate a reference frequency which is 6x the incoming clock, i.e. 180 MHz in this case.

Using that, and adjusting the frequency setting to work at 180 MSa/sec, we get a much better approximation:

Still not perfect, but by analyzing the FFT of this signal, we can find out what’s going on:

This output signal is made up mostly of a 10 MHz sine, with another peak at 90 MHz.

Unfortunately, the output circuit on this board isn’t working yet (this is what probably threw me off when trying this circuit before), so I can’t test the effect of the 60 MHz low-pass filter yet. It won’t filter out the 30 MHz residue visible in that last picture, but should definitely reduce the frequency components over 60 MHz.

Ok, all the important bits seem to work – I “only” need to troubleshoot that analog back end a bit more.

Update – I found the problem: the SMD trimpot was fractured, i.e. no contact. I’ve replaced it with a fixed 220 Ω resistor for now – this brings the output to ≈ 680 mVpp (or 350 mVpp into 50 Ω) – the sine wave output is now considerably cleaner, but several of the frequency peaks ≥ 90 MHz are still present:

I suspect that the 30 MHz clock is “feeding through” somewhere, perhaps better decoupling would avoid that.

The quickest way to describe what Fritzing is about, is to quote the first paragraph on their site:

Fritzing is an open-source initiative to support designers, artists, researchers and hobbyists to work creatively with interactive electronics. We are creating a software and website in the spirit of Processing and Arduino, developing a tool that allows users to document their prototypes, share them with others, teach electronics in a classroom, and to create a pcb layout for professional manufacturing.

It has been around for a while, evolving and improving at a steady pace. Here’s a screen shot of the main window:

The neat bit is that support has now been added for the JeeNode, as well as support for the main two prototyping / extension mechanisms, the JeePlug and the Carrier Card. Just search for the “JeeLabs” keyword:

So now you can use Fritzing to design and document projects, and create printed circuit boards for them!