Jan Czochralski And The Silicon Revolution

July 22, 2020 by Dan Maloney 39 Comments

If you were to travel back in time to the turn of the previous century and try to convince the average person that the grains of sand on just about any beach would be the basis of an industry worth hundreds of billions of dollars within 100 years, they’d probably have thought you were crazy. Aside from being coarse, rough, and irritating, sand is everywhere, and convincing anyone of its value would be a hard sell, unless your interlocutor was a real estate visionary with an appreciation of the future value of seaside property and a lot of patience.

Fast forward to our time, and we all know the value of the material that comes from common quartz sand: silicon, specifically the ultra-purified crystals of silicon that end up as the wafers we depend on to build the circuitry of life. The trip from beach to chip foundry is a long and non-obvious one which would not have been possible without the insights of an undistinguished Polish student and one-time druggist who discovered the process that made the Information Age possible: Jan Czochralski.

Continue reading “Jan Czochralski And The Silicon Revolution” →

Pulling A Crystal By Grinding It

May 10, 2020 by Jenny List 29 Comments

If you own a radio transmitter, from a $10 Baofeng handheld to a $1000 fancy all-band transceiver, setting the frequency is simply a case of dialing in where you want to go. A phase-locked-loop frequency synthesizer or a software-defined radio will generate your frequency, and away you go. There was a time though when synthesizers were impossibly complex and radio amateurs were faced with a simple choice. Use an LC oscillator and put up with drifting in frequency, or use a crystal oscillator, and be restricted to only the frequencies of the crystals you had. [Mark Erdle, AE2EA] modified a 1950s broadcast AM broadcast transmitter for the 1.8MHz amateur band, and his friend [Andy Flowers, K0SM] thought it needed its crystal back for originality rather than the external frequency source [Mark] had provided. He documents the process of modifying a crystal oven and moving a crystal frequency in the video below the break.

A crystal oven is a unit containing the crystal itself alongside a thermostatic heater, and in this one, the crystal was a 1970s-vintage hermetically sealed HC6 device. He modified the oven to take a socket for older FT243 crystals because the quartz element can easily be accessed. [Andy] picked a crystal as close as he could find below the required frequency. He then ground it down with very fine grit on a glass plate, reducing its mass and thus its resonant frequency. We’re taken through the process of getting it close to frequency, but sadly don’t see the etching that he uses for the very last stage. At the end of the video, we see a QSO on the transmitter itself, which is something of an oddity in an age when AM on amateur bands has been supplanted by other modes for decades.

If you’re curious about the transmitter there’s a video thread following its restoration, and if the guts of older radio gear interests you then take a look at this aircraft receiver lovingly brought back to life.

Continue reading “Pulling A Crystal By Grinding It” →

Gorgeous Clock, And Not A Line Of Code In Sight

March 4, 2020 by Donald Papp 41 Comments

[Harry] dropped us a note to let us know about his completed CMOS clock project, and we’re delighted that he did because it’s gorgeous. It’s a digital clock satisfyingly assembled entirely from hardware logic, without a single line of code. There are three main parts to this kind of digital clock: ensuring a stable time base, allowing for setting the time, and turning the counter outputs into a numerical display.

Keeping accurate time is done with a 32.768 kHz crystal, and using CMOS logic to divide that down to a 1 Hz square wave. From there, keeping track of hours and minutes and seconds is mostly a matter of having counters reset and carry at the appropriate times. Setting the clock is done by diverting the 1 Hz signal so that it directly increments either the hours or minutes counter. The counter values are always shown “live” on six 7-segment displays, which makes it all human-readable.

The whole thing is tastefully enclosed in a glass dome which looks great, but [Harry] helpfully warns prospective makers that such things have an unfortunate side effect of being a fingerprint magnet. Schematics and design files are provided for those who want a closer look.

This clock uses a crystal and divider, but there’s another method for keeping accurate time and that’s to base it off the alternating current frequency of power from the grid. Not a bad method, albeit one that depends on being plugged into the wall.

A Radio For The Apocalypse

November 15, 2019 by Jonathan Bennett 54 Comments

There’s been a spate of apocalypse related articles over the last few weeks, but when I saw an AM radio made from a hand-wound coil and an oxidized British penny, I couldn’t help but be impressed. We’ve covered foxhole radios, stereotypical radios that are cobbled together from found parts during wartime.

This example uses a variable capacitor for tuning, but that’s technically optional. All that’s really needed is a coil and something to work as a diode. Surprisingly, copper oxide is a semiconductor, and the surface oxidation on a penny is enough to form a rudimentary diode. Though, note, not all pennies have that necessary coating of copper. If a penny has green oxide, it’s likely a candidate.

Need a quickly cobbled together AM radio? Have some wire and a penny? Yeah, watch the video below the break, and you’ll know how to make it happen. When the apocalypse comes, you’ll thank us.

Continue reading “A Radio For The Apocalypse” →

Crystal Oscillators Explained

December 8, 2018 by Al Williams 13 Comments

We’ve read a lot about oscillators, but crystal oscillators seem to be a bit of a mystery. Hobby-level books tend to say, build a circuit like this and then mess with it until it oscillates. Engineering texts tend to go on about loop gains but aren’t very clear about practice. A [circuit digest] post that continues a series on oscillators has a good, practical treatment of the subject.

Crystals are made to have a natural resonant frequency and will oscillate at that frequency or a multiple thereof with the proper excitation. The trick, of course, is finding the proper excitation.

The post starts with a basic model of a crystal having a series capacitance and inductance along with a resistance. There’s also a shunt or parallel capacitor. When you order a crystal, you specify if you want the resonant frequency in series or parallel mode — that is, which of the capacitors in the model you want to resonate with the inductor — so the model has actual practical application.

By applying the usual formula for resonance on the model you’ll see there is a null and a peak which corresponds to the two resonance points. The dip is the series frequency and the peak is the parallel. You can actually see a trace for a real crystal in a recent post we did on the Analog Discovery 2. It matches the math pretty well, as you can see on the right.

Continue reading “Crystal Oscillators Explained” →

The Crystal (Testing) Method

October 20, 2018 by Al Williams 19 Comments

It used to be any good electronics experimenter had a bag full of crystals because you never knew what frequency you might need. These days, you are likely to have far fewer because you usually just need one reference frequency and derive all the other frequencies from it. But how can you test a crystal? As [Mousa] points out in a recent video, you can’t test it with a multimeter.

His approach is simple: Monitor a function generator with an oscilloscope, but put the crystal under test in series. Then you move the frequency along until you see the voltage on the oscilloscope peak. That frequency should match the crystal’s operating frequency.

Continue reading “The Crystal (Testing) Method” →

Transistor Logic Clock Gets Stacked Up

September 6, 2018 by Tom Nardi 10 Comments

A couple years back we covered a very impressive transistor logic clock which was laid out so an observer could watch all of the counters doing their thing, complete with gratuitous blinkenlights. It had 777 transistors on 41 perfboards, and exactly zero crystals: the clock signal was extracted from the mains frequency of 50 Hz. It was obviously a labor of love and certainly looked impressive, but it wasn’t exactly the most practical timepiece we’d ever seen.

Creator [B Brett] recently wrote in to share news that the second version of his transistor logic clock has been completed, and we can confidently say it’s a triumph. He’s dropped the 41 perfboards in favor of 9 professionally fabricated PCBs, which this time around are stacked vertically to make it a bit more desktop friendly. The end goal of a transistor logic clock that you can take apart to study is the same, but this “MkII” as he calls it is a far more refined version of the concept.

In addition to using fewer boards, the new MkII design cuts the logic down to only 283 transistors. This is thanks in part to the fact that he allowed himself the luxury of including an oscillator this time. The clock uses a standard watch crystal at 32.768 KHz, the output of which is converted into a square wave through a Schmitt trigger. This is then fed into a divider higher up the stack which uses flip flops to produce 1Hz and 2Hz signals for use throughout the rest of the clock.

In addition to the original version of this project, we’ve also seen a beautiful single-board wall mounted version, and even a “dead bug” style one built from scraps.

Continue reading “Transistor Logic Clock Gets Stacked Up” →