Another weekend another weekend die-shot! Zeptobars have this beautiful blue-and-orange image of a DS2401 silicon chip.
It has some really interesting features including patent numbers on the die:
And these paths leading to the ROM at the center of the silicon:
The full-size image is 47+ megapixels, so you can really navigate around it up-close if you wish. You’ll also note the iButton trademark stamp, alluding to this silicon came from or was intended for an iButton device.
Having hacked away with high voltage for many years I’ve ended up using a large number of very different high voltage sources. I say sources and not power supplies because I’ve even powered a corona motor by rubbing a PVC pipe with a cotton cloth, making use of the triboelectric effect. But while the voltage from that is high, the current is too low for producing the necessary ion wind to make a lifter fly up off a tabletop. For that I use a flyback transformer and Cockcroft-Walton voltage multiplier power supply that’s plugged into a wall socket.
So yes, I have an unorthodox skillset when it comes to sourcing high voltage. It’s time I sat down and listed most of the power sources I’ve used over the years, including a bit about how they work, what their output is like and what they can be used for, as well as some idea of cost or ease of making. The order is from least powerful to most powerful so keep reading for the ones that really bite.
You’ve no doubt encountered this effect. It’s how your body is charged when you rub your feet on carpet and then get a shock from touching a door knob. When you rub two specific materials together there’s a transfer of electrons from one to the other. Not just any two materials will work. To find out which materials are good to use, have a look at a triboelectric series table.
Materials that are on the positive end of the table will become positively charged when rubbed against materials on the negative end of the table. Those materials will become negatively charged. The further apart they are in the table, the stronger the charging.
This would be considered an electrostatic power source because charge is accumulated on surfaces. Being insulating materials, that charge can’t move around.
The amount of charge transferred between the materials per unit of time is small meaning that the current available is small. You won’t be powering any heavy loads with this, but the corona motor powered this way turns at around one revolution every 5 seconds and can be stopped with the light touch of a finger. You already have a feel for the power from getting mild shocks from touching doorknobs. This is of course an easy power source to make.
A Wimshurst machine is also a high voltage/low current power source. It consists of two counter-rotating disks, usually rotated with a hand crank. The disks have metal sectors on them that are spaced apart. The charging occurs where the neutralizer brushes contact the sectors as they pass by. That charge is then removed at collectors on the left and right edges of the disks and is usually accumulated in Leyden jars (capacitors) and across a spark gap where a spark occurs when the voltage is sufficient to break down the air in the gap.
But if you’re making use of the Wimshurst machine then you’re usually not producing sparks. In the photo you see wires going from the Wimshurst machine to a ball cyclotron, making the balls in it rotate around inside the bowl.
The voltage with this one is limited by your losses in the Leyden jars and spark gap and your load. That’s why efforts are made to have everything be well rounded. The spark gap also limits the voltage and with this one I’ve produced sparks around 3 inches/7 centimeters long.
The current is indirectly determined by the disks’ diameter. That’s because larger selectors will produce more charge than smaller sectors. Also, the faster the disk turns, the more sectors will pass by the collectors per second and so more charge will be available.
They aren’t too hard to make. I find that the trickiest parts are to find or make the pulleys needed for transferring the hand crank rotation to the disks. The disks can be acrylic which you can cut with a scroll saw or laser cutter, and often small Wimshurst machines are made using CDs.
Van de Graaff Generator
From the outside, a Van de Graaff generator looks like a big ball, or dome, on top of a vertical tube and more stuff at the base of the tube. While that dome is hollow, inside the tube is a belt on rollers. The stuff at the base of the tube includes a motor to turn the rollers and belt. The outer surface of the belt is charged by a combination of the same triboelectric effect we spoke of above, and some nearby sharp pointed brushes. That charge is transferred to the outer surface of the dome.
The amount of charge that can accumulate on the dome is limited only by its diameter. A smaller diameter dome can be thought of as a sharper object than a larger diameter dome. Sharper objects have stronger electric fields surrounding them, which break down the air more easily, taking charge from the object. The big Van de Graaff pictured is rated at around 400kV and the small one is around 80kV.
They’re still a low current source, as the current is produced by the triboelectric effect at the belt and rollers and mechanically transported by the movement of the belt. Wider belt and rollers and a faster rotation gives higher current.
They’re medium hard to make. Since the triboelectric effect is involved, the rollers and belt have to be the right combination of materials. For a small one the dome is often a soda can and for a large one it’s often made using metal salad bowls or a large garden ball.
Moving on to the higher current power sources, an often used type is a circuit using a flyback transformer and one or more transistors. As expected, the flyback transformer has a primary winding which induces a current in the secondary winding. However, there’s also a feedback winding which at the same time shuts down the transistor which stops the current going to the primary. This causes the magnetic field to collapse and a large high voltage spike to appear at the secondary. Since there’s now no more current on the feedback coil, the transistor turns on again and the cycle repeats.
Flyback transformers can be bought online but can also be salvaged from old CRT TVs and CRT PC monitors. They’ll most often have high voltage diodes built-in after the output of the secondary winding, which means the output is DC.
I built mine into the small cube shown above. You can see the nice continuous arcs you can get from it. I’ve also powered a Jacob’s ladder. Mine produced around 20kV output with a high current.
Flyback transformer plus Cockcroft-Walton voltage multiplier
If you’re lucky enough to find a flyback transformer with no built in diodes like the one shown here, then at the output you can add a Cockcroft-Walton voltage multiplier circuit. This multiplier consists of capacitors and diodes that take the flyback’s alternating output and smooth it out to flat DC but while multiplying the voltage over some number of stages. The number of stages simply depends on the number of sets of capacitors and diodes you add on. Each added stage increases the output voltage.
The voltage will have been stepped up, but the current will be lower than without the multiplier. It will still be high though, high enough to provide sufficient ionization to make a lifter fly.
You can make your own multiplier boards or you can buy multipliers. The ones you can buy are usually called triplers since they have three stages. They’ll raise a 20kV input to 30kV, also sufficient for flying a lifter.
An almost out-of-the-box source of these types of power supplies is to use an old CRT PC monitor. Simply remove the high voltage wire going into the cathode ray tube and use that wire. I do find that long sparks will damage these monitors easily so be sure to include around 240 kilohms of at least 2 watt rated resistance in series with the output.
These are the interesting high voltage sources with which I have experience. But I’d love to hear about your own high voltage hacks in the comments below. I’d also enjoy hearing questions or ideas on using or building high voltage supplies so don’t be shy.
If you’ve been to an apartment complex with a locked front door, you’ve seen the buzzer systems. You press the corresponding button for the apartment you want and can talk to the resident. They can press a button to unlock the door briefly, and then you go up to their apartment and they don’t have to come down to let you in. But what if you’re the resident and you want to go for a run without your keys jingling in your pocket? What if you want to open it using just your smartphone?
I knew this was a silly problem, and everyone I told about it thought that for the amount of time and effort it might save, it was hardly worth it.
How fast can I put this together using only parts I have around the apartment? Turns out about 2 hours.
The most critical step was hacking into the buzzer system. Two screws and a razor blade later and the plate was separated from the wall. Markings on the PCB made it easy to find on the Internet. Looking at the PCB and online schematic, there was no DC power to use, but I did know that the switch connected two wires which resulted in a buzzed door. This was a job for a low voltage relay, the normal solution when you want to press a button but don’t know what the two wires do and want to interfere with the electronics as little as possible.
Next was the wireless component. I had a Particle Photon lying around in my box of dev boards, which was super convenient. I got that set up with the Particle app from the Play store. Then I installed IFTTT, which would let me control it eventually. Then I set up a Particle account and an IFTTT account. I plugged in the Photon, connected to it from my phone, and told it how to connect to my wireless. A few seconds later it had rebooted and was showing up online. Now I could write code in the Particle web app, flash it to my Photon over WiFi, and I was good to go.
I started with an example blink app (web-connected-led.ino) and flashed it to my Photon to make sure it worked. Then I just needed to make it turn on for a second, then turn off, so I created a function to do that specifically. Done. The entire code portion took only a few minutes.
I found a relay. Relays take a lot of power to switch, and this one switches at 5V, but the Photon GPIO is at 3.3V. I wouldn’t be able to drive the relay straight from the output pin. No problem, this is a job for an N-Channel MOSFET. I only had a surface mount one, but I also have mad soldering iron skills.
I hooked up the MOSFET and relay, and I threw in a pull-down resistor so that when powered on there wouldn’t be any weirdness and the MOSFET will stay low and not switch unintentionally. Finally, I put in a 1/8″ audio connector so that I can easily attach it to the wire from the wall plate(which was a repurposed audio cable).
Dinner and Internet Outage
For some reason my Internet went out for an hour, and it was the perfect time to get some dinner.
Back to work, in IFTTT, I connected it to the Particle channel. I created a recipe that would call the function I had written on my Photon whenever I texted myself the word “opensesame”. It’s really amazing how well the Particle channel interacts with my code on my Photon. I could flash the firmware and the IFTTT recipe creation would immediately know what functions were available.
Then I noticed Do, which is part of IFTTT but drops off the IF part. After a quick download I had the ability to have a shortcut button on my home screen which when pressed would call the function on my Photon. No more text messages, just a shortcut.
A quick test and the process worked. I could hear the relay click on then off. I was confident that this solution would work when I plugged it into the buzzer.
Unfortunately, there wasn’t power near my wall buzzer, and I didn’t want to change batteries regularly. I spent an unnecessarily long period of time figuring out how to fish the wire through the drywall and down and to the other side of the wall, where it went into another room that had close access to an outlet. I ended up making a hole in the drywall about 1/2″ in diameter at the bottom of a closet. This was the only damage to the apartment, and it is easily fixed.
The wire fished, I just had to screw it into the terminals on the wall plate. It was the PT and T terminals, and you can clearly see that these two terminals have traces to either side of the switch helpfully marked “DOOR”.
I brought my phone (and keys just in case) out to the front door, pressed the button, and after about 5 seconds, there was the buzz that let me in. The next morning, I went for my run without my jingling keys.
I’m not concerned about security. The Photon is connected to my home router via encrypted WiFi, and the IFTTT and Particle integration uses OAuth.
There is some delay between when I press the button and when the Photon gets the message and buzzes the door. It turns out it would probably be faster for me to just use the key to open it.
Yes, this relies on the Internet being up, so it’s a little worrying that I had an outage while I was working on the project. I won’t rely on this all the time, but it’s a good gimmick.
I’ve never used IFTTT before, so I’m really happy with how quickly this hack came together and how easy it was to do, especially with the Particle channel.
We’ve had other buzzer hacks over the years, like the morse code buzzer, but this one was an exercise in speedy implementation and integration with existing services, showing how easy it has become to make IoT out of anything.
Last weekend was Hamvention, the place you want to be on the third weekend in May. It is the world’s largest gathering of amateur radio enthusiasts, and an exceedingly large flea market containing all sorts of electronica.
The booths of Hamvention include a few notable Open Hardware folk, but for the most part, you’re looking a few big booths from Yaesu, an entire section dedicated to everything ARRL, and a few pop-ups from the usual suspects. Rigol was there, showing off their test equipment and selling the DS1052E oscilloscope for far more than it’s worth. The Rigol Zed is a much better buy, anyway.
As with any gathering of hams, antennas are everywhere. The largest by far was the tower at right. With a little more equipment, this antenna could do a moon bounce. It’s a shame the moon was full this weekend, and everyone went to bed early.
Giant antennas and an amateur radio trade show notwithstanding, the biggest draw is the flea market. You’re looking at about two football fields worth of parking spaces, filled with cars, tents, and collapsible tables and the strangest electronic devices you’ve ever seen. What was that like? Read on below.
Almost anything you could ever want can be found at Hamvention. The flea market, like the attendees, hasn’t really caught on to the digital age yet, so you won’t find bare ATmegas and PICs, or even a few random EPROMs. I couldn’t find a single tube of 74-series logic. If it’s old school you want, or just some interesting things enclosed in glass, Hamvention is for you.
By far the largest single stall at the flea market was Mendelsons. Mendelsons is a Dayton institution and gigantic store filled to the brim with surplus and liquidation inventory. You can get everything from a traffic signal to pallets of laundry detergent, and of course Mendlesons brought out the electronic goods to Hamvention. Just about everything you could want was there, and properly organized, to boot.
Test Equipment and Random Gear
If you need a scope that’s a bit better than what the $400 Rigol can handle, your best bet is Hamvention. Here, everything, from the 1970s Tektronix 465 to the 1980s Tek 2225, to some crazy late-80s, early-90s HP gear. There were a surprising number of spectrum analyzers, and more than enough analog meter to satiate any steampunk aficionado.
The prices? On Friday, they’re about what you would expect. On Saturday, the prices start dropping but with more attendees, competition is increased. Hamvention closes at 1PM on Sunday, and these guys don’t want to carry all this stuff home. That’s where the bargains are, if everything hasn’t been snatched up in the previous two days.
Radios and Paraphernalia
You didn’t think Hamvention would have a ton of radios, right? Everything was there, from beautiful 1920s receivers in cases made of endangered wood, to fantastic plastic 1940s receivers, to ham gear from the 70s.
The Hamvention Takeaway
Hamvention is the premier amateur radio meetup – the biggest in the world – and the people pulling their pickups into a parking space and unloading their shack are the most knowledgeable people on the planet. They have the best stuff, and they’re trying to get rid of it. Need a tube that’s not a 12AX7, 6L6, or EL34? It’s at Hamvention, probably for a fair price.
Hamvention was great, a perfect example of a great swap meet, and a lot of fun, too. If you’re around Dayton, you need to check it out. It’s more than worth driving hundreds of miles and getting a vastly overpriced hotel room. All the best stuff is here, and it’s all for sale.
Sometimes you start building, and the project evolves. Layers upon layers of functionality accrue, accrete, and otherwise just pile up. Or at least we’re guessing that’s what happened with [Varun Kumar]’s sweet “Surveillance Car Controlled by DTMF“.
In case you haven’t ever dug into not-so-ancient telephony, Dual-tone, multi-frequency signalling is what made old touch-tone phones work. DTMF, as you’d guess, encodes data in audio by playing two pitches at once. Eight tones are mapped to sixteen numbers by using a matrix that looks not coincidentally like the old phone keypad (but with an extra column). One pitch corresponds to a column, and one to a row. Figure out which tones are playing, and you’ve decoded the signal.
Anyway, you can get DTMF decoder chips for pennies on eBay, and they make a great remote-control interface for a simple robot, which is presumably how [Varun] got started. And then he decided that he needed a cell phone on the robot to send back video over WiFi, and realized that he could also use the phone as a remote controller. So he downloaded a DTMF-tone-generator app to the phone, which he then controls over VNC. Details on GitHub.
Yeah, that’s right: VNC over WiFi controls an app on the phone that makes two tones out the audio jack, is decoded by a DTMF IC, and the (parallel) binary output is fed into an Arduino that serves as the brains of the bot.
In the end, it looks Rube-Goldbergian, but using the phone’s audio out as a control signal is actually pretty clever. The phone-to-Arduino interface is the tricky part, after all. DTMF is an easy standard to work with and it avoids having to go through the hassle of a USB or Bluetooth device, for instance.
We love DTMF. With sixteen signals, it’s just a lot cooler than the Kansas City tape standard and derivatives, and there’s a whole (by now partially defunct) infrastructure built up around it. Bring on the DTMF hacks!