Counting Seconds (Video)

It's not necessarily riveting, but here is a video which shows the tube counting from 0-9 and back again at 1Hz.

I am using a different tube than I intend to use in the future. Mostly that's because it came with a convenient solderable socket. Anyway, this was made possible by the use of a binary counter (SN74F163A) and a Quad-NAND gate (74F00) pulled from an old oscilloscope I tore down. Thanks to some help from a few people over at allaboutcircuits.com!

Sure, the wall outlet signal reduced to 1Hz from my previous post seems to work great, but the added "complication" of having to use a NAND gate with the counter I purchased makes this method a bit cumbersome. I will likely purchase or try to find another counter. Once I work this out, I will post updated schematics.

share on: facebook

Success! Let the counting begin

I have been struggling with a particular issue for several weeks now. Even though I get a seemingly perfect TTL (0-5V) pulse every time the 60Hz AC signal inverts, the IC counters all put out erratic signals which are not a simple divisor of the 60Hz signal. I thought originally that I had fried the IC's, but I tried several types of counters and ended up with the same result. I tried putting in capacitors almost everywhere, hoping that I could solve this by smoothing out the signal. I even set it up so an transistor made a more aesthetically-pleasing square wave (I decided to leave the transistor in).

After being infuriated for ages about this issue, I finally stumbled upon a forum post which noted that the rise of the signal from the 60Hz signal may not happen as smoothly as expected. In other words, small fluctuations could cause the IC to get tripped up and start pulsing faster than it should. Apparently, I was lucky the first time when I set up the initial timing circuit.

The solution: a Schmitt trigger.

http://talkingelectronics.com/projects/OP-AMP/OP-AMP-2.html

It is a positive feedback that pulls the signal up once you hit a particular threshold. The type I chose was an "Inverted Schmitt trigger". These are often used to debounce a switch. I happened to have a quad op-amp IC laying around from another project. After hooking it up with three resistors, I got a perfect 10Hz output from the counter/divider I thought I had fried!

All soldered in:

Here is the new schematic:

Moving onwards!

share on: facebook

Home-Made Enclosure - Main Board

I am waiting for the IC's to arrive from Digikey, so I decided to start working on the enclosure. After finally purchasing a Dewalt router so I can use the 50 new router bits my parents gave me for Christmas, I set out to build the main enclosure out of wood. I believe I have cherry-laminated plywood (correct me if I'm wrong), which I carefully beveled at 45 degree angles on the inner edges. After assembly using glue and finishing nails, I filled in the nail holes and sanded it down. Personally, I am really impressed with myself. This job took about 4 hours, but I bet I could get it down to maybe 2 hours or so if I were to do it again.

The dimensions are 4"x7"x2" and it is open underneath. When the circuitry is complete, I will make a bottom plate for the board to sit on and drill the holes for the wires. I'm still trying to decide what finish the box will get. By the way, my wife has convinced me to make the small nixie tube boxes (where the tubes will actually be located) out of wood as well. I might chronicle the construction of those pieces.

share on: facebook

Lighting a Tube!

I have finally taken some pictures of a Nixie tube lit up using the power-supply I built.

Note the new AC transformer. I will be sure to update the schematic once I get the new IC's.

share on: facebook

Annoyances with Power Conversion

In the past post, I gave the full schematic of the nixie clock, but I did not completely explain the resistor that goes from the "neutral" lead to ground. That was simply because I wasn't sure what it was for!

Turns out the transformer is not a 9VAC transformer, but a 24VAC transformer instead! The resistor is voltage dividing the output to 9V. This wouldn't be a problem, except that when pulling higher currents from the transformer for, say, turning 9VDC into 170VDC, the current prefers to pass through the resistor instead of powering the device!

Basically, I would read no voltage from the transformer when I hooked up the Boost Converter. Arrgg.

Before realizing the purpose of the resistor, I removed the resistor and gave it a shot. Yay, I get 170VDC from the boost converter, but I found out a few days later that I ended up frying the decade counter IC's with 12+ VDC. Whoops!

So, what did I do? First I thought I would wind a new coil out of the existing one. I found this awesome tutorial on how to change the output voltage of the secondary coil by reducing the number of windings. After chiseling away the laminated segments of the core one by one, unwinding the mess of wire for the secondary to count how many turns there were, rewinding a fewer number of turns, and finally jury-rigging back the laminated segments, I got a nasty looking temperamental ugly stepchild of a transformer which jiggled and rattled at 60Hz on my desk like a buzzing bee! Yeah, it outputs 12VAC, but there is no way this thing is going in the final clock unless it ends up being an alarm feature. Yeah, I learned how these transformers are constructed, but in reality, I think I wasted about 3 hours of my time.

In defeat, I went out and bought an appropriate 120VAC to 12VAC transformer from RadioShack for around $3. I will post pictures of the reconditioned device soon.

share on: facebook

AC Signal Conversion Schematic

I finally had a bit of time and I went ahead and finished the schematic for AC signal conversion. The schematic shows how I get a DC signal from 120VAC at 170V, 9V, 5V, and an oscillating 5V signal at a rate of exactly 1Hz. Please let me know if you see any errors. By the way, I read a while back that this method of converting AC to DC is supposed to be inefficient, but I can't say exactly why. If anyone has any suggestions to get a more efficient 9VDC signal, I would really appreciate it.

Next: need to alter the already-made power supply to incorporate this new transformer and circuits.
share on: facebook

Moving to the dark-side: AC-signal conversion

I mentioned in the previous post that I need to count the 60Hz oscillation from the wall AC in order to obtain good timing for my nixie clock. The problem is that I have never messed with AC electronics before, so I had to start with the basics.

I found an old AC to 9VDC wall adapter and tore it apart. I was surprised to find out how simple the design was, although I seem to remember that this method of converting to a DC signal is extremely inefficient. All it contained was a transformer with 120VAC in and 9VAC out. One of the leads on the 9V output was attached to a diode in serial (1N-4004, 1Amp) and the other was attached to a 620 Ohm 1/2 Watt resistor in serial.

This revelation was spectacular! This transformer is completely and utterly unregulated, which means that the positive lead (one with the diode) oscillates somewhere between 9V and 0V at a frequency of 60 Hz with the AC input with respect to the ground lead (one with the resistor). Only with a capacitor added in serial would this signal be smoothed out.

I set out to split the positive lead into two components - one that has a steady and regulated 9VDC output (to be converted to 5VDC) and the other which oscillates between 9V and 0V. For the 9VDC, I added a 330uF (16V) capacitor in parallel to ground. I converted that to 5VDC in the same fashion as I did for the high voltage power supply. For the oscillating signal, I left out the capacitor and divided the voltage, using resistors, down to 5V (I tested this on the 9VDC signal first). I'll put up the schematic of this soon. Now I have a steady 5VDC voltage and an oscillating 5V to 0V signal at a frequency at 60Hz!

The previous image shows the components used for the 5VDC and oscillating 5V to 0V signal on the right (everything to the right of the red diode). I added another capacitor after the MOSFET just to make sure everything was properly smoothed out. The yellow diode to the far right indicates that the device is powered. The rest of the stuff (red diode and to the left) is used for counting.

I used two NTE4017B decade counting circuits which conveniently let you count between 1 and 10. Go to www.alldatasheet.com to find information on this circuit. The first one (right) counts up to 10, which means that a pulse is emitted from the circuit at a frequency of 6 Hz. The second one (left) counts up to 6, which means that a pulse is emitted from the circuit at a frequency of 1 Hz. The red diode flickers at 6 Hz and the left yellow diode flickers at 1 Hz. The two resistors on the far left do nothing (sorry, I forgot to remove them before I took the pictures!)

Again, sorry, I will submit schematics soon. I only put this together yesterday.

Until next time!

share on: facebook

Timing is Everything

BRIEF RANT: What is wrong with Radioshack and Fry's electronics!? I understand that if you are like most of my friends you will roll your eyes at what I'm about to say, but if you are an avid electronic hobbyist, I hope you understand my frustration. The problem is that I can't find any stupid crystal oscillators! As I mentioned briefly in my previous post, they provide very accurate pulses at a high frequency that you can use to provide very precise timing. They are ubiquitous in hobby radio applications (here is why I am appalled at Radioshack) and are necessary to run programmable circuits, which are widely available at Fry's, WHO DOES NOT SELL CRYSTAL OSCILLATORS! So, if you want to get every component at one store EXCEPT the one you desperately need, go to Fry's.

Well, at least I thought I needed a crystal oscillator...

Turns out you can get extremely accurate timing based off of the 60Hz AC signal from your wall outlet. Apparently the power company will oscillate the AC signal at exactly 60 pulses a second every minute, every hour, and every day of the year. It's supposed to be one of the most accurate methods of keeping good time. I have seen a few examples of circuits using this method of timing on Hackaday.com, but I felt they left out a few important points. Hopefully the next few posts will clear this up. The only issue is that my nixie power supply runs off of 9VDC. So I have decided to put a transformer directly onto the board, meaning I will have direct access to the oscillating signal. Unfortunately, I have never-before tinkered with an AC signal.

See you next time!

share on: facebook

Nixie Tube Power Supply (Part 2)

Sorry about the lack of posts lately. I have been crazy busy with work, not to mention the three vacations I have had back to back (believe it or not, that can be a bit tiring--but don't get me wrong, I would do it all over again). Anyways, enough about me, I have pictures of the power supply!

That large black cylinder is the inductor (in case you were wondering). Look at the previous post for a schematic. I got almost every single part from Digikey. Aside from the inductor, I received enough parts to make almost five of these power supplies with a final cost of around $40-$50. Not too bad if you ask me. Buying in bulk obviously reduces the prices even further. Not shown is the 9VDC power supply, which I am currently borrowing from my other nixie clock.

I thought originally that I had purchased a high-voltage 2.2uF capacitor (C4), but it turns out I missed it when I did the ordering. The large gray rectangular box is a 0.68uF capacitor rated at 310V. I yanked it from a broken computer power supply I had sitting around. I would have been pretty angry if I hadn't found it... crisis averted! Turns out the 0.68uF capacitor works just fine instead of the 2.2uF cap.

The back-side doesn't look too pretty, but it's a project board, what do you expect?? I'll upgrade to PCB (printed circuit board) eventually!

So, success! I get an output of 170VDC (adjustable by adjusting the potentiometer--the blue rectangular box) out of the red wire (red means DANGER!). I also get a convenient 5VDC output I can use for the integrated circuits out of the yellow wire. Black is common for both outputs. I'll see if I can't eventually post a picture of the nixie tube hooked up. I know this works from experience, but it would be nice to show it on the blog.

Next, I need to figure out a way to get the timing on the clock right. This has been bothering me for a bit since I would like to avoid using PICs (programmable integrated circuits), but the only reliable way to get a timed pulse that will be accurate over months or years with a DC input is to use a crystal oscillator. These puppies put out thousands to millions of pulses per second... bringing that down to 1 pulse per second will take a large number of divider circuits. Ok, I'll get back to you on this! 'Till next time!

share on: facebook

Nixie Tube Power Supply

Sphere's Nixie Tube Page specifies that the IN-14 tubes require a 170-180VDC power supply and a 33kOhm resistor in series (before the anode). Most AC to DC power supplies provide between 3.3VDC to 15-20VDC. I also want to have 5VDC for logic circuits. For the first nixie clock (Nixie 1.0), I purchased a boost converter kit, something like this one. For me, it was a "black box" when I soldered the components on, but I knew I needed to figure out a bit more about how it works if I intended on making my own from scratch.

I manually copied the circuit diagram into Eagle Cadsoft (you can get a limited version for free which should work great for this project from here). Below is a picture of the circuit diagram I drew up. If you need help with using Eagle (...and you probably will because it is not intuitive), take a look at some youtube videos. There are comprehensive tutorials out there... that's the way I learned how to use it.

So, this schematic looks daunting, and to tell you the truth, I'm not completely certain what every part does, but I will try my best to explain what I do know. Some of my information came from this helpful website.

I have +9VDC coming in up top. The C1 capacitor is there to smooth out the DC input in case there is any ripple. This may not be necessary, but it can't hurt. A point of warning, make sure that all of the capacitors are rated for the appropriate voltage.

The inductor (L1) is the business end of things and is hooked up at one end directly to the 9VDC and the other to a miriad of other components. The purpose of the other components is to switch on and off the flow of current in order to maintain a high potential (160-200VDC). After the diode (D1), we have 160-200VDC. So what are all of the other components?

The C3 and C4 capacitors are high-voltage capacitors which are there to buffer the output voltage. The integrated circuit (IC) is a 555 timer. The 555 timer is there to rapidly switch the transistor (Q1) on and off. When the transistor is closed (passes current), the inducted current goes to ground. Once switched open, the current is forced to pass through the diode. If this process is repeated at a rate which is faster than the decay time of the current running through the inductor, you get a potential build-up passed the diode. I will talk a bit more about how the 555 timer works in general later, but right now you have to believe me that many of the remaining components are just there to ensure the switching is rapid.

Lastly, the voltage has to be regulated, or the voltage will keep rising to no end. This is done using the resistors R3, R4, and R5. R4 is adjustable, so you can adjust the final voltage. This resistor series acts as a voltage divider, allowing the transistor (T1) to begin conducting once we reach 160-200V. Once T1 conducts, the voltage is dropped, and increases the frequency of the 555 timer cycle, providing active feedback.

Now that I know how this works, I can begin soldering! See you next time.

share on: facebook

The Plan

Ok, so Nixie 2.0 needs to be different from all of the other nixie clocks out there. I'm tired of seeing the same thing all the time. People make neat boxes and enclosures, but they sometimes lack that coolness factor.

After thinking about it for a while, I thought it would be neat to have modular units for the tens hours, hours, tens minutes, and minutes. I admit, I'm getting this idea from a ThinkGeek product, but I don't think the final product will look anything like their Matrix Cube clock. I made a simple sketch, which is shown below:

Yeah, it's crude, but I hope you get the general idea. There will be a base which contains all of the business-end things--the power supply, timer circuits, and buttons. Each of the nixie tubes will be contained in a very small box which will be wired to the base. All of the boxes for each of the tubes will be able to nest into a depression in the base, or they can be moved around at will. I wonder if I need to put something heavy at the bottom to make sure they don't tip over. I have thought about maybe even using a monitor cable as the method of supplying the wires to the tubes, allowing me to unplug each nixie tube box if I want to exchange it, or it may allow me to put the nixie tubes in confined areas, hiding the base.

The point here is that I want the base to be as skinny as possible, so I will try to compact all of the components into as small of an area as possible. I really have no idea about the location of the buttons, the style and material of the box, or how I might add the neon lamps. All I know now is that I need to make as small of a power supply as possible, on the cheap, from scratch.

share on: facebook

Nixie Tubes!

Nixie Clock 2.0 has some pretty neat inspiration -- a plethora of nixie tubes. My older brother gave me a bunch of Russian IN-14 tubes. Take a look at this website, they have an insane number of tubes for sale. He got them on ebay still attached to a display rack:

Of course, I have already taken a few of them off of the rack (desoldering these was quite a chore). You can see on the far right that they come with a few neon lamps, which might be good for the colon between hours and minutes? All of these items take 170VDC and each of the nixie tubes has an anode and 12 cathodes (one for each of the digits, 0-9, and the last two are for decimals before or after the digits).

So, I think you can probably already get an idea of the issues associated with making a clock out of these tubes. They take very high voltage (although very low current) and require switching between each cathode in order to display any one of the digits. I also need 5V for switching using integrated circuits. I would like all of this to be in a compact design. Speaking of design... next post. See you then!

share on: facebook

Nixie Clock 2.0

So, it's time to make a nixie clock! To those who don't know what nixie tubes are, take a look at the wikipedia page: http://en.wikipedia.org/wiki/Nixie_tube. My older brother gave me a bunch of these over the past year. They come in a number of types and sizes, but they all have a similar orange-red glow to them. They were used as an early display technology, but now people mostly use them for making retro displays such as clocks.

I have already made a nixie clock. A picture of the clock is shown below (thanks Lance for taking the B&W film picture of me). I learned a great deal about electronics from the project, but I used an Arduino (user-friendly programmable input/output device) for the project and purchased a kit for the power source. It does have a lot of features, like two alarms, temperature reading (shown below in degrees celsius), 12-hour/24-hour format, sleep setting (where it turns the tubes off at night), and an auxiliary input which does nothing right now. Still, I wanted to do everything from scratch and not rely on convenient technology.

So, this blog will now be dedicated to updating the progress of making the nixie tube clock completely from scratch, but without the fancy extras. I will start with a 9VDC power supply and make all of the components from either available or purchased resistors, integrated circuits, capacitors, transistors, etc. The next post will explain what my design will be and I will follow with updates on everything from the power supply to controlling the tubes and constructing the actual clock.

I might also show how I made the first nixie clock (Nixie 1.0). I'm pretty proud of it.

Enjoy!

share on: facebook