The Ultimate Alarm Clock.

I absolutely hate being late for things.  I find it to both be rude to people who may be waiting for me, as well as in the case of my job it is just plain disrespectful and unprofessional.  Sure, things happen that are outside of mycontrol, weather, traffic, etc.  I guess I have to live with that, but if it is correctable I would like to engineer the problem out. 

One of the reasons I am sometimes late is because on rare occasions I oversleep.  Waking up in the morning is the bane of my existence.  If it were up to my circadian rhythm I would go to bed at 3-4AM and wake at 12-1pm.  I simply just function better nocturnally.  On work nights I do try to get to sleep at a respectable time, so one of the things that occassionally fails me is my alarm clock.

When it comes to alarm clocks I realize that either or both of the following is true... A)  There really is no perfect alarm clock on the market and/or B) I am entirely way to picky and obsessive when it comes to alarm clocks.  Over the years I have tried several different alarm clocks and have found fault with all of them.  I am currently using the RCA RP5430 (pictured)

The problem with this current unit is I think RF noise from my cell-phone (which also resides on my headboard at night) is injecting interferring noise into the clock causing adverse effects.  I have actually seen the clock act like it is possessed (with the numbers on it flickering about randomly).  This has resulted in the time becoming changed and/or the alarm times to change, or the alarms to be disabled etc. 

It is time that I set out to build the ultimate alarm clock  This project is longer term as I have other projects to complete before hand, however this post should serve as the seed for this idea.  Below are the list of features that I would like to incorporate.  Feedback is always welcome.

  • Snooze button on top
  • A "hard snooze deadline"  a time which the system will no longer let you snooze and the only option is to wake up and turn off the alarm(see below about turning off)
  • A "cognitive puzzle" should be required to turn off the alarm, this could be entering a random number that would displayed on the display into a keypad.  This would require the user to wake up enough to do this instead of fumbling around for the button half asleep and disabling the alarm.
  • "intelligent snooze"  I like to snooze for at least 30minutes prior to waking up.  I would like a voice announcement to say something like "You have t-minus 20 minutes until wake up"
  • When the deadline alarm goes off (the time which you can no longer afford to sleep otherwise you will be late) the alarm needs to be annoying enough as to indicate the severity of the situation, yet not so annoying as to cause me to want to throw the clock (this might be difficult)
  • No Radio, I don't use sleep radios or wake radios, so I feel no need to implement this
  • Shielded from RF Interference.  I'll line the damn thing with lead if that's what it takes!! (don't email me about lead, I am not literally going to use lead, rather proper ground shielding)  I am thinking I will start with the concept of a faraday cage.

These are just a few of my thoughts, I will post more as I think of them.  I think this is very much doable.  The alarm noises could be implemented through an arduino .wav or .mp3 shield.



Advanced Servo Mixer powered by Arduino.

I have recently began playing around with RC planes and helicopters and like most things I delve into I find limitations on things I want to do but I am unable because the products offered for the masses are limited.  This is typically because they tend to sacrifice complexity for ease of use.  In this case I require a servo mixer that can do conditional mixing.  The following video might illustrate my needs a little more clearly.

Currently I have been able to find 2 types of commercial servo mixers on the market.  The first I'll call a "dumb" type which simply takes 2 inputs and mixes it to 2 outputs.  This isn't to say the product is bad, it definitly works well for what it does, I am just calling it "dumb" to say it has no configurability.  The second type is made by Desert RC.  This mixer is a little better in that it is programable and can be configured for custom mixes, but you are limited to using their configuration utility and there are limits.  While the desert RC mixer is a fantasic product it will not work for my needs.

I have decided to try and build my own servo mixer that will be based on Arduino.  The type of mixer I am making would never work for the masses because to configure it involves writing code.  By writing code, I will have total control over how the mixer process input commands and passes them to output commands.  In my case I am sacrificing ease of use for complexity.

Here is a video overview of the current draft of the schematic I am working on. (Best to watch in fullscreen 1080p in order to see all the schematic symbols)

As always feedback is welcome.


Circuit Board Business Cards

I had this neat idea to promote my blog site (that you are reading right now) as well as demonstrate some of the capabilities I have mastered to date.  I can't say the idea is all original, I was inspired by reading the instructables article here by an author known simply as 'Sponges'  located here

I guess could have fabricated it out of my house the way he did, but I have really taken a liking to eagle cad and batchPCB, and my design was going to be a bit more complex.  My project would have more LEDs as well as a more advanced microchip.  Unfortunately this has driven my cost up a little higher then his, but for the features I wanted to include, this was acceptable.  Also, I am also not yet familiar with the PIC family of microprocessors.  The purpose of this project was to hone my SMT soldering skills.  The first board I assembled wasn't too bad, but by my 3rd, I think i now have the technique pretty much mastered.  Here are some pictures of the boards in various phases of assembly.

Here is a video demo of one of these in action.

These boards were part of a small test run to mainly ensure I didn't have any design flaws. I have amended some of the text on the card, mainly changed the url to since I now have that domain. I am going to commission Gold Phoenix PCB Fab house to make up some of these with blue soldermask and white silkscreen. Also, stay tuned, in a few months I am going to unveil an alternate version of these with slightly different electronics on it. These will be a "Geek/Hacker" variant. I just finalized the design last night, I ordered a small run for test, and if they test out, I'll order a production run.

I need to clean up my source code a bit, but If there is any interest, I will publish all the source material time permitting.


Arduino based Large Digital Clock with 7 segment display

I've been working on a prototype for a large clock I want to make.  This clock will feature 6.5inch digits for the Hours, Minutes & Seconds and will feature 4inch tall displays for the Month & Day readout.  Being as I am going to hang this clock up at work, It will likely be high up on the wall and out of reach.  So when I started thinking about how to program this clock, I thought wireless would be the perfect solution.  With the xbee radios from digi being so cheap and easy to implement this seemed like the perfect fit.  I then started thinking about how else I could pimp this thing out and I thought it would be nice to have a remote viewer on the "clock setting device".  This screen will display real time information being fed from the clock on the wall as well as have intuitive menus for setting it.

To realize all of this,  I wanted to proof out as much of my code as possible on the small scale before implementing it on the final design.  Besides,  I need to wait for some of the circuit boards I designed for this to come in from China (about 4-6 weeks).

Here are a few pics of the small scale prototype, work in progress.



Here is a youtube video showing an overview of the project...

Feb282011 Submission - Megaphone "Portable Dog Killer"

Disclaimer:  First, before everyone goes writing me about how vile of a person I am because I made a device that kills helpless animals, please read the entire post.  This device does not actually kill or harm dogs or any other living creature in any way shape or form.


Recently I had a problem in the neighborhood whereby a couple of "aberrant K9's" were getting lose because the irresponsible owner would not repair the hole in his fence that was allowing them to escape.  The problem got so bad one day when returning from the grocery store, I found myself trapped in my car under seige by these animals.  Every time I tried to exit my car they would approach me aggressively. Not one to test fate the police were called, but by the time they responded, the animals were no where to be found.  It was not uncommon for these animals to get loose several times a week.

One day I was listing to the Security Now! podcast and Steve Gibson was describing a story from when he was a kid when he constructed what he called a "Portable Dog Killer" (here is a link to the mp3 of that podcast). While he didn't give specifics of the circuit, he did mention that it contained a 555 timer and a piezoelectric speaker.  Using my imagination I could piece together what he did.  His results with his homemade gadget were spectacular.  I began to think...   What if I took that concept and fed it in to a megaphone.  Megaphones are perfect... They are designed acoustically to be loud.  I quickly ordered up the biggest, loudest megaphone I could find (without breaking the bank).

Now I should say I am an amateur with electronics.  I am not an engineer or an expert by any means.  I can copy a formula for frequency and plug in my own values (such as the formula for freq), but beyond that I am lost.  So my first pass at bread-boarding failed because the output of the 555 could not drive the cone directly.  I had read that mosfets can be useful for driving high current devices so I wired one in... Success.  This is still a work in progress.  The name "Portable Dog Killer" is a a tribute to Steve Gibson.  On a quick side note, please check out his site at I highly recommend listening to the podcast linked above,  it is supremely entertaining and inspiring for any aspiring maker.



Video of me testing out the breadboarded  circuit

(ATTENTION JUDGES This video is of an early prototype version posted here to show my progress with this.  Please see the 2nd video below for the final version of the contest submission)


Circuit Boards arrived from BatchPCB, They are great... I order 2 they send me 3!!  They do this all the time, they sometimes need to fill up a panel, so they send you the extras free of charge.

What a value!


Here I am exploring the limits of warranties hours after ownership on this new Megaphone....

Assembled board in the Megaphone.

Close up of a knob I installed to vary the output frequency (from 8HZ to 22KHZ)

Here is a picture of the finished unit (looks basically like a generic megaphone)

Here is a video of me bench demoing the unit (sorry, no video of me testing this on actual animals)

***ATTENTION JUDGES*** The video below is the current version and represents the contest entry.

Screen-shots of The Spectrum Analyzer.  The First is Frequency turned down to about 8k, the 2nd, turned up to 19k.


Fortunately Animal Control dealt with the Dogs problem before I had to use this, so my development on this has been put on hold for now.  But should the hounds of hell come knocking on my door, Ill be ready.

What does the future hold?

Well for one I would like to make it louder.  I will probably reinstall the original megaphones circuit via a switch. In addition to that, I had worked out an arduino based frequency counter (picture below).  I had thought of doing a digital read out of the frequency.  I suppose if I do, I could have the arduino generate the tone as well,  but what is the fun in that?  The 555 is so perfect at what it does, it has an infinite number of uses and I am glad I had the opportunity to share this one.  I look forward to learning a lot more of what it can do via the submissions to this contest.  Thank you for your time.

Picture of Arduino based Freq counter test using an earlier version of the board (pre-mosfet)

If you want to reach me you can feel free to email me at tminnick at gmail or you can find me on twitter here @TomMinnick