ME Hax

Super Ultra High Power Telephoto Lens Hack

2011-07-16T23:31:00.000-07:00

Last night was a full moon and it happened to be a very clear night. There were a couple stray clouds meandering across the night sky to which my dad alerted me and suggested I grab my camera. I grabbed my shiny new Nikon D3100 and headed out to grab some shots of the moon, but as I only had the kit 18-55mm Nikkor lens, I was not planning on any close ups. This is what I started expecting:

I make it outside just as the clouds are passing but grabbed a few pics anyway, as it was nice out and still looked cool. My dad mentioned how back in the day, people would go out and shoot entire rolls of film of the moon with it in different locations to use later in enhancing other pictures. That seemed pretty cool, but we discussed the miracles of the GIMP and other things with cut and paste editing. Anyway, the idea of actually looking at the moon came up and I mentioned how awesome it looked through my spotting scope. shortly thereafter I made a trip inside and grabbed my Gander Mountain 12-50x scope and a tripod. At 50x, the moon entirely filled the eyepiece of the scope with great detail.

Since the camera was still outside, I decided to see how well it worked if I just put the camera behind the scope and looked through it. It actually worked fantastically well. Though it was very difficult to hold still and keep the moon in the frame, I managed to grab a couple awesome shots such as the one below

With the great success that that experiment led to, I promptly set about the task of making a more stable lens hack platform. For this platform, I planned on just using a board with some 1/4-20 bolts to hold everything in place and to allow the camera to be positioned up high enough so that it would be on axis with the scope. initially I had only placed a single hole for a bolt for each the scope and camera, but found that I needed to adjust the distance between them more than I had anticipated, so one of the holes became a slot with the help of a 1/4" router bit and the drill press (I was far too impatient to bother setting up the router).

To provide a way to attache the mount to my tripod, I drilled a 3rd hole and counter bored it slightly and pressed a 1/4-20 nut into it. This later proved to not work so well as it was terribly not balanced, so I later adjusted it. When replacing it, I also filed some notches into the bolt and Guerrilla Glued it into place. Hopefully the notches and expanding action of the glue will keep it snug this time.

When it all gets put together, it looks something like this:

and can take pictures like these:

mailbox ~300'

Green thing ~90'

Overall, though the image quality is nowhere near fantastic and does not fill the whole sensor, it is still quite excellent for the price. It does not respond well with small apertures as there is major vignetting, which is improved greatly with wider apertures. There also is a lot of chromatic aberration present, especially if the camera and scope are slightly off axis. I believe that CA is still present when using the scope on ts own as it really isn't designed for this use... In the final pic of the moon above, I removed the red and blue channels and desaturated it, to remove large amounts of CA

But yeah, reasonable very long lens for maybe $2 in parts(or entirely scrap if you have a proper basement)? Sounds great to me =]

These few pics are hosted in 14.2MP at my Picasa
also, check out my Flickr... for some more interesting stuff but limited hosting and such...

Arduino Stroboscope

2011-02-21T00:06:00.000-08:00

I was watching TV a bit ago and playing with my model wind turbine from my fluids class and was having a good time playing with aliasing. From there, the obvious answer to my problem that I didn't really have was to build a stroboscope. This little tangent has resulted in a stroboscope capable of measuring rotational speeds from 577 to 30,000 rpm.

Before I get too far into the code and such, a brief description of a stroboscope is in order. A stroboscope is essentially a pulsed light at a known frequency. when the frequency of a rotating object matches that of the stroboscope, it appears to stand still. It is possible for the object to be moving at a multiple of the strobe frequency as well. for more info check out the interwebs or something.

My Arduino was already set up with 6 LEDs connected to ports 3, 5, 6, 9, 10, and 11 so I used that as a starting place. Since I don't have any drivers or expertise yet in dealing with led/lcd displays, I decided that the serial out would work for taking measurements.first I started trying to use the tone function, however, the 50% duty cycle was not adequate for my needs. The next step was to use a time based funtion that would be independent of the time it took to run the code. This was solved using millis(); however this did not produce adequate results as will be discussed later.

A potentiometer was used to control the strobe rate running off pin A0. all 6 LEDs were used as brighter is better. The basic circuit is about a follows:

What really makes it tick is the code. Its split into 2 sections because if you use only millis(), resolution at high speeds is terrible. I then tried using only micros() but then the lowest frequency was 31hz or about 1800 rpm. I split the code so that for high speed, it uses a version based on mircos() and low speed uses millis(). This split allowed for good resolution to about from 9.7Hz to 300Hz.

In terms of displaying the speed, I used a simple serial output to read the strobe rate as well as the rotational speed in rpm. To avoid delaying the program unnecessarily, the speeds are only reported 5 times per second. I had also noticed that having the serial port set to 9600 baud would cause errors in code execution. In order to handle data transfer, the baud rate was set to 115200. With the higher baud rate, everything seems to run smoothly.

Overall it was a neat little project that could prove to be very useful later on. Here is a little video that kind of shows it working, though that is a bit difficult to convey. It would have been nice to have an accurate stroboscope for some of my other projects, and now that I do essentially have one, I'm sure it will find some more interesting uses.

I have included a video that somewhat shows it working, but for best results, you would actually be using it. That and the video that my camera takes is horrid...

Its pretty cool, It even has a variable delay for how long to leave the LEDs on so it always will produce a sharp image with a constant brightness.I was quite surprised at how well the arduino was handling the math for how sensitive it was.

If anyone has an idea of something else cool to use it for, don't hesitate to mention it =]

And obviously some code =]

**EDIT: This is the newer code.. The new code uses micros()exclusively using an unsigned long for 17-2,500Hz so 1,020rpm to 150,000rpm The older code can be found here**

/*
  Analog input, serial output
 
 Reads an analog input pin, uses this value as a time delay for a stroboscope.
 Also sends frequency info back via serial.
 
 The circuit:
 * potentiometer connected to analog pin 0.
 Center pin of the potentiometer goes to the analog pin.
 side pins of the potentiometer go to +5V and ground
 * LEDs connected from digital pins 3,5,6,9,10,and 11 to ground
 
 created 2/20/2011

 modified 2/21/2011 
 by Steve Krave
 .
 
 */

//sets constants
const int analogInPin = A0;  // Analog input pin that the potentiometer is attached to

// Variables etc...
float sensorValue = 1;       // value read from the pot
float hold = 1;              // hold time between iterations (sort of)
float holdFrequency = 1;
unsigned long microsPrev = 0;       // set up value for timing       
unsigned long microsCurrent = 0;    // used for timing purposes
unsigned long millisSerial =0;      //used for serial output timing
float frequency = 1;         // stores frequency
float rpm = 0;              // rpm

void setup() {
  // initialize serial communications at 115200 bps
  // lower speeds can delay execution
  Serial.begin(115200); 

  //initialize pins as outputs
  pinMode(3, OUTPUT);
  pinMode(5, OUTPUT);
  pinMode(6, OUTPUT);
  pinMode(9, OUTPUT);
  pinMode(10, OUTPUT);
  pinMode(11, OUTPUT);
}

void loop() {
  // this while is a workaround for timer limitations using micros vs millis
  sensorValue = analogRead(analogInPin);   //take potentiometer reading
  hold = sensorValue*55 + 400;             //add some time to scale and for code execution
  microsCurrent = micros();                // collect current time
  if (microsCurrent > microsPrev + hold){  //set up timing loop
    frequency = 1000000/(microsCurrent - microsPrev);    
    microsPrev = microsPrev + hold;            //set up millis for delay stuff

    //set all 6 LEDs high (they were already plugged in)
    digitalWrite(3,HIGH);
    digitalWrite(5,HIGH);
    digitalWrite(6,HIGH);
    digitalWrite(9,HIGH);
    digitalWrite(10,HIGH);
    digitalWrite(11,HIGH);
    delayMicroseconds(.02*hold);  //leave LEDs on long enough to see

    //set all LEDs low
    digitalWrite(3,LOW);
    digitalWrite(5,LOW);
    digitalWrite(6,LOW);
    digitalWrite(9,LOW);
    digitalWrite(10,LOW);
    digitalWrite(11,LOW);

    if(millis() > millisSerial + 200){ 
      millisSerial=millis();    //reset timer
      holdFrequency = 1000000/hold;    
      //do some math to find values
      rpm = frequency * 60;

      // print the results to the serial monitor:
      //      Serial.print("sensor  " );      //print sensor data for debug
      //      Serial.print(sensorValue);
      //Serial.print("holdFreq = " );                       
      //Serial.print(holdFrequency);
      Serial.print("  freq = " );                       
      Serial.print(frequency);      
      Serial.print("  rpm = ");      
      Serial.println(rpm);
    }
  }
}

6 channel serial LED Dimmer w/ Arduino

2011-02-17T14:12:00.000-08:00

While continuing research for senior project we came across the LanBox DMX controller. It essentially takes a hex string and converts it to DMX. Since the LanBox is expensive and we only have one, I wanted to make something that could function similarly for prototyping applications.

What I came up with was using all 6 PWM pins on the Arduino and assigning them intensities through a serial string beginning with "A"

Since allowable intensities varied between 0 and 255 I had the code break down the serial string into chunks of 3 characters. I dont know a whole ton about hex, but it would have been ideal.

So if you send the string "A255000255000122200"
channel intensities would be

The idea is that a whole string of all the instrument (LED) values would be sent in some sort of loop to control dimming operations so that minimal to no code would have to reside on the dmx device improving reliability of the system.

In the meantime, this is the code that I have come up with ideally I would like to be able to send information only on certain channels with some sort of identifier.

The exact code that is currently in use can be found after the break

/*
 this is a chunk of stuff to fade 3 leds from serial read
 
 */

//Callout LED pins
int ledA = 3;
int ledD = 5;
int ledC = 6;
int ledE = 9;
int ledF = 10;
int ledG = 11;

double A_lvl =64;
double D_lvl =64;
double C_lvl =64;
double E_lvl =64;
double F_lvl =64;
double G_lvl =64;
char kind_of_data;

void setup() {
  Serial.begin(115200); 
  analogWrite(ledA, A_lvl);         
  analogWrite(ledD, D_lvl);   
  analogWrite(ledC, C_lvl);
  analogWrite(ledE, E_lvl);         
  analogWrite(ledF, F_lvl);   
  analogWrite(ledG, G_lvl);
}

void loop()  { 
  while(Serial.available() > 0)
  {
    kind_of_data = Serial.read();
    if (kind_of_data == 'A' ) {

      delay(1);
      double A100 = Serial.read()- '0';
      delay(1);
      double A10 = Serial.read()- '0';
      delay(1);
      double A1 = Serial.read()- '0';
      A_lvl=A1+10*A10+A100*100;

      delay(1);
      double D100 = Serial.read()- '0';
      delay(1);
      double D10 = Serial.read()- '0';
      delay(1);
      double D1 = Serial.read()- '0';
      D_lvl=D1+10*D10+D100*100;

      delay(1);
      double C100 = Serial.read()- '0';
      delay(1);
      double C10 = Serial.read()- '0';
      delay(1);
      double C1 = Serial.read()- '0';
      C_lvl=C1+10*C10+C100*100;  

      delay(1);
      double E100 = Serial.read()- '0';
      delay(1);
      double E10 = Serial.read()- '0';
      delay(1);
      double E1 = Serial.read()- '0';
      E_lvl=E1+10*E10+E100*100;

      delay(1);
      double F100 = Serial.read()- '0';
      delay(1);
      double F10 = Serial.read()- '0';
      delay(1);
      double F1 = Serial.read()- '0';
      F_lvl=F1+10*F10+F100*100;

      delay(1);
      double G100 = Serial.read()- '0';
      delay(1);
      double G10 = Serial.read()- '0';
      delay(1);
      double G1 = Serial.read()- '0';
      G_lvl=G1+10*G10+G100*100;  

      analogWrite(ledA, int(A_lvl));         
      analogWrite(ledD, int(D_lvl));   
      analogWrite(ledC, int(C_lvl));
      analogWrite(ledE, int(E_lvl));         
      analogWrite(ledF, int(F_lvl));   
      analogWrite(ledG, int(G_lvl));

      Serial.print("A: ");
      Serial.println(A_lvl);
      Serial.print("D: ");
      Serial.println(D_lvl);
      Serial.print("C: ");
      Serial.println(C_lvl);
      Serial.print("E: ");
      Serial.println(E_lvl);
      Serial.print("F: ");
      Serial.println(F_lvl);
      Serial.print("G: ");
      Serial.println(G_lvl); 
    }
  }
}

DIY temp controlled fermentation chamber from broken fridge.

2011-02-05T17:11:00.000-08:00

(reposted from last year from Homebrewtalk)

So a couple weeks ago, my fridge broke. Since I live in an apartment, the landlord had a new one brought

3 flights of stairs by his crew. I saw an opportunity in this as it took those guys around 20 minutes to bring the new one up. I asked them if I could keep it and they said "sure go ahead." I now had this dead fridge out in my 30º sun room that I needed to get working on. I already had a thermostat in my closet, but the closet has been proving too leaky to keep at a reasonable temp. So it worked out really well.

Here's what I did.

I went to home depot to get a soldering iron so I could start piecing this thing together with my old Honeywell classic mercury thermometer. On the clearance rack, there was a Honeywell non-programmable digital thermostat for $15 from $30. Score!

I proceeded to tear it apart so I could move the thermistor inside the fridge. If I had the option, I didn't want to have to open the fridge to check or change the temp.

Getting the thermistor off was easy, I had a hard time getting it back on as I had no way to remove the old solder so I couldn't get the leads through easily and when they did punch through, they took the solder pads with them. I also need more practice on smaller things, I'm used to soldering larger things like PA speakers and cables

After adding the jumper to the thermistor, I punched some holes in the fridge and ran the cable outside.

My roommate Gigawatts and I had built a control box type thingy with an older thermostat to keep the closet warm while out of town so we just adapted it to this build. It contains a 12V wall adapter, a 12V relay, a male IEC connector (like on your computer power supply) and a standard Edison plug for whatever 120V appliance we wanted to switch. I added an ethernet port to the back of it so it didn't have to be hard wired into my thermostat.

Since this fridge was outside essentially and temps were dropping into the 20's, the batteries weren't holding out so we used one of the spare pairs in the ethernet cable to send 3V to the thermostat so the batteries were no longer needed. The 12V to 3V circuit is in the bottom left of the box, near the blue tape.

For heating the fridge, I had planned on using the defrost element from the evap coil in the freezer, but at 450 watts, it was creating an extremely hot layer at the top of the fridge and I was getting temp swings of 12º as it started mixing down. I am looking for a small 120V fan that I could stick in there to mix things up, but until then, I will be using a 100 watt incandescent lamp. Its still plenty powerful to keep the fridge warm, but I would feel better with something that wont shatter if it gets wet or something. This is what I ended up with. It will probably hold 2 standard ale pails, but the fermenter in there now is quite a bit wider and shorter than my other fermenters.

Closed loop evaporative wort chiller

2011-02-04T21:39:00.000-08:00

A port of my post from Homebrewtalk.com found at http://www.homebrewtalk.com/f51/closed-loop-evaporative-wort-chiller-217094/

Reading the forums, I noticed that a lot of people were feeling somewhat guilty about all the water that they go through chilling their beer. As a good engineer interested in thermodynamics, I decided that I would take a shot at tackling this problem. Since many of the people with the water issues were in warm climates with fairly warm water temperatures tended to be in the southwest, I decided that an evaporative chiller would be the easiest and most effective way to go.

My cooling system consists of the following parts connected in this order
1: High head pump (115' head, 650GPH))
2: Wort chiller (I used an immersion chiller but a counterflow or other isolated chiller would work as well)
3: Spray head (constructed out of PVC, 1/8" hole with flat surface to disperse water, I used a 14 Ga. copper wire)
4: Spray Column (used some circular duct from an old furnace install)
5: Reservoir (An old trash can in my case, I cleaned it, but it doesn't really matter)

The idea is to pump the water through the chiller and disperse it into a fine mist while moving massive amounts of air to evaporate as much water as possible. Each pound of water evaporated removes about 960 BTU from the environment in the form of a temperature drop. With evaporative cooling, it is possible to cool the working fluid down to the dew point so that means you can cool below the outside temperature which can be very significant in dry climates.

I have some more pics and details of the rig following

The rig worked very well, however it should be noted that we were working in 20 degree weather. We were able to cool 5 gallons of beer from 180 to 80 in about 6 minutes which is on par with what I was able to achieve with straight well water. We did have a failure in the system as one of the hoses obtained from the trash essentially burst in use but lasted long enough for some results.

A pic of the spray head

A pic of the system in action during the first test run. notice the ice. During the first test, it brought 4 gallons of water from 60F to a tub of slush in 7 minutes

Some other relevant info:
With the mist nozzle on, flow rates of 4GPM were obtained. this was using a 1/8" orifice.
Estimated airflow would be between 400 and 600CFM without any fans etc...

A couple videos of it in action

Hopefully I can dig up a new hose and some better fittings and improve upon it a bit this spring. I am quite curious as to how it will perform in higher temperatures

Prototype for Arduino control of Midi Dimmer pack

2011-02-04T20:57:00.000-08:00

For my ME Senior Project this year, I have been assigned to work with the Galena Historical Society in Galena, Illinois. My Group has the task of designing three different exhibits for the museum. One exhibit involves the creation of a holographic type projection of Ulysses S. Grant and his wife Julia. The second exhibit is an interactive exhibit designed to highlight the famous Peace in Union painting, which will possibly be using this Arduino control, and a third hypothetical exhibit consisting of a rotating/translating display case. Write ups or more details on the display case and holographic projection may be posted later.

For the illuminated painting, the idea is to have a touchscreen with a digital version of the painting. Guests will be able to click on the individual figures to learn more about everyone contained in the painting. When a character is selected, they will be highlighted on the actual painting on the wall either by the addition or subtraction of lighting. Users will be able to navigate through various layers of details for each individual as they see fit, with the lighting following along.

The team is currently investigating several forms of lighting control, ranging from the standard Arduino to a Phidgets controller or perhaps a LANBox DMX controller. Ideally we will be able to create smooth fades between preset scenes without a huge investment.

Being a group of MEs, we aren't the most familiar with interactive museum exhibits so it has been an interesting process so far. We have a couple prototype interfaces in VB and html/php that we have been working with. At this point, it looks like we are going to got he VB route as it seems a little easier to talk to serial devices.

On the actual hardware control side, we picked up a few Arduinos to play with and try to start making some headway. Currently we have some little circuits rigged up to just trip relays so they stand as a fallback to actual commercial control solutions. As a next step, dimming was the logical option. My roommate Gigawatts had an old Midi controlled dimmer pack that we had been playing with for lighting control in our apartment and the necessary cables and stuff to hook it to an Arduino. I spent most of the night figuring out how to get midi to make stuff happen. I first went through playing with a button to change the light that was on. Although that was fun a moderately useful, I was more concerned about the quality of fades that could be obtained with MIDI as I am not terribly familiar with it as a lighting control interface. The biggest concern was that since we will be driving LEDs with it was that it would look choppy. I ended up sending MIDI values from 0x00 to 0x90 with a 20ms pause between steps. I suppose the resolution should not really have been a concern as 20ms is about one full cycle at 60Hz.

Anyway, This is what I ended up with for a circuit and such:

And this is the code that I have:

/* Midi light control
 
 Basic code to fade a light channel up and down via midi
 Initially used with Leprecon LD-360-MIDI dimmer pack
 
 Designed for:
 Bradley University Mechanical Engineering Senior project
 in cooperation the Galena Historical Society
 
 
 The circuit:
 * digital in 1 connected to MIDI jack pin 5
 * MIDI jack pin 2 connected to ground
 * MIDI jack pin 4 connected to +5V through 220-ohm resistor
 
 By Steve Krave
 Created 2 Feb 2010
 
 find more at http://MEhax.blogspot.com
 
 This code is based on the midi tutorial found on the main Arduino page
 http://www.arduino.cc/en/Tutorial/MIDI
 */

int dim = 1;      // dim controls whether to fade up or down  
int intensity = 0;  // Controls the intensity, a value from 0 to 127
void setup() {
  //  Set MIDI baud rate:
  Serial.begin(31250);
}

void loop() {
  intensity = intensity + dim;
  //Set Intensity Value. Min 0x00, max 0x90
  //int value = 64 * (1 + sin(millis()/500));
  noteOn(0x90, 0x06, intensity);
  //controlls fade rate (ms) per step
  delay(20);
  //sets to fade out when max intensity is reached
  if (intensity >= 0x90) dim = -1;
  //set to fade on when intensity hits 0
  if (intensity <= 0x00) dim = 1;
}

//  plays a MIDI note.  Doesn't check to see that
//  cmd is greater than 127, or that data values are  less than 127:
void noteOn(int cmd, int pitch, int velocity) {
  Serial.print(cmd, BYTE);
  Serial.print(pitch, BYTE);
  Serial.print(velocity, BYTE);
}

For a final result of:

On the video there is a set of regular old incandescent Christmas lights plugged in as well as this is a fairly cheap string of LEDs and will not dim properly without them.

Overall, I am quite pleased with the MIDI solution, though it does eat up the main serial port of the Arduino. Later on if we continue to use this method, it will probably require a software serial port but that not a major issue. It will of course take triggers and run the desired fades instead of a loop or any of that nonsense, but that's fairly trivial overall.

Death Ray (automatic smoke ring cannon)

2010-10-13T20:12:00.000-07:00

While working on senior project the other day, I decided I wanted to build a death ray. I'm not quite sure why exactly, but it seemed like a spectacular idea. I have built smoke ring launcher things in the past but they were always operated by hitting the back membrane or whatever. This time, I thought it would be cool to use a speaker cone as the back diaphragm instead of the bottom of a garbage can or something.

Lucky for me, back in high school I managed to obtain a couple busted old 15" PA drivers and had built a very rough prototype subwoofer enclosure for it. I was feeling adventurous at the time so I built it with a passive radiator. I built it like crap so it sounded about the same so it was sitting in my basement all alone. The hole where the passive radiator went was quite useful as no holes needed to be cut in the enclosure. I took out the driver and flipped it around to expose the full surface area of the cone to the working side.

With the driver mounted, all that was left was to cut a hole in a piece of cardboard on the front and to get some power to the driver. I started off with a hole the size of a masking tape roll cut centered in a square of cardboard. I generated some waveforms in audacity to feed the driver through an old amp I had laying around. I found that using the top half of a sine wave at a high amplitude worked best. Also, having a reasonable amount of space between pulses helped clean up the smoke rings. I finally settled on using a 40hz wave with spacing between 0.2 and 1 second.

After playing around with that a bit, I cut a final orifice in a piece of 1/2" ply the size of a Minwax can. In all it worked fairly well and shit smoke rings about 12'. It probably would shoot them farther, but it was somewhat windy out and there was a fence in the way. I thing it would be awesome to use propane instead of smoke and shoot flaming rings of fire, but it may not be the smartest thing to do.

Here is a quick video of it in action, it sounds funny, but my little camera just couldn't pick it up properly.

While I had the sub and tone generator out, I decided to test for the brown note. Overall, it didn't do much, but I really felt funny afterward

Worlds Easiest Hack

2010-10-04T16:02:00.000-07:00

Last winter, I had a refrigerator die on me, so naturally I tore everything out of it that still worked. This included some relays, a light socket, a heating element, and of course the little switch that turns the light on when the door is open.

About 5 minutes later, it was wired up to the socket that was just obtained with a short chunk of speaker wire. Cool. My roommate Gigawatts and I had some networking stuff in one of our closets with no light so we threw this little device in there with a little bit (really a lot) of hot glue. Here's a quick pic and a short, but awesome video starring Gigawatts' left hand.

Phone Activated Door Hax

2010-10-04T15:40:00.000-07:00

This was a quick hack to buzz myself or guests in without using my swipe card or pressing the buzz button on the panel. The Idea was to use the ringing signal on a phone line to trip a relay, therefore opening my door. Ringing phones were not an issue as there are no land line phones connected in the apartment.

After taking some measurements and talking about the project with the phone tech at my school, I figured out what I was going to do. After some research, I determined that phones typically ring in the 90 to 120 VAC range. I measured mine to be around 200, sometimes over 300, Still not a big deal. The other thing though is that there is always 10VDC sitting on the line for when the phone isn't ringing but you're talking etc... To deal with that, I used a pair of non polar capacitors in parallel, as a single .5µf capacitor wasn't passing enough current to trip the relay.

From here, I connected a bridge rectifier chip, which in turn was connected to my 12V relay. I soldered a 47µf 35V cap across the terminals to smooth out the DC. It may not have been necessary, but I had it laying around and I figured at 25Hz I might as well.

Diagram of the circuit I used

Hodge Podge of parts and hot glue. I love that stuff, plus, I don't have any perf board.

My pretty (yeah right) wiring job

The intercom/door thingy, easy enough to deal with.

Final packaging of the opener. It's crammed in the top of a ranch salad dressing neck with lots of hot glue. I used a chunk of Cat5 to connect to the phone line. My apartment has RJ45 connections so it works fine.

I packaged it all up and stuck it in the wall and its working great. Everything tucked away and connected. I cut a small hole for the cable in the bottom of the plastic so it looks better/involves fewer broken intercoms.