Note: any camera that accepts a TTL LOW signal to capture an image will work but you will need to figure out the connection on your own.
Step 1: First Things First
-Skills: (cause chicks dig guys with skills)
2.5mm plug (a wired cell phone headset is your best source)
Case for the circuitry, Altoids tin is almost too small
CD4093 -or- CD40106
100k pot, preferably log (audio) taper
2x 100k resistor
SPST switch for power
SPDT switch for sec/min (i used a single on-off-on DPDT for power and range)
9V Battery Clip
All component values are non-critical. I built this device using only items in my shop/ junk bin at the time. The plug can be a Tip-Sleeve or a Tip-Ring-Sleeve (2 or 3 contacts) this will only alter your build slightly.
A 1M linear pot and a 120k resistor will give you more accuracy, but you will have to turn it backwards. Check the last step for more info.
Step 2: Schematic
-Power connections are not drawn. Connect positive power to pin 14 and ground to pin 7 of all chips
-I highly recommend building this on a breadboard before you commit it to solder.
-Any supply voltage from 3 to 15 volts should work, but might mess up your timing. A well regulated supply is recommended if you don't use a battery.
-You can use any CMOS inverting Schmitt trigger. I used the 4093 because I had it. Don’t forget to tie any unused inputs to the supply or ground, CMOS chips can oscillate and cause problems when the inputs float.
-A log (audio) taper pot will give you a more accurate interval setting at the lower end and will save you a headache when you are trying to set a time between 1 and 10 seconds. Looking at the front of the pot, use the two leftmost lugs for your connection. Again, check to last step for an alternative.
-Some capacitance across the supply is a good idea, especially if you are going to run your circuit off of non-battery power, a .1uF film or ceramic in parallel with any larger electrolytic is probably overkill, but will work.
-Schmitt trigger oscillators are not very accurate. Supply voltage will affect the frequency. My 100k pot was actually 96k and my 330uF cap was 3 paralleled 100uF caps and my timing came out to almost exactly 1-60 sec/min. If you are worried about accurate timing, play with your values and a stopwatch to get it right.
-If your circuit seems to be working, but won’t trigger your camera, try increasing the size of the .22uF cap. This should increase the trigger time, allowing your camera enough time to realize it needs to take a picture.
-You can use LEDS hooked up to the outputs of the 4024 through 1k resistors to visually confirm that your circuit is working. Your circuit should trigger in time with Q0 on the seconds setting and once the bits “walk” out to Qs 3, 4, 5 and 6 on the minutes setting.
Step 3: Construction
I elected to use an RCA plug on the other end of my cable, allowing the use of a manual switch (this is a great way to minimize vibration in a long range shot with a slow shutter speed) or the intervalometer or maybe a shot triggered by sound or light in a future circuit.
Use whatever construction method you prefer. If you are not worried about ugliness, you can take a page out of my book and go all Rambo with the hot glue gun.
For me, an Altoids tin (spearmint) just barely fit all the guts of the circuit with room for a 9v battery. Your mileage may vary.
As soon as I got the guts in the case, my circuit started tweaking out. I had to add a cap in parallel with C2 to make it trigger reliably. The schematic reflects this change. Also, the last quarter of my pot's rotation was shorted out, limiting my circuit to 40 sec/min. I suspect this is because it is a cheap log taper. I added an extra cap in parallel with C1 to try and compensate.
Step 4: Use
After that, simply hook it up to your camera when you feel like making a time lapse.
You shouldn't have to worry about battery life, my circuit draws 3.5 mA peak, but I think it's close to 500uA average draw. Your camera will probably die before your intervalometer.
The first interval after power-on will be longer than those that follow, this is a limitation of relaxation oscillators and is more trouble than it’s worth to fix. This will be most noticeable when you turn on the circuit while it is set to 60 seconds, it shouldn’t be a problem most of the time, though.
The first cycle after you change your timing may be faster than you expect. For example, if you are set on one minute and turn the dial to 60, 30 seconds after it last triggered, it will trigger about 30 minutes after you turn the dial. The next time lapse will be 60 minutes.
Step 5: Possible Modifications
An easy change is to add a momentary N.O. switch between the input of U1D and ground or the output of U1D and the positive supply. This would let you trigger a capture at any time. It would also reset any binary counting that has occurred, interacting strangely with timing on the seconds setting and effectively restarting the interval on the minutes setting.
The mod I would most recommend is replacing the pot with a 1M linear pot in parallel with a 120k resistor. In this arrangement you would want to use the two rightmost lugs (looking at the front of the pot) and operate it backwards from normal operation. Connecting to the leftmost lugs would be like using a reverse-log pot and is the opposite of what you want, but test your timing before doing anything permanent, just to make sure. For more info on custom controlling the taper of a pot check out the Elliot Sound Products site.
For the adventurous out there I have added a schematic that adds another 2024 for dividing the minutes signal by 60 to extend your timing to 1-60 hours! You will need at least a DP3T switch, probably easiest to find in a rotary form. This circuit is unbuilt and untested. You may need to tweak C3 to get it to fire properly. Running off of AC power is definitely recommended here. Let me know how this works if you decide to build it.