There are a few posts on this site that make use of a sound activated flash trigger. The following is a (hopefully clear) description of the final version of the trigger used and a circuit diagram for anyone wishing to construct their own. So bear with me while I fumble out some words…
The initial version of this trigger was a very simple single transistor amplifier, a piezo drum and an SCR. It performed well for quite a while but the need to drive the flashes directly (rather than from eBay radio triggers) and to have a system of delay and debounce led me to constructing a Mark II. I also did a more permanent job of the circuit by vero-boarding and casing it.
A few terms:
Delay: The time between a sufficient sound being received and the flash sync signal being sent.
Debounce: The dead period after the flashes fire during which no further signals are allowed pass. Prevents unwanted double exposures.
Gain: The amount amplification applied to the microphone signal, aka sensitivity.
There are a few circuits like this out there on the net but overall I found the quality of the designs quite poor – some almost like they had never actually been built. So I aggregated a few of the designs, added some of my own stuff and came up with this. Rest assured that if you build this correctly it will work as I have a completed one myself. It is a little over engineered and several components could possibly be discarded but I desiged it with best practices in mind and following all the guidelines in the ICs datasheets.
In the BOM I have indicated Jaycar part numbers for the various non R/C components. However the parts are quite common and I am sure they all can be sourced from any decent supplier or from Digikey/Farnell.
Pictures of my own implementation follow to give an idea of how you might case the circuit. The case I used is a simple Jiffy type box which has the advantage of a battery compartment for the 9V cell required to power the circuit. Makes for a nice clean unit. The pots are miniature types and linear. Finer tuning in the low end could be achieved if they were logarithmic – I guess it depends on your requirements as to weather that would be useful.
This is the circuit on veroboard inside the case. The C5/R2 Zorbel network are the only components not visible as they are on the back of the board having being added as an afterthought. The wires visible run to the pots, switches and connectors. The red tape insulates the circuit from the switch terminals which could potentially touch the board due to their depth.
For input I just stuck my test mic into an old solder tube as shown below but any electret will work. Most cheap PC microphones are of this type.
And finally all squared away and labelled. For output you can either get some cheap eBay 3.5mm to PC Sync cables (or make your own) or run one of the outputs into a radio trigger transmitter. Though the later presents some problems for high speed photography – see post and comments here. However the flashes are connected they must be used in their LOWEST power setting. This ensures the minimum flash period which is essential for high speed. Both Canon and Nikon flashes are capable of producing approx 1/30000s bursts on their lowest setting.
If you are going to have a shot at building it I would recommend initially doing it on a breadboard. Get the amp working first – you can substitute the R3 resistor with a speaker and see if the mic is being reasonably amplified. Keep in mind this is not a high end audio amplifier! As long a there is no oscillating/reverberating or high levels of distortion you are probably good to go. I used a old PC speaker to test with.
After you get the amp going reliably (which has the most potential to cause trouble being analog) proceed to the delay/debounce which is mostly logic. When building the amp try to keep it as compact as possible – long tracks and cables are never an amps best friend.
Have fun and feel free to ask any questions you might have.