Jump to content
Sign in to follow this  
Bevanj

How to make a cheap, power efficient LED trigger

Recommended Posts

Posted (edited)

Hi, 

As some of you are aware, I was making LED trigger boards. Sadly, I've been super busy with my day job, and haven't really had much time to make them any more. Sorry to those that may have tried to contact me in regard to them. 

In true Wetpixel form, here is the basics of how I have been making them, and how they work. The core of the circuit involves some simple logic gates, a RC timer circuit, and a NPN transistor. Plus resistors, caps, and a diode. 

This design, provided the right components are picked, will work for years, and take tens of thousands of photos off a single set of button cell batteries. 

There are plenty of schematics out there that simply hook a hotshoe output to a transistor to drive the low side of the LEDs. However, there are some serious downsides to doing that. If you pick a shutter speed of say 1/200th of a second, the LEDs will light for 1/200th of a second - let alone if you pick an even longer shutter speed. This is incredibly wasteful of valuable battery energy. Far far shorter LED pulses are enough to trigger a fibre optic based strobe, and set it off, making for far far more efficient use of the small amount of energy in button cell batteries.  

Into the nitty-gritty. If you look at the drawing, there are two types of logic gates used.

One is a schmitt inverter - these simply invert a signal. If the input is high, the output will be low. If the input is low, the output will be high. To note though - is the schmitt trigger on the input. This causes the detected input threshold from low to high to take place at quite a specific voltage, and the transition from high to low to take place also at quite a specific voltage. 

The other is an AND gate - if both inputs to this gate are high, the output will be high. In all other input scenarios (either is high, or both are low), the output will be low. 

Some basics...
D1 provides input polarity protection for if some numpty puts the batteries in back to front. 
C1 and C2 provide some input voltage stability. 

After the hotshoe, you have R3. This is a pull down resistor. It simply makes sure that the input to R4 is pulled to ground if the hotshoe contacts are open. 
R4 simply provides a bit of resistance to the input to Schmitt inverter #1. It is probably not needed, given the incredibly high input impedance of the Schmitt inverter. 
Should the hotshoe be open, the input to SI#1 will be low, and due to how the gate works, SI#1 output (labeled as Point#1) will be low. The opposite is true. Should the hotshoe close, SI#1 input will be high, and SI#1 output (Point 1) will be low. 

R5 and C3 form the heart of how this circuit works. They form what is known as an "RC Timer."  If the output from SI#1 goes high, it takes a while for C3 to charge through the resistance of R5. Likewise, if SI#1 output goes low, it will take a while for C3 to discharge through R5 to ground through SI#1. Thus, there is time lag between SI#1 output and what is happening at C3, and in turn the input of SI#3. This time lag is defined by the choice of values of C3 and R5, and can be finely tuned. 

SI#3 and SI#4 were used simply to get a non-inverting output output off the RC timer circuit. Two inverters = a non-inverter. But, in this case... with a schmitt trigger. I went this way, because I already needed to order schmitt inverters for SI#1. If you want to use a non-inverting schmitt buffer in place of SI#3, and SI#4, that will do the exact same thing. 

Sooo... actual operation. 
If the hotshoe is open, SI#1 output will be high. Provided enough time since the last shot, C3 will be fully charged. SI#3 and SI#4 will be providing HIGH to one side of the AND gate. 

Bam! A photo is taken. The hotshoe contacts close. SI#1 input goes high. S#1 output goes low. At the same time, the RC timer starts discharging, and SI#2 input goes low. SI#2 output goes high, and VOLIA, both inputs to the AND gate are high! The NPN transistor turns on, and the LED turns on! Once this has been in this state for long enough, C3 will have discharged enough for SI#3 input to go low, and in turn through SI#4, for the AND gate lower input to go low. This means LEDs will turn off, even before the hotshoe opens, and after an amount of time defined by the selecton of R5 and C3!

When the hotshoe opens again, if you chase the logic, the upper side of the AND gate will instantly go low. The RC filter will take a while to charge (plenty before a camera is ready for another shutter actuation). The lower output to the AND gate will be HIGH, and we are ready to take another photo! 

How cool is that? The flash time of the LEDs can be fine tuned down to microseconds simply by the choice of resistance of R5, and the capacitance of C3.   

Provided you pick the right logic gates that have incredibly low quiescent current, and are 6V tolerant, along with the right type of capacitors that have low leakage, this circuit can sit dormant on a set of batteries for YEARS with no use! 

Hopefully this is useful to somebody! 

Happy photo taking, 
Bevan 

For those interested: 
https://en.wikipedia.org/wiki/Schmitt_trigger
https://en.wikipedia.org/wiki/RC_time_constant

EDIT: I'll scan the drawing when I have access to a scanner. 

20210807_203542.jpg

Edited by Bevanj
  • Like 1
  • Thanks 1

Share this post


Link to post
Share on other sites
Posted (edited)

The AND gates I was using are NC7S08M5X
The SI gates I was using are NC7S14M5X 

I would recommend X7R ceramic capacitors. They have very low leakage. 

I'll have a crack at finding some through-hole equivalents, and putting the design together for a small-ish board that people can order off Oshpark, or JLCpcb, and solder at home with basic soldering skills. 

Bevan 

Edited by Bevanj
  • Like 1

Share this post


Link to post
Share on other sites

It is very kind of you to share this on Wetpixel - much appreciated.

Share this post


Link to post
Share on other sites
5 hours ago, Bevanj said:

The AND gates I was using are NC7S08M5X
The SI gates I was using are NC7S14M5X 

I would recommend X7R ceramic capacitors. They have very low leakage. 

I'll have a crack at finding some through-hole equivalents, and putting the design together for a small-ish board that people can order off Oshpark, or JLCpcb, and solder at home with basic soldering skills. 

Bevan 

What is the recommended LED to use, how important is this component selection?

Share this post


Link to post
Share on other sites
Posted (edited)

I was using Cree C503C-WAS-CBBDB231 LEDs.

I think it is of far more importance is getting good quality fibres which have low refractive losses. 


The TC74HC132AP is very appealing. It has low quiescent current, and a NAND can be configured as an inverter, or an AND gate. This whole thing can be made using two. 16pin DIP versions. Too bad there's not a 6ch version of this IC rather than a 4ch version - if there was, it could all be done with one IC, some resistors, and capacitors. 

 

Edited by Bevanj

Share this post


Link to post
Share on other sites

Chris, Elias put me on to these to use as replacements for a Subal LED trigger that had failed:

660 nm, 5 mm, from  LEDSupply.com. 

They work perfectly for the UW-Technics TTL/HSS initiator

 

Share this post


Link to post
Share on other sites

Hi Bevan:  When you fill in the details, would you please, for us electronic beginners, provide a parts list?

Diode D1
Resistors R1 thru R5
Capacitors X7R type:  C1 thru C3
Schmitt non-inverting or other  SI 1 thru 4, (S1 NC714M5X, ?2, 3, 4)
AND gate (? NC7S08MX5)
battery type and number.  

Many thank you's

Share this post


Link to post
Share on other sites
Sign in to follow this  

Sponsors

Advertisements



×
×
  • Create New...