The Strobe Light

The page is about all the my attempts, failures, and success of trying to develop my own strobe light circuit. One of the biggest obstacles was the power source as I wanted to power it from a batteries instead of a wall outlet. The various sections describe the different circuits which were attempted.

Throught the process of making this strobe circuit several different circuits were found and considered. I mixed parts of them to develop a final working circuit. Here are some of the circuits and why they were or weren't choosen.

This is the original circuit found in the book of circuits. It originally caught my interest because of the idea of powering a strobe off of such a low votlage. I was interested in learning about the voltage conversion and how the strobe light itself worked. However, radioshack no longer sold the coil or the transformer. This posed some big problems. I was able to find the coil online through someone who happened to have extras that he was selling for a very reasonable price. However, i could not find a transformer that was the same as this one. However, I found a “simliar” one.

I used the simliar transformer to try and create this circuit. It was a 220 to 12 V transformer. So it had about the same turns ratio as the original transformer. However, in this circuit the resistance and inductance of the transformer were very important. In the end I was able to get the circuit to work slighty by using a pot for R1 and adjusting it to get the output oscillations to be large enough. The problem with this however was that the oscillations became unstable and the voltage grew until it stopped oscillating all together. Thus this circuit was determined to not work but was important in starting me thinking about this project.

When I recieved the trigger coils this circuit was on the back of the package. So i built this and powered it from the high voltage DC supply shown in the voltage converters section. I was able to get it to flash by hand triggering a switch and could get it to quite a fast trigger rate. It was the first proof that this circuit might actually be possible to make.

Once I got a strobe circuit working I wanted to be able to make a remote trigger and began thinking i could use a triac or SCR for the triggering. This is a circuit that I found that would do just that control using an SCR and opto isolated triac. However, this sort of triggerin method was used with both a triac and SCR without any luck. The final trigger design will discuss the issues and how the final design was reached.

This used a G13599 inverter transformer from Electronic Goldmine. Only the power conersion part of the circuit was used. The actual strobe still used the radioshack circuit. This circuit worked but had a drawback of extremely slow charging time. The charging was acceptable for pulsing once every few seconds but too slow to even pulse once a second. So even though the circuit worked it was not practical.

After all of these disappointments in voltage converters Teknick was consulted. His idea was to create a flyback converter from a boost circuit. Using a commericially avaliable boost chip we could replace the inductor with a coupled inductor and use the secondary as a step up for the high voltage side. Originally a Max733, 15V boost chip was used. The coupled inductor was hand made from a toroid. The primary had 8 turns and the secondary had 80. Resistance and inductances were 18mΩ, 18.5 µHand 1Ω, 1.8 mH respectively. However, because of the internal feedback circuitry and low voltage capabilities the max output acheived was 90V.

So the MAX629 was then selected. It is a 28V boost with adjustable feedback. One problem is that it was an SOIC so it needed to be soldered to a surfboard before it could be used for prototyping. The feedback was done on the primary side for simplicity and with some minor changes such as a star ground the circuit began to work and was able to be regulated rather well. Zener equilalent to 27.3V were placed on the primary side output to prevent the Lx pin from reaching too high of a voltage due to the flyback voltage reflections. This also limited the maximum output which you could achieve which provided some safety. The circuit was then used to power the strobe and charged faster than the inverter circuit. This provied a low voltage powered power supply. The circuit diagram is part of the final circuit diagram below.

Now that I had the power source I just needed an automated triggering system. It is possible through the use of a neon bulb or diac and RC network to make a control like that used in a light dimmer. However, I wanted to make a way to control the strobe from a microcontroller.

Now that I knew I wanted to be able to trigger the circuit with a microcontroller I knew that I would need some sort of isolation such as an optoisolator. Since I was using DC and not AC i figured I could use a H11D1 high voltage opto-transistor. However, when i tried it it didn't work. I later determined that the turn on wasn't fast enough to give the quick pulse needed for the trigger coil to work. So I tried to use the circuit from the internet that used the opto isolation triac. I was very excited at first when I saw it pulse. However, it didn't pulse after the first one. I investigated and later realized that it was because there was too much current flowing through the triac because the coil turns to a short at DC. Since the triac was now seeing DC it wouldn't turn off, this was a problem. So I combined both the SCR and the opto-transistor, the opto-transistor would turn off and thus the SCR would not be triggered anymore. It worked! Because of the low amount of current that actually flows the SCR was not latched and would turn off as soon as the transistor turned off. Now I could trigger it remoteley with a low voltage. (The circuit diagram is part of the final circuit diagram below.)

The final part of the puzzle was something to control the new trigger circuit and this would be a microntroller. A MC9S08QE8 to be exact. I had seen in the demo for the QE32 that there was part of the program which used the pot and ADC to change the duty on a PWM. So I took that idea and using the Process Expert and Programmable Pulse Generator (PPG) bean to generate my signal. I used a pot to the ADC and then determined which output frequency should be given for each input. Then I tested the program on just a flashing LED and it worked. As you will see in the circuit diagram a NMOS was added to provide extra isolation and to take the current draw load off of the microcontroller. This completed the part needed and they just needed to be put together.

Yes this is finally the moment you have been waiting for the final circuit. In the final circuit you can see the flyback, trigger, and strobe circuits. The strobe circuit is literally just the the radioshack one with the switch replaced by my trigger circuit. The flyback is on the right and you can see the pot for the feedback circuit, the inductor, and output diode. After the output diode, the capcitor smooths the signal and creates the DC. This powered the strobe which wsa triggered externally by the microcontroller. A 10 µF or 22 µF could be choosen depending on the speed and intensity needed with the 10 µF being faster but less intense. Thus I finally had my completed strobe circuit.

Note: After the original circuit was made the microcontroller stopped functioning. Upon investigation it was found that it was drawing an extremely large amount of current (700+ mA). By adding a 100nF bypass capacitor from VCC to ground the current was brought down to around 10mA. Thus a 100nF and 10μF capcitor were added to the power supply (batteries in this case) for the microcontroller.