Xenon flash circuit

[maxgreen]

Hi everyone,

I am currently trying to build a xenon flash circuit with a voltage input of 6v or 12v, that I can connect to my optical slave flash unit.
The xenon flash circuit should not be a too complex circuit since I am new at this, but I of course know all the basics.
I also want to be able to fit the circuit into a plastic cover, just like a normal external camera flash, and therefore the unit should be as small as possible.
Does anyone know where I can find schematics on such a unit or maybe give me guidelines on how to start building the unit.

I found the following circuit **broken link removed**, that seems to be very simple, but I am not able to find the inverter transformer anywhere besides their website, and this company has closed until late december. So is possible to build a similar circuit, just with another transformer that is possible to buy elsewhere?


Thanks in advance.

 

[Blueteeth]

The little transformer is by far the hardest part to find, as it is usually custom made by the manufacturer of flash equipment - there are no real 'standard' off-the-shelve parts, unless you are willing to try and roll your own

Funnily enough I was going to recommend an ebay shop, but its the ebay store for that very company! so its also closed until late December..

Other options would be to buy a low voltage 'strobe' circuit. It will have to be modified so it doesn't constantly trigger, and so it can be connected to your slave trigger. Note, ebay has *many* mounted flash units second hand, most will run off 4xAA's (6v) and can be quite powerful. These are generally cheap, but reliable, and would pretty much be a pre-built solution - albeit one you don't have much control over.

So really, its either design from scratch (DIY transformer, not easy because you have to get the right core material and work out the windings, and its not a case of 1:25 turns ratio..) or modify an existing circuit for your needs. I'd go with the latter

 

[dougy83]

The inverter circuit looks like a basic blocking oscillator (ala Joule Thief). See attached circuit (modified) from their website. The drawn-on coils should provide the correct function. The dots indicate the winding 'start' end of the coil. The capacitor I added is optional.. leave it out for now.

The coils, namely Primary, Secondary and Feedback (P, S, F) will need turns ratios to ensure that the output voltage is reached (not difficult as it's a flyback topology), and that the transistor rating are not exceeded.

We also don't want the collector voltage to exceed 30V or the base voltage to go too much under -5V (for 2SD882 transistor). So, we want 400V out.. say S=200 turns on a small ferrite toroid (maybe 20mm OD or so), then P=10 turns (for 20V across the coil & 25V to collector w/ 5V input) and F=~4 turns. Done.

This style of inverter is very forgiving, and no values are that critical: the oscillator runs at a frequency based on the transformer and/or transistor saturation characteristics.

 

[dr pepper]

Chemelec who's a member of this forum has one on his website and it uses just an inductor no need to wind a tranny, the voltage produced then goes through a multiplier, here:

**broken link removed**

 

[maxgreen]

Thanks for your replies.

Blueteeth, actually I found that same user on ebay, and as you said, it is the same company The thing is I need to design this transformer from scratch, since I am planning to have my unit in production. So I either need buy an transformer where I can use it for commercial use or built my own. I of course would prefer to buy one since it would save me some time, but it doesn't seem like it is that easy to find.

Dougy83, according to you words, it sound as if it is easy to build this and very straight forward? I haven't tried to build inverter circuit before, but is really just as you mention?

Thanks dr pepper, I will try to contact Chemelec. I think that I will be able to use his design and just remove the strobe functionality.

 

[Blueteeth]

Ahh for production! In which case it could be tough to find an OEM for such transformers. Whilst there are many companies in India and china kicking out these things, because they are made specifically for a manufacturer they don't have much in the way of internet 'advertising', and so not commercial. The only transformers I've seen available specifically designed for flash charger circuits have a turns ratio of roughly 1:12 and are used for Linear technology's line of flyback converter chips. They don't have a feedback winding which means you'll need to drive it as a current-limited boost converter (can of worms).

As far as I know, almost all of these circuits are self-oscillating flybacks, using a feedback winding and a single transistor, this provides a current source to the capacitor being charged, so the turns ratio isn't absolutely critical, a ratio of 1:10 would be fine. But that means the inductance of the primary really is critical as this determines the frequency. That frequency changes as the voltage on the capacitor rises (know that high pitch whine that increases in pitch as it charges?).

What kind of power are we talking here? Professional setups can use multiple capacitor chargers for a capacitor bank, so they provide a quick succession of flashes - each cap providing the power for a single flash, and can draw well over two amps from a battery back. Others, like those found in disposable camera flashes run off a single AA, and draw little current. I'll admit the company you linked in your first post looks to have some good kits, with a range of power though-put options.

The boost converter/voltage multiplier circuit linked above looks ok. It may not be as efficient, and require many more parts but... all those parts are widely available! Using the formula's for the 555 timer (resistor and cap values to on/off times) you can determine the peak current in the inductor, and so the current rating you'll need. I would change the diodes in the multiplier to ultra-fast versions, such as the UF4005/7 (available and just as cheap) which should increase efficiency by a small ammount. And make modifications to it so it stops charging when the cap is full.

 

[4pyros]

I am planning to have my unit in production.

How reliable does it need to be?
Can you use a LED strobe?

 

[dougy83]

Dougy83, according to you words, it sound as if it is easy to build this and very straight forward? I haven't tried to build inverter circuit before, but is really just as you mention?

Yes, it is that easy. I built a couple when I was a teenager using random cores (different torroids and ferrite rods) and they worked well (i.e. they charged a cap quickly).

Depending on how many units are in your 'production' run you'll have different avenues to obtain your parts incl. transformers. You can get 1000 transformers medium-sized custom made in China or India for a couple of dollars each; it'll be cheaper for smaller transformers. The quality that I've seen is very good, and the companies that my old work used to deal with really knew what they were doing.

For preliminary testing, you can just wrap some wire around a torroid, or run a simulation or two.

The only transformers I've seen available specifically designed for flash charger circuits have a turns ratio of roughly 1:12 and are used for Linear technology's line of flyback converter chips. They don't have a feedback winding which means you'll need to drive it as a current-limited boost converter (can of worms).

Because the voltage across the primary coil is proportional to that across the secondary coil, it's easy to sense the output voltage without a dedicated supplementary feedback circuit. IIRC they do use a cycle-by-cycle current-limit, but this is common for many SMPS controller ICs.

These ICs are very good at what they do... google 'photo flash charger ic' for a bunch from different manufacturers.

My ex-company used one in their designs.. I'm not sure which manufacturer though.

 

[maxgreen]

The power I need produced is about 300-400v. I found the following compenent after dougy83 post: https://dk.farnell.com/diodes-inc/zxsc440x8/ic-photo-flash-charger-smd-msop8/dp/1226474RL
The only thing about this is that it only charges a capacitor of 80
uF and I think this might be too low. My xenon tube is 50 ws 300-500 volt, then I would probably need a 400 volt 1000uF capacitor or does it have to be this big?
Several of you mentioned that for production it would be a lot cheaper and easier to have a custom made transformer, and that is probably what I would go for when I go into production.

4pyros you mention a LED strobe, but can the light produced by an LED even get close to what a xenon produce? I made a small test with LED while ago and my conclusion was that it wasn't bright enough. It does have to be very reliable. Do you have a LED circuit in mind that you think might work?

Blueteeth, actually I haven't thought of a capacitor bank, but when you mentioned it, I realized that I need into my circuit. Though I would only need 1. Thanks

 

[ronv]

How fast is your base unit? To get the 50 ws fast enough with the low voltage may be a problem. You can use a bigger cap but it will take longer to charge it with a fixed voltage and current.

 

[circuitspecialists]

schematic for Xenon flash circuit

There is a kit that I think will provide what you need. The kit is KIT 162 and the complete schematic is posted on line. I will try to attach the pdf file to this post. It is titled kit162_manual.pdf

 

[4pyros]

4pyros you mention a LED strobe, but can the light produced by an LED even get close to what a xenon produce? I made a small test with LED while ago and my conclusion was that it wasn't bright enough. It does have to be very reliable. Do you have a LED circuit in mind that you think might work?

I general xenons are not thatreliablee. They will only last so long depending on how hard they are driven. They may last only a few hundred hours.
LEDs can be bright but you may need more than one and overdrive them slightly but they should still last thousands of hours.

 

[Blueteeth]

Yes, it is that easy. I built a couple when I was a teenager using random cores (different torroids and ferrite rods) and they worked well (i.e. they charged a cap quickly). ...For preliminary testing, you can just wrap some wire around a torroid, or run a simulation or two.

The core, and the number of turns will determine the inductance, as well as saturation current. With a feedback winding for a self oscillating converter/charger it'll run at the frequency determined by this inductance, voltage input, and voltage output. Whilst it'll probably work with a toroid, efficiency, current consumption (and therefore charge time) will vary greatly with core material and windings. Good for trial and error and for a prototype, but a hassle to have any sort of control over battery life, charge time etc.. Almost all of these transformers will have a 'gap' in the core where it can store more energy. So if you roll your own, you'll need to buy pre-gapped cores.

Because the voltage across the primary coil is proportional to that across the secondary coil, it's easy to sense the output voltage without a dedicated supplementary feedback circuit. IIRC they do use a cycle-by-cycle current-limit, but this is common for many SMPS controller ICs.

I meant feedback windings that are used in single transistor self-oscillating flyback converters, not for measuring capacitor voltage, but to determine what the field is doing in the transformer. But you're right, you can sense the caps output voltage via the primary peak voltage.

To the OP:
Its true, they are essentially boost converters with added circuitry specifically designed for flybacks (isolated boost converter) to charge capacitors. Unlike a boost converter though, it doesn't regulate output voltage quickly, as the output cap is expected to be large, instead it just stops charging when a preset output voltage is achieved. In simpler circuits like the ones from that store, and in most camera's, because the charger is self oscillating, its remarkably simple, and its efficiency, power consumption (charge time) is determined by the transformer characteristics rather than circuit control.

I've found with my own flash cricuits/cap chargers that its either a simple and cheap circuit, with critical parts like the transformer - fixed characteristics, but does the job well. Or, one that uses clever control, with many more parts, but relaxes the constraints put on the transformer/inductor design. Obviously for mass production, simple/cheap is better so I think you're going down the right road here. If of course you want to make it very efficient, small, or have some digital control over it, then IC's specifically designed for the task would be handy. For the most part though, anyone using a flash just wants to trigger it, and then expects to wait until its charged without having a countdown or charge monitor or anything, simple is better.

That component you found for charging 80uF, the size of the capacitor doesn't matter to the charger - these things just pump current into caps, so as ronv said, it'll just take longer to charge. Its not *quite* as simple as 'double the capacitance = double the charge time' because as it charges, it takes more current. But the charger doesn't care. If you want faster charging using an IC driver, you will need to increase the current limit, or increase the input voltage.

Sorry for the long posts, years ago I was intrigued by these little converters found in disposable camera's and spent far too long working out what was happening and building my own :/ (wasted youth eh?)

Just remembered I had this on my PC, great little app note!
https://www.electro-tech-online.com/custompdfs/2012/11/dn25.pdf

 

[dougy83]

In addition to the above variables you mentioned, the frequency is very dependant on the maximum current that can flow through the transistor and transformer primary (due to saturation of the transformer core, the transistor characteristics or coil resistance) -- because when the current stops increasing, the feedback winding current is reversed and the transistor is turned off. It's this positive feedback in conjuction with the max current limitation that allows the circuit to operate at all.