intro, and some hints on voltage multipliers wanted and other kinds of stuff / High Voltage / Forums

Forums

4hv.org :: Forums :: High Voltage

« Previous topic | Next topic »

intro, and some hints on voltage multipliers wanted and other kinds of stuff

Move Thread

LAN_403

OJD

Sun Jun 10 2012, 01:47AM

Registered Member #5192 Joined: Wed Jun 06 2012, 08:05PM
Location:
Posts: 2

Hello, I'm in the process of designing and building a photo flash generator (the studio variety) and would appreciate some help. English isn't my native language so please bear with me if you find any lingual problems.

First of all you might wanna warn me, high voltage and deadly charges in the capacitors, ok got that. I've been repairing and building flash devices for quite some time now so I know the basics (i.e. always monitor cap charge and remember to discharge before medling, a 8 hrs numbness in my forearm taught me that once). I now wanna go deeper and make something more useful. I will not build this entirely myself, my dear step brother who is more electronically enclined than me (and who is/was registered on this forum and recommended it to me) will help me and make sure I don't make any once-in-a-lifetime mistakes :)

Ok, that's me and my experiences in the field of HV electronics, the question probably a lot of you have been asking yourself now is why on earth would you want to build/design something you can already buy? Well, the answer is simple, bang for buck. I live in Europe, the prevailent brands here are profoto, bowens, hensel and broncolor. Profoto, hensel and broncolor deliver what I want but for a very very high cost, bowens are not really up to todays standards and are as well bloody expensive... There are some american brands that are on an acceptable price level but only do 115 VAC and with a dull flash head sortiment (is that even an english word? hmm), and hence practically useless to me.

I've made a box diagram showing how I'm thinking about this generator. First a primer on what I'm actually trying to do, for those whose photographic experienced is limited to less than commercial level photography:

Studio flashes comes in two varieties, the monohead and the pack/head kind. A monoflash is a unit which contains the generator and flash tube in the same box so to speak and the pack/head variety is a setup where you use a loose box (the generator) generating a charge in a cap bank, you can connect a head or more usually several heads utilising the charge in the cap bank to discharge the flash tubes in the heads. The generator also usually contains some kind of control electronics, controlling distrubition and power of the different heads. It is the generator I am trying to build now. The heads are not very advanced, they usually just contain a tube, a halogen modeling light, a small ignition cap and the ignition coil.

Alright, so to understand my small diagram you'll have to understand two things. The 'color' (in quotation marks because color is a very advanced subject) of the light from a xenon flash tube is determined by a lot of things, but the two easiest to alter is voltage (the voltage over the tube, not the ignition voltage) and flash duration. A lower voltage renders the spectrum towards the yellow/red (a 'warmer' light). A shorter flash duration renders the spectrum towards the blue, because the flash tube does develop it's complete cycle so to speak (struggling with my english vocabulary...). The intensity of the flash is determined by two factors (really a hell lot of more, but two that are electronically easy to manipulate), energy (amount of charge in the cap bank) and flash duration.

Ok, so here is my block diagram:
7170773961 C8e2941e94 B

As you probably know, we have 230 VAC here in Europe, most resedential estates have 10 A fuses (except the UK) so that's my limit here. I want to go from mains to some sort voltage multiplier. From the multiplier I want to be able to control voltage (as this can manipulate both flash color and total energy in cap bank, hence intensity). With the set known voltage I want to charge my capacitor bank. Be able to distribute the potential charge over a number of sockets (to witch my heads are connected). When I feel time is ready I want to use my ignition circuit to discharge, and control the duration with some sort or timer/IGBT circuit (and the flash duration controls both the intensity and color, hence the little feedback loop back to voltage control, if I want to change duration and maintain color). The modelling light box is just a circuit controlling halogen modelling lights, and to make sure they go off when discharging the flash tubes and that the different halogen lights follow the set power distribution). Alright I hope my weird ideas are as clear to you as it was to me when I drew this.

I have been researching this for quite some time now and I have found a lot of resources, problem is they are pretty vague generally. There is a russian forum dedicated to studio flashes. Problem is that it's in russian (mr google translate is only so helpful) and that it is pretty scarce in participation. Even if you do find anybody willing to help they usually just throw you a schematic without any explanation. Not really a safe way to go through with things. Since spare time is a rare commodity for me I will go through this project box by box, and I'm now starting to look in to voltage multipliers. If you do have any good references on basic voltage multiplier design please share, all I find is really bad stuff by companies promoting their own limited products and wikipedia. Not one mentions on how to actually control voltage or current.

Ok, back to the topic, voltage multipliers! I have come to the understanding that most studio flashes operate around 350-450 VDC. There are of course a number of ways of turning my 230 mains voltage to that. Viable solutions seem to be

1. Transformer
2. switched transformer
3. Voltage multiplier

The ordinary transformer will be way too heavy for the currents I will be operating at so they are unfortunate out of the question. Switched transformers are interesting but I find it hard to find any good design notes for the currents and voltages I will be operating at. Most design instructions deal with down transforming mains into low voltage. I have built a switched HV generator for an on camera flash using a TL494, two mosfets and a coil I took out of a PC power supply, so I have some modest experience from flybacks. However, not really interested in the extra weight and not really keen on coiling my own custom coil, and I have no idea on coil design either. Ok, so the only option left (unless there are any other that I have overlooked) is the voltage multiplier. This is commonly used in commercial flash units. Both in the US and Europe. I have been looking at a few circuit schematics, but have no idea on how they work. I know the potential dangers with multipliers, i.e. no isolation from mains. I also know about double isolation of the unit and so on to minimize risks with the multiplier.

Things I will have to consider is charge time. I want to be able to charge the cap bank as fast as possible with no risk of damaging the caps, minimal heat buildup and under the limit that the 10A mains breaker impedes. The optimal thing would two build a circuit with two different options, slow charge and fast charge, where fast charge is close to the current limit (or limit set by caps) and slow charge offers the option for weaker mains.

Here is a schematic I found on the russian studio flash forum (couldn't find the thread from which it came from, it was referenced elsewhere, sorry for the extreme lame schematic, not mine as stated):
7355963970 9625d07b0a B

This was referenced as 'fast charge, 1 sec' and it didn't state how much capacitance it was charging at that time. What I get from this schematic is the voltage multiplier stage (my designation of the upper left corner), the zener diodes controlling the voltage and the thyristor controlling the outgoing current. I've looked at a schematic from bownes and they look very similar, with the exception that they had two multiplying stages for slow charge and fast charge, with different caps and resistors but the same kind of diodes. Why is the two transistors connected so weird?

So my questions:

1. Is the above scheme a viable one?
2. How does different caps and res' in the multiplying stage affect current, or do they?
3. How is it easiest possible to control current? By thyristor as the above scheme any other way?
4. Can I control voltage by any other way than zener diodes? Could I control voltage by switching to another set of zener diodes?
5. What is the fastest way of charging capacitors in a safe manner? How much current can they take?

For those interested I can mention that I plan on using 21 capacitors from nippon, rated at 1000 uf and 450VDC a piece, giving me a total of approx 1600 J at 390V (which will probably be my upper limit). I'm thinking that a battery of smaller capacitors will give me a higher possible out current (to the flash tube that is) than few bigger ones. Or so I'm told at least. My thought is to divide the cap bank in three parts, one which is half of the bank, and two units which would represent a quarter of the bank each, so 1/2, 1/4 and 1/4. These would be split up on three sockets giving me a set 1:1:2 light intensity ratio by default.

Thankful for any hints and tips. This can go down in two ways, either I will be subject to a hail of mockeries, tauntings and warnings. Or, which I hope, I will be given some great advice. I hope 4hv can give me the latter. :) Thanks for reading.

Sulaiman

Sun Jun 10 2012, 10:39AM

Registered Member #162 Joined: Mon Feb 13 2006, 10:25AM
Location: United Kingdom
Posts: 3141

I'm still working out how the regulation works, but here are a couple of answers

"Why is the two transistors connected so weird?"
... they are used as 'zener' diodes
the base-emitter junction of bipolar transistors 'breaks down' when reverse biased,
for 2N3904 at about 6V minimum according to a Fairchild datasheet, probably around 7V to 10v in practice.

So my questions:

1. Is the above scheme a viable one?
2. How does different caps and res' in the multiplying stage affect current, or do they?
3. How is it easiest possible to control current? By thyristor as the above scheme any other way?
4. Can I control voltage by any other way than zener diodes? Could I control voltage by switching to another set of zener diodes?
5. What is the fastest way of charging capacitors in a safe manner? How much current can they take?

1. not sure of all the details yet,
but basically the main thyristor cannot be turned on when the output voltage is higher than the zener voltage,
and the thyristor T2 cannot turn on to trigger the main thyristor when the output is greater than the zener voltage,
so the circuit looks ok to me.
2. the output current is almost proportional to the capacitance in the multiplier stage
3. the charging current is mainly determined by the mains voltage and frequency, and the 'multiplier' capacitance.
4. you can switch in other zeners or 'short-out' additional zener diodes in series with the existing zeners.
5. since the 'flash' capacitors will be used for high current discharges,
much greater than the 10A available from the supply,
the charging current is not an issue for the 'flash' capacitors.
The limits are
- the 10A mains supply
- the current through the 'multiplier' capacitors (C1,C2)
- the current through the 'multiplier' diodes. (D1,D2,D3,D10)
- the current through the main thyristor (T1)
- the current through the input resistors (R1,R2)

This design will draw 'pulses' of current from the mains supply near the peaks of the mains voltage (100 'pulses' per second with 50Hz ac)
This would probably not pass European certification as a 'consumer' device.

OJD

Sun Jun 10 2012, 06:28PM

Registered Member #5192 Joined: Wed Jun 06 2012, 08:05PM
Location:
Posts: 2

Sulaiman wrote ...

I'm still working out how the regulation works, but here are a couple of answers

"Why is the two transistors connected so weird?"
... they are used as 'zener' diodes
the base-emitter junction of bipolar transistors 'breaks down' when reverse biased,
for 2N3904 at about 6V minimum according to a Fairchild datasheet, probably around 7V to 10v in practice.

So my questions:

1. Is the above scheme a viable one?
2. How does different caps and res' in the multiplying stage affect current, or do they?
3. How is it easiest possible to control current? By thyristor as the above scheme any other way?
4. Can I control voltage by any other way than zener diodes? Could I control voltage by switching to another set of zener diodes?
5. What is the fastest way of charging capacitors in a safe manner? How much current can they take?

1. not sure of all the details yet,
but basically the main thyristor cannot be turned on when the output voltage is higher than the zener voltage,
and the thyristor T2 cannot turn on to trigger the main thyristor when the output is greater than the zener voltage,
so the circuit looks ok to me.
2. the output current is almost proportional to the capacitance in the multiplier stage
3. the charging current is mainly determined by the mains voltage and frequency, and the 'multiplier' capacitance.
4. you can switch in other zeners or 'short-out' additional zener diodes in series with the existing zeners.
5. since the 'flash' capacitors will be used for high current discharges,
much greater than the 10A available from the supply,
the charging current is not an issue for the 'flash' capacitors.
The limits are
- the 10A mains supply
- the current through the 'multiplier' capacitors (C1,C2)
- the current through the 'multiplier' diodes. (D1,D2,D3,D10)
- the current through the main thyristor (T1)
- the current through the input resistors (R1,R2)

This design will draw 'pulses' of current from the mains supply near the peaks of the mains voltage (100 'pulses' per second with 50Hz ac)
This would probably not pass European certification as a 'consumer' device.

Thanks for the great reply, this made my, otherwise rainy and gray, day.

So basically I could 'up' the caps in the multiplier until I'm happy with the recharge time? And for a switchable multiplier with slow and fast charge options I could just use two different set of caps in the multiplier? Thanks for the hint on the shorting out of zeners. I will have to incorporate some sort of discharge of the cap bank when switching though, otherwise things could get complicated.

I don't know what you mean by "'flash' capacitors". If you are implying that I'm using what is sold as 'flash caps' on ebay and so on, that isn't the case, these are ordinary high quality nippons (I wanted sprague, like the commercial units, but alas to expensive). If you are impying the capacitors used for the flash then you are right :)

Don't worry about CE markings, this will be for my own use. Not a commercial consumer product. :)

Here is a link that is interesting:

Although this generator only runs on 115 VAC and is a bit to uncontrollable for my use, and almost always to powerful. Here is the multiplier module of that generator:

Note the big resistor limiting the current to the capacitors. Is that only to keep the current within mains fuse limits?

Moderator(s): Chris Russell, Noelle, Alex, Tesladownunder, Dave Marshall, Dave Billington, Bjørn, Steve Conner, Wolfram, Kizmo, Mads Barnkob