Using Auto Flash for Portraits...

Discussion in 'Casual Photo Conversations' started by hjoseph7, Jul 26, 2022.

  1. Can you use Auto Flash(the old Thyristor technology) to take portraits ?
    A Thyristor(Auto) flash has a photocell in the front to measure light reflected back from the subject. It is calibrated to ISO and aperture and shuts off the flash pulse when it's seen enough.

    With that said, I have 2 Vivitar 283 flash units with Thyristor technology, If I use one flash as the Main and the other as Fill would I get perfectly exposed portraits since the flash is doing all the work ?
    No need to figure out flash meter readings, Guide numbers, flash to subject distance, aperture, ETTL, iTTl whatever. since the flash is doing all the work
    Am I correct ?

    The reason why I ask is that I don't have a live subject with me to test my theory right now...
     
  2. It would work reasonably well but you can use yourself as the subject. Since you have the 283 so that I think it's OK to do that. You can set the flash for the same aperture as your lens for the main and smaller aperture (larger f number) for the fill. If I use a flash or flashes with variable power in manual I would rather do that.
     
  3. Anuthing, a broom, will serve as test subject. Your checking lighting, not pose, right.

    Fill needs to underexpose a bit. If you set the fill to a smaller aperture than the lens, it will put out more, not less light. So you'll want to do the opposite of what Bebu suggested. When using, say, f/8 on the lens, set the fill flash to f/5.6 or f/4.
     
    Last edited: Jul 26, 2022
  4. You're right! thanks.
     
  5. Yeah it’s pretty simple, like q.g. said, set the flash to expose about a full stop under your primary light source. It takes a little experimenting and may change from one situation to another but you can preview and adjust as needed. After a while it will come naturally and you won’t really have to think about it.

    Rick H.
     
  6. As an amateur, I only ever take 'in-context' portraits at work/home/leisure locations to support articles. I very rarely use a flash and I've never invested in multiple off-camera flashes. Or spent much time learning how to use a flash ;). I don''t feel confident using one. On the few occasions I use an (on-camera) flash, it's always manually dialed down and bounced off a wall or a reflector.

    I'd never heard of a Thyristor(Auto) flash so just out of curiosity I looked it up. I came across this article on Thyristor(Auto) flashes (and how best to use them) which you might be interested in. In places, it goes into a lot of detail but it also seems to give some practical tips. I have no idea whether it's useful but I thought I'd just share it.
     
  7. All portable flashes have thyristor to control flash power whether automatically via the flash sensor or TTL, iTTL, etc.. or even variable manual power they all make use of thryristor. This is not always true with studio flash.
     
    mikemorrell likes this.
  8. Sorry about the hair splitting but wasn't the special thing about thyristors that they preserve the unused capacitor charge? Dad bought an Agfa auto-flash lacking that feature and needing full recharge time, no matter how weak the exposure.

    On topic: Do as you like. Getting tweakable RAW files should be possible. Don't(!) expect miracles! Auto flash is a tech level below TTL flash, which also doesn't always expose everything absolutely right. For perfect reproduceable results stick to manual flash setting. I would not dare to assembly line shoot a broad ethnic rainbow of corporate headshots with auto flashes, assuming that my backdrop will look the same in batch processed files.
     
  9. Yes the early units simply wasted the remaining charge.
     
  10. I'd say it depends on your setup.
    If there's little overlap between key and fill, such that the key flash only 'sees' it's own light, and the same for the fill, then you should get a good enough exposure. (Let's leave 'perfection' for an alternative Universe, eh?)

    Issues might arise if the fill light falls directly on the sensor of the key, or vice-versa. But on the whole it works well enough in practise, as long as you're aware of the above and avoid such a situation.

    Let's take Loop lighting as an example, where the key is to one side, and usually a hard-ish light with its sensor directly pointed at the subject. No problem on its own, but a frontal fill flash from the camera position, even well diffused, might cause the key's sensor to cut off early and give a weak key and not the lighting ratio expected from the respective aperture settings.
    In other words, it's trial and error time. But once a good balance is found, the exposures should be consistent, even with a change of subject.

    Personally, I like to add a hard, possibly snooted or flagged, hair light or kicker from above and behind the subject. Auto exposure of any sort tends not to work well with such a light, and it nearly always needs manual control and override.

    Another issue that might be found; is if using a flash that doesn't allow the sensor to be pointed separately from the flash itself. I.e having a 180/270 degree swivel head. This will preclude bouncing the flash from a brolly or other surface. Vivitar 28x speedlights need not apply for this job!
    IME, the simpler and 'lower' level of tech gets it right more often than the over-complex and incomprehensible algorithms of TTL flash exposure.
     
    Last edited: Jul 28, 2022
  11. As well as I know it, the early auto flash units had an additional flashtube inside.
    When it was time to shut off the flash, they fire the hidden tube to dump the charge.

    That is all within the 1/1000th of a second or so that the flash is active.

    For a thyristor (SCR) flash, there is an SCR in series with the flashtube, with its gate active when the flash is fired.

    There is another, smaller, flashtube inside. When it is time to turn off the flash, this tube is fired, which then puts current the opposite direction through the SCR, until it turns off. There is also a capacitor to limit the charge that can go from the second flashtube.

    As for the actual question, though, it isn't so obvious what happens when two such flashes go off at about the same time.
    Each sensor will see the light from the other one. I suspect that the light doesn't balance so easily between the two,
    but maybe close enough.
     
    Jochen and Ken Katz like this.
  12. I believe the pre thyristor flash units secondary flash tube was called a "squelch tube" or "squelch circuit" (dim memories from about 1/2 a century ago). The first flash unit I used was a Metz 202. A big honking bracket attached / external lead battery type unit with 5 AE settings and a low/hi power switch. Since it would dump any excess capacitor juice into the squelch tube, you could reduce recycle time by setting it at low power if you were close up. I also used a Metz 402 later on, which was similar in appearance but had the famous thyristor circuit, eliminating the need for the low/high power switch. The 202 had a better quality of light IMHO.

    I was told that using 2 autoflash units at the same time would not work well, but I never got the chance to try it. Frankly with digital, I would set the flash units on manual and adjust levels by eye / histogram. Unfortunately, the 283 does not have a manual setting without getting the applicable accessory gizmo for it. My Vivitar 283 is sitting in an old camera bag, both of which are gathering dust in a closet for the last few decades
     
  13. Never seen that complicated arrangement.

    The circuit Metz first used in their thyristor (SCR) controlled flash was much simpler. It was an SCR in series with a small value, high voltage ~ 3.3uF / 400v capacitor. The two are wired in parallel with the flash tube. When the flash needed to be quenched the SCR was triggered, thus momentarily diverting the sustaining current into the small capacitor, and extinguishing the flash. Simple, but very effective and efficient, and with no secondary flash tube involved.

    I've seen the old quench-tube cicuits as well, and the secondary dump-tube is a short, fat, black-painted thing looking more like a lightning protector. You wouldn't recognise it as a flash tube unless you knew its purpose.

    Nowadays there are high voltage, high current insulated-gate bipolar transistors (IGBTs) that can be very rapidly switched on and off to allow TTL control 'Morse code' to be transmitted by the flash tube, and to stutter the flash for high-speed focal-plane shutter synch. Controlling the flash duration is now a fairly secondary function.
     
  14. The one I described is pretty much the Vivitar 283.
    Flash tubes are very convenient high speed, tough, high current switches.

    Before the thryistor, there was the thyratron, which is like a vacuum tube, except gas filled.
    And for higher currents and speed, there is the ignitron.

    High speed high current switching is never easy, and especially you have to watch
    how much inductance there is around them. I believe SCRs can switch off faster than
    flashtubes, but it might depend on the gas.
     
  15. SCRs are avalanche devices, and cannot be turned off by any gate voltage once triggered. They only turn off once the current reduces below some 'holding' value, or if they're short circuited. They cannot be gated off.

    Obviously, short circuiting an SCR requires the use of another switch of some kind across it, and the whole exercise becomes pointless. However, this didn't stop Vivitar from designing such a stupid arrangement - Screenshot_20220811_081318.jpg
    As can be seen, the thyristor is placed in series with the flash tube! And really serves little purpose, because a quench tube is used to turn it off. The quench tube places a diode and 3.9uF capacitor across the SCR, and might just as well have been put in parallel with the flash tube itself. Or the thyristor used instead.

    Today is the first time I've seen the schematic for Vivitar's so-called 'Thyristor Controlled' flash. It's really no such thing. It's a quench-tube controlled flash.

    What a crazy and redundant circuit! :confused::p
     
  16. Do you have the Metz schematic?

    Anyway, how fast you turn off a bipolar transistor depends on how fast you can such all the minority carriers out of the base.
    For an SCR, it is also how fast you suck the minority carriers out.

    So, when QT-1 fires, it sends a pulse of current backwards through the SCR, hopefully until it is off.
    Actually, it is more interesting than that. As FT1 fires, it seems that SCR1 will be off.
    A pulse of current goes through D8, C15, and eventually to the gate of SCR1, turning it
    on just as the flash current is rising.

    When QT1 fires, which is presumably a lot smaller (and faster!) than FT1,
    it dumps charge through C15 and D8 backwards into SCR1, sucking out
    minority carriers.

    If C1 is 1100uF, and C15 is 3.9uF, less than 1% of the full C1 charge is available,
    but it should only discharge until SCR1 is off.

    Then all the ionized atoms in the flashtube have to find electrons again until it is off.

    It will be interesting to see how the Metz flash works.

    From your description, the SCR has to zero the voltage across the tube,
    which it can do if there is enough series inductance, and the pulse is
    fast enough. That is, the dI/dt is big enough. And as with C15 above,
    there is capacitance to limit the length of the pulse.

    But it has to be long enough for the tube to deionize, and be able
    to hold off the full voltage. As well as I know, that is longer than it
    takes to turn off SCR1 above. And for that time, it has to keep the
    current away from the tube, hopefully a small fraction of the flash time.
     
  17. No, it doesn't Glen. When the quench tube fires it puts the 3.9uF capacitor (+ diode) in parallel with the SCR and momentarily diverts current through the capacitor. Thereby starving the SCR of its holding current and turning it off.

    In that Vivitar circuit the SCR never really works in gated mode. It simply acts as an avalanche diode. But I agree with your addition of 'hopefully', since the circuit looks more cobbled together than designed.
    No, but I've been inside an early 45 series Metz in order to add a trigger-voltage lowering circuit. I was curious about the 400v 3.3uF polyester capacitor next to the flash tube, and traced it back to a thyristor.

    In effect the Metz circuit simply replaces the old quench-tube function of shorting the flash tube. But it does it by triggering an SCR when the flash needs to be turned off. The SCR places a discharged low-inductance 3.3uF capacitor across the tube. Thus momentarily starving the plasma of sustaining current and extinguishing the flash. Much simpler and direct than the Vivitar circuit, and using the SCR in true gated mode.

    It's also a self-resetting circuit; since the 3.3uF capacitor can only absorb a small amount of energy before a voltage balance is established between itself and the SCR and flash tube. So the SCR quickly becomes starved of holding current too. There's a resistor across the polyester capacitor to keep it normally discharged when the SCR is in the off state.

    The rest of the Metz circuit is pretty standard inverter, trigger and pulse-timing stuff.
     
    Last edited: Aug 12, 2022
  18. OK, the way I think of it, and also the 283 designers, is that an opposing current is added to the SCR current, such that the result is zero.
    (Since it won't conduct the other way.)

    The result is, as you note, the same amount of current going through QT1.

    With the charge on C15, and if the SCR did conduct the other way, then current would flow the other way.

    If you look at it as diverting current, then it will never go the other way, and you can't be sure it will get to zero.
    That is, C15 starts charged one way, and ends up charged the other way.
     
  19. The complication with describing it that way, is figuring out how much current it is.

    Actually, that is important in figuring out the whole thing. How much current goes through the flashtube?

    One limit is L1, which limits how fast the current can change, but also there is inductance and resistance from the wiring itself.

    So, how much current is going through the flashtube, that you have to divert through the 3.3uF capacitor?
    But also the discharge current will increase as the flashtube voltage decreases.
    So, I suspect that there is enough inductance to slow down the increase.

    Also, as the capacitor charges, the voltage will increase.

    If it increases too fast, the tube won't be completely deionized before the voltage is high enough.

    Fluorescent lamps operating on AC don't completely deionize when the voltage crosses zero.
    There are enough ions to start conducting again the other way.

    Presumably designers figure all this out, in more detail than we do here.

    It might be, though, that if you add gas besides xenon, you can speed up the deionization.
     
  20. There's always going to be inductance, but it's fairly insignificant compared to the deliberate capacitance of 3.3uF. And that's why Metz wire both the thyristor and capacitor directly adjacent to the flash tube on the same board.

    We can go deep into the theory with different degrees and views of circuit operation almost ad-infinitum. That doesn't in the least affect the fact that both Vivitar's and Metz's circuits work adequately.
    It's just that the Metz circuit is far more elegant and logical.
    That really doesn't matter much.
    A short-circuit is a short-circuit, and whether it dumps the current in 1 nanosecond or 1 microsecond will make very little difference to a 200uS flash pulse.

    That photographic flash tubes can be made to switch on and off at micro-second time scales is empirically proven. The theory doesn't matter unless you're designing such a circuit.

    But there are good and cost and component-efficient designs, and then there's Vivitar's.:eek:

    P.S. I was once asked to design a flash-tube triggering circuit by a theoretical physicist. He insisted I use a 200 amp (continuous) capable thyristor that he'd blagged a sample of from a sales rep. I protested that the self-capacitance of the huge device would definitely not help its longevity when exposed to a rapid pulse. However I grew tired of arguing and used the monstrous thing just to keep him quiet.
    It failed and went short-circuit after a few weeks of light usage.
    The much smaller 8 amp rated device I replaced it with ran for years, and is possibly still working.
     
    Last edited: Aug 13, 2022

Share This Page

1111