Photographic Strobes

There are a lot of dearly held beliefs about strobes out there, but let me try to dispel the myths and the mysteries.

Is there a relationship between Guide Number and the electrical energy in watt-seconds?

Yes there is. Here is the formula. It can be off by as much as a stop because not all strobe zoom lenses are equal.

where E is the watt-seconds stored in the capacitor, and EFL is the effective focal length of the lens. (For a full frame 35 mm camera.)

For engineering Geeks: This page gives the set of assumptions and how the above formula was calculated.

Here is a table of Guide Numbers for a number of photographic strobes with different focal length settings at ISO 100.

I have also plotted several commercially available photographic flash guns as symbols to see how they fare against the formula. Please note that this is a log-log scale.

The top dark blue line represents a very aggressively focused flash gun set for a 135 mm focal length lens. This is a modest telephoto lens. There are a few flashguns with this capability and some are shown as circles on the plot, with Guide Numbers from 45 to 55.

The gold curve is for 85 mm EFL lenses. The black curve is for 50 mm EFL lenses.

35 mm is the normal "design to" focal length, and most strobes unless otherwise specified are rated for a 35 mm focal length lens. Along this curve, you can see built-in flashes and flashes in disposable cameras are shown in the light yellow shaded area, having guide numbers from about 10 to 14.

The orange area represents low cost flash guns with guide numbers up to about 20 or so.

The pink area represents small flash guns, but possibly with a number of features such as adjustable field of view, and possibly IGBT (see below) circuitry.

The green area represents premium flash guns with guide numbers around 30 at 35mm and up to 60 at 125 mm focal length.

The light turquoise colored area is the area occupied by the "potato masher" flash gun. I happen to have a Nissin 4500GTE, so I plotted it on the graph.

Strobe Circuitry

In 1970 I got an SLR 35mm camera (Ricoh) and an electronic flash (Prinz.) Two years later I was aware that the strobe was bad for my camera's contacts (more about this later) so I modified it with an LASCR across the sync contacts, which not only saved the contacts in the camera, but also made the unit into a slave flash. This unit stored about 13 watt-seconds, or joules when the ready light came on and eventually came to a full charge at 26 joules. The Guide number in feet for ASA 64 film was 64. In the process I reverse-engineered the strobe. I don't have a picture of this strobe that I know of, so I found a similar model online for the picture.

About that time, I bought a Rolei flash. It has three features. It was very small and light. It had a hot shoe. It had auto exposure (via a quench tube.)  It reached a grand total of 9 joules at full charge.

About then, the buzz in flash photography was the auto-thyristor flash. A thyristor is the generic name given to SCRs, Triacs, and Quadracs. Being a struggling college student, I could not justify buying one of the new energy conserving flashes. By luck, Honeywell, who had a few years earlier invented auto-exposure flash photography, had run into a lot of competition. I was able to acquire several boards capacitors, reflectors and sensors from a surplus parts sale. I reversed engineered the units and built an auto flash.
 

Eventually, the inability to aim the flash separately from the sensor was limiting for bounce flash. I bought a unit from Spiratone called a Strobemaster. I actually still have this unit, but the battery holder contacts eventually broke from the stress of putting batteries in it all the time. I should probably say that Spiratone was able to repair my unit for a reasonable price after the hot shoe connector was damaged when it when through a loop-the-loop amusement park ride attached to the camera. I sometimes was unhappy with Spiratone, but they were quite reaonable with this flash.


 I next bought a unit from Kmart which was auto thyristor and had a zoom fresnel lens to adjust the spread angle.

Next I bought a Nissin 4500gte potato masher. It has everything including a safe trigger circuit compatible with modern cameras. It also has about 4 times the power of my earlier flash units. I also have several studio flash units which run on 110 volts. With all of the reverse engineering, I thought I would share a little about how these puppies work.

My most recent flashgun is a Yongnuo YN-565EX.

                                                My flash guns

I have mentioned the energy for a strobe being stored in a capacitor. Let's take a look at the simplest possible strobe circuit.

Here we have a 300-500 volt battery as a power supply which charges up C1 when the power switch is turned on. The Xenon lamp is connected directly to the C1, which is charged up to 300 volts. No the lamp does not light up. Why? Because 300 volts is not enough to start the discharge and light off the lamp.

Trigger Needed

Instead, the xenon gas needs to be ionized first. This is accomplished with a 5000 volt pulse outside the tube. Notice that C2 charges up slowly to about 200 volts (R2/R3 make a voltage divider that leaves 200 volts at their junction.) When the camera shutter sync contact is made, the energy in C2 (about 2 millijoules) rushes through the transformer terminals 1 and 4, causing a large voltage at terminal 5 which ionizes the xenon gas. This allows C1 to discharge through the xenon lamp. Once the lamp starts conducting, its impedance is about 2 ohms, but it changes in a complicate way with voltage so that more current flows at lower voltage. The "negative resistance slope" makes gas discharge lamps unstable, requiring a ballast for steady-state operation. Since all we want is a pulse of light, we do not need to be concerned about this.

So there you have it. The simplest possible photographic flash. This is pretty much what you got when you bought an electronic flash in the 1960s.

Drawbacks

After a little thought, you will probably notice several things that are not essential, but make the flash unit much more user friendly. Let's try to point these out one at a time, and show solutions to these inconveniences.

Lack of a "Ready Light"

When the power switch is thrown, the unit is not ready to operate immediately. For one thing, C2 must be charged up through some relatively high resistance. So we wish to have a ready light come on when C2 should be charged up. This is what the circuit looks like:

R4 and R4 form a voltage divider so that as C1 charges, the voltage on the neon bulb increases. The neon bulb requires about 90 volts to fire, so the voltage on C1 must be at least 120 volts for the neon bulb to be lit. Obviously this is not a full charge, but in this case, it is enough to fire the strobe. This is why many photographers will wait several seconds after the ready light comes on to use the strobe. A ready light does not mean a full charge.

For reference, here is the reverse engineered circuit of my very first flash, the Prinz Universal Tri-light.

If you ignore the battery and the transistor circuit, looking at only the 110V AC input, it is very similar to the circuits we have already seen. It has a two transistor oscillator (they are completely wired in parallel, so one larger transistor would have worked)  with a step up transformer to generate the 300-350 volts needed to fire the flash tube.

Auto exposure

Our strobe so far dumps nearly all available energy from C1 through the xenon tube. Suppose that this is too much light for the exposure. What does the photographer do? The only compensation that a photographer could use was to change the aperture stop to control the exposure. This generally precluded a bounced flash shot with any accuracy. So what if we point a sensor at the subject, but allow the flash to be pointed at a ceiling of a wall? We could get an auto exposure if the flash gets turned off as soon as enough light is received at the sensor. This was first accomplished with a quench tube:

The quench tube circuit looks complicated, but it really isn't. It operates very much like the xenon tube, but it has a lower impedance, so when it fires, most of the remaining energy in the main capacitor is diverted to the quench tube, depriving the xenon tube of any more energy. The quench tube is triggered just like the xenon tube, with a high voltage pulse from XFR2. The energy pulse comes from C6, and is driven by the CdS photocell. When the gate of the SCR is low enough, the SCR fires and sends the pulse through the transformer. But what are the extra resistors and capacitors for?

Remember, out trigger circuit has to work while the main tube is flashing. How much current is flowing? Well 300 volts through 2 ohms is about 150 amperes. That is a lot of current, so the ground in the strobe circuit is actually a number of volts above the normal ground level while all of the current is flowing. Therefore, we have to isolate the quench tube trigger circuitry from the normal ground to keep the pulse voltage from interfering with its operation.

Let's look at an actual circuit from about 1970.

My little Rolei's circuit is shown here:

We now have the essential elements of a fully functional flash. (We'll get to the modern flash with all of its power options, remote triggering and TTL in a moment.)

Let's briefly talk about some of the deficiencies in our circuit so far.

1. The power supply is not practical. 300 volt batteries are dangerous and costly. We solve the problem by building an oscillator that sends current through another step up transformer to a higher voltage. That way our battery can be 1.5 volts - 9 volts. We will look at some more low voltage battery power supplies in a moment..
   We have just seen this type of circuit in the Prinz strobe. We have also seen that you can use a voltage doubler and rectifier to get the 110 line voltage in the right ballpark.
   
    
2. So far, we have shown that the trigger current goes through the camera sync contacts. This trigger current was powerful enough to erode or in some cases destroy the camera synch contacts. All newer cameras will accept only a maximum of 6 volts and a few milliamps of current. Higher voltages will destroy a newer camera. The circuits you are seeing should never be used with a modern digital camera!
3. The exposure control we have shown so far wastes the remaining energy in the capacitor when the exposure is complete. This requires a maximum recycle time to recharge the capacitor for a new exposure, and wastes battery life.

We will address these deficiencies by improvements in their associated circuits.

Power Supplies

I am a bit lazy so I am going to borrow some graphics that others have posted on the web. Here is a power supply represented in text form (a bit hard to read, but you can do it.)

                                                                       J2-2
                                             +--------------------------o HV+---+-----------------------------
                                             |  S2 Flash +-----------+          |      R4
      P1-1  S1 Power            o  T1        | Intensity |     R2    |          +----/\/\-----+---+
   BT- o---+---/ --+-------------+     o     |      High o   330,2W  |               4.3M     |   |    
           |       |              ):: +------+-+-------+->o---/\/\---+                        /  +++ IL1
           |       /        D 15T )::(         |   Low | o     R3    |            Ready R5 +->\  |o| NE2 
           |    R1 \          #20 )::(         |       +--+---/\/\---+             Cal. 5M |  /  |o| Ready
           |   150 /   +---------+ ::(         |          |  10K,1W  |                     |  \  +++ 
           |       \   |           ::( O 1950T |          |          |               R6    |  |   | 
           |       |   |        o  ::( #46     |      C2 _|_     C3 _|_         +---/\/\---+--+---+    
           |-      +---|---------+ ::(         |   260uF ---  260uF ---         |
     BT1   _       |   |          )::(         |    350V  |    350V  |          |
    2.4V  ___      |   |    F 15T )::(    D1   |    D2    |          | J2-1     |
    Sub-C  _       |   |      #30 ):: +---|<|--|-+--|<|---+----------+--o HV- --+--------------
     NiCd ___      |   |     +---+       BAY90 | | BAY90
           |+      |   |  Q1 |                 | |
           |       |  C \|   |                 | |
           |   C1 _|_    |---+                 | |  O = Output
           | 47uF --- E /| AD136               | |  D = Drive
           |  10V  |   |   (PNP)               | |  F = Feedback
           |       |   |                       | |
      P1-2 |       |   |                       | |
 COM  o----+-------+---+-----------------------+ |
                                                 |
      P1-3                                       |
 CCAC o------------------------------------------+

Reference

Vivitar Auto 253 Power Supply

                   AC   D3
               +-o IN o-|>|--+
      S1       |             |
    DC  AC     |      D1     |X D2                                   Flashlamp
  +--o  o +-o  o +----|>|----+--|>|--+----------+-------+-------------------+
  |    /..|.. /  |           |       |          |       |               FL1 |
  |   |  +|  |   |           | LT1   +          |       |                   |
  |   |  ___ +---|------+   +++   ::(           \       \                  +|
  |   |   _      |      |   |o|   ::( L1        / R5    / R3               _|_
  |   |  ___ BT1 |      |   |o|   ::(           \ 1.2M  \ 3.3M            | | |
  |   |   _  9V  |      |   +++   ::(           /       /       Trigger  ||   |
  |   |  -|      |  C3  |    |       +          |       |                ||   |
  |   +---+ T1 +-+--||--+    |       |          |       |  C2      T2 +--||   |
  |         ::(3    220 |    \ R2    |          |       +--||--+   ::(   ||   |
  +-------+ ::(     pF  |    / 1,2M  | Energy   |       | .047 |   ::(   || _ |
  |       2)::( 118     |    \       | Storage  |       |  uF  +-+ ::(    |_|_|
  | R1 <.1 )::(         |    |       | 380 uF   | Ready |         )::(      |
  / 4.7K   ):: +---+    |    |      +| 350 V   +++      | Shutter )::(      |
  \    +--+ ::(5   |    +----+     __|__       |o| IL1   |-       )::(     -|
  /    |  4 ::( .2 |         |     _____ C1    |o|      | S2   +-+    +-+   |
  |    |       +--------+    |       |         +++      |      |        |   |
  |  |/ C      1   |    |    |      -|          |       +------+--+-----+   |
  +--|    Q1       |    |    |       |          |       |         |         |
  |  |\ E 2SB324   |  +_|_C4 |       |          |       / R4     _|_ C5     |
  |    |           |   ___   |       |          |       \ 3.3M   --- 100 pF |
  +----|-----------+  - | 10 |       |          |       /         |         |
       |  Inverter      | uF |Y      |          |       |         |         |
       +----------------+----+-------+----------+-------+---------+---------+

Please note that it is customary in American schematic diagrams to show resistors as wiggly lines but in England it is customary to show them as long open rectangles. Also coils in the US are shown as several loops, while in the UK that are shown as solid rectangles.

Panasonic PE-280C

Reference

Here is the full schematic of the Honeywell Strobonar 480S unit that I got in the form of spare parts.

(The following is from: Kevin Horton (khorton@tech.iupui.edu))

Note in particular the old Vivitar 253 was also line powered from a 220 volt AC line. Have 110 volts? No problem, use a simple voltage doubler.

The trigger

We noted that the trigger in early flashes dumped the current for the trigger transformer right through the camera contacts. One way to get around this is to run that current through an SCR, and only let the small current of the gate go through the camera. I modified my first flash unit in 1976 to do this, only I went a step further. I wanted to also use it as a slave flash, so I used an LASCR. It worked well and prevent further damage to my old SLR camera.

This material is from http://www.mts.net/~wrpa/Quench.html

Basic flash circuit without high voltage generation. For speedlights the photoflash capacitor is charged to about 300 volts. When a flash tube conducts the flash tube's resistance could be about 2 Ohms. So for a brief time you could have 150 Amps going through the circuit.
     

When the trigger circuit is closed by the camera's sync connection (hotshoe or PC circuit) the charge on the small capacitor in the primary side of the Trigger transformer steps up the voltage to about 5,000 volts on the secondary side. (green path) This voltage ionizes the gas in the flash tube making the gas conductive. With the gas conductive the charge on the photoflash capacitor discharges through the flash tube and produces light. (red path)

With the discharge through the photoflash capacitor voltage keeps dropping. At about 60 volts the flash tube extinguishes. When the flash tube starts to conduct it can not be turned off. It can only self extinguish when the voltage drops too low.


Simple flashes that have only a full power output operate this way. They may have some additional electronics to reduce the trigger voltage. Most studio strobes are like this as well.

The following was used in earlier Auto hotshoe flashes. A sensor was used to determine when there was enough light for an exposure and having sensed that point the remainder of the charge still on the photoflash capacitor was dumped to a Quench tube, throwing away the remainder of the charge.

The trigger circuit is triggered to generating the ionizing voltage causing the flash tube to conduct (red path). Once the exposure value was reached the Quench tube (also requiring a high voltage ionizing voltage) provided an easier path (closer electrodes) and diverted the current away from the flash tube. 


While the Quench Tube allowed exposure control it was not a very efficient circuit. The remaining charge on the photoflash capacitor was wasted. Every time the flash was fired the recharge time would always be as long as when a full power flash was called for.

A "thyristor flash" had an electronic switch circuit that would be able to interrupt the current through the flash tube. With a series switch in series with the flash tube the discharge of the capacitor could be interrupted and the unused charge still on the photoflash capacitor could remain there and not be wasted as was done with the quench tube circuit.

The Thyristor circuit required a more complex circuit. (a simplified version is shown below). The thyristor (Silicon Controlled Rectifier - SCR) was in some ways like a flash tube. Once triggered it would continue conducting until there was not enough current available to keep it turned on. You could trigger it but you could not shut it off. So a clever technique was used to momentarily divert the current from the main path SCR in order to allow it to turn off.


As with the other circuits above the process starts with the trigger signal generating the voltage to ionize the gas in the tube to make it conductive. A small signal was created to trigger the SCR in series with the flash tube. So with the flash tube turned on and SCR2 turned on this allowed the charge on the photoflash capacitor to follow the red path shown below.

Once the sensor picked up enough reflected light from the scene the quench signal turned on SCR3. This allowed the current to follow the blue shown below. This starved SCR2 in the main path and it turned off. But because of the small capacitor C in series between the bottom of the flash tube and the top of SCR3 that current path was only momentary until capacitor C got charged up. The blue path was choked off and the remainder of the unused charge on the photoflash capcitor remained there.

Thyristor circuits are now obsolete since Insulated Gate Bipolar Transistors IGBT became practical for flash circuits. Unlike the thyristor/SCR the IGBT has a switch input rather than a trigger input. A control signal is required to turn it on and keep it on. Once the control signal is removed the IGBT will turn off.


This is about as complicated as it gets. The trigger signal would come from the microprocessor in a current flash. The above Thyristor examples left out a number of parts to simplify the explanation.

 

Insulated Gate Bipolar Transistor The following material is from a Fairchild application note.

This represents a modern flash block diagram. It does not show the receiver for signal from the camera, the control for the motorized Fresnel lens to set the zoom angle, the auto focus illumination assist beam, or other high end functions.

This is a typical flash circuit using a Fairchild isolated gate bipolar transistor. Please note that IGBTs are not currently suitable for high power strobes.

This is a demo flash that Fairchild engineers built to demonstrate that their IGBT is suitable for a photographic strobe use. Note that at 290 μF the maximum strobe energy is 12 watt-seconds, and is is more likely 9 watt-seconds. This is not much more energy than one of the built-in strobes in the camera body. It is definitely a lightweight.

How an SCR is quenched.

From Doug Smith

http://home.comcast.net/~dougsmit/bounceflashtoys.html

http://photocamel.com/forum/lighting-technique/29010-home-made-flash-bracket.html

http://photocamel.com/forum/lighting-technique/29713-flash-diffusers-cheaper-answers.html

I shot a test series showing the effects of several versions of home made flash diffusers. For some time I have been expressing the opinion that any diffuser is better than none but this series does show differences in terms of degree of shadow and amount of fill on faces.


They are named for the recycle bin trash used to make them:

 

Direct	Bounce	Cut	Solid
BBQ	Syrup	Paper	Bubble

Direct: Direct flash on camera
Bounce: Bounce flash no diffuser
Cut: Milky plastic diffuser with bottom removed (baby juice bottle)
Solid: Milky plastic diffuser with bottom intact (baby juice bottle)
BBQ: Clear, thin plastic bottle (Barbecue sauce bottle)
Syrup: Clear bottle with silver reflector on back (Pancake syrup bottle)
Paper: Plain white paper reflector behind flash
Bubble: Bubble wrap bag
Except for the direct flash image, all were taken with the flash head pointed up.

I own NO commercial diffuser so I can't answer how this test would differ if it included a Fong or Omnibounce. My total expense on the project is less than either of the above and I got the contents of these bottles as a bonus. I learned a lot in the process. I would love to see someone post a similar series using all of the commercial units.

Overall my favorite is the baby juice bottle but I flop back and forth between the cut and uncut one according to the subject. I should have done these in portrait mode to make the shadows more evident.

Solid (left) Cut(right)

Both shown here were made from plastic bottles that once contained Delmonte brand baby apple juice. They are soft plastic and milky. Properly cut, they slide easily over the flash head and get rotated 90 degrees to hold firmly in place. One (on the ball head) also had the bottom of the bottle cut out so more of the flash escapes upward to bounce off of the ceiling. This works better in a room with a white ceiling. The bottle (on camera) with the bottom intact throws more light (in proportion) forward and can cause more of a shadow behind the subject but it works better if the ceiling is not good for bouncing.

Syrup

The diffuser shown on the macro rig [...] is very specialized made from a clear (pancake syrup) bottle with a piece of silver tape on one side to throw more light down on the nearby subject. It wastes a lot of light but softens the shadows compared to direct flash. It was made just before I shot the image so it happened to be on the flash at that time. I'm not sure it offers any advantage over the plain bottles shown on my page. Part of the 'game' is trying enough things to see which have merit and which are 'also ran'. Any diffuser (including the $50 commercial ones) give better results than direct flash.

Bubble

Paper

Milk (left) BBQ (center and right)

Universal Bounce Diffuser

This type of diffuser does nothing but bounce some light off the walls and allow some to go to the subject. If the walls were painted black, its only effect would be to reduce the brightness of the flash.

Here I am using a diffuser that I got designed by Gary Fong, but, it does little more than a Universal diffuser. It seems to be more efficient than the milky plastic type, but I have not done any real tests to see.

I also found these same items in the Spiratone 1991 Studio Lighting Catalog, under the name " air-flector" for $14.95

	Direct builtiin flash on the camera body.
	Direct flash handle mount ("potato masher")
	Direct flash with wide angle diffuser
	Bounced flash, 45 degress up to ceiling, no swivel.
	Bounced flash to ceiling. 45 degrees up and swiveled 90 degrees to the left.
	Bounced flash to ceiling. 45 degrees up and swiveled 90 degrees to the left with flip-it card.
	Bounced flash to ceiling. 45 degrees up and swiveled 90 degrees to the left with flip-it card. Card intercepting minimal light.
	Bounced flash to ceiling. 45 degrees up and swiveled 90 degrees to the left with flip-it card and flip-it diffuser.
	Home made giant bounce card

This jibberish is the opinion of Bill Otto and does not necessarily represent the views of any photographic organization.
Copyright © 2013  [Bill Otto]. All rights reserved. Certain graphics have been lifted from other web sites. If I did not give you credit, please email me.
Revised: April 08, 2015 .