High speed flashes.

[Sceadwian]

Was browsing around today, and found a DIY page for high speed photography. I've always known the flash duration was the limiting factor for this type of photography, but this page has some actual oscilloscope images of the flash duration as measured from a photo diode from a decentflash and I was trying to understand how these waveforms are related to the circuitry.

The Sigma EF-500 flash - Light duration vs. output power | DIYPhotography.net

Starting with the slowest one it seems pretty straight forward, the initial rising lump is the start of the arc in the flash tube which peaks as the arc is fully formed and becomes fully conductive, and slowly sloping drop is consistent with a capacitor discharge curve, so regardless of the power level the flash is charged to the same voltage, but with shorter flash durations it seems that there's a timed cutoff (quiet abrupt) of the discharge. The 1/128th power level doesn't even get to the point where the arc is fully formed.

What in this circuit are they using to so abruptly stop the discharge? On the highest speed one's there seems to be a slight knee but this looks like parasitic capacitance. I'm wondering because the voltage of a typical flash is a few hundred volts from a relatively large capacitor and the cutoffs I see in these scope shots are nearly like hitting a wall. I can't even rule out that the scope shots knee curve for rise/fall times on the highest speed settings are from the photo-diodes junction charge.

 

[cowana]

I'm under the impression that the light stops when the arc can no longer be sustained - and that's a pretty sharp cutoff.

Andrew

 

[Sceadwian]

cowana... I already stated exactly that. Do you have any insight into the circuitry that is being used to cut off this discharge so sharply which is the question I was asking?

 

[cowana]

In cheap flashes at least, there is no cutoff circuitry - simply when the voltage falls below the voltage needed to sustain an arc, the arc collapses and stops giving out any light. That happens very quickly. Must cheap flashes just use a triac to trigger the discharge, and then the actual xenon tube stops it.

 

[Sceadwian]

How do you explain the brick wall? The fact that the values measured are consistent with the original discharge up until the cutoff point?
The graphs depicted show a very obvious capacitor discharge curve, SOMETHING is stopping this normal discharge, my question was what?

This was not a cheap flash, a market value module was probably several hundred dollars at the time of purchase.

 

[cowana]

The voltage falls below what is needed to sustain an arc in the xenon, the arc can no longer be generated, so it collapses and the light output falls to zero.

At that's how cheap ones work - I've never done any work on expensive ones before.

 

[Diver300]

To trigger a flash tube, a very large voltage is needed to ionise the gas. That is usually capacitively connected by a wire around the outside of the tube and the voltage comes from a pulse transformer. In cheap flashes the capacitor is directly connected to the tube, and there is nothing to stop the arc so it carries on until the capacitor is discharged.

To stop the light, the current has to be stopped. That was done with a thyristor and a commutator circuit . I don't know if there are better technologies now. I guess that a Gate Turn Off, or GTO thyristor, would be used.

 

[Sceadwian]

The graphs show they definitely use something that cuts off the discharge, that's what I want to know about.

 

[Sceadwian]

GTO thyristor looks like it might fit the bill, that's a SHARP cutoff though.

 

[Reloadron]

This may be the answer. Scroll down the page to Variable power intensity and look at the how different caps are used for different settings of power. Though, that is one sharp cutoff. Scrolling way down to the comments there is mention of timing the SCR or an IGBT.

Ron

 

[Sceadwian]

Would you please look at the graphs again?
The rise curves and peaks are IDENTICAL for EVERY single last image if you overlay them at scale.

This by every frame of that graph screams it's a single cap discharged through the tube with some time delayed cutoff. All I want to know is how that cutoff works.

 

[Reloadron]

If you scroll down that link and muddle through the comments they mention SCR (won't work) and IGBT. I did see a circuit where a thryistor was used. That made sense. I vote with a thyristor for the scope images presented. They are sharp.

Ron

 

[RMMM]

Remember, it may not be only the circuit.. It may also be HOW it is tested. A photo-diode oscilloscope, is a photo diode connected to the BNC connector.

So what you are seeing IS NOT scoped from the circuit. It is SCOPED from the "receiving" photo diode.. And that is not a very sharp cutoff for a diode.

 

[crutschow]

To abruptly terminate a flash, you could connect a second SCR across the storage capacitor. Firing this SCR would rapidly discharge the cap and terminate the flash.

 

[RMMM]

That would be a feat..

You would need a big route to ground to "bleed" that cap off mighty quickly. A few small, fast scr's could prolly handle it.

Or mosfets to disrupt the current to the flashbulb. A proper mosfet driver can have 19ns turn off times.. Probably faster..

 

[KMoffett]

I'm with Sceadwain. I saw a patented circuit where a second SCR is placed in series in series with the flash tube. It was triggered on at the same time as the trigger-transformer SCR. To cut of the flash, a third SCR shorts a small capacitor, charged with a reverse voltage, across the A/K of the second SCR, cutting it off. At least that's how I remember it.

Ken