Checking focal plane shutter fast speeds with a fluorescent tube

[snusmumriken]

I have read* that the shortest speeds of a typical 35mm focal plane shutter can be checked using a fluorescent light as a source. Allegedly, three diagonal lines should be visible at 1/1000 sec, two at 1/500. sec. So the principle is similar to the famous Leica shutter speed tester. I have tried this, and although I do catch a glimpse of the fluorescent tube, I don't see any lines at either speed.

I presume the idea is that fluorescent tubes flicker. But I thought the flickering reflects the frequency of the AC current, which is ~50Hz here in the UK. 50Hz would be 0.02 seconds per cycle, and of course 1/1000 sec is much briefer at 0.001 sec. So I don't understand how this could work anyway.

Can anyone cast any light on this? (Sorry, pun not intended.)



* The source was a pdf that someone posted here on Photrio a while back. It is Section 11 of some book or other.

 

[4season]

Electronic ballasts and LED retrofits won't necessarily flicker at 50 Hz.

 

[snusmumriken]

Electronic ballasts and LED retrofits won't necessarily flicker at 50 Hz.

Could you expand a bit on that for a non-electrician? The lamp I tried has one of those cylindrical plug-in starters, so I don't think it has electronic ballast. Deffo not an LED retrofit.

 

[Chan Tran]

I don't know if it would work. But if it does work I have an idea of use the LED and have it flashes at a controlled rate. It's fairly easy to flash the LED.

 

[snusmumriken]

I don't know if it would work. But if it does work I have an idea of use the LED and have it flashes at a controlled rate. It's fairly easy to flash the LED.

As far as I can see, it would require the LED to flash at 3,000Hz. Is that possible?

 

[Alan Edward Klein]

Has anyone every tried a video mode that records quickly like my Sony RX100 that has a highest speed of 960fps?

 

[reddesert]

The relevant number is the focal plane shutter curtain travel time, which is roughly the sync speed, typically 1/30 - 1/125 sec depending on camera. The high speeds are made by a narrow slit. So if the sync speed is slower than the nominal light on time of ~ 1/100 sec (~ half a cycle), there should be banding in the image, for the same reason that electronic flash used above the sync speed exposes only part of the image. You might not see this banding by eye well due to persistence in the lightbulb or your eye. However, I don't see how one would get diagonal lines without some sort of slit or light source traveling cross-wise to the shutter, which is how the Leica tester works.

 

[Chan Tran]

As far as I can see, it would require the LED to flash at 3,000Hz. Is that possible?

Yes! I have seen LED flashes at 40,000 times per second

 

[snusmumriken]

You might not see this banding by eye well due to persistence in the lightbulb or your eye.

Hmm, of course there is some kind of coating inside the tube, precisely to make it more persistent.

However, I don't see how one would get diagonal lines without some sort of slit or light source traveling cross-wise to the shutter, which is how the Leica tester works.

That makes me think. Presumably any pulse waves travel along the tube? So perhaps I need to orient the camera so that the tube runs top to bottom of the shutter? I’ll nip out to the darkroom and try this shortly in case it makes a difference.

I’m not holding out great hopes - it’s the principle that interests me, and whether one could make a Leica-style tester using modern electronic components.

 

[snusmumriken]

Has anyone every tried a video mode that records quickly like my Sony RX100 that has a highest speed of 960fps?

Yes, someone posted images in another thread a few weeks back. But it’s this particular approach that I wanted to explore here.

 

[Sirius Glass]

Electronic ballasts and LED retrofits won't necessarily flicker at 50 Hz.

Or 60 Hz in North America

 

[snusmumriken]

That makes me think. Presumably any pulse waves travel along the tube? So perhaps I need to orient the camera so that the tube runs top to bottom of the shutter? I’ll nip out to the darkroom and try this shortly in case it makes a difference.

It didn't.

 

[DWThomas]

[ ... ]I presume the idea is that fluorescent tubes flicker. But I thought the flickering reflects the frequency of the AC current, which is ~50Hz here in the UK. 50Hz would be 0.02 seconds per cycle, and of course 1/1000 sec is much briefer at 0.001 sec. So I don't understand how this could work anyway.

Can anyone cast any light on this? (Sorry, pun not intended.)

* The source was a pdf that someone posted here on Photrio a while back. It is Section 11 of some book or other.

The "flicker" of an old-time fluorescent tube driven by AC via a magnetic ballast is actually at twice the line frequency. The current flow, via mercury vapor in the tube, is bi-directional and the light is loosely proportional to the current.

In my "flaming" (literally! ) 1950s youth I modulated DC current through a 20 watt fluorescent tube using a hefty tube ("valve") used to drive horizontal scans in TVs. A TV power transformer with a voltage doubler provided about 700 volts DC for a power supply.** I managed to receive music across the basement using a selenium solar cell that was sold in hobby stores, mounted in a parabolic reflector, to pick up the signal. I have no idea what a frequency response may have looked like, but a relatively decent amount of treble came through. So likely the frequency response is not an issue.

One old classic used to be that to take a screen shot off a TV any shutter speed faster than 1/30 sec would have uneven results. But that's in CRT days, all bets are off in this century.

A moving LED being pulsed at a known rate would be fun.

** "Don't try this at home"