Electronic Flash Testing and Equipment?

[Ralph Javins]

Good morning, APUG;

Anyone know where to find information on how to test photographic electronic flash units for output level and color temperature?

I have photographic flash light meters that can test the light output of strobe lights or electronic flash units with Xenon gas discharge flash tubes. There is also a flash color meter that can determine the color temperature in "Degrees Kelvin" of the light output. Is this enough to determine the functionality of a photographic electronic flash unit?

The photographic flash light meters will tell us if the the flash tube and the electrical power circuits are generating their rated power output to give full rated light output. The flash color meter can tell us if the flash tube is aging and its color temperature has begun to shift downward toward the red end of the visible spectrum. Is this enough to reliably tell the current state of health of a photographic electronic flash unit?

Does this seem to be a reasonable approach for testing of our electronic flash systems? And, yes, I did make a search of the past postings for testing of electronic flash systems. There are references to ways of testing light output, the use of 18% gray reflectance cards illuminated by the flash units, and some similar things, but no commentary on testing for color temperature.

 

[RalphLambrecht]

Since I do exclusively B&W, color temperature is not much of an issue for me, but color temperature will get warmer with the age and use of the flash tube (I assume). The other thing I like to test for is the flash duration. That can be useful to know for a variety of reasons.

 

[Chan Tran]

A Flashmeter and a color flashmeter should be fine to test the electronic flash. Grey card is not neccessary as these meters are basically incident meters. Testing for flash duration would need different instrument all together. I am thinking of a very fast photocell and then measure its output via an oscilloscope.

 

[Ralph Javins]

Good morning, Ralph Lambrecht and Chan Tran;

Your comments about flash duration are well taken, and have been considered also. That is easy to do with an Oscilloscope or a Frequency Counter in Interval or Period Mode and a photo-transistor or photo-diode such as the Texas Instruments 1N2175, which has a response time well into the Megacycle range, and it would allow us to measure the flash duration in microseconds. One nice thing about the oscilloscope is that we can configure the test to show us the shape of the light output curve; something a frequency counter cannot do. Flash Duration is easy to measure, but I am not sure how to interpret that information as to its relationship to the working of the flash unit and projections for its future life.

 

[Chan Tran]

I wouldn't do it with a counter though as the intensity of the light isn't constant. It peaks rapidly and then fall off a little less rapid. One has to see it the graph on the scope to tell when you consider that the light fall off too low to consider it as part of the duration.

 

[ic-racer]

The only way to know for sure about the color balance is to test wth the slide film you will be using and some typical subject matter. In terms of rated output, so what if it is a little low, the bulbs can be pretty expensive. The output will drop as the bulb ages, but as long as it fires consistenty, I'd be happy unless you are selling it or doing some scientific experiement where you need to know the output.

 

[lynux]

Good morning, Ralph Lambrecht and Chan Tran;

Your comments about flash duration are well taken, and have been considered also. That is easy to do with an Oscilloscope or a Frequency Counter in Interval or Period Mode and a photo-transistor or photo-diode such as the Texas Instruments 1N2175, which has a response time well into the Megacycle range, and it would allow us to measure the flash duration in microseconds.

Q1: are you using the photo-diode in voltage mode, or in current mode?
(the response times might be different)
Q2: flash duration on a digital storage scope sounds ok, if you are using a
period/frequency counter, you need to define a precise triggering level, this
might be very difficult to set?

if its getting too technical, you can also reply per pm. thanks.

 

[Sethasaurus]

I was reading about flashes the other day and came across this site:
Dead Link Removed


I use a phototransistor connected to the mic input on my laptop soundcard to measure shutter speeds. This typically gives me two peaks (one inverted at the point where the transistor stops/starts conducting).


So, I guess if you're interested in measuring the 'T.1' time of a flash, you could use a photodiode or something with a predictable response to give you a waveform that looks like the one on the site above.


According to that info, the AlienBees B400 at full power gives a T.1 duration of 1/2000s. The waveform should be easily sampled by the average soundcard ADC and you'll be able to measure it in audacity or a similar wave editing app.

 

[Mainecoonmaniac]

Just shoot some chrome film of a Macbeth color checker, process it and look at the transparency on a daylight balanced light box then work from there. Some high end strobes will look more neutral. Some strobes will look bluish. I assisted some photographers that geled their strobes to warm up the color temperature. No need to buy special gear. If you want to do it the cheats way, use a digital slr. Open the image in Photoshop and use the eye dropper on a middle grey patch and read the RGB value. Ideally all values should be the same. If the red value is low, chance are the strobe is bluish.

 

[Chan Tran]

I was reading about flashes the other day and came across this site:
Dead Link Removed


I use a phototransistor connected to the mic input on my laptop soundcard to measure shutter speeds. This typically gives me two peaks (one inverted at the point where the transistor stops/starts conducting).


So, I guess if you're interested in measuring the 'T.1' time of a flash, you could use a photodiode or something with a predictable response to give you a waveform that looks like the one on the site above.


According to that info, the AlienBees B400 at full power gives a T.1 duration of 1/2000s. The waveform should be easily sampled by the average soundcard ADC and you'll be able to measure it in audacity or a similar wave editing app.

I think the sampling rate of the sound card or 44 Khz isn't sufficient to have a meaningful graph of the flash duration. I doubt it at 96Khz like some cards are capable of.

 

[Ralph Javins]

Q1: are you using the photo-diode in voltage mode, or in current mode?
(the response times might be different)
Q2: flash duration on a digital storage scope sounds ok, if you are using a
period/frequency counter, you need to define a precise triggering level, this
might be very difficult to set?

if its getting too technical, you can also reply per pm. thanks.

Good morning, Lynux;

For the measurements we are taking in this application, the difference in the timing resulting from the way that we are setting up the test (voltage mode or current mode) will not be that significant. We are looking at something that will be measured in the region of a few milliseconds -- single digits -- and the response time of the phototdiode is at least in the microsecond range, if not nanoseconds. That is at least three (3) orders of magnitude different than the minimum response time of the testing unit element. If we are looking at the 10% level for our trigger and measurement points, we are definitely in milliseconds. If we are looking at the 50% level, then we will probably be down in the one or two millisecond range with an electronic flash unit at full power with a full dump from the flash capacitor. If we are looking at a partial power flash, then we may be down around 50 microseconds probably at the shortest. Then we might start looking at the effects of the timing capabilities of the components we are using. The 10% point and the 50% point seem to be the two levels where the photoflash people measure the duration of their electronic flash units. In this application, I do not think that we are approaching any real limitations of our test equipment. I know that my oscilloscope is rated to go to 200 Mc, so that will not be a limitation in showing the light output curve.

In any case, I am still interested in suggestions or references on how to test these things. I have worked in electronics all my life, but I admit that I am not really familiar with testing techniques for electro-optic devices.

I do admit that I am going by my memory of the specifications from the middle 1960s for the response time of the 1N2175, but I have not yet been able to locate a full data sheet for the photodiode. My TI Data Books from that time period disappeared with the death of a marriage.

 

[Sethasaurus]

You will easily be able to measure a pulse of the order of milliseconds with your average soundcard.
You will amost certainly be able to capture and measure the trace on a halfway decent scope.

The old quote holds true - "Can't never could".

 

[Nicholas Lindan]

If you want to measure time accurately with a photodiode you are going to have to reverse bias it and measure the current.

A 10K resistor for current sensing and a 9v battery for reverse bias will work, as will many other arrangements.

I don't know of any working EEs who still maintain a bookcase of old data books.

A '2175 is an old, old part. You might want to try something a bit more modern like a BPW34 [not all that modern, TTtT] or an OP950 if you want something faster.