Fluorescent Banding in DIY light vs CoolLight

B

brynsank

Active member
I have built a 10x55w biax fluorescent light based on Fulham Workhorse ballasts (operating at 50-60kHz) - see attached image2012-05-17 18.46.01.jpg. I also own a 6x55w CoolLights StudioFlo (the dimmable version) and have tested the two side by side. See this video for the results (shot on Panasonic HPX250): https://vimeo.com/43185825

You can see that there is banding from my DIY light at any shutter speed that is not a multiple of 60 (besides 1/30, of course). It is only barely detectable at 1/100, but it's there. 1/120 is completely clean though. However on the CoolLight, there is no banding at ANY shutter speed - why is this, and what can I do to achieve this?

The bulbs in my DIY light are KinoFlo TrueMatch 5500k. In the CoolLight, they are the same bulbs as it was shipped with.
 
brynsank said:
I have built a 10x55w biax fluorescent light based on Fulham Workhorse ballasts (operating at 50-60kHz)
Click to expand...

And that's your banding problem. The CoolLights units use an electronic ballast running at much higher rate (something around 40,000 Hz IIRC). So to the camera, the CoolLights units look like continuous light -- at any frame rate, and any shutter speed. If you want this performance, you need a ballast that will fire the lights at a much higher rate.
 
Bruce, thanks for your quick response. However is not 60 kHz equal to 60,000 Hz and thereby faster than the 40,000 Hz frequency of the CoolLights ballasts? This is what is confusing me - according to the specs they are firing at 150% the rate of the CoolLights, but yielding poorer results.
 
Beyond specs, testing is the number one thing to consider when choosing a ballast for such a fixture. Just because a ballast lists a high frequency rate doesn't mean that all the bases are covered and you're home free. I've tested lots of ballasts including some from Advance which are very good and they only run at 25Khz and exhibit no such problem.

I have heard that there is a ripple problem of some sort in these Fulham ballasts, more than one has reported it. Its just an example of what could be done wrong in a ballast but not necessarily detected. 50 - 60,000 hz is more than enough to not feel there is no flicker to the human eye. The ballast was not specifically made for video/film use so testing of the possibility of some kind of sub-frequencies rippling and resonating may not have been considered an issue by the designers.

How well does that ballast do what it says it does is the main question. Specs can be wrong, etc. Testing is the only thing that in the end can tell you whether it will work for this kind of application. By the way, it only takes 1 bad ballast in a multi-tube fixture to make the whole thing seem like it doesn't work. 1 ballast causing flicker could make the whole thing seem as if it was flickering. Isolation of each circuit would ultimately eliminate that as a possibility.
 
Last edited:
Interesting, thanks for your response Richard. The light has 5 ballasts, each with its own switch so I can test each one individually. I had originally hoped to purchase the 2x55w ballasts listed on your website (http://www.coollights.biz/t5ho-universal-voltage-ballast-p-98.html), but it says they are currently out of stock. I went with the Fulhams based on online research. Do you have a recommendation for an alternative I could try? I'm looking to buy about about 15-20 of them to build a number of other fixtures. This first light is my test case.
 
The last time I did major tests on a lot of ballasts was about 7 years ago. I liked the Advance and Universal back then among others. A lot of ballasts I never considered because they were 110v only or 220v only types and we always wanted universal voltage input from the start. That eliminated Osram and Philips which was a pity. That's also why I never considered Fulham for example and it was never part of our tests, they were 110v only.

The problem with drawing on that information from the past is that things can change in 7 years. Just look around and buy one and test before you buy more. We normally sell ballasts but we're out of stock on them at the moment.
 
Last edited:
Very interesting. I'm going to go test them out individually right now. Thank you for taking the time to respond.
 
brynsank said:
Bruce, thanks for your quick response. However is not 60 kHz equal to 60,000 Hz and thereby faster than the 40,000 Hz frequency of the CoolLights ballasts? This is what is confusing me - according to the specs they are firing at 150% the rate of the CoolLights, but yielding poorer results.
Click to expand...

My bad -- I read that as 50-60 Hz. As in, line frequency. You're right, that should easily be sufficient for your needs. Hmmm....
 
Well I just tested them individually and they all yield about the same amount of banding, so Fulham Workhorse are officially struck of my list of options. Interestingly, after I was done with the test I turned on the ceiling lights (T5 fluoros) and tested those out. No banding whatsoever, at any shutterspeed. And here I was thinking that the regular residential stuff wouldn't be sufficient... I guess I'm off to Home Depot to see what they have there.
Thank you all for your input.
 
You must log in or register to reply here.
Back
Top