D.I.Y. 900-watt CF Lighting Units

[Martin Catt]

Just finished a pair of these for use in an upcoming documentary project. They've been a spare-time construction project for about three months. Each light has nine 100-watt equivalent daylight-balanced CF lamps. A pair of aluminum 9x9" baking pans nested together form the main housing and reflector assembly. Nine surface-mount phenolic sockets mount to the rear pan, with holes drilled through the second to pass the screw bases of the lamps.

The hard part was making sixteen custom hinges for the barn doors. I looked, but couldn't find a suitable hinge, so I ended up spending a Saturday making them from aluminum angle stock.

The yoke mounts are machined from 1" aluminum bar stock, and the yoke is bent from 1" x .125" aluminum extrusion. The mount from the yoke to the light stand is also machined from 1" aluminum bar.

A standard electronics project box houses the switches, mounted on four stand-offs from the rear pan. This gives a convenient place to wind the 20' electrical cord, and keeps the switchbox cool. I wasn't sure how hot the unit would get, so I decided to play it safe. Both units have been on for over 30 minutes, and only get warm to the touch.

There are two switches, one for on/off, and for high/low. High uses all nine lamps; low uses only five lamps, giving roughly a 1-stop range. The lamps are not dimmable, but the intensity can be tweaked a little by removing individual lamps as needed.

Rough guess of cost is about $50 each for materials, lamps not included. They fit perfectly on a pair of Novatron light stands left over from my still-photo days.

Regards;
Martin

[Joe Bowey]

I'm glad I came upon this post. I had just changed the lightbulbs in my ceilinf fan to cfs from costco and noticed it not ony being brighter but the color on my table seemed more natural. So I was thinking of making a light set also. I am doing a doc and really can't afforad nice lights so this might be a option.
How does it look on camera? I use the jvdhd100 camera. Also have you made a soft box for it?
Excellent DIY.

[Richard Andrewski]

Nice job. Did you have the yoke mounts made somewhere else or do you have some machining tools? Looks like an expensive job.

We engineered a part called the "Universal Yoke Adapter" to make things like this easier on people...

http://www.coollights.biz/female-uni...pter-p-42.html

Picture is also enclosed below. We made them in 1/4-20 and 5/16-18 sizes. These just basically rivet or screw to the side of your case with the outer two mounting holes, then add a clamping bar or T handle which screws into the inner threaded hole and a rubber washer for some added grip and braking action and you have a great yoke mount. Here's a video describing their use:

http://www.coollights.biz/wordpress/archives/22


A few other suggestions for some wanting to try a project like this.

Those spirals function best with a bit of room around them. If they are too close to each other, the light output is not as efficient as with some space around them. Thus, use a bigger baking pan.

For the barndoors, I showed a method to make these in my Cool Lights Video Magazine earlier in the year:

http://www.coollights.biz/wordpress/archives/21

Just use around 22 to 24 gauge and make a template.

One way to make this unit its own "softbox" would be the following additions (although it is already very "soft" because of the surface area). First, get some thin gauge mirrored/polished aluminum and put that on the doors with rubber cement. Alanod makes the best grade known as "Miro" but just check with them and find where you can buy a small quantity for your project(s):

http://www.alanod.com/Miro/index.html

We use Miro in our reflectors and it really is the best material. Just cut it to the exact size of the doors. If the aluminum is too hard to find, the next best thing is to use reflective mylar like this--which is a bit easier to work with too:

http://www.hydroponics.net/c/54

Once the barndoors are coated with such a surface, they become your intensifiers or added reflectors then. You've just increased the surface area emanating light now. There are some gaps though in between the barndoors which make it not a perfect larger surface.

Going a step further, and making some aluminum corner pieces that velcro on the barndoors thus making them one continuous area is the best idea. The inner surface of these corner pieces should also be coated with the mirrored aluminum or mylar too of course. Now you've made up for the gaps in the larger surface area and have a very intense and uninterrupted light emanation area.

Next, have someone you know who sews, make up a square or rectangular flozier out of some diffusion silk like this:

1/2 stop silk: http://www.rosebrand.com/product747/...id=1&info=Silk

Full stop silk:
http://www.rosebrand.com/product1277...id=1&info=Silk

A flozier is basically just a square or rectangle of such material with elastic around the edges so it will stay put on the barndoors. You may find though that with CFL spirals such as the 30w to 40w N:Vision brand from Home Depot that there's not much need for added diffusion.

Finally, with an adequate amount of switches, you have your poor person's "dimmer" to bring down the intensity such that extra diffusion wouldn't be necessary for that use either. Hope this helps.

[Martin Catt]

Whew! After over a dozen views, SOMEBODY finally comments. I was worried you all were too busy snickering.

Joe: No softbox yet. As Richard said, the light is "soft" by nature, coming from a distributed source, rather than a pinpoint source. I've got a full set of umbrellas and bounce cards from the studio flash kit, and I expect I'll adapt them as needed. Because the lights run much cooler than I even expected, I can just clothespin scrim or diffusor material across the barn doors, or hang a cloth in front of the light as needed.

Richard: Pretty much made everything myself. I have a small lathe and milling machine (tabletop sized), plus free access to a full-sized Rockwell engine lathe and milling machine at a friend's house. Expensive? Well, as far as machining time, think astronomical. The materials came out of the scrap bin.

These were essentially a one-off prototype with individual parts machined twice to make a pair. Parts were milled out of solid bar stock, and yes, it's extremely over-engineered. If I decide to make a few more, I'll go with casting aluminum parts (I've also got a small foundry setup). These were to fill a short-term need and find out where the weak points in the design were. The point is, when I'm on a shoot, the last thing I have to worry about is the lights coming apart or not working.

The barn door design is up for modification after I field-test it for a while. I went with black doors specifically after trying bright aluminum. Shiny doors work well on a focused source, but with a broad source like this, it actually scattered more light where I didn't want it. These are made from .040" aluminum, bead-blasted, primed, and painted flat black on one side. I want to be able to cut the light spill without having to use bunches of gobo's.

Baking pans unfortunately only come in standardized sizes, 9x9" being one of them. I would have preferred a 12x12", but they don't make them in that size, and the cost to make one in materials and time didn't look like a good return on investment. There IS a standard-sized serving tray used in food service that's rectangular and about 9x18", plus it's stainless steel. I may take a swag at another set of lights using those if I decide I need more.

The first design I considered used nine of the rotary-switch socket fixtures, one for each bulb, with the switches coming out the back of the pan. I decided it looked a little too cheesey, despite the ability to switch individual lamps, and went with the hi/low setup. Bat-handle toggle switches are fairly indestructable, and should hold up to being dragged around from location to location.

Actually, I want to make an aluminum eggcrate to help "throw" the light a little better. For now, though, it's nice to have 1.8kw of light on demand.

Martin

[Greg Boston]

Nice job, Martin! The only suggestion from me would be to paint the exteriors flat black also. That eliminates any unwanted reflections and I feel it would give them an additional touch of professional appearance.

-gb-

[Gary Moses]

Wouldn't it have been a lot easier and less expensive just ot buy one of these?
http://cgi.ebay.com/Continuous-Light...QQcmdZViewItem

I bought 2 and have been using them for almost a year. It has 2 circuits, one I use for 5500 degree CFL's and the other circuit for 3500 CFL's.
Just a thought.
Gary

[Richard Andrewski]

I would swear I've seen other larger sized baking pans at a restaurant supply. I used to live in Houston and there was Ace Mart on I-10 which had a lot of different sizes I think. I used a round one for a similar idea to this before for fitting in a softbox.

On your casting setup, is that sand casting you're using? Sounds like you're setup for some serious metal work there. I guess you need to do quite a bit of machining on the cast parts if they're sand cast types right? I've become pretty familiar with all these casting types now and while die casting is the ultimate, it's also the most expensive to setup for. But, I wonder about plaster casting for aluminum and how the quality and detail is for that of the finished pieces. Also, I understand that spin casting can be interesting for some small parts but is mostly only good for metals like zinc.

[Brian Luce]

Originally Posted by Gary Moses

Wouldn't it have been a lot easier and less expensive just ot buy one of these?
http://cgi.ebay.com/Continuous-Light...QQcmdZViewItem

I bought 2 and have been using them for almost a year. It has 2 circuits, one I use for 5500 degree CFL's and the other circuit for 3500 CFL's.
Just a thought.
Gary

I don't understand, you can use one spiral bulb for two different temps?

How's the build quality?

[Gary Moses]

Bruce,
There are 2 different types of bulbs you can buy at home depot/Lowes, nvision daylight (5500 kelvin) and warm (3500 kelvin), each bulb puts out 100 watts and uses about 23 watts.
There are 2 circuits in the unit I bought, one has 5 sockets and the other has 4 sockets. I put 5 daylights in the 5 socket, the bulbs put out 1300 lumens and 4 warm in the 4 sockets, each bulbs put out 1600 lumens. It closely balances the light output. 6500 lumens for Daylight and 6400 for the warm.

[David Fleming]

Just thoguht I would mention that Britek makes CF fixtures and sells them with sofboxes and stands quite resonably. The set I have uses four 42watt CF bulbs with individual switches and a 24" x 24" softbox with diffusion silk.

Everything packs down into a faily compact duffle bag, and I can usually get all three lights assembled and running in about 15 minutes. My impresson owning them for about a year and using them on 7 or 8 projects is that they are not built for extreme duty service, but if you take care of them they will work well.

Most everyone I have used them with has been curious but supprised with the results. If you need an inexpensive soft lighting solution I would definatelly check em out.

[Martin Catt]

Originally Posted by Greg Boston

Nice job, Martin! The only suggestion from me would be to paint the exteriors flat black also. That eliminates any unwanted reflections and I feel it would give them an additional touch of professional appearance.

-gb-

I'd considered that. These are prototypes, per se, and there were a lot of unanswered questions as to how well they would work once finished. The big question was heat management. The double-pan design gives excellent convection cooling, so even after an hour or so, they only get warm. Oddly enough, the barn doors get hotter than the main chassis, though not surprising since they're painted black. I originally went for one side black/other side natural beadblast so I could flip'em over if I wanted a more reflective surface. Having an unpainted side lets them shed heat better than if they were painted both sides.

It's also a cosmetic issue. The outside of the barn doors will get banged and scratched up. The matte beadblast finish hides the dings better than straight black. Now, if I was REALLY dedicated, I'd go out and get them black anodized.

Martin

[Martin Catt]

Originally Posted by Gary Moses

Wouldn't it have been a lot easier and less expensive just ot buy one of these?
http://cgi.ebay.com/Continuous-Light...QQcmdZViewItem

Gary

Sure, but it wouldn't have been near as much fun. Actually, the materials cost for the pair I made would have been the same for a single unit like you've shown (when you add in shipping, etc.). The unit you've listed also doesn't come with barn doors, plus I wanted a yoke mount for balance. The Novatron flash heads use an hinged post like yours, and I've always found them to be top-heavy and off-balance for certain setups. When you buy, you get what's available. When you build, you get what you want.

Martin

[Martin Catt]

Originally Posted by Richard Andrewski

I would swear I've seen other larger sized baking pans at a restaurant supply. I used to live in Houston and there was Ace Mart on I-10 which had a lot of different sizes I think. I used a round one for a similar idea to this before for fitting in a softbox.

Well, I confess I didn't look to hard, though I do have connections in the restaurant supply industry. The 9x9" pans hit the price point I wanted and I could get them at Wal-Mart. I've toyed with the idea of a video ring light using a ring cake pan, though I can't see my having any real application for it any time soon. If I was shooting industrial or medical footage and needed a well-and-even-lit closeup, maybe.

Originally Posted by Richard Andrewski

On your casting setup, is that sand casting you're using? Sounds like you're setup for some serious metal work there. I guess you need to do quite a bit of machining on the cast parts if they're sand cast types right? I've become pretty familiar with all these casting types now and while die casting is the ultimate, it's also the most expensive to setup for. But, I wonder about plaster casting for aluminum and how the quality and detail is for that of the finished pieces. Also, I understand that spin casting can be interesting for some small parts but is mostly only good for metals like zinc.

Yeah, sand casting. I do "small" stuff, limited to about 4 pounds aluminum or so. Picture of my foundry at:
http://backyardmetalcasting.com/guest_martin.html

How much machining is needed depends on the "look" you want. I usually just machine the critical mating surfaces and leave the rest natural. It gives a softer, more rounded, organic look to the piece. The finish depends on the coarseness of the casting sand. Finer sand gives a smoother finish.

Die casting is great if you need hundreds or thousands of pieces, and can spread out the cost. Some of the zinc-based die-cast alloys are actually stronger than cast iron.

I'm guessing that when you say plaster, you mean investment casting. Regular plaster tends to explode when you use it as a casting mold. Investment casting lets you do really complicated shapes that can't be die or sand cast, but you need a wax original for every part you want to cast. It's great for short runs of high-value items. Jewelers use it a lot. Centrifugal or spin casting isn't limited to zinc. It goes all the way up to refractory metals like platinum.

Sand casting falls in between the two. It's great if you want to do just a few things, or you can grind out a hundred or so. I haven't touched the foundry for a year, so I'm looking for an excuse to melt something.

Martin

[Richard Andrewski]

Didn't mean to hijack the thread but I love this kind of stuff. Here's an excerpt from a site describing what I'm talking about by plaster casting for aluminum:


"In this method, either a permeable (aerated) or impermeable plaster is used for the mold. The plaster in slurry form is poured around a pattern, the pattern is removed and the plaster mold is baked before the casting is poured. The high insulating value of the plaster allows castings with thin wads to be poured.

Minimum wall thickness of aluminum plaster castings typically is 1.5 mm. Plaster molds have high reproducibility, permitting castings to be made with fine details and close tolerances. Mechanical properties and casting quality depend on alloy composition and foundry technique. Slow cooling due to the highly insulating nature of plaster molds tends to magnify solidification-related problems, and thus solidification must be controlled carefully to obtain good mechanical properties.

Cost of basic equipment for plaster casting is low; however, because plaster molding is slower than sand molding, cost of operation is high. Aluminum alloys commonly used for plaster casting are 295.0, 355.0, C355.0, 356.0 and A356.0."



The above came from this site:


http://www.key-to-metals.com/Article59.htm

Actually sounds like a good compromise for getting lots of small details without going to die casting. I would also imagine that some of the really hard stone plaster varieties like that used by dentists to make molds might work better in high heat situations.

The spin casting also sounds really interesting to me:

http://www.tekcast.com/

But on that site, they claim its mostly good for use with casting zinc and not higher heat metals like aluminum.

We'd love to find a great alternative for casting low numbers of parts a hundred at a time, but getting high detail and precision. The cost of making a die casting mold is really high, even here in China. I looked at sand casting but it can't be used to reproduce really great detail and thin parts of a model.

[Richard Andrewski]

Originally Posted by Martin Catt

Sure, but it wouldn't have been near as much fun. Actually, the materials cost for the pair I made would have been the same for a single unit like you've shown (when you add in shipping, etc.). The unit you've listed also doesn't come with barn doors, plus I wanted a yoke mount for balance. The Novatron flash heads use an hinged post like yours, and I've always found them to be top-heavy and off-balance for certain setups. When you buy, you get what's available. When you build, you get what you want.

Martin

I think this is the point for some. Many, like me just do it for the fun and challenge or stretching their craftsmanship. In the early days of fluorescent lighting for media production however, companies like Cool Lights or some of the others didn't exist and you did have to pay big money to get a fluorescent unit. So DIY was the best alternative then for those with smaller budgets. Today it is getting much harder with all the available alternatives to consider it cost effective to do DIY however if that's your main reason for doing it.