A cheap 1200W dummy load using halogen bulbs

[intabits]

I recently built this "Cheap and Cheerful" dummy load based on 24x 12V/50W halogen bulbs.
The bulbs can be wired any which-way to provide a dummy load capable of dissipating up to 1200 watts.
Because of the non-linear behavior of bulbs, and their time dependence, this is not suitable for any sort of dynamic use, but does provide a flexible high power static load.
It has come in handy a number of times already.


Here it's handling 33V @6A with 12 bulbs wired as 2P6S


And the other side, showing the terminal strips that allow the bulbs to be wired in any required configuration



Video of making and using it

[IDEngineer]

Nice! We use traditional incandescent car headlights as a quick ~60W 12VDC load around here. (We do a lot of 12VDC design work.) Nothing as formal as your setup, though. Very nice DIY solution. Kudos! 

[Wolfgang]

Another sick idea would be some inserts from ladies hairdryers. Their light is not as good, but otherwise they are not bad as loads. 

[coppercone2]

I like that you used a protective cage. Just make sure it can't set something on fire from radiant heat.

Commercial ones I have seen will space the bulbs out more and elevate the cage higher, basically leaving the bulbs facing upwards and make a fully perforated shell. This will reduce the light output (nice for the operator) and increase protection.

It is possible that the longevity of the bulbs will be effected if they are getting too hot from each other.
https://www.nema.org/Policy/Environmental-Stewardship/Lamps/Documents/LSD%201%20T-H%20Lamps%20v2_4%202003%20C6.pdf

I never saw it with that type of 'heatsinked' hologen bulb though, only normal incandescents, but the equipment I dealt with probably had lead paint on it. Also they usually seem to be painted a medium dark greyish blue color.

[intabits]

I expect that running it at full capacity would be fairly rare, and of course one would take extra care when doing so.

I have had it running quite hard, face down so that the bulbs are only an inch above the table. The table did get quite warm, but nowhere near burning. The "dichroic" property of the bulbs is supposed to direct the IR out the back of the bulb, maybe that was happening.

In practice, for me at least, I don't expect to run it anywhere near 1200W very often, or for long periods.
And so far, the power supplies under test have been the only parts of the system running near their temperature limit. 

[coppercone2]

how does your resistance variation/power draw variation look like over time with that? does it stabilize quickly?

[intabits]

I'm not sure exactly what you mean, but with a constant voltage input the current settles quickly and then stays there.

That is, there doesn't seem to be any sort of long term variations, just the short term effects of bulb heating/cooling time and positive tempco, resulting from a significant variation in any of voltage, load, or current capacity.

As I mentioned it's not so good in a dynamic situation. So for example, from cold, the sudden application of voltage causes a heavy inrush current that then subsides to a constant steady current within a a couple of seconds.

But if the power source has current limiting, there is a strong possibility of wild oscillations, and I've had that happen once. It usually doesn't, as I'm generally winding up the input voltage relatively slowly via a variac, which avoids such violent current surges.

[coppercone2]

I have been meaning to build one of these the halogen bulbs look like a good choice since you can fan cool a ducted heatsink nicely and actually route most of the heat out a window.

[m3vuv]

ive always thought of using a couple of stainless steel or copper rods in a bucket of salt water as a dummy load,ive electro etched pc boards before,they can sink some current.

[coppercone2]

I can see the lab turning into a humid mess. Don't let salt water boil in your lab the aresol will carry and deposit into your equipment and encourage corrosion.