Re: Incandescent lamps as sources of heat

Electric fire bars are probably cheaper and much more robust for bigger loads. US 100v bulbs are a bit more efficient than UK 240v ones.

I have used light bulbs as a heat source for terrariums.

The ratio of light to heat goes up at lower voltages too.

If you put a lamp in a box, no light escapes and 100% becomes heat.

Incandescents are very nonlinear.

What sorts of resistances and powers do you need?

We make programmable electronic loads, resistive and constant-current and resistive+inductive.

Some big mosfets on a heat sink can be an adjustable load, constant-current or constant-resistance. Add a bridge rectifier for AC.

An incandescent filament is fast. You can audio modulate one usefully.

Let the light escape and don't worry about it.

But I really don't understand what you are trying to do. Do you need a dummy load, or are you trying to heat some space?

I'm not sure that is correct. The energy input to the surfaces will start instantaneously - but they will be slow to heat up because the actual amount of energy input is small and their thermal mass is large. The same energy flow carried by convected air would actually take longer to heat them because the air would have to be heated first to establish the circulation.

The ones we used had wide-pitches spirals of nichrome wire stretched back and forth between porcelain insulating plates. If the fan failed they would take a minute or two to glow red hot, by which time a bimetallic thermal cutout would have disconnected them. The danger wasn't the nichrome or the porcelain melting but the effect of high temperature on the motor windings and the possibility of starting a fire if dust had collected near the element.

If you want the elements to operate without the fan-assisted air flow (which is quite gentle in a large industrial heater) just put pairs of elements in series and lay them horizontally so that they are convection cooled. They are unlikely to overheat at one quarter of their rated power.

For testing the power dissipation capabilities of a large die-cast box, I just bolted a couple of aluminium-cased power resistors onto it. From those results I was able to calculate the largest size of audio amplifier that could be built into that box without needing a separate heat sink.

"100% efficient!" at turning electricity into heat. Glow orange in use.

This is a traditional radiant electric fire with modern 600W fire bars:

Convenient as indestructible 1kW loads at mains voltage.

Filament bulbs last a lot longer if you put a small power resistor in series to limit the inrush current into the cold filament.

You can surround them with a big plastic or ducting tube.

Halogen/tungsten lamps produce >95% of their luminous output in the infrared.

Model?

RL

In theory you could measure the thermal conductivity of the different materials or combinations of material the heat energy would have to pass through, then calculate the overall temperature gradient to whatever ambient you set as your maximum. The problem with large spaces and objects is that poor convection and conduction give very significant differences in temperature between various parts.

Stirring the air will distribute the temperature more evenly and bring down the hot spots, but it will still hit a limit imposed by the conductivity of the various layers the heat has to travel through. You also have to take into account the certainty that a fan will fail during the lifetime of the equipment (and may put an end to that lifetime). Convection 'chimneys' are much more reliable as long as the equipment is only used the right way up.

Even with the most careful calculations, things can be disrupted by unexpected events. I once left an automatic battery charger, in the corner of a Portakabin, supplying the batteries running a P.A. system for a large event. When I came back to check, I found it was underneath a pile of coats and almost too hot to touch. The internal temperature sensing had shut down the output to a level which it could sustain, so there was no damage but the batteries weren't happy. When I removed the coats, it cooled down and returned to normal operation.

The moral of that story is to assume the equipment will eventually overheat from some unforseen cause and make sure it degrades gracefully without damage.

There is a tendency, left over from the days of valves and expensive power transistors, to try to extract the maximum output from a single device. Two transistors in parallel have half the dissipation and half the thermal resistance to the heatsink of a single transistor, so the slight extra complication of using two devices is more than compensated by the four-fold reduction in the temperature gradient.

Do a search for G4 halogen 20W 2 pin bulb. These run on 12 volts and were (maybe still are?) popular for under-cabinet kitchen lighting. I know 10W's were available, not sure about other sizes. Not the standard screw-in Edison base, but widely available.

buck a-piece from off-shore'

For laods, when the real thing is too expensive or unobtainable, Use generic HVAC heaters and duct fans. standard sizes etc. If you want calibration, wire your own using nichrome heater coiled wire.

Maybe maybe maybe. What's your problem?

Stick your heaters in a wooden box, the average surface temperature rise of the outer box wall, above ambient, will be:

• / - 5%

from cigarette box to two-man carry shipping container, in free air.

Don't forget to measure each centimeter accurately, for averaging. It's sort of cool knowing what each centimeter is actually doing, and may give you some useful packaging ideas.

This is, of course, reversible. You can reliably estimate power dissipation of a surface are with known aveerage rise.

What you can't tell, is what the spot temperature of the internal radiator. That's modelling for you. The smaller the source gets, the higher it's temperature in order to dissipate the power being wasted, and the stupidder the modeling and modeller becomes.

I've had to demonstrate this too many times in the past, to enjoy explaining the miracle coefficient ' 1 '.

It's drawn from and confirmed by those demonstations and from ripple current rating charts of phsically dimensioned, thermally restricted, computer grade aluminium elctrolytic caps. From before the days of built-in fans for cooling for reliable equipment.. ..........................

Thermal rise of surrounding air is fixed by it's 'specific heat'.

Move volume of such air elsewhere, by any means you desire, to evacuate volume area requiring cooling. Rates are measurable.

Add conditions codicils or flambools as you will. principals will be the same.

What would you want? You'd follow the mfr's instructions, or draw from your own experience/observation. There's probably a second hand book discarded from the local library or tech college somewhere

You're not designing a clothes closet or an airing cupboard. Those are really tricky.

I didn't recommend for or against them, and have no intention of doing so. You implied that you didn't know of any readily available halogen bulbs under "hundreds of watts" and I gave you an example of a 20W bulb type that I know is readily available. Anything further is beyond my interest and entirely up to you.

I am normally the only user, so the problem doesn't arise. I designed it so that with a 30% overload, the casing temperature would rise to 50 C in a 20 C ambient. Above that temperature the operator woulld burn their fingers on the controls, so would be inclined to take some action to reduce the load. (It is a combined mixer and P.A. amplifier, so the operator would be constantly handling the controls.)

Even at that case temperature the output transistors would have their junctions well below the maximum-allowable temperature because they are run in paralleled pairs, so the thermal resistance between the junctions and the heat sink is much lower than it would be with a single device. The drop in current gain with output current is also reduced, so the driver stage has less work to do.

I once had to use it badly overloaded, in emergency, for an evening dance in a large hall. Before we started I went to the rubbish area and tore a flap off a large cardboard box. Two of us took it in turns to sit in the 'wings' all evening, fanning air past the casing to prevent it overheating. It survived the event but one of the dancers told me the next day that he had seen what we were doing and had spent the whole dance with his fingers crossed for us.

In an amplifier with separate inaccessible heat sinks I would incorporate some form of thermal shut-down or power reduction. One excellent method is to use a thermistor as one element of a voltage divider in the signal path; as the temperature rises, the gain starts to reduce. I precede it by a soft clipper which doesn't operate on normal input levels. If the user tries to restore the excessive output by winding up the signal level, the clipper starts to operate and the output begins to sound distorted.

Under 'field' conditions, this is a reasonable compromise because the system doesn't sudenly stop working. A bit of soft clipping can actually make the amplifier sound as though it is delivering more power than its undistorted rating.

With portable P.A. work it is highly unlikely that the operator will bother to read any installation instructions or even be aware that they exist. I've never seen a P.A. operator consult a handbook (or its online equivalent). The equipment is put in a place that is convenient

- if it doesn't work there, it is deemed faulty.

Even with permanent installations I think the best you can hope for is to specify a maximum ambient temperature and the total wattage to be dissipated - that covers your back when it breaks down. Unless this is to be a major industrial installation, I doubt if anyone will bother to read what you have written (or understand what it means in practice).

Your best bet is to assume the worst and leave lots of thermal headroom for when even-worse-than-the-worst happens..

It sounds as though you are thinking of dumping kilowatts of heat into a domestic environment - or are you looking at two possible different uses for the equipment: single domestic units and bulk industrial ones?

...leaving the customer and hundreds of punters with no P.A.?

Always carry at least two spares and be prepared to improvise or repair on the spot if things go wrong.

That appears to be unique to the HVAC industry (at least in the UK) and to structural engineering - everything else, apart from major engineering projects, appears to be rule of thumb.