Simple Thermal Power Measurement

[ion]

Incandescent lamps would appear to be a good candidate for power output measurement. Since they come in all power levels from "grain of wheat " types that draw less than 30 milliwatts (1.5 V @ 20 mA) to giant 1kW line operated lamps, we have the ability to match our lamp to the expected output of the device under test "DUT".

Lamps not only provide a true rms function by integrating pulses  into heat, no matter what the crest factor and regardless of positive or negative value. They also provide a quick visual assesment of output level.

We need first to calibrate the lamp bulb envelope for temperature vs. power input using a known source e.g power supply or mains depending on type. A thermocouple fastened (e.g. high temperature epoxy or ceramic cement or taped firmly) to the top surface of the lamp can read temperature which is correlated to power input at several input power points for characterization. Ambient must be norted and subtracted from each reading for normalization. Plotted on a graph we can now look up the temperature over ambient and convert the reading to power when hooked to our DUT.

Another method is to feed the output into a sensitive D'Arsonval movement meter the scale of which is directly calibrated in power. For this it is good to use a second ambient sensing thermocoupe in inverse series which nulls out ambient. In this way we need not consider or subtract out ambient (as in the prior method) since it is automatically subtracted electronically.

For quick and easy calibration from the mains, we can use full power direct connection from the mains and half power with a diode in series to determine two points and characterize by drawing a straight line between the points. Also, considering "no power input" as a point we have a total of three points that can be curve fitted. The more points that can be used the better. Lamps can be wired in series to reduce power e.g. 4 lamps = 1/4 power in each lamp. Always measure the current and voltage to determine the actual power while characterizing your "standard" lamp

The same methods can be used with a resistor in place of the lamp and extra sensitivity at low levels can be had if our resistor is insulated from ambient.

IR thermometers can be used to measure bulb or resistor temperature, but I recommend that the test spot on the top of the emitter be painted flat black with high temperature paint to normalize emissivity. Then a test stand should be used to mount the IR gun and emitter such that they remain at a fixed distance and spot position.

Attached is a sample graph for a 60 Watt 120 volt lamp measured with a thermocouple taped to the top of the bulb in free air (unstirred) and corrected for ambient.

I will be updating the graph with measurements using an IR thermometer for comparison.

[Grumage]

Dear ION.

Your post could not have come at a better time !!

Please find attached a schematic that Verpies dropped us at OU .Com.

He suggested a GOOD vetting and my own thought was that this forum contains some of the best minds available.

Cheers Grum.

[ion]

Dear Grumage

Yes, the use of a multiplier IC is an excellent way to do power measurement, and I am familiar with this method and like it very much.

I just wanted to describe  something for the average guy who would not have the ability to build with or use multiplier IC's.

IMO, both methods have their place and are valuable.

Thank you, Grum and Verpies.

Attached is another method where power is determined by using a "null" against a known power input, again a very simple bridge balance method.

[Grumage]

Dear ION.

I am all for simple !!   

That is why I built this. See attached. It is a 60 W 240 V lamp calibrated with a variac and wattmeter. Shining onto a small PV cell into a series calibration pot and micro ammeter.

However it does have its drawbacks but it is far better than as Verpies put's it "eyeballing" !!

Cheers Grum.

[ion]

Thank you Grum, that is another good method.

My field was thermocouple instrumentation so I tend to use them.

Readily available Copper and Iron wires could be used for TC's, purity not that important since we are not comparing to a table of mV , just using relative output.

PV's have sufficient output and can also drive a D'arsonval meter directly (like older light meters), by which you can print the Watts on a scale behind the needle, no batteries needed for a stand alone instrument.

I think I have beat the subject of thermal measurement to death on my bench and will give it a rest.