Calibration needed for LPM

[gillza]

The amount of heat generated on the TEC is nothing any heatsink can't handle... If anything the fan will possibly keep it too cold & throw it off. Most LPM's are calibrated at aproximately 22-24 degrees celcius & the fan will cool it too much...

That's a possibility not 100% accurate

Thanks for the explanation.

 

[zartoop]

Could you explain why? Im genuinely curious. Wouldn't it help dessipate heat from the heatsink and assist in "pumping" the heat from the cold side?

For a TEC based sensor a fan is arguably unnecessary at the powers (<3w) that are typically measured in hobbyist lasers. Sensors for measuring high power lasers do have cooling fans.

There are a number of considerations that might make a fan a good idea. I left the fan in place as it came with the heatsink. I haven't used it yet.

A TEC outputs a voltage as a function (very nearly linear for some depending on quality, materials and construction) of the temperature difference between its faces. So if we hold one face at a constant temperature and apply (or remove) thermal energy to the other face to induce a temperature difference a voltage will be generated.

A heatsink is a heat exchanger. It exchanges heat by air convection, radiation and conduction. Air convection is the largest exchange mechanism and is greatly increased by forcing large air volumes through the heat exchanger. The heatsink is used to maintain a constant temperature of the TEC face.

Ideally we want to have the constant temp face as just that = constant temp during the setup, calibration and laser testing. There are a number of things that will prevent that happening in an ideal manner. If we cannot have an ideal constant heatsink temperature then we wish to reduce the heatsink temperature change as much as possible;

1) Heat conduction (albiet very small as a TEC has a high thermal resistance) across the TEC will induce a temperature change in the heatsink.
A heatsink with a lower thermal resistance (TR) will change less in temperature for a given heat input. The typical sensor heatsink (TO-3P and similar) has a TR of about 7.5C/W. The heatsink I have used has an approx TR of 1C/W in convection and <0.5C/W with fan assist. For example let's assume that we are testing a 1 watt laser and that 5% of the thermal energy absorbed by the TEC is conducted to the other face by thermal conduction. A TO-3P will rise in temp by 5% of 7.5C = 0.37C. The larger heatsink without fan assist, 5% of 1C = 0.05C and with fan assist 0.025C. In short the fan and lower TR heatsink will rise less in temperature which equates to a lower error due to conduction induced temp change. In practice this may not be worth compensating for as even a 0.37C difference may only mean a couple of millivolts error in 1V. That's an error of 0.1% to 0.3% for the TO-3P and <.001% for the more efficient heatsink.

2) Ambient conditions will affect the temperature of the heatsink.
The ambient conditions to be controlled are ambient temperature, radiation, convection and conduction = The temp of the room, air currents, sunlight etc will affect the temp of the heatsink and the TEC. In practice it is important to reduce all these effects. Make sure the test room is at a constant stable temperature, no direct sunlight, lights on or off remain the same, room heating or cooling is off and no air currents. For the TEC: shield it from direct radiation/light except that being measured (radiation); prevent air circulation in the TEC region (convection); ensure the 'input' face of the TEC is not touching anything else (conduction). For the heatsink: ensure the radiation (UV, visible and IR) falling on the heatsink remains constant and that the heatsink is shiny i.e. not black as that will only enhance it's absorption(radiation); ensure the heatsink is thermally isolated at its contact points i.e. rubber feet (conduction). Where a fan may really come into its own is with convection. With natural convection the heatsink can only dissipate heat via local convection air currents due to air movement in the room, heatsink induced air currents and that same air is affected by heat from your hands, breath, power supply or other equipment you have lying in close proximity. i.e. the heatsink will only transfer energy to the 'local' air and that air is 'locally' affected. By using the fan you are exchanging heat with a much larger air volume which in effect 'averages out' ambient temp variations.

3) The ambient temp at calibration will unlikely be the same as that during settling time and laser test measurement.
In short this is not really directly effected by fan or not. Essentially you want to take measurements at the calibration temp and where this is not possible then apply correction factors. A liquid cooled heatsink can be used to maintain an almost perfect heatsink temp. If you are testing in an environment more than ~7C away from the calibration temp then you should change the environment.

Thermal inertia also plays a role and I will detail this when I have more time.

So ... long story short. Is a fan really necessary? Not really. Does it make things more accurate? Yes. Is the level of accuracy improvement worthwhile? That's up to you but probably not as there are many other factors that will affect accuracy orders of magnitude greater than having a fan or not.

 

[zartoop]

The amount of heat generated on the TEC is nothing any heatsink can't handle... If anything the fan will possibly keep it too cold & throw it off. Most LPM's are calibrated at aproximately 22-24 degrees celcius & the fan will cool it too much...

That's a possibility not 100% accurate

Whilst you are generally correct there may be a slight mis-understanding where you say the fan may keep the heatsink 'too cold'. The heatsink should ideally maintain a constant temperature and that constant temp should be close to, if not at, the calibration temp. A fan will never cool the heatsink below ambient temp.

 

[cheech226]

up until today all of my research in this hobby has been on lasers themselves. i bought a laserbee early on because i just Had to know the numbers, and really didn't care how it worked. today i put a lighter to a heatsinked tec and watched the voltage rise! the posts above answered a lot of the questions that immediately sprung to mind. thanks.
plenty of fun ahead!