need help finding a suitable coating for a homemade thermopile sensor

[shrad]

hi guys

I'm in the process of designing a small power meter, but it is difficult to find a suitable coating to absorb the visible spectrum entirely

I tried a fine deposit of airfloat charcoal powder with alcohol, which was neat when it dried, but it seems to absorb only the lower part of the spectrum, near the blue region

I had no voltage change when pointing red wavelengths at the thermopile, when 405nm was clearly absorbed

I have access to common household dyes as water-resist pens, paint, and very fine charcoal dust, as well as fluorescein, rhodamyn and methylen blue

would it be easily achievable? I'm also tempted by the wheatstone bridge + resistor approach, or designing my own thermocouple sensor with copper foil and another metal, but it is hard to tell which method would best suit my needs and have a homogeneous reading from 400nm to commonly used IR wavelengths, like maximum 1500nm

if you have any idea, please help me

[mixedgas]

Any flat black paint. if you have Krylon over there the krylon flat black is your brand of choice. So is soot from a match.

Steve

[shrad]

I also tried soot from a candle, does it make any difference? is it important if the peltier faces are gold-coated?

I'll clean the peltier face again and try pinewood soot

thanks for always being so helpful

[shrad]

I also have decided to design a FET amplifier, diferential if the pair of FETs I have are of close enough specifications... it should provide sufficient gain to measure 0.1-1 mW range values (I measured 3mA change with a 2.4mW pointer so it should be sufficient to efficiently drive an opamp)

[andythemechanic]

I once used candle soot on an uncoated peltier element, and it worked very well (Until I hit it with a pulsed Nd:Yag and had a white peltier again). Did it not work for you?

Andreas

[shrad]

yes it works now! but I have issues with the peltier voltage dropping when connected to an opamp at very low readings, so I have to make a high impedance voltage amplifier and I have no fet opamp in my drawers

I have to do a bit of search to find one, or I'll do a small transistor amplifier... the gain doesn't have to be too great so I think I'll stick with that solution

I'm using a 8*6mm TEC with gold plated faces potted to a small aluminium heatsink, and it seems to react very fast (less than 10 seconds for the value to settle)

[andythemechanic]

Maybe an intrumentation amp would be the right thing here. For example the AD620. Very high input impedance and low offset voltage and drift. Not cheap though.
Shrad, where did you get this tiny TECs? I am looking for such small ones.

Andreas

[shrad]

from mebe3d on ebay, who is also selling some coherent verdi and avia parts some times

he doesn't have any left, but chris (hesc_photonics or whatever in the future) was selling some a pair of weeks ago, and maybe he still has some

the best use would be for optical crystal regulation though, as you may easily solder a copper mount on it with a bit of indium or other low melting temperature alloy

the best thing I could do to measure powers accurately would be to buy a decent thermopile head from ebay, but as I haven't found a bargain yet, I'm still looking to build one myself

a friend has already built a nice powermeter from such a TEC device, which is very stable and accurate, and that would be a nice replacement for a commercial unit