DIY Thermal LPM for under $50

[Meatball]

So... is it bad if my output is only giving me a -9V reading or so?

There is no change upon pot adjustment.

 

[MarioMaster]

So... is it bad if my output is only giving me a -9V reading or so?

There is no change upon pot adjustment.

I'd say yes since there's no negative voltage rail in this circuit. I'd recheck your connections.

 

[Meatball]

Lol that's what I thought, but I also figured something in the range of +-9V is huge.

Mario, would you be able to send me the original schematic? There's something about the one posted by that other guy that doesn't seem right...

Are there some specific measurements I can be looking for, to help determine what I wrong? My connection checking is not finding anything!

 

[jib77]

Make sure you are not connecting net intersections that do not have a dot on them ... I see two on my schematic.

 

[Meatball]

Got it to read a reasonable voltage now...
I'm useless when it comes to wiring pots. I didn't even ground the 5K, but for some reason I just hooked it to the other side of the resistor! lol

Just waiting for the adhesive to set up, and I'll get down to calibration.

Which pot is the gain pot? The 5K?

EDIT: When it's not responding to a laser on the TEC.. that's bad right??

 

[jib77]

5K is the Zero, 10K is the Gain.

 

[Aaron]

i also cant get the 5 kohm res and the 2kohm i can only get 1kohm and 10kohm
and what type of TEC do i need is this ok 00711-5L31-03CA what sort of voltage max???
cheers

Get it from Virtual Village for only 8.99

 

[jib77]

Those TEC's will not work very well, way too big ... you need a small one with high number of thermal junctions(the metal fingers on the inside). This is the one to use: >>link<<. There is a reason its $38 ... you don't want to skimp on this part.

 

[Xplorer877]

Those TEC's will not work very well, way too big ... you need a small one with high number of thermal junctions(the metal fingers on the inside). This is the one to use: >>link<<. There is a reason its $38 ... you don't want to skimp on this part.

Thats what I used and it works great!!!
The reading will take about 10-12 seconds to stabilize when measuring my 950mW 445nm. That's actually really good.

The bigger the plate is the longer it will take the laser to heat it up. Which means it will take more time for the thermal difference between the plates to stabilize and thus the reading will take longer to stabilize.

-Tony

 

[qumefox]

Cool thanks buddy.

Another thing I just thought of is that the TEC will generate a voltage proportional to the difference between the two ceramic plates. So the hot side would be the side that the laser shines on and the cold side would be attached to the heat sink. So a consistent reading will also rely on the same ambient temperature of the room at the time it is calibrated. The Laserbee would also be affected by this. It might mean that the reading may shift quite a bit depending on the temp of the heatsink.

If the room was 60 deg and you got 100mW lets say, then if you test again in 80 degs with the same laser the difference would be -20 degs resulting in a reading less than 100mW. It could be significant if it was a really hot day.

-Tony

Ok.. I know this particular post was over 6 months ago, however, I just now read the whole thread and saw no correction of this..

This information is false. Ambient temperature will have very little affect on the readings. (Well, unless the temp is changing WHILE your measuring. My laserbee is very susceptible to air currents) TEC's are differential devices. It's the difference in temperature between the two sides is what causes the Vout in this application. The ambient temperature of the device is irrelevant to an extent so long as both sides of the device are the same temperature when zeroed. A heat source of a fixed power will heat one side by the same amount regardless of it's initial temperature.

For example. Say 100mW of laser power would heat one side of the TEC 1 degree higher than the other. If the ambient temperature is 50F.. You would have 50F on the cold side, and 51F on the hot side.. 51-50=1, which is the important part.. this 1 degree is what causes the Vout.

Compare it to an ambient of 80F with the same situation.. 81-80=... you guessed it.. still a 1 degree temperature differential, which means the Vout of the tec would be the same as it was at 50F.

 

[Xplorer877]

^Yup. After some test with my meter I came to the conclusion that the ambient temp doesn't matter.
It was just a thought on my part.

-Tony

 

[HIMNL9]

Well, ambient temp does matter only if you don't left the sensor stabilize ..... the heatsink after the plate is for "hold" the temp of the "cold" side when you read the power, after all, and have a thermal gradient much much slow thatn the "hot" reading side ..... if you carry a meter from a room with a temperature to another with different temperature, the reading side change its temperature much quick than the rest, so it can false your reading ..... same for air currents and for "touch" the heatsink, or placing it on a metallic surface with temperature different from the ambient one.

Is for this reason that lab protocols incudes 30 minutes "temperature stabilization" interval, before read the power ..... we can't do it, sometimes, and after all, we don't need lab accuracy, but still some precautions (no air flows, insulation, a bit of waiting for thermal stabilization) are mandatory, if you want an affordable reading.

 

[Wolfman29]

Alright... so I don't understand the schematic you posted, Jib. I want to build this (and I know this is a necro post, too), but I don't understand where the 7th lead from the LM358N goes. And are IC1A and IC1B the same chip, just different parts of different circuits? And what about the GNDs? Are the GNDs different for those two circuits?

Oh, and what's the cap value of C1? .1 what? Is it .1uF? If so, that's awesome - then I have a working understanding of this and I can start accumulating parts.

 

[bluuray]

doesnt go anywhere.. there connected to GND i think
and the IC1A and IC1B are just names so he knows which ones which.. i assume..

 

[MarioMaster]

Hey guys thanks for digging this thread back up. I'll clean the schematic up a bit and post a circuit board layout if people are interested in making one.

 

[Wolfman29]

No problem, I guess =p

Hey, for the resistors and trimmers in this circuit, the power flowing through them won't be significant, right? So I can just go with low-wattage parts, yeah?