A (potentially easy) $25 DIY USB Data-Logging LPM?

rhd said:

I think you'd destroy the photocell way too quickly.

Click to expand...

They don't make high temp photocells or anything?
What if you ran it through a filter like safety glasses... do they work linearly? or is there some weird physics behind it?

@Guy:
I actually did that, and using some math and some stuff, I got reasonable estimates, or at least correlations.

Just throwing out ideas... *from someone who has NO experience at this*...

If you want accuracy from a "cheap" homemade kluged-together idea... (but one that is BRILLIANT and seems to work)... you probably need to eliminate as many "wild variables" as possible.

(Guyfromhe) "Safety Filter" probably = some hi-quality ND (neutral density) filter. BUT can one GUARANTEE that the effect of a particular ND will be equal across ALL wavelengths? (shrug). What about IR, and for 405's? An ND might not be specifically designed to "work with" near-ultraviolet. For IR, unless the maker of the ND has specific data, who knows how it will react?

-----

Variation #1

Large photocell... I'm thinking... one of those LARGE circular ones. Mount it on some kind of LIGHT (white) background. CLEARLY mark the extents of the photocell (the "target area").

Use it in a dark or almost-dark room.

As a practice, you de-focus the laser until it spreads out to the extent of the photocell surface.

MAYBE the lower radiation-per-whatever will prevent melting the cell.

-----

Variation #2

Sampling.

Spread the beam (defocus) out to some width at a certain distance... in your experiments you will *standardize* this.

Then *sample* a part of that spread beam with a small mirror... a dentists mirror comes to mind. Bounce this fragment of radiation to your sensor.

MAYBE after a LOT of experimentation, calculations, calibrations, a cheap DIY LPM can come out.

I've actually looked into this cause I thought it would be nice to make my own data logging LPM. Was inspired by the 2.5W Laserbee

Hate to say it guys but this thing couldn't come close.

2degree range on that dude is way too terrible of accuracy. Almost to cheap to even use as a temperature reader! This is especially the case for the precision we would want. Say you want.... -+1mW resolution you would need a device that would raise to 2000C per W .... and i'd have a max of a 1W resolution(unless you don't mind going to 4000C per w...) However, there could be a few problems at 2000C =P... and not to mention the fact that the meter can't read that high. Something like this perhaps could work as a terrible measuring device for 5-50W range devices at best.... So all of you that have those go ahead and try it out =/

Hasn't anyone noticed the thermocouple on the DX link...
It is only one (1) thermocouple junction and it's surface
is quite small... (see attached pic with red circle)
It is a common thermocouple sensor for PID controllers
or industrial Thermometers. They sell them cheap on eBay..

K Thermocouple Probe -50 to 250°C 1M thermometer sens | eBay

You would need to focus the entire Laser's beam under
test onto that small surface to get any accurate readings...

I suppose you could attach a larger target to the thermocouple
bead but then you would run into the target/heat/dissipation
problems encountered in this DIY LPM project...

http://laserpointerforums.com/f42/diy-laser-power-gauge-26332.html#post460104

and here is more info on another inexpensive Thermal DIY LPM...

http://laserpointerforums.com/f42/s...er-using-ir-thermometer-26341.html#post512759


Thermal experiments with Lasers can be done by anyone just
using a standard bulb type or digital household or cooking
thermometer... (been there... done that)


Jerry

random person said:

I've actually looked into this cause I thought it would be nice to make my own data logging LPM. Was inspired by the 2.5W Laserbee

Hate to say it guys but this thing couldn't come close.

2degree range on that dude is way too terrible of accuracy. Almost to cheap to even use as a temperature reader! This is especially the case for the precision we would want. Say you want.... -+1mW resolution you would need a device that would raise to 2000C per W .... and i'd have a max of a 1W resolution(unless you don't mind going to 4000C per w...) However, there could be a few problems at 2000C =P... and not to mention the fact that the meter can't read that high. Something like this perhaps could work as a terrible measuring device for 5-50W range devices at best.... So all of you that have those go ahead and try it out =/

Click to expand...

Can you cite some sources? Specifically, the math to justify your 5-50W range comment. I'm not disagreeing, but I'd like to understand how you calculated that figure.

rhd said:

Can you cite some sources? Specifically, the math to justify your 5-50W range comment. I'm not disagreeing, but I'd like to understand how you calculated that figure.

Click to expand...

My calculations may be wrong... however... lets see....


Say we want 1mw resolution and a 1W peak. That means our meter would accurately need to read 1000 steps. Now the meter is rated for 2 degree accuracy so that would define out step size... (ideally you would want a 4degree step size to account for less error). So we need 1000 steps with 2degrees celcius per step. Thus we need a material that heats up to 2000C when 1W is applied.... Which...simply doesn't work?

As for my 5-50 comment, its realistic to assume that power (1W++), the temp rise on devices would be large enough that it can offset the meter's hideous accuracy. Use say a crummy heatsink ..say 10C per 1W. So a 5W laser would read 50C and a 10W laser would read 100C.... something the meter could easily read ideally... The meter might work then However... thats just tooo hot and not good for most paints. However, decent for a rough calculation.

Laserbee has the right idea using the thermal couples because those are more sensitive.

That's where you've made a mistake.

The resolution (or "steps") aren't 2 degrees, they're 0.25 degrees.

Accuracy =/= resolution. Those are separate figures.

rhd said:

That's where you've made a mistake.

The resolution (or "steps") aren't 2 degrees, they're 0.25 degrees.

Accuracy =/= resolution. Those are separate figures.

Click to expand...

Then we shall agree to disagree... although the resolution may be up to .25 degrees it is pointless if that resolution exceeds the meters accuracy. At work we define our outputs signals in terms of our accuracy, not in terms of our resolution. Although we may have a 16bit ADC we can't possibly state a voltage output based upon the true resolution of that ADC because the noise floor is atleast 50bits out of the 16384 decimal bits...ect. If you neglect your accuracy and simply go by resolution your readings are flawed. In general you always want your step size to far exceed your accuracy for realistic readings. Basically the same concept as sig figs

This thread makes me wish I was an electronics wiz. It would be cool if this could be made to work. Anyone know what kind of temperatures lasers would raise the sensor up to? Say using a 1W

Helios said:

This thread makes me wish I was an electronics wiz. It would be cool if this could be made to work. Anyone know what kind of temperatures lasers would raise the sensor up to? Say using a 1W

Click to expand...

Like I said earlier... take a household or kitchen thermometer
and shine a laser on the sensor... You can do the test yourself..



Jerry

Random person:

My bathroom scale has an accuracy of +/- 3 lbs.

However, it can measure the difference between me, and me holding a 0.25 lb flash light.

That's resolution. To say that the resolution = the accuracy (+/-) in such a case is absurd.

rhd said:

Random person:

My bathroom scale has an accuracy of +/- 3 lbs.

However, it can measure the difference between me, and me holding a 0.25 lb flash light.

That's resolution. To say that the resolution = the accuracy (+/-) in such a case is absurd.

Click to expand...

Hmm.... i'm seeing your point but I'm still agreeing to disagree. While your weight scale at home may be able to read something below its accuracy, at that point you still have to question the true accuracy of that reading given the tolerance. Where as if you took a scale that was designed for 1lb max with say a .01 accuracy, you could safely assume your reading is accurate up to .1lb ..ect. So yes this would work for very high high powered lasers and a very rough approximation for lower powered lasers. Say I take the worse rated T0-220 heatsink on digikey and thermally adhere it to that sensor: http://search.digikey.com/scripts/DkSearch/dksus.dll?Detail&name=294-1049-ND , I still only got a 41.7C per W. That would make 41.7 / .25 --> 166 steps between 0 and 1W with + - ~7% accuracy. Generally, if I were to data-log I would want high precision + accuracy. All this coming from my ignorant, narrow minded EE design perspective. This idea wouldn't be for me, however, not to say its not for you