Which budget Laser Power Meter for measuring telescope mirror reflectivity?

[Tatsu]

I am an amateur astronomer and would like to test the reflectivity of my telescope mirrors. My idea is to use one of my laser pointers (nominal 20mW and 100mW) and measure the intensity before and after the reflection in the mirror. Typical mirror reflectivities in astronomy are 96-99%, which means that the laser power meter (LPM) should have an accuracy of at least 1%. Is there any affordable LPM on the market that would do the trick? Affordable in my case would mean max around 250 US$.
I have the possibility to buy a second hand Sanwa LP-1 / Sper Scientific 840011.
rebrand.ly/nrdep6
Will this LPM be able to do what I want?

 

[brendon7358]

Would be way too unreliable. Lasers measure more of a difference than that measuring them 1 minute apart. I don't know if the laser is unstable or if it's the LPM. Maybe it could be done with a lab laser and an lpm in an extremely temperature controlled environment, but with just regular hobbyist stuff no way.

 

[Tatsu]

Brendon, thank you for your near immediate reply. I am not sure if I understand you correctly. Are you saying that temperature shifts from one measurement to the other - in the environment or within the laser - will create an error margin far above the intended accuracy of the measurement?

 

[brendon7358]

Brendon, thank you for your near immediate reply. I am not sure if I understand you correctly. Are you saying that temperature shifts from one measurement to the other - in the environment or within the laser - will create an error margin far above the intended accuracy of the measurement?

Yes, the LPM measures based on how much heat the laser produces on the thermopile. And also the laser itself is inherently unstable due to voltage, and temperature fluctuations. You can overcome this to an extent with a lab laser though.

 

[Tatsu]

Lab laser means a whole different price point for sure...
I was hoping that the LPM I mentioned didn't use the heat method for measuring. I read there are other methods such as the one used in
the Newport 820, which can be had at low price off E-Bay. Would something like this be able to help me? Or are you basically saying that irrespective of the measurement method used, the error margin wll be too large in comparisn to the desired accuracy of the measurement?

If you happen to know an LPM, which you think can serve me, I would appreciate any advice in this direction. I would also consider a device at a somewhat above my self imposed financial limit.

 

[Anthony P]

I have a Sperr and some known mirrors, and a couple calibrated/ certified lasers. I will try it and get back to you when I have time.

 

[RA_pierce]

While I can't comment on the LPM aspect, I have to agree with brendon about the reliability of most lasers we build and use here.
Your best bet is to go with a low power diode laser since these will be relatively stable. However, if the power output is low, you will need an LPM with very high resolution if you are to detect a 1-4% loss to the mirror.
DPSS lasers are not ideal because they are more sensitive to temperature than diodes. Additionally, higher power diode lasers produce a lot of waste heat which can result in fluctuations in power and shifts in wavelength.

A 200 mW 650 nm laser diode would be my first choice for something inexpensive and stable over its operating temperature but then you will need to reliably measure a loss of 2-8 mW.

If you can obtain such a laser, it may be possible to accomplish what you want.
My naive approach would probably go something like this:

1. Measure output of laser every ~2 seconds for 2 minutes.
2. Allow laser to cool to room temperature for several minutes.
3. Measure output of laser every ~2 seconds after telescope optics.
4. Repeat as many times as necessary to obtain reliable averages for your control (without telescope) and experimental (with telescope) runs.

It would also be a good idea to randomize the order of your control and experimental runs with a coin flip to ameliorate bias.
Some other considerations that may be important:

• distance the beam travels before reaching the LPM should be standardized. This means you will need to make sure the beam travels the same distance when it does not bounce off the mirror and when it does.
• the telescope mirror will distort the beam profile - the mirror is slightly concave, so you can expect the beam diameter to be smaller after reflection (before the beam waist). This shouldn't be a big deal, since a smaller beam will more easily fit onto the LPM sensor. It could be an issue if the beam expands substantially before it gets to the sensor. Any light that is clipped due to the beam not fitting on the sensor will bias your results.
• more power will provide better accuracy but may result in noisier measurements and poses higher risk of injury or damage to your equipment.

Typically, the power of a diode laser will slowly drop during continuous use unless the laser features a feedback and power control circuit and/or temperature control.
By plotting the results from each of many runs and doing a quick t-test, you may be able to get a good idea of what the loss to optics is.

I think this sounds like a really cool project and I would be interested to see if this works with hobby equipment.

Do you have access to standard protocols for doing these tests?

 

[paul1598419]

I am unaware of ANY LPM that has an accuracy of 1% or less. That includes professional ones. You can likely get a better measurement using some other method. I'm sure this has been done before so a search is your best option. Not all LPMs use thermopiles. Some lower power ones use optical sensors, but the accuracy is still limited.

 

[Anthony P]

Just a FYI thing. In the 4mW range, the Sperr has a resolution of .001mW. I will be using a Spectra-Physics HeNe at .75mW, target mirror will be a brand new Thor Labs 70/30 beam splitter. I have no idea what kind of accuracy or precision to expect, but I should have time to try later today.

 

[lasersbee]

If you are doing relative measurements the Optical Sensor
Sper could work for you. If you want actual NIST traceable
power measurement accuracy then the Sper LPM will not
cut it.
BTW... the Sper can only measure up to 40mW.

Jerry

 

[Cyparagon]

Accuracy and repeatability aren't the same, and OP needs the latter. For instance, if a meter is off by 5%, but always off by 5%, it would be a better choice than a meter that was within 1%, but sometimes varies as low as -0.5% or as high as +0.8%.

A respected PL member has some insight which I certainly concur with, but wishes to relay this anonymously, so I will share the gist here:
"He needs a silicon-based detector for reliability and accuracy, and a He-Ne laser for output stability (after a reasonable warm-up). Forget using the diode laser pointer".

A low power diode with no optical feedback might have stability close to that of a HeNe with the proper driver and after a long warmup, but if you're got an optical sensor that would be easy to confirm with your setup. Use an optical sensor or ask a professional.

 

[Alaskan]

First thing which came to mind for me was also the silicon-based detector.

 

[Anthony P]

Cyp really hit it on the head. Hand held diode lasers are useless for this experiment ( or nearly any other). Precision is far more important than accuracy. It's not like anyone would take one single reading and call it good enough.
I did complete a set of experiments and here are the results... draw your own conclusions.

Laser1= Spectra-Physics .5mW HeNe
Laser2= Melles Griot 7mW HeNe polarized
Meter= Sperr 840011 with silicon diode sensor
mirror= Thor Labs BST 10R 70/30 beam splitter

With laser1, and meter set to 4mW range, hold max, 10 sec exposure; standard deviation came in at 1.72%. Raw measurements had a mean of 67.1%R which is very nearly identical to the Thor Labs data for this mirror, unpolarized, 45 AOI.

With laser2, meter set to 40mW range, max hold, 10sec; deviation came in at .0004%( Did not allow for Sig Figs). raw mean of 72.94. This is nearly 10% off of expected values... accuracy vs precision?

My personal conclusion is that this set up can in fact give you amateur "ball-park" indications of a mirrors reflectivity, but professional level readings would require professional equipment.

 

[BowtieGuy]

This looks like it would do the job nicely, but I can't imagine what it would cost.

PRONTO-Si - Low power portable laser power meter

Discover our PRONTO-Si laser power meter. This portable device is ideal for measurement up to 800mW. Gentec-EO manufactures high accuracy laser beam measurement devices.

www.gentec-eo.com

 

[bertha]

While I can't comment on the LPM aspect, I have to agree with brendon about the reliability of most lasers we build and use here.
Your best bet is to go with a low power diode laser since these will be relatively stable. However, if the power output is low, you will need an LPM with very high resolution if you are to detect a 1-4% loss to the mirror.
DPSS lasers are not ideal because they are more sensitive to temperature than diodes. Additionally, higher power diode lasers produce a lot of waste heat which can result in fluctuations in power and shifts in wavelength.

A 200 mW 650 nm laser diode would be my first choice for something inexpensive and stable over its operating temperature but then you will need to reliably measure a loss of 2-8 mW.

If you can obtain such a laser, it may be possible to accomplish what you want.
My naive approach would probably go something like this:

1. Measure output of laser every ~2 seconds for 2 minutes.
2. Allow laser to cool to room temperature for several minutes.
3. Measure output of laser every ~2 seconds after telescope optics.
4. Repeat as many times as necessary to obtain reliable averages for your control (without telescope) and experimental (with telescope) runs.

It would also be a good idea to randomize the order of your control and experimental runs with a coin flip to ameliorate bias.
Some other considerations that may be important:

• distance the beam travels before reaching the LPM should be standardized. This means you will need to make sure the beam travels the same distance when it does not bounce off the mirror and when it does.
• the telescope mirror will distort the beam profile - the mirror is slightly concave, so you can expect the beam diameter to be smaller after reflection (before the beam waist). This shouldn't be a big deal, since a smaller beam will more easily fit onto the LPM sensor. It could be an issue if the beam expands substantially before it gets to the sensor. Any light that is clipped due to the beam not fitting on the sensor will bias your results.
• more power will provide better accuracy but may result in noisier measurements and poses higher risk of injury or damage to your equipment.

Typically, the power of a diode laser will slowly drop during continuous use unless the laser features a feedback and power control circuit and/or temperature control.
By plotting the results from each of many runs and doing a quick t-test, you may be able to get a good idea of what the loss to optics is.

I think this sounds like a really cool project and I would be interested to see if this works with hobby equipment.

Do you have access to standard protocols for doing these tests?

Agree your opinion

 

[Singlemode Laser]

You don't need an accurate power meter if you want to measure the reflectivity of a mirror, just a precise one.

The weapon of choice here is a large photo diode, reverse biased (for a highly linear response) and a resistor to convert current to voltage. Photodiodes don't have the thermal problems and low sensitivity of thermal sensors. You can easily detect micro watts of power. No need to calibrate if only relative measurements are performed (like in your case). A multimeter or oscilloscope you can easily get 4 digits and 6-8 are possible with averaging.

Just divide the voltage you get after the mirror by the voltage you get before the mirror = reflectivity of the mirror for the wavelength used. Done this dozens of times myself.

Singlemode