Radium vrs Radion

Centerline · Dec 8, 2017

So I'm just about to pull the trigger on a spectroradiometer. After speaking with a support person at Apogee I believe the SS110 should do the job (allow me to break down the spectral bands nm gradient by nm gradient). Does anyone have any input here @Dana Riddle - pull the trigger?

Thanks

Da Ghost · Dec 10, 2017

Taggging along.

Centerline · Dec 10, 2017

OK, @saltyfilmfolks - pull the trigger on the spectroradiometer

saltyfilmfolks · Dec 10, 2017

Centerline said:
OK, @saltyfilmfolks - pull the trigger on the spectroradiometer

You did it ?
The ss110?

Centerline · Dec 10, 2017

saltyfilmfolks said:
You did it ?
The ss110?

Not yet but my finger is very itchy. I need someone besides the tech at Apogee to confirm that this device is going to meet my needs - give me the ability to measure the individual nm gradients.

saltyfilmfolks · Dec 10, 2017

Centerline said:
Not yet but my finger is very itchy. I need someone besides the tech at Apogee to confirm that this device is going to meet my needs - give me the ability to measure the individual nm gradients.

Yea, good idea.
There wa one Dana reccomended to
@Grey Guy I belive.

JaimeAdams · Dec 10, 2017

You are running the Radium at 250W, not on an M80 ballast at 280W or an HQI which I believe is around 270W, is that accurate? From my understanding that will change the color temp, maybe others such as @hybridazn or @saltyfilmfolks or @Dana Riddle could comment on that.

Dana Riddle · Dec 10, 2017

JaimeAdams said:
You are running the Radium at 250W, not on an M80 ballast at 280W or an HQI which I believe is around 270W, is that accurate? From my understanding that will change the color temp, maybe others such as @hybridazn or @saltyfilmfolks or @Dana Riddle could comment on that.

Yes, the ballast will make a difference, even if it's the same brand/model. I know Sanjay did a lot of testing with ballasts but I don't recall exactly where he published. My advisor on ballasts was a Georgia Tech engineer.

Dana Riddle · Dec 10, 2017

Centerline said:
So I'm just about to pull the trigger on a spectroradiometer. After speaking with a support person at Apogee I believe the SS110 should do the job (allow me to break down the spectral bands nm gradient by nm gradient). Does anyone have any input here @Dana Riddle - pull the trigger?

Thanks

As much as I like Apogee products, you might want to consider other options. Choosing a spec that uses fiber optic patch cords will give you a lot more flexibility and allow underwater measurements, coral reflectance, absorbance and fluorescence. There's such a spec that's priced at $900 US, and I've heard good things about it. Let me do some checking...

Dana Riddle · Dec 10, 2017

Here’s one for $399. Less than $900 with all the bells and whistles. @kevensquint - You have one of these don't you? Comments?

https://www.pasco.com/spectrometer/

Ocean Optics STS fiber optic spectrometer. $1500 plus patch cord and software (~$300.) 1nm resolution depending upon slit size.

http://www.shop.spectrecology.com/STS-VIS-Sub-miniature-fiber-optic-spectrometer-STS-VIS.htm

Don’t know much about this one, but the price is OK:

http://www.shopstellarnet.com/low-cost-green-wave-spectrometer/

Centerline · Dec 10, 2017

JaimeAdams said:
You are running the Radium at 250W, not on an M80 ballast at 280W or an HQI which I believe is around 270W, is that accurate? From my understanding that will change the color temp, maybe others such as @hybridazn or @saltyfilmfolks or @Dana Riddle could comment on that.

Actually I just ordered a M80 from Hamilton. And your right - it would have changed the output a tad.
https://hamiltontechnology.com/index.php?route=product/product&product_id=180

Centerline · Dec 10, 2017

Dana Riddle said:
Here’s one for $399. Less than $900 with all the bells and whistles. @kevensquint - You have one of these don't you? Comments?

https://www.pasco.com/spectrometer/

Ocean Optics STS fiber optic spectrometer. $1500 plus patch cord and software (~$300.) 1nm resolution depending upon slit size.

http://www.shop.spectrecology.com/STS-VIS-Sub-miniature-fiber-optic-spectrometer-STS-VIS.htm

Don’t know much about this one, but the price is OK:

http://www.shopstellarnet.com/low-cost-green-wave-spectrometer/

WOW! Thanks for taking the time to check these out Dana - I appreciate your help as I dont know what I dont know

. Reading the docs on the Ocean optics now. Unless Im able to get that 1nm resolution the experiment itself makes no sense.

Dana Riddle · Dec 11, 2017

Centerline said:
WOW! Thanks for taking the time to check these out Dana - I appreciate your help as I dont know what I dont know . Reading the docs on the Ocean optics now. Unless Im able to get that 1nm resolution the experiment itself makes no sense.

You're welcome! Please keep us posted. One comment - I'm not sure what the scope of your experiments are, but is 1nm resolution really necessary? There's going to be some noise in the signal and this results in 'spikey' readings. There are a couple of ways to compensate - averaging the number of measurements will give a smoothed reading. Another way is to 'boxcar' the reading. In short, a measurement of 1nm resolution often gets smoothed out to give a 'pretty' picture. If you plan to make comparisons, it might be best to standard your measurement parameters, such as integration time, averaging, and boxcar smoothing. The technical staff of Ocean Optics is very good, and I suppose others are as well, but let me know if I can be of any assistance. One last comment - competition usually brings prices down. Not so with Ocean Optics, the price of a USB spec comparable to the two I own has gone up by 50% over the years, and their software by 100%. On the other hand, I've had one of the specs for almost 18 years, and it is rock solid.

Centerline · Dec 11, 2017

Dana Riddle said:
You're welcome! Please keep us posted. One comment - I'm not sure what the scope of your experiments are, but is 1nm resolution really necessary? There's going to be some noise in the signal and this results in 'spikey' readings. There are a couple of ways to compensate - averaging the number of measurements will give a smoothed reading. Another way is to 'boxcar' the reading. In short, a measurement of 1nm resolution often gets smoothed out to give a 'pretty' picture. If you plan to make comparisons, it might be best to standard your measurement parameters, such as integration time, averaging, and boxcar smoothing. The technical staff of Ocean Optics is very good, and I suppose others are as well, but let me know if I can be of any assistance. One last comment - competition usually brings prices down. Not so with Ocean Optics, the price of a USB spec comparable to the two I own has gone up by 50% over the years, and their software by 100%. On the other hand, I've had one of the specs for almost 18 years, and it is rock solid.

I have assumed that the 1nm resolution is important as I'm pretty sure there will be parts of the spectrum missing with the LEDs - a diode that produces 420 will not produce at 425 - correct? So lets say a Radium has a small peak at 425 and the LED doesn't - is this important?
Ill be ordering the Oceans Optics meter today - thanks again for the advice and regardless of the price its still a good bit less expensive than the Apogee

Dana Riddle · Dec 11, 2017

Centerline said:
I have assumed that the 1nm resolution is important as I'm pretty sure there will be parts of the spectrum missing with the LEDs - a diode that produces 420 will not produce at 425 - correct? So lets say a Radium has a small peak at 425 and the LED doesn't - is this important?
Ill be ordering the Oceans Optics meter today - thanks again for the advice and regardless of the price its still a good bit less expensive than the Apogee

LED's output are bell curves, so you could have a resolution of several nm and still have a pretty good idea of the spectral quality. Comparing spectra should be a breeze regardless of output - which brings up a point: Bandwidth measured by a spec varies, so when speaking to an OOI engineer you'll be asked to specify if you want to measure ultraviolet/visible radiation or visible/infrared. Hope this helps. Any questions, I'm here.

Centerline · Dec 11, 2017

Dana Riddle said:
LED's output are bell curves, so you could have a resolution of several nm and still have a pretty good idea of the spectral quality. Comparing spectra should be a breeze regardless of output - which brings up a point: Bandwidth measured by a spec varies, so when speaking to an OOI engineer you'll be asked to specify if you want to measure ultraviolet/visible radiation or visible/infrared. Hope this helps. Any questions, I'm here.

And my answer should be ultraviolet/visible. Correct?

Dana Riddle · Dec 11, 2017

Centerline said:
And my answer should be ultraviolet/visible. Correct?

I looked at the STS and the only option I see is 350-800nm, which is unusual (there were many options on the USB2000.) This should allow you to see LEDs' UV and chlorophyll fluorescence, so, yeah, it would work for me. Slit size is important. The larger the slit, the less the resolution but works in very low light situations (but if you decide to go with a narrow slit for 1nm resolution, the software should allow a high integration time for use in low light.)

Centerline · Dec 12, 2017

Dana Riddle said:
I looked at the STS and the only option I see is 350-800nm, which is unusual (there were many options on the USB2000.) This should allow you to see LEDs' UV and chlorophyll fluorescence, so, yeah, it would work for me. Slit size is important. The larger the slit, the less the resolution but works in very low light situations (but if you decide to go with a narrow slit for 1nm resolution, the software should allow a high integration time for use in low light.)

I'm so out of my depth here lol and sooner or later your going to get tired of all these questions lol.

With respect to low light and 1nm resolution, I don't want to anything that is going to effect the veracity of LED results but I do want the granular data. I have been a high level database/systems/software engineer for the last 27 years and I crave granular data (Yes, I just publicly admitted I have a small fetch for data sets). So I guess I need to ask the techs at OOI if the software will allow for the high integration time.

Re-reading your comments above I note you mentioned LED output are bell curves. So does this mean that a 420 nm emitter may actually be dipping into 410 or climbing into the 430 range, peeking at 420? I have always thought that the emitters had a distinct spectrum profile and that was it.

Dana Riddle · Dec 12, 2017

Here is a graph showing LED emissions (blue, green, yellow, red.) As you can see, LEDs peak at a given wavelength, but the bandwidth can be rather large. It is for this reason that Forward Width Half Maximum (FWHM) is often reported - this is the bandwidth where intensity is above 50% of maximum. Ocean Optics software OceanView should allow integration times of your choosing - microseconds to seconds.