Spectrometer ratings?

[Nonya]

Have you shared your results?

The Hopoocolor will read down to 350nm, but it is more expensive; however, given availability, the Spectryx option would be a good thing for the hobby. Besides GHL and Kessil, most LED lights don't have UV diodes, so looking below 400nm isn't needed.

Have you compared the Hopoocolor to the UV and fluorescent spectra? If so, how well did it match up? Was it off near the 350 end?

 

[oreo54]

Again, you feel the need to critique something you've never touched.

How about that French spectrometer? How accurate is it across the vis spectrum? What kind of resolution? That might be an actual contribution to the thread.

Already did that. Somewhat

Spectrometer ratings?

I bought a Thunder Optics spectrometer from France awhile ago. Using this software: https://www.effemm2.de/spectragryph/ It was adequate for my purposes. Accuracy? Resolution? How did you calibrate it?

www.reef2reef.com

 

[AKReefing]

Already did that. Somewhat

Spectrometer ratings?

I bought a Thunder Optics spectrometer from France awhile ago. Using this software: https://www.effemm2.de/spectragryph/ It was adequate for my purposes. Accuracy? Resolution? How did you calibrate it?

www.reef2reef.com

Can you do a little better? Maybe overlay your readings with the fluorescent and/or UV spectra for a visual comparison? Just a request. Does it match precisely, is it total junk, or maybe somewhere between?

 

[AKReefing]

Thanks. I was about to request that someone delete the whole thread. Very few of the responses seem to address the original post. I'm still a newb to this site. Is it always like that here?

 

[telegraham]

What @oreo54 is sharing is simply direct commentary, and I agree with them. I've not read their comments as a hit piece. Like, not at all. As for what you're using, it's cool, but lack of availability makes it unattractive to me; but that's easy for me to say since I have several spectrometers.

Hoping you test some lights and share the data. Doesn't really matter what spectrometer you have (as long as it's not a PARwise). They all do a good enough job.

Happy testing!

 

[telegraham]

So here's the UV chart. It appears the Spectryx drops off somewhere just below 400nm.

I missed this. Thank you! Will compare. Does your software allow for Excel export? If I've already asked, sorry.

 

[AKReefing]

I missed this. Thank you! Will compare. Does your software allow for Excel export? If I've already asked, sorry.

I saw the same request from someone else a few days ago. I've been too busy with chores to check it out. I have to connect the meter to the laptop to get the software to function. It locks up otherwise. It's on my list.

The white fluorescent overlay is in https://www.reef2reef.com/threads/trying-out-a-new-spectrometer.1028010/page-2 See post #34. As with the UV test, the white fluorescent peaks appear to be pretty much an excellent match to the reference graph. I just with it had some capability below 400nm. I'm not willing to pay the extra just to get that range right now.

There are a couple of used OceanOptics USB 2000+ and Spectral Products SM240/6 spectrometers out there, but they're priced over what I'm willing to spend just to get below 400nm. They probably qualify as lab grade, but not wallet grade.

 

[DaJMasta]

Thanks for the input, but I made a simple request: "For those of you who own one or more spectrometers, please list them, describe how you'd rate it/them, and why."

Please note that I didn't ask for data sheets, how much they spent, whether they work in a lab or perform scientific research, etc. Can we PLEASE try to stick to that?

You're missing the point of what I said, your "simple request" isn't a particularly useful way to pose the question you probably want answered, but fine:

I own a couple of Ocean Optics (now Ocean Insight) spectrometers and a couple of monochromators. All of them would be prohibitively expensive for a hobbyist to buy new and require other equipment to be used in a reef tank - at least an optical fiber and a cosine corrector - and then would need to be calibrated for that (I've got a set of tungsten ribbon lamps with calibration data, but they're from the 80s so even with the low hours of usage the irradiance calibrations I can do at home aren't necessarily traceable to national/international standards.) Even then, it's a laboratory production to setup all the parts and computer to take a measurement in a meaningful, repeatable way.

I can recommend most of the Ocean Optics line I've worked with (USB2000, USB4000, HR4000, Maya2000 Pro) in terms of size and capability, and their software is easy to use and is fully featured, but you will be shelling out for any of it (software included), they come in dozens of configurations per model (plus custom configurations), and any of them would be overkill for hobbyist lighting measurement. Older USB2000s (in particular, with their old logo) will only operate with older operating systems, because of a limitation in their USB chip's firmware, so getting a somewhat newer version (bigger logo on it) means you can use it with a modern OS or in a VM. They're also ill suited to working near saltwater, so getting the sensing probe and suitable fiber are a pain. Any of the mentioned spectrometers would not give you more than an approximation of the light intensity at the input of the cosine corrector until they've been calibrated with a reference source, so none of them would be suitable for measurements as a standalone product, all of them need accessories and calibration with them in place.

A spectrometer is a scientific instrument, so the manufacturer providing detailed specifications and the operating conditions they are valid for is essential for its usefulness, and a spectrometer sold without specifications (at least typical, if not guaranteed/calibrated) is really not of all that much use, so a discussion of it experientially will really only tell you something if your experience differs from rated performance. The majority of available spectrometers are also configurable platforms, so the spectral bandwidth, resolution, sensitivity, and off axis/out of band rejection is going to vary based on configuration - another reason why a discussion of "is spectrometer x better than spectrometer y" isn't that valuable of a discussion to have. I've got a USB4000 that's configured for the near UV to near IR band with a spectral resolution of something like 5-7nm FWHM, but the very same spectrometer can have a different grating to cover almost any band from around 180nm UV to 1100 nm NIR, and with narrower bands you'll get finer spectral resolution because of the math inherent in the configuration. An inline lens with the fiber or wider entrance slits can increase sensitivity, but both decrease spectral resolution - it's a tradeoff you have to decide on for your application. Then you can move to a larger format design like the HR4000 - the same interface, software, and basic configuration options, but because of the longer path length and slightly different optical elements, the FWHM for a similar bandwidth configuration as above is something like 2nm - 3x better or more.

You probably don't need it to be that accurate in terms of wavelength nor in terms of spectral resolution, since reef tank applications are generally going to be for characterizing lighting rather than spectroscopy or applications where narrow peaks just don't matter as much. What you really need is consistent amplitude sensitivity, and by definition that's going to need to be calibrated because of how nonlinear the detectors used are (take a look at the spectral response for this line sensor - similar idea to what will be in any USB spectrometer: https://www.hamamatsu.com/us/en/pro...e-sensor/for-spectrophotometry/S11639-01.html ), so again the important feature to look for in a spectrometer for reef lighting is going to be its irradiance calibration.

Something like the Seneye Spectra is better suited to the application in terms of usability, though I can't speak to its performance myself. LiCor also has some field-use instruments (not as many spectrometers, but them too) which could be an option, but generally speaking, the stuff that will be convenient to use in and around saltwater are going to be boutique manufacturers making measurement equipment for ocean research and you are going to be shelling out thousands to even get your foot in the door. Again, it's hard to recommend a specific unit or brand when you're sort of at the mercy of availability to be able to even afford one or when one is a physical spectrometer, one is a remote sensing head, one is a handheld device, etc.

From the thread last linked, it looks like you're trying to evaluate UV performance for a spectrometer you have? Again, the datasheet is the answer as it should give you a QE or at least relative efficiency curve across its operating band (and if one isn't present shorter than 400 nm, it either hasn't been characterized or can't be counted on and shouldn't be used for a real measurement because of that). If you want to try and evaluate out of band performance on your own, you need a source with a known response (calibration lamp) to map it against and a test fixture which can keep out stray light (at least from the specra posted, your fluorescent lamp plots sure look like they have a lot of background light bleeding in.)

So maybe the better choice than asking for recommendations for the few people with them across the hundreds of models from dozens of manufacturers in thousands of configurations is to find out what you're trying to do and prioritize with your measurement, and then maybe look for recommendations of how to get that/make it better. There are dozens of options to do what you want, minimum, but the majority of them are going to be difficult to get a hold of or aren't well suited to workflow around a tank.

 

[AKReefing]

You're missing the point of what I said, your "simple request" isn't a particularly useful way to pose the question you probably want answered, but fine:

The reality is that I'm a mere reef hobbyist in a reefkeeping forum just asking what other reef hobbyists have and what they like and dislike about them. The answers could very well be of use to me--as a non-scientist or industry professional.

It was, and is, a simple request that you interpreted differently.

Let me explain why I care only about wavelength accuracy and resolution. My plan is to build an LED fixture that maximizes the coverage between 400 and just over 500nm. 380 to 400 would be a bonus, but having the capability to measure that small range isn't worth the extra hundreds of dollars. It would be great if I could somehow set it up so that the FWHM pointsfrom 400 to 500nm intersected, not for photosynthetic response, but for fluorescence. Out of that effort I've decided that a spectrometer could be useful. I've looked a a LOT of spectrometers online, most of which are cost prohibitive. Out of that search/research, I've learned a bit, and was wondering about other hobbyists' experiences. Nothing more than that.

I've got a USB4000 that's configured for the near UV to near IR band with a spectral resolution of something like 5-7nm FWHM

I'm happy with mine, which appears to be sub-10nm resolution at the FWHM points, particularly in the 400-500nm range. That's more than good enough for my needs.

You probably don't need it to be that accurate in terms of wavelength nor in terms of spectral resolution,

I do. See above.

since reef tank applications are generally going to be for characterizing lighting rather than spectroscopy or applications where narrow peaks just don't matter as much.

I don't need or want characterization. I need wavelength accuracy and resolution. Again, please see above.

From the thread last linked, it looks like you're trying to evaluate UV performance for a spectrometer you have?

Luxdium recommended that I use a mercury vapor-based UV sterilizer with the Spectryx to check accuracy. I did and it looks like I was within 1 nm of the mercury peaks above 400nm (404-577nm). Surprising. Prior to that I compared it to the white fluorescent bulb standard peaks.

Again, the datasheet is the answer as it should give you a QE or at least relative efficiency curve across its operating band (and if one isn't present shorter than 400 nm, it either hasn't been characterized or can't be counted on and shouldn't be used for a real measurement because of that). If you want to try and evaluate out of band performance on your own, you need a source with a known response (calibration lamp) to map it against and a test fixture which can keep out stray light (at least from the specra posted, your fluorescent lamp plots sure look like they have a lot of background light bleeding in.)

Mine doesn't have a data sheet, but matched both the UV sterilizer and a white fluorescent bulb spectrum extremely well. I can't see why an average reefer would need more.

So maybe the better choice than asking for recommendations for the few people with them across the hundreds of models from dozens of manufacturers in thousands of configurations is to find out what you're trying to do and prioritize with your measurement, and then maybe look for recommendations of how to get that/make it better.

I didn't ask for recommendations, just personal evaluations. Perhaps I should have asked for them to answer from a reef hobbyist's POV.

There are dozens of options to do what you want, minimum, but the majority of them are going to be difficult to get a hold of or aren't well suited to workflow around a tank.

Let's hope smart people take measurements away from the tank, or use a long fiber.

Sorry if this is a bit disjointed or redundant, but we're at the 2 minute warning.

 

[DaJMasta]

I don't need or want characterization. I need wavelength accuracy and resolution. Again, please see above.

I've said my piece, there's no use belaboring the points, but as a clue to why you definitely do not need a tiny FWHM: take a look at the spectral output chart for any LED you're considering, UV included. Unless you're trying to make a fixture out of laser diodes, those peaks are no where near as narrow as those found in fluorescent lamps, so being able to discern extremely narrow peaks is not actually going to offer you any additional performance.

If your peak is 10-30nm wide and varies by 10nm between batches, you need neither spectral resolution nor wavelength accuracy unless you're going to be binning your own LEDs.

 

[Reefing Madness]

For those of you who own one or more spectrometers, please list them, describe how you'd rate it/them, and why.
Theres is the OPs original question, really is no need for the bickering going on.
thanks

 

[AKReefing]

For those of you who own one or more spectrometers, please list them, describe how you'd rate it/them, and why.
Theres is the OPs original question, really is no need for the bickering going on.
thanks

Thank you. My word would have been "pedantic".

 

[AKReefing]

So I got a reply from the "maker". They've posted two on eBay. The look is different. I don't like that they say 350nm is typical, which it definitely isn't for mine.

I complained to the maker about the price increase. They dropped it to $250 for now.

 

[AKReefing]

I missed this. Thank you! Will compare. Does your software allow for Excel export? If I've already asked, sorry.

Finally got around to testing this. The default file is a space delimited text file. I just measured a set of LEDs sold on Amazon as 405nm. The peaks come in at 405-406, which gives me confidence in the Spectryx, LEDs and the truthfulness of the seller, who also has other LEDs in the UVA range spaced 10nm apart. These LEDs have a wide spike, above 40nm at FWHM. The UV sterilizer at about the same point was well under 10nm, proving the LED is definitely not narrow band. That could prove useful.

I graphed the text file and noticed it actually records data all the way down to 200nm--not that it actually detects it. Above 300nm it actually shows what could be detections. The GUI doesn't seem to show anything below 370, and the chart stops at 350. Could just be a problem with the GUI if it's actually showing good data in the text file. I'll give the UV sterilizer another shot tomorrow and see what the text file says. Could be interesting.

Here's the 405nm LED text file and the graph created from the text file exported to Excel, graphed and saved as a JPG..

 

[Nonya]

Finally got around to testing this. The default file is a space delimited text file. I just measured a set of LEDs sold on Amazon as 405nm. The peaks come in at 405-406, which gives me confidence in the Spectryx, LEDs and the truthfulness of the seller, who also has other LEDs in the UVA range spaced 10nm apart. These LEDs have a wide spike, above 40nm at FWHM. The UV sterilizer at about the same point was well under 10nm, proving the LED is definitely not narrow band. That could prove useful.

I graphed the text file and noticed it actually records data all the way down to 200nm--not that it actually detects it. Above 300nm it actually shows what could be detections. The GUI doesn't seem to show anything below 370, and the chart stops at 350. Could just be a problem with the GUI if it's actually showing good data in the text file. I'll give the UV sterilizer another shot tomorrow and see what the text file says. Could be interesting.

Here's the 405nm LED text file and the graph created from the text file exported to Excel, graphed and saved as a JPG..

The UV sterilizer doesn't register at all below 400nm. I'm wondering if it actually has anything below 400.

 

[DaJMasta]

The UV sterilizer doesn't register at all below 400nm. I'm wondering if it actually has anything below 400.

It's probably emitting at shorter wavelengths than can be detected by the spectrometer you're using. While UV begins at 400nm, until 360nm or so, it's far UVA - basically blacklight - and has almost negligible sterilization effect except at very high intensity. That means UV bulbs intended for sterilization use much shorter wavelengths - typically 254nm - which is much farther outside of the visible band. Unless your spectrometer is specifically made for its detection, it's outside of the detection range for normal silicon based sensors, so even if the graph were to stretch to that short of a wavelength, the overall efficiency of turning light intensity into signal would be so low it would barely be represented. Fluorescent bulbs are also generally emitting fairly narrow spectral lines, so you don't get as much bleed through in other bands that may be easier to detect.

Depending on the spectrometer's design, it also could be that they just filter out the light. Removing out of band interference helps noise floor and can effect accuracy, so it could be there's a long pass filter built into the design that simply blocks UV components to make the intended range work better and prevent damage from wavelengths it can't discern.

 

[Nonya]

Yeah. I my buy a few of the 380nm LEDs as an experiment.

 

[AKReefing]

Yeah. I my buy a few of the 380nm LEDs as an experiment.

What spectrometer do you have?

 

[Nonya]

What spectrometer do you have?

I don't have one at the moment. Somebody in our local group does, though. He also has some kind of UV sterilizer, but I think it's the submerged and enclosed type, so I don't know if he can even get to the light source.

 

[Assirem]

Accuracy? Resolution? How did you calibrate it?

Hello,
I also have one spectrometer of Thunder Optics.
I got it a year and half ago, it is very stable for now, and the resolution is less than 0.5nm, it is terrific for this price and the calibration file comes with the spectrometer in USB flash memory.
I am using it too to monitor my aquarium lamps.
Assi