Forum der Vereinigung der Sternfreunde

Forum of the German Amateur Astronomy Association
It is currently 22. April 2019, 21:58:47 PM

All times are UTC+02:00




Post new topic  Reply to topic  [ 26 posts ] 
Author Message
PostPosted: 23. October 2015, 09:16:36 AM 
Offline
Meister
User avatar

Joined: 11. April 2007, 10:47:33 AM
Posts: 384
Location: Sydney
Hi guys,

I was always wondering why the efficiency of my little single-mode spectrograph is not quite as high as expected. The spectrograph (RHEA) I am talking about uses an R2 31.6 g/mm echelle grating from Thorlabs (GE2550-0363) as main disperser. The overall efficiency from the fiber input to the focal plane was never higher than about 14 %, when launching light from a HgAr source through the system. Here I was using the (green) 546 nm emission line only. Well, the line was not falling directly on the blaze peak of that particular order but more than 20-22% total efficiency was anyway not feasible.

So, what could that else be? The throughput of my collimator (which I use in double-pass), the prism-cross disperser and the injection optics perform within the specs. But what about the echelle grating? There is basically no graphs figures available stating the performance of that grating. As I did not find any further details/specs/measurements of this particular grating I did my own using a 658 nm laser, a photodetector and the grating in Littrow configuration. So, by turning, not tilting, the grating, (i.e. changing the off-plane angle gamma) this changes the diffraction angle beta, which basically means my laser source is sweeping over the blaze peak which tells me the maximum efficiency.

(From the grating equation we know, incident angle, grating constant, wavelength (here laser) and echelle order are constant (when remaining in the same order). By changing gamma, beta changes which sweeps the emission line through the order (is that correct explained???)

The obtained result is quite depressing …. 38% efficiency only. What the hell???? Gratings which can be purchased on the Newport website (Richardson Gratings) perform way better with efficiencies beyond 60%, depending on the master, sometimes up to 80%!!!!

So, why are these off-the-shelf gratings do perform so bad? The answer is obvious when doing an optical surface analysis, by means looking a bit closer at the problem. The attached figure shows an image of what we call an ideal echelle profile and another image of the actual Thorlabs echelle. Light in an echelle spectrograph is striking the smaller facets (like I draw the rays) and gets diffracted backwards. Well, not all of the light in the Thorlabs echelle does this. You can clearly see a chamfer where we would expect a perfect sharp edge. So, what do we learn from that? If you are striving for maximum efficiency, do it properly and don’t save on the wrong end.

I have an identical grating laying around which I will test in the next days. The other one could be of course an outlier but if the second grating shows a similar performance (comes a different batch) we know at least what’s going on here.

Have a nice weekend!

Cheers
Tobias


Attachments:
echelleComparison.jpg
echelleComparison.jpg [ 239.49 KiB | Viewed 4858 times ]

_________________
Dein Projekt. Mach es fertig, bevor es Dich fertig macht!
Top
   
PostPosted: 23. October 2015, 10:15:33 AM 
Offline
Meister
User avatar

Joined: 31. July 2006, 16:43:32 PM
Posts: 3325
Location: Köln
Hi Tobias, your investigation is very interesting! It shows the importance to check all different optical elements in detail.
Quote:
(From the grating equation we know, incident angle, grating constant, wavelength (here laser) and echelle order are constant (when remaining in the same order). By changing gamma, beta changes which sweeps the emission line through the order (is that correct explained???)
I believe, this explanaition is correct.
Quote:
Well, not all of the light in the Thorlabs echelle does this.
Some time ago Ulrich Waldschläger scrutinized the surface quality of our R2 31.6 g/mm echelle grating from Thorlabs with a microscope at Bruker Nano in Berlin (unfortunately I do not know the item number). The result is attached (see page 252 in our book - Note that we also discuss the influence of the angle of incidence on the Echelle efficiency in chapter 6.3.3 on page 248). There seems to be a chamfer at each grating line but certainly not as prominent as in your grating. I do not know if this is a general production difference or just a coinsidence. We also have a Zeiss grating and it might be interesting to see its quality. It might be possible that a perfect line geometry can not be manufactured (remember the production process) but I fully agree that this is an important issue.

Another potential issue is the positive and negative aberration as we describe in figure 6.14 on page 252.

Cheers, Thomas


Attachments:
echelle_2000.jpg
echelle_2000.jpg [ 151.74 KiB | Viewed 4848 times ]
Top
   
PostPosted: 23. October 2015, 17:30:48 PM 
Offline
Meister

Joined: 13. August 2009, 17:04:50 PM
Posts: 263
Location: Maienfeld, Schweiz
Hello Tobias,

in my Echelle I also used a Thorlabs grating with blaze angle 63° and 79L/mm. I measured the efficiency by bypassing the echelle grating with a plane mirror and comparing the signal strength of the spectra with and without echelle grating (flux/A). The estimated echelle efficiency was calculated at 37%, close to what you have measured. Thorlabs did not specify the efficiency at that time (2012). The low efficiency was quite disappointing and I used the spectrograph for some spectra in the "low resolution mode" bypassing the echelle grating. Enclosed you find a sketch of my setup. I did not measure the actual profile of the grating but assumed that rounded corners were responsible.

Regards, Martin


Attachments:
high low resolution echelle.pdf [105.49 KiB]
Downloaded 78 times
Top
   
PostPosted: 23. October 2015, 18:16:19 PM 
Offline
Meister
User avatar

Joined: 19. August 2006, 11:45:20 AM
Posts: 2719
Location: 66907 Glan-Münchweiler
Hi, Tobias and Martin,

rather interesting observations !!

since Martin demonstrate the same result, I estimate, the low efficiency of the thorlab gratings seems to be the clear reason. Unfortunately I do not know, which grating is built in in the eshel from shelyak. Perhaps anybody knows this.......

regards
berthold


Top
   
PostPosted: 23. October 2015, 22:41:29 PM 
Offline
Meister
User avatar

Joined: 01. August 2006, 16:08:56 PM
Posts: 3806
Location: 68163 Mannheim
Quote:
Some time ago Ulrich Waldschläger scrutinized the surface quality of our R2 31.6 g/mm echelle grating from Thorlabs with a microscope at Bruker Nano in Berlin (unfortunately I do not know the item number). The result is attached (see page 252 in our book -
Hallo Ullrich,

ist der relativ gleichmäßige Schmutz in den Rillen durch die Präparation für das Elektronenmikroskop verursacht? Oder war der wirklich schon vorher auf dem Gitter?
Der nimmt ja auch einiges Licht weg (da müsste mal der Dachkehrer zum Einsatz kommen :D ).

_________________
Herzliche Grüße / best regards

Lothar

https://lotharschanne.wordpress.com/


Top
   
PostPosted: 24. October 2015, 19:21:05 PM 
Offline
Dauernutzer
User avatar

Joined: 20. December 2012, 00:14:47 AM
Posts: 130
Location: Berlin
Hi Lothar,

No! Typical preparation for SEM investigations and just for isolators is a carbon (or gold) coating to get a ground contact for the electrons and to avoid charging effects on sample surfaces. Due to the Al-reflection layer no coating was necessary and of course forbidden from the beginning.
I was also suprised by the residues in th edge of the grid. I am not so familiar with the production procedure of these parts, but I would expect a replica technology from a master piece like a disk record. In this case it could be a residue from the separation agent. But for this purpose is it to porose. Maybe the reflectivity is not so effected as it seems. The cross section of the grid shape is not a perfect angle or edge and the residue are placed deep in the edge. Means the remaining smal face of the edge is all we really need for the purpose.

Regards,

Ulrich


Top
   
PostPosted: 24. October 2015, 21:57:01 PM 
Offline
Meister
User avatar

Joined: 31. July 2006, 16:43:32 PM
Posts: 3325
Location: Köln
Ulrich is right! It is highly unlikely that the material in the deep part of the grating line has any influence on light diffraction because of the grating shadow effect. So, it does not matter what material it is.

I wonder, though, if the chamfer shape on top is really different to the shape of Daniel's grating. From the microscope image and due to projection effects this is difficult to say.

Cheers, Thomas


Top
   
PostPosted: 25. October 2015, 04:34:24 AM 
Offline
Meister
User avatar

Joined: 11. April 2007, 10:47:33 AM
Posts: 384
Location: Sydney
Quote:
Ulrich is right! It is highly unlikely that the material in the deep part of the grating line has any influence on light diffraction because of the grating shadow effect. So, it does not matter what material it is.
I think it depends a bit on how the grating is operated. In the case of a near-Littrow configuration with theta=0 and gamma > 0 there might be some light losses expected since the full width of the facet is used. In the other configuration (in-plane) where gamma =0 but theta >0 this might be not the case since parts of the facet are shadowed by how the beam approaches the facets (see attached figure). I assume there might be some undercut (case A) applied or smaller angle used between horizontal and vertical grating facet (case B) but I am uncertain about the shape and its dimension.

Btw. I measured the other 31.6 g/mm grating and got only 26% efficiency.

Cheers
Tobias


Attachments:
grooveShadowing.jpg
grooveShadowing.jpg [ 85.75 KiB | Viewed 4743 times ]

_________________
Dein Projekt. Mach es fertig, bevor es Dich fertig macht!
Top
   
PostPosted: 25. October 2015, 12:01:25 PM 
Offline
Meister
User avatar

Joined: 11. December 2009, 21:26:32 PM
Posts: 1263
Location: 14478 Potsdam
Hey Folks,

do not mix up comparison in this respect between (measured) efficiency curves by Richardson of the master grating and the resulting efficiency of replicated gratings after, maybe, 20 replications.
The efficiency curves are only representative for a given production line in the sense of spectral response. ONLY the master and maybe the second and third master will perform similar in the sense of efficiency. And there are basically no masters available to us. Even those from the Perkin-Elmer spectrographs sometimes offered are replicated. Also the échelle for PEPSI (840x214 mm²) is replicated!

ciao,
Daniel

_________________
Daniel P. Sablowski

https://www.kunstmann.de/buch/axel_hack ... 42000/t-2/
https://dpsablowski.wordpress.com
https://github.com/DPSablowski
http://www.aip.de


Top
   
PostPosted: 25. October 2015, 13:30:57 PM 
Offline
Meister
User avatar

Joined: 31. July 2006, 16:43:32 PM
Posts: 3325
Location: Köln
Hi Tobias, the "undercut" remains invisible. Hence, this is speculativ. Better not taking it into consideration as long as we do not know. There is some material but it remains unknown and might introduce light loss. Your argument with Near-Littrow is correct, of course. The only way to get an answer for the gratings is a direct request to Thorlabs, I presume.
Cheers, Thomas


Top
   
PostPosted: 03. November 2015, 05:12:24 AM 
Offline
Meister
User avatar

Joined: 11. April 2007, 10:47:33 AM
Posts: 384
Location: Sydney
FYI I purchased two echelle gratings from Newport/RG with the following specs:

(1) 25x50x6 31.6G 63.9d Grating, Substrate: Zerodur, coating AL, price: $360 (overstock)
(2) 25x50x6 31.6G 63.9d Grating, substrate: Zerodur, coating Ag, price: $720, lead time: 4-6 weeks

Both gratings are replicas from the MR152 master (peak efficiency >80%). I will measure the performance and keep you updated on the achieved results.

Cheers
Tobias

_________________
Dein Projekt. Mach es fertig, bevor es Dich fertig macht!


Top
   
PostPosted: 04. November 2015, 09:17:35 AM 
Offline
Meister
User avatar

Joined: 31. July 2006, 16:43:32 PM
Posts: 3325
Location: Köln
Hi Tobias, do you have respective efficiency curves available? This might be interesting to see the efficiencies some degrees beside the Blaze maximum.
Cheers, Thomas


Top
   
PostPosted: 04. November 2015, 15:17:38 PM 
Offline
Meister
User avatar

Joined: 31. July 2006, 16:43:32 PM
Posts: 3325
Location: Köln
Hi all! Today I got an answer from Thorlabs about their Echelle efficiencies.

They say (in German) that the theoretical efficiencies for this grating type is of the order of 35%. Hence, the measurements from Tobias make sense.
They also say that Echelle gratings always show significant lower efficiencies compared to standard gratings.
The material on the line bottom is probably dust from the manufacturing process. Its composition is unknown.
The "plateau" area on top of the lines is on the inactive side, hence, of no impact.
The active line flanks are produced by "line-by-line compression" where each grating line is compressed after each other. Hence, this is not a master replica. This very process introduces the chamfer on top of each grating line.

My personal summary:
A 35% efficiency is not normal but small, especially if Tobias' second grating has only 26%. This might be the result of a very cheap price resulting from a possibly quick and cheap manufacturing process (line-by-line compression). Dust in the lines is a complete "no-go" and not acceptable. The Newport gratings introduced by Tobias are probably a much better choice.

Cheers, Thomas


Top
   
PostPosted: 04. November 2015, 16:03:47 PM 
Offline
Meister
User avatar

Joined: 11. April 2007, 10:47:33 AM
Posts: 384
Location: Sydney
Hi Thomas,

Thanks for the update on the Thorlabs specs! Let's see how it goes with the Newport echelles.

Cheers!

_________________
Dein Projekt. Mach es fertig, bevor es Dich fertig macht!


Top
   
PostPosted: 04. November 2015, 17:27:19 PM 
Offline
Meister
User avatar

Joined: 31. July 2006, 16:43:32 PM
Posts: 3325
Location: Köln
Now I checked the information on the Thorlabs website at http://www.thorlabs.de/newgrouppage9.cf ... E2550-0363 . The information available is not very deep. Without information about the efficiency I can not recommend paying almost 200 Euros. Thomas


Top
   
PostPosted: 05. November 2015, 03:28:32 AM 
Offline
Meister
User avatar

Joined: 11. April 2007, 10:47:33 AM
Posts: 384
Location: Sydney
Quote:
Hi Tobias, do you have respective efficiency curves available? This might be interesting to see the efficiencies some degrees beside the Blaze maximum.
Cheers, Thomas
I have attached one of my quick n' dirty measurements. Please note that this was done with my "fliegender Aufbau". I did not measure the turning angle as I was interested in the peak efficiency only. Maybe that helps ... somehow


Attachments:
oldEchelleEfficiency.png
oldEchelleEfficiency.png [ 88.42 KiB | Viewed 4474 times ]

_________________
Dein Projekt. Mach es fertig, bevor es Dich fertig macht!
Top
   
PostPosted: 05. November 2015, 09:16:40 AM 
Offline
Meister
User avatar

Joined: 11. December 2009, 21:26:32 PM
Posts: 1263
Location: 14478 Potsdam
Hi Tobias,

good to know some absolute measurements. I had also the feeling that something is eating up the efficiency. This would lead to a factor of 0.5 for the efficiency with respect to the manufacturer specification of the master.

Good work Tobias, as always!

Daniel

_________________
Daniel P. Sablowski

https://www.kunstmann.de/buch/axel_hack ... 42000/t-2/
https://dpsablowski.wordpress.com
https://github.com/DPSablowski
http://www.aip.de


Top
   
PostPosted: 05. November 2015, 11:46:37 AM 
Offline
Meister
User avatar

Joined: 31. July 2006, 16:43:32 PM
Posts: 3325
Location: Köln
Hi Tobias, it would be important to see the curve for efficiency vs. angle of incidence (how many degrees per datapoints). Note the attached figure 6.16 from page 252 in our book. Already for (should be similar for ) the efficiency drops to 70% of the maximum efficiency. This is very important for the design of an echelle and needs to be considered not to loose too much light. For instance, if we take your extremely bad echelle efficiency of 26% and apply an angle of incidence of, say, 10 degrees, the final efficiency for the echelle grating alone is of the order of 10%! So, could you perhaps give such information?

Cheers, Thomas


Attachments:
EPSvsTHETA.jpg
EPSvsTHETA.jpg [ 94.68 KiB | Viewed 4451 times ]
Top
   
PostPosted: 05. November 2015, 12:35:38 PM 
Offline
Meister
User avatar

Joined: 11. April 2007, 10:47:33 AM
Posts: 384
Location: Sydney
Hi Thomas,

That's right. But ... I need to redo the whole measurement, maybe with the new grating. I am using the echelle in near-Littrow configuration with off-plane angle gamma somewhere between 1.5-2.5° (gamma is wavelength dependent here as I am using a prism XD).

Cheers

_________________
Dein Projekt. Mach es fertig, bevor es Dich fertig macht!


Top
   
PostPosted: 05. November 2015, 13:01:09 PM 
Offline
Meister
User avatar

Joined: 31. July 2006, 16:43:32 PM
Posts: 3325
Location: Köln
Of course, I understand the necessary effort. Maybe later, if possible. I wonder how you estimate the peak efficiency for . A pure Littrow configuration is impossible. How do you do it?


Top
   
PostPosted: 05. November 2015, 23:40:51 PM 
Offline
Meister

Joined: 13. August 2009, 17:04:50 PM
Posts: 263
Location: Maienfeld, Schweiz
Hi Thomas,

as Tobias says, he will measure at an angle theta = 0° gamma = 1.5 -2.5° This is as close to Littrow as perfect, since the efficiency does hardly vary with the angle gamma ( see e.g. here: http://spektroskopie.fg-vds.de/pdf/pyo2003.pdf fig 14 and 15 (for an R =2 and R=2.75 echelle). This is also the reason why I chose theta = 0 for my echelle spectrometer (gamma approx. 7°). The efficiencies shown in the paper are calculated for an ideal grating, measured efficiencies are expected to be lower (much lower for Thorlabs echelle gratings :( )

Regards, Martin


Top
   
PostPosted: 06. November 2015, 06:08:08 AM 
Offline
Meister
User avatar

Joined: 11. April 2007, 10:47:33 AM
Posts: 384
Location: Sydney
This is what I assume: The reason why the efficiency drops for an increasing in-plane angle theta is due to shadowing of the vertical grating facet, similar to a chamfer or dirt that drops the efficiency (see figure above). On the other hand, an increasing off-plane angle gamma does not shadow the vertical grating facet and thus the efficiency should remain somehow constant. Why the efficiency slightly increases (as shown in Fig. 14, p. 31) is not obvious to me ... maybe it has something to do how the beam is projected on the echelle.

Btw. the measurement was done with an off-plane angle larger than 5 degree, but in the final instrument gamma will be much smaller.

Cheers

_________________
Dein Projekt. Mach es fertig, bevor es Dich fertig macht!


Top
   
PostPosted: 06. November 2015, 13:50:47 PM 
Offline
Meister
User avatar

Joined: 31. July 2006, 16:43:32 PM
Posts: 3325
Location: Köln
Martin, Tobias, you are right of course! Never mix up and , what I unfortunately did here.
Cheers, Thomas


Top
   
PostPosted: 18. December 2015, 04:48:04 AM 
Offline
Meister
User avatar

Joined: 11. April 2007, 10:47:33 AM
Posts: 384
Location: Sydney
Okay, my gratings have arrived (took a fair bit longer than anticipated) ... will keep you updated with measurements as soon I have transferred the delicate substrate into a nice enclosure.

Cheers,
Tobias


Attachments:
DSC_0684.JPG
DSC_0684.JPG [ 134.78 KiB | Viewed 3967 times ]
DSC_0683.JPG
DSC_0683.JPG [ 169.54 KiB | Viewed 3967 times ]

_________________
Dein Projekt. Mach es fertig, bevor es Dich fertig macht!
Top
   
PostPosted: 24. December 2015, 06:38:46 AM 
Offline
Meister
User avatar

Joined: 11. April 2007, 10:47:33 AM
Posts: 384
Location: Sydney
Ok, here we go. Got some time to do a quick measurement of both gratings. The aluminium coated echelle has an efficiency of around 62-65% and for the silver coated grating I got 71-73%.

Merry Xmas!

Tobias


Attachments:
DSC_0689.JPG
DSC_0689.JPG [ 149.33 KiB | Viewed 3841 times ]
DSC_0686.JPG
DSC_0686.JPG [ 153.82 KiB | Viewed 3841 times ]
DSC_0688.JPG
DSC_0688.JPG [ 153.37 KiB | Viewed 3841 times ]

_________________
Dein Projekt. Mach es fertig, bevor es Dich fertig macht!
Top
   
PostPosted: 15. September 2018, 08:56:09 AM 
Offline
Meister
User avatar

Joined: 11. December 2009, 21:26:32 PM
Posts: 1263
Location: 14478 Potsdam
Hi Tobias,

could you give some details about your measurement setup to measure the efficiency?

I have the problem that the center of orders of a 79 g/mm grating (R2) are much brighter compared to 31.6 g/mm (R2) used in same spectrograph. Both gratings are from Thorlabs.

cheers,
D

_________________
Daniel P. Sablowski

https://www.kunstmann.de/buch/axel_hack ... 42000/t-2/
https://dpsablowski.wordpress.com
https://github.com/DPSablowski
http://www.aip.de


Top
   
Display posts from previous:  Sort by  
Post new topic  Reply to topic  [ 26 posts ] 

All times are UTC+02:00


Who is online

Users browsing this forum: No registered users and 1 guest


You cannot post new topics in this forum
You cannot reply to topics in this forum
You cannot edit your posts in this forum
You cannot delete your posts in this forum
You cannot post attachments in this forum

Search for:
Jump to:  
cron
Powered by phpBB® Forum Software © phpBB Limited