Spectroscopic analysis of a Lunt CaK filter
-
- Almost There...
- Posts: 1054
- Joined: Mon Jun 04, 2012 8:13 pm
- Location: London, England
- Has thanked: 42 times
- Been thanked: 1487 times
Spectroscopic analysis of a Lunt CaK filter
I was inspired by Christian's excellent analysis of his Alluxa CaK filter (viewtopic.php?t=37280) using his Solex spectroscope to try the same technique with my homebrew CaK filter (viewtopic.php?t=13804). The filter consists of a Baader K-line filter, followed by the yellow filter from a Lunt B1200 module, followed by a polarisation filter. Normally I use this behind a Herschel wedge with a refractor (or without the wedge with a solar Newtonian).
In general, the filter gives acceptable results (see download/file.php?id=9028&mode=view) but it does not give as high contrast as my CaK PST (see download/file.php?id=14487&mode=view).
To do this experiment, I took our foamboard Solex spectroscope off its telescope and mounted the CaK module on the front. This was easy since the T-adapter which holds the CaK filters just took the place of the 2-inch nosepiece that normally goes there. I mounted the contraption on my Vixen Polarie so it would track the Sun.
First an image taken with direct sunlight with no filter. This is essentially highly collimated light (approximately f107).
Then with the CaK filter in place:
With Photoshop, I overlaid the two images (using exclusion) to get a rough idea of the how the bandwidth and position of the filtering looked with respect to the spectral line. My initial impression was the bandwidth did not look too bad but the filter line was definitely not at the centre of the CaK spectral line.
Using the program RSpec, I could extract the data from the images and plot using Excel.
Dividing the transmission spectrum by the solar intensity gives the true filter transmission profile.
The x-axes in the two charts are in pixels, but knowing the distance between the K and H lines, it is easy to convert pixels into angstroms. My estimate is the bandwidth (FWHM) of the filter is 2.52 angstroms and the offset is 1.10 angstroms (to the red). I believe the nominal bandwidth of the Lunt filter is 2.4 angstroms, with no specification for the offset. Given some of the uncertainties of my measurement method, I would say the bandwidth is "within spec" of the manufacturer, but the offset is quite considerable and surprisingly high. I believe the nominal value of the CaK PST bandwidth is 2.2 angstroms, not much different than the Lunt CaK filter. I suspect why the CaK PST gives higher contrast images than the Lunt CaK filter is likely due to the offset. My guess is if I were able to do the same test on the CaK PST, I would find it was better tuned to the centre of the CaK spectral line.
As I have mentioned elsewhere, one of the strengths of the spectroheliograph is not only the narrower bandwidth but also the ability to image at the minimum of the spectral line (or at a chosen offset). Unlike H-alpha etalons, CaK filters are difficult (or even impossible) to tune.
In general, the filter gives acceptable results (see download/file.php?id=9028&mode=view) but it does not give as high contrast as my CaK PST (see download/file.php?id=14487&mode=view).
To do this experiment, I took our foamboard Solex spectroscope off its telescope and mounted the CaK module on the front. This was easy since the T-adapter which holds the CaK filters just took the place of the 2-inch nosepiece that normally goes there. I mounted the contraption on my Vixen Polarie so it would track the Sun.
First an image taken with direct sunlight with no filter. This is essentially highly collimated light (approximately f107).
Then with the CaK filter in place:
With Photoshop, I overlaid the two images (using exclusion) to get a rough idea of the how the bandwidth and position of the filtering looked with respect to the spectral line. My initial impression was the bandwidth did not look too bad but the filter line was definitely not at the centre of the CaK spectral line.
Using the program RSpec, I could extract the data from the images and plot using Excel.
Dividing the transmission spectrum by the solar intensity gives the true filter transmission profile.
The x-axes in the two charts are in pixels, but knowing the distance between the K and H lines, it is easy to convert pixels into angstroms. My estimate is the bandwidth (FWHM) of the filter is 2.52 angstroms and the offset is 1.10 angstroms (to the red). I believe the nominal bandwidth of the Lunt filter is 2.4 angstroms, with no specification for the offset. Given some of the uncertainties of my measurement method, I would say the bandwidth is "within spec" of the manufacturer, but the offset is quite considerable and surprisingly high. I believe the nominal value of the CaK PST bandwidth is 2.2 angstroms, not much different than the Lunt CaK filter. I suspect why the CaK PST gives higher contrast images than the Lunt CaK filter is likely due to the offset. My guess is if I were able to do the same test on the CaK PST, I would find it was better tuned to the centre of the CaK spectral line.
As I have mentioned elsewhere, one of the strengths of the spectroheliograph is not only the narrower bandwidth but also the ability to image at the minimum of the spectral line (or at a chosen offset). Unlike H-alpha etalons, CaK filters are difficult (or even impossible) to tune.
- marktownley
- Librarian
- Posts: 42274
- Joined: Tue Oct 18, 2011 5:27 pm
- Location: Brierley Hills, UK
- Has thanked: 20442 times
- Been thanked: 10248 times
- Contact:
Re: Spectroscopic analysis of a Lunt CaK filter
Interesting analysis Douglas, many thanks.
You should be able to tilt the filter by a predetermined amount to compensate for the offset and bring it on band?
You should be able to tilt the filter by a predetermined amount to compensate for the offset and bring it on band?
http://brierleyhillsolar.blogspot.co.uk/
Solar images, a collection of all the most up to date live solar data on the web, imaging & processing tutorials - please take a look!
- Carbon60
- Way More Fun to Share It!!
- Posts: 14209
- Joined: Wed Mar 07, 2012 12:33 pm
- Location: Lancashire, UK
- Has thanked: 8418 times
- Been thanked: 8165 times
Re: Spectroscopic analysis of a Lunt CaK filter
Very interesting. Thank you, Douglas.
Stu.
Stu.
H-alpha, WL and Ca II K imaging kit for various image scales.
Fluxgate Magnetometers (1s and 150s Cadence).
Radio meteor detector.
More images at http://www.flickr.com/photos/solarcarbon60/
Fluxgate Magnetometers (1s and 150s Cadence).
Radio meteor detector.
More images at http://www.flickr.com/photos/solarcarbon60/
-
- Way More Fun to Share It!!
- Posts: 2150
- Joined: Sun Jun 14, 2015 4:46 pm
- Location: France
- Has thanked: 1 time
- Been thanked: 2714 times
- Contact:
Re: Spectroscopic analysis of a Lunt CaK filter
Very interesting Douglas. Thanks for sharing.
Just some thoughts. Depeding on the mechanical rigidity of the SHG body, there might be some shift in wavelength between the solar spectrum image and the image with the the Ca K filter. On my side, I have to register the two spectra by a few pixels. I use the lines visible in both spectra.
For example, have you check that the first small dent on the red side of Ca K is at the same X position on the image with the solar spectrum and on the image with the Ca K filter ?
You've probably have done this already. This is just to check.
Just some thoughts. Depeding on the mechanical rigidity of the SHG body, there might be some shift in wavelength between the solar spectrum image and the image with the the Ca K filter. On my side, I have to register the two spectra by a few pixels. I use the lines visible in both spectra.
For example, have you check that the first small dent on the red side of Ca K is at the same X position on the image with the solar spectrum and on the image with the Ca K filter ?
You've probably have done this already. This is just to check.
Christian Viladrich
Co-author of "Planetary Astronomy"
http://planetary-astronomy.com/
Editor of "Solar Astronomy"
http://www.astronomiesolaire.com/
Co-author of "Planetary Astronomy"
http://planetary-astronomy.com/
Editor of "Solar Astronomy"
http://www.astronomiesolaire.com/
- Montana
- Librarian
- Posts: 34563
- Joined: Mon Oct 17, 2011 5:25 pm
- Location: Cheshire, UK
- Has thanked: 17672 times
- Been thanked: 8793 times
Re: Spectroscopic analysis of a Lunt CaK filter
Very interesting, you can easily see it is just missing the centre line, I would be very interested too if you could just tilt it slightly to get it on band.
At Solarscope, Helmut showed me the plot of halpha etalon and how they purposely have it set shifted slightly to the right, this way you can tilt the filter and manually get it just perfect on the centre line. If it the other way you couldn't tilt it back. This is why tilt tuned etalons can never be tuned into the red wing.
Alexandra
At Solarscope, Helmut showed me the plot of halpha etalon and how they purposely have it set shifted slightly to the right, this way you can tilt the filter and manually get it just perfect on the centre line. If it the other way you couldn't tilt it back. This is why tilt tuned etalons can never be tuned into the red wing.
Alexandra
-
- Way More Fun to Share It!!
- Posts: 2150
- Joined: Sun Jun 14, 2015 4:46 pm
- Location: France
- Has thanked: 1 time
- Been thanked: 2714 times
- Contact:
Re: Spectroscopic analysis of a Lunt CaK filter
Another effect is the temperature drift, when the filter is heated the CWL drifts to the red side. When it is cooled, it drifts to the blue side.
For example, PST CA K temperature coefficient is 0.005 A/°C according very old spec. Not sure whether this was the actual value.This would mean 0.05 A/10°C swing, which is quite good.
Coronado SolarMax Ca K is said to be better than 0.05 A/°C, or 0.5 A/10°C.
For example, PST CA K temperature coefficient is 0.005 A/°C according very old spec. Not sure whether this was the actual value.This would mean 0.05 A/10°C swing, which is quite good.
Coronado SolarMax Ca K is said to be better than 0.05 A/°C, or 0.5 A/10°C.
Christian Viladrich
Co-author of "Planetary Astronomy"
http://planetary-astronomy.com/
Editor of "Solar Astronomy"
http://www.astronomiesolaire.com/
Co-author of "Planetary Astronomy"
http://planetary-astronomy.com/
Editor of "Solar Astronomy"
http://www.astronomiesolaire.com/
-
- Almost There...
- Posts: 1054
- Joined: Mon Jun 04, 2012 8:13 pm
- Location: London, England
- Has thanked: 42 times
- Been thanked: 1487 times
Re: Spectroscopic analysis of a Lunt CaK filter
I went back and checked on this to make sure and I did not find any problem with the registration. The filter is extremely light and does not put much weight on the collimator (which itself is very stiff). I actually put a "dummy" adapter on the front for the Calcium spectrum, more just to match the lighting conditions but it also would match any mechanical load.christian viladrich wrote: ↑Fri Jul 29, 2022 7:02 am Just some thoughts. Depending on the mechanical rigidity of the SHG body, there might be some shift in wavelength between the solar spectrum image and the image with the the Ca K filter.
-
- Almost There...
- Posts: 1054
- Joined: Mon Jun 04, 2012 8:13 pm
- Location: London, England
- Has thanked: 42 times
- Been thanked: 1487 times
Re: Spectroscopic analysis of a Lunt CaK filter
After I did this, and found the quite sizable red shift, it occurred to me that perhaps this was an intentional feature by Lunt. It might be that when they mount the filter in their housing, they introduce a specific tilt to bring it on band. I had a look at the original housing and there is not an obvious tilting mechanism, but there are some rotational marks. So perhaps there is a clever but simple system whereby they rotate the housing to tilt the deliberately off band filter.marktownley wrote: ↑Fri Jul 29, 2022 1:32 am You should be able to tilt the filter by a predetermined amount to compensate for the offset and bring it on band?
I should be able to add a little tilt and redo this experiment. Does any have an estimate how much tilt is required?
-
- Way More Fun to Share It!!
- Posts: 2150
- Joined: Sun Jun 14, 2015 4:46 pm
- Location: France
- Has thanked: 1 time
- Been thanked: 2714 times
- Contact:
Re: Spectroscopic analysis of a Lunt CaK filter
Great ! So, we have this 1.1 A offset. This is not too bad for a 2.4 A filter.thesmiths wrote: ↑Fri Jul 29, 2022 2:30 pmI went back and checked on this to make sure and I did not find any problem with the registration. The filter is extremely light and does not put much weight on the collimator (which itself is very stiff). I actually put a "dummy" adapter on the front for the Calcium spectrum, more just to match the lighting conditions but it also would match any mechanical load.christian viladrich wrote: ↑Fri Jul 29, 2022 7:02 am Just some thoughts. Depending on the mechanical rigidity of the SHG body, there might be some shift in wavelength between the solar spectrum image and the image with the the Ca K filter.
Christian Viladrich
Co-author of "Planetary Astronomy"
http://planetary-astronomy.com/
Editor of "Solar Astronomy"
http://www.astronomiesolaire.com/
Co-author of "Planetary Astronomy"
http://planetary-astronomy.com/
Editor of "Solar Astronomy"
http://www.astronomiesolaire.com/
-
- Way More Fun to Share It!!
- Posts: 2150
- Joined: Sun Jun 14, 2015 4:46 pm
- Location: France
- Has thanked: 1 time
- Been thanked: 2714 times
- Contact:
Re: Spectroscopic analysis of a Lunt CaK filter
Here is the relation between the tilt of the filter (in radian) and the wavelength shift :
http://astrosurf.com/viladrich/astro/in ... mula-2.JPG
n is the effective index of the filter. We can assume something between 1.4 and 1.6.
Christian Viladrich
Co-author of "Planetary Astronomy"
http://planetary-astronomy.com/
Editor of "Solar Astronomy"
http://www.astronomiesolaire.com/
Co-author of "Planetary Astronomy"
http://planetary-astronomy.com/
Editor of "Solar Astronomy"
http://www.astronomiesolaire.com/
-
- Almost There...
- Posts: 1054
- Joined: Mon Jun 04, 2012 8:13 pm
- Location: London, England
- Has thanked: 42 times
- Been thanked: 1487 times
Re: Spectroscopic analysis of a Lunt CaK filter
Thanks, Christian. Using that formula, I get a number near 2 degrees. That's in the regime which is not too small and not too large. I will have a think about the best way to accomplish a tilt.christian viladrich wrote: ↑Fri Jul 29, 2022 5:35 pm Here is the relation between the tilt of the filter (in radian) and the wavelength shift :
-
- Almost There...
- Posts: 1054
- Joined: Mon Jun 04, 2012 8:13 pm
- Location: London, England
- Has thanked: 42 times
- Been thanked: 1487 times
Re: Spectroscopic analysis of a Lunt CaK filter
I disassembled my CaK filter stack and put it back together in a way where I could manually apply some tilt to the Lunt yellow filter. The Lunt filter is sandwiched between two pieces of thin black foam board, held together with matte black Washi masking tape. This is held by friction inside the T-adapter. The Baader K-line filter is screwed on to the front of the adapter. In normal use, the T-thread would go right into a ZWO camera and the end with the Baader filter would go into a 1.25" Lunt Herschel wedge.
I took a spectrum of the Calcium lines with the SHG as above and then put the filter adapter on and took another spectrum with the CaK filter. By toggling between the two images, I could see that I had applied too much tilt and now the transmission peak was too far to the blue. I straightened out the filter a bit and repeated the process. On the third attempt, I was able to visually see that the transmission line of the filter was close to being on top of the darkest part of the CaK line. I then decided to stop and do some analysis.
The image above is basically what you can see by toggling the images on the computer screen during observations. Using RSpec and Excel, I plotted the spectral lines and the transmission spectrum. Without even trying very hard, I was able to get very close to being on band (compare to the first attempt in the earlier post above).
I didn't have a chance to do any imaging with the filter as it was too late in the day. I'll probably put back a little more tilt to get really on band and then see if the imaging contrast has improved.
Lastly, I did a rough estimate of the bandwidth and it seems tilting the filter increased the bandwidth from around 2.5 angstroms to around 3.0 angstroms. So roughly speaking, shifting by 1 angstrom has a cost of about 0.5 angstrom of bandwidth. I did the acquisition of this data quite quickly because of non-ideal weather, so I will try to redo this to get a more accurate bandwidth measurement (it seems a little on the wide side). It was a little cloudy so this might have led to less than ideal collimated light falling on the filter. It looks like the focus also wasn't as good as in my previous attempt, which would have broadened the lines. But I am encouraged to discover that tuning the filter is fairly straightforward.
I took a spectrum of the Calcium lines with the SHG as above and then put the filter adapter on and took another spectrum with the CaK filter. By toggling between the two images, I could see that I had applied too much tilt and now the transmission peak was too far to the blue. I straightened out the filter a bit and repeated the process. On the third attempt, I was able to visually see that the transmission line of the filter was close to being on top of the darkest part of the CaK line. I then decided to stop and do some analysis.
The image above is basically what you can see by toggling the images on the computer screen during observations. Using RSpec and Excel, I plotted the spectral lines and the transmission spectrum. Without even trying very hard, I was able to get very close to being on band (compare to the first attempt in the earlier post above).
I didn't have a chance to do any imaging with the filter as it was too late in the day. I'll probably put back a little more tilt to get really on band and then see if the imaging contrast has improved.
Lastly, I did a rough estimate of the bandwidth and it seems tilting the filter increased the bandwidth from around 2.5 angstroms to around 3.0 angstroms. So roughly speaking, shifting by 1 angstrom has a cost of about 0.5 angstrom of bandwidth. I did the acquisition of this data quite quickly because of non-ideal weather, so I will try to redo this to get a more accurate bandwidth measurement (it seems a little on the wide side). It was a little cloudy so this might have led to less than ideal collimated light falling on the filter. It looks like the focus also wasn't as good as in my previous attempt, which would have broadened the lines. But I am encouraged to discover that tuning the filter is fairly straightforward.
- marktownley
- Librarian
- Posts: 42274
- Joined: Tue Oct 18, 2011 5:27 pm
- Location: Brierley Hills, UK
- Has thanked: 20442 times
- Been thanked: 10248 times
- Contact:
Re: Spectroscopic analysis of a Lunt CaK filter
Excellent analysis Douglas. Looks like we were both doing similar things today.
My first image is my 'default' CaK filter - very pretty etc, blah blah.
CaK-FD-60mm-f10-ZWO183mm-bw by Mark Townley, on Flickr
Next picture tilting the filter to bring onto centreline. I think it looks a lot more like one of your SHG images.
CaK-FD-60mm-f10-zwo183mm-off-axis-tuned-bw by Mark Townley, on Flickr
Much to explore with all of this! Just need more sunshine!
Mark
My first image is my 'default' CaK filter - very pretty etc, blah blah.
CaK-FD-60mm-f10-ZWO183mm-bw by Mark Townley, on Flickr
Next picture tilting the filter to bring onto centreline. I think it looks a lot more like one of your SHG images.
CaK-FD-60mm-f10-zwo183mm-off-axis-tuned-bw by Mark Townley, on Flickr
Much to explore with all of this! Just need more sunshine!
Mark
http://brierleyhillsolar.blogspot.co.uk/
Solar images, a collection of all the most up to date live solar data on the web, imaging & processing tutorials - please take a look!
-
- Almost There...
- Posts: 1054
- Joined: Mon Jun 04, 2012 8:13 pm
- Location: London, England
- Has thanked: 42 times
- Been thanked: 1487 times
Re: Spectroscopic analysis of a Lunt CaK filter
Hi Mark, that's a dramatic difference. What's your process to bring the filter onto the centreline? Don't you actually have a double stack?marktownley wrote: ↑Mon Aug 01, 2022 11:11 pm Next picture tilting the filter to bring onto centreline. I think it looks a lot more like one of your SHG images.
- Bob Yoesle
- Almost There...
- Posts: 994
- Joined: Thu Mar 08, 2012 7:24 pm
- Has thanked: 540 times
- Been thanked: 807 times
Re: Spectroscopic analysis of a Lunt CaK filter
For comparison, I had Mark Wagner do an evaluation of one of my legacy PST yellow dichroic CaK blocking filters (also apparently hard coated like the Alluxa):
Tilting also brings this filter on-band at its advertised nominal 2.2 A FWHM. The transmission is not as high as Christian measured for the Alluxa filter, but I'm hoping the Alluxa-PST combination will make for a good double-stack. If the profiles are Lorentzian with 1.5 and 2.2 FWHM, the resultant FWHM would be about 1.1, with a 10% width of around 2.6 A and a 1% width of 5.4 A. Depending on the pre-filtering, adding an additional PST CaK might also be possible with good exposure times...
Bob
Tilting also brings this filter on-band at its advertised nominal 2.2 A FWHM. The transmission is not as high as Christian measured for the Alluxa filter, but I'm hoping the Alluxa-PST combination will make for a good double-stack. If the profiles are Lorentzian with 1.5 and 2.2 FWHM, the resultant FWHM would be about 1.1, with a 10% width of around 2.6 A and a 1% width of 5.4 A. Depending on the pre-filtering, adding an additional PST CaK might also be possible with good exposure times...
Bob
Diagonally parked in a parallel universe.
Curiosity is the father of knowledge; uncertainty is the mother of wisdom.
Dark-Sky Defenders
Goldendale Observatory
Curiosity is the father of knowledge; uncertainty is the mother of wisdom.
Dark-Sky Defenders
Goldendale Observatory
- marktownley
- Librarian
- Posts: 42274
- Joined: Tue Oct 18, 2011 5:27 pm
- Location: Brierley Hills, UK
- Has thanked: 20442 times
- Been thanked: 10248 times
- Contact:
Re: Spectroscopic analysis of a Lunt CaK filter
Hi Douglas.thesmiths wrote: ↑Tue Aug 02, 2022 12:59 amHi Mark, that's a dramatic difference. What's your process to bring the filter onto the centreline? Don't you actually have a double stack?marktownley wrote: ↑Mon Aug 01, 2022 11:11 pm Next picture tilting the filter to bring onto centreline. I think it looks a lot more like one of your SHG images.
With the disk below, I realised by moving it around the (big) frame of the ZWO183, when it is in the corner of the frame it is off axis by a couple of degrees. You can see the disk visibly tune and change appearance as you do this.
Soooo, I need to think how I want to achieve this tilt, but on axis. My obvious solution is to use Apollo Laskys skybender, which will give me a live tuneable way of doing it, or, my other way is to shim the filters to get this. A 1mm shim should give a couple of degree tilt across the face of the filter which should put me in the ball park.
I'd originally been using a double stack CaK filter - 2 yellow CaK PST filters like in Bobs picture. Yesterday I finally got round to trying a triple stack (i've had a third CaK PST for while). A bit like etalons there is an order to the filters that works, and an order that doesn't, yesterday I finally got this sussed out. I have very little loss in light comparing triple to double stack.
Todays job is to 'reconfigure' the CaK module for the 3 filters, in true fashion it is damp and grey so testing is very unlikely, and family commitments mean it is likely the weekend before I can revisit it (assuming clear skies!)
Lots of tinkering to do!
Mark
http://brierleyhillsolar.blogspot.co.uk/
Solar images, a collection of all the most up to date live solar data on the web, imaging & processing tutorials - please take a look!
-
- Almost There...
- Posts: 1054
- Joined: Mon Jun 04, 2012 8:13 pm
- Location: London, England
- Has thanked: 42 times
- Been thanked: 1487 times
Re: Spectroscopic analysis of a Lunt CaK filter
That's also an interesting demonstration why a high f-ratio light beam is important with these types of filters.marktownley wrote: ↑Tue Aug 02, 2022 9:50 am I realised by moving it around the (big) frame of the ZWO183, when it is in the corner of the frame it is off axis by a couple of degrees. You can see the disk visibly tune and change appearance as you do this.
I should add that that's one nice advantage I've experienced with the SHG approach. It's not really that sensitive to the f-ratio of the telescope so you can get a very narrowband using a relatively low f-ratio (e.g. f6 or f7), enabling a full-disk image with a not very big sensor. It's on the opposite extreme to the Quark in that regard (very high f-ratio required, only a narrow field of view).