Jump to content


Photo

Single Vs. Double Stack

  • Please log in to reply
84 replies to this topic

#51 BYoesle

BYoesle

    Aurora

  • *****
  • Posts: 4947
  • Joined: 12 Jun 2004
  • Loc: Goldendale, Washington USA

Posted 31 December 2012 - 02:16 AM

Thank you very much George - these are very useful graphs and FWHM values!

As Joe noted, it looks as if the tail suppression is the more significant aspect compared to reduced bandwidth, although going from 0.7 to 0.45 A (FWHM - 0.5 on the verticle scale) with double stacking two etalons seems to give more “bang” compared to triple or quad stacking. As with most things it seems to be the law of diminishing returns... ERF removal would seem to be a must - hacksaw or no hacksaw.

I would think these shapes might possibly change a little bit for the solid etalons, although I don’t think they’re as susceptible to as much dimming with decreased bandwidth as via stacking multiple etalons...

Do you have any idea(s) as to why tilting a pair of DS etalons off-band would produce a dimmer image verses a brighter image? Going off-band would seem to allow more continuum in, and tilting an etalon should widen the bandpass (at least if it’s a Lorentzian), which would seem to be a double-whammy for giving you a brighter disk.

The front etalon (and attached D-ERF) was “pre-tilted” as usual to remove ghost images, and then the whole DS assembly was tilted to go off-band. Could it have something to do with how the D-ERF’s dielectric coatings act when tilted? They seem to be more reflective at an angle verses straight on (normal to the filter) - yet the tilting is not that great... total of a few degrees at the most.

Or would it more properly be inherent to the etalons themselves? For example, do their different initial tilts make them behave as if they are “shifted” rather than identical, thus producing the dimming you noted earlier for both Lorentzian and Gaussian transmissions? Would tilting them in unison somehow exacerbate the “shift- dimming”?

There certainly appears to be no significant brightening on-band, dimming off-band, or vice-versa, with Joe’s video via a hi-res diffraction grating. So it would seem to have something to do with the behavior of etalons, tilting, coatings, etc...

#52 marktownley

marktownley

    Cosmos

  • -----
  • Posts: 8603
  • Joined: 19 Aug 2008
  • Loc: West Midlands, UK

Posted 31 December 2012 - 06:34 AM

I’m not sure I see much benefit to triple stacking compared to double stacking


Hi Bob et al,

I've had a bit of a play round with triple stacking. My results were very contrasty. This is with the TS40.

Posted Image
TS Full disk bw by Mark Townley, on Flickr

Banding and sweetspotting issues become an absolute pain and is tricky to get everything just right, so I agree that double stacking is a better and easier alternative. I would have explored this more this year but we've had the wettest year for over 100 years and solar moments have been at a premium...

Mark :)

#53 George9

George9

    Messenger

  • *****
  • Posts: 459
  • Joined: 11 Dec 2004

Posted 31 December 2012 - 09:46 AM

Do you have any idea(s) as to why tilting a pair of DS etalons off-band would produce a dimmer image verses a brighter image? Going off-band would seem to allow more continuum in, and tilting an etalon should widen the bandpass (at least if it’s a Lorentzian), which would seem to be a double-whammy for giving you a brighter disk.

Bob, I don't have a good explanation. My assumption was that tilting the pair inadvertently shifted them with respect to each other. E.g., if you tilt the front one to avoid the ghost (and you optimized the tilt for the best view), then tilting the pair could either untilt the front one with respect to the sun or further tilt it depending on the direction. My tilting double stack now belongs to a friend, so I can't try it, but I would just experiment with single stack tilting to note the behavior, and then add the second element to the already tilted first element, optimizing brightness. In other words, I suspect you had to do some fiddling to get your on-band double stack optimized, so to be fair you need to fiddle to optimize the off-band double stack.

For my current PT double stack, off band (continuum) is a little brighter than on band (H-alpha), but not enormously so.

George

#54 BYoesle

BYoesle

    Aurora

  • *****
  • Posts: 4947
  • Joined: 12 Jun 2004
  • Loc: Goldendale, Washington USA

Posted 31 December 2012 - 09:50 AM

Excellent Mark!

Banding and sweetspotting issues become an absolute pain and is tricky to get everything just right...


Looks like you nailed it :waytogo:

Any comments about prominences with TS? Jesus' image looks as if they lose a little intensity compared to SS and DS, but again processing needs to kept the same for valid comparisons - what’s your impression?

#55 BYoesle

BYoesle

    Aurora

  • *****
  • Posts: 4947
  • Joined: 12 Jun 2004
  • Loc: Goldendale, Washington USA

Posted 31 December 2012 - 10:51 AM

My assumption was that tilting the pair inadvertently shifted them with respect to each other. E.g., if you tilt the front one to avoid the ghost (and you optimized the tilt for the best view), then tilting the pair could either untilt the front one with respect to the sun or further tilt it depending on the direction.... but I would just experiment with single stack tilting to note the behavior, and then add the second element to the already tilted first element, optimizing brightness. In other words, I suspect you had to do some fiddling to get your on-band double stack optimized, so to be fair you need to fiddle to optimize the off-band double stack.

George


Thanks George - that makes a lot of sense. :bow:

I'm looking forward to that experiment when it gets a little warmer - handling a frozen DS etalon assembly with my bare hands was a little unnerving. Fiddling with two frozen etalons and an separate ERF bare handed would probably result in a visit to the hospital ER...

#56 Bill Cowles

Bill Cowles

    MacGyver

  • *****
  • Posts: 2843
  • Joined: 16 Apr 2006
  • Loc: Utah

Posted 31 December 2012 - 03:09 PM

Love your setup and images! :coldday:

Bill

#57 Spectral Joe

Spectral Joe

    Vostok 1

  • -----
  • Posts: 181
  • Joined: 28 Feb 2011
  • Loc: Livermore CA

Posted 31 December 2012 - 11:56 PM

With the recurring question of how to set filters (single, double, or any multiple) onto H alpha, I have a suggestion. Get a source of hydrogen emission that you have control of. You can get a hydrogen discharge tube (commonly referred to as a "spectrum tube") for between $20 and $40, depending on the vendor. You can get the power supply and mount for $150 to $200, or you can build your own. So, $170 to $240 sounds like a lot? Some people spend that on eyepieces. With the spectral source there is no question where you are tuned. This seems to be an issue with double stacking, I see images that clearly show chromospheric network features that are see in the wings of H alpha, the stack is tuned red or blue of the line core. As the passband gets narrower (with a properly tuned double stack) the details seen in the line wings fades, and different details are seen, ones that were hidden before. Tilt tuned systems may have trouble finding a tilt value that doesn't give annoying reflections, pressure tuning is better for stacked systems for that reason. But even pressure tuned systems will show a wavelength variation across the field, such are the woes of filter systems. Oh, one last thing before I get off my soapbox - go lightly with the Photoshop or Registax rolling pin, bakers know that overworking the dough or batter can render the result less tasty.

#58 George9

George9

    Messenger

  • *****
  • Posts: 459
  • Joined: 11 Dec 2004

Posted 01 January 2013 - 02:21 PM

Interesting suggestion, Joe. And it's always fun to have another device to work with. I guess you could learn how to optimize the filters, but you would presumably still need to do in-the-field tweaking (e.g., as temperature changes or as air leaks slowly out of the PT).

George

#59 Spectral Joe

Spectral Joe

    Vostok 1

  • -----
  • Posts: 181
  • Joined: 28 Feb 2011
  • Loc: Livermore CA

Posted 01 January 2013 - 02:50 PM

The thing is portable, and can be taken to the scope. One of the drawbacks to etalons is not knowing exactly where you're tuned, this is a way to find out. An yes, monitoring etalon conditions is a great idea, and achievable for the gadgeteer.

#60 BYoesle

BYoesle

    Aurora

  • *****
  • Posts: 4947
  • Joined: 12 Jun 2004
  • Loc: Goldendale, Washington USA

Posted 03 January 2013 - 02:33 AM

:thanx: Bill

With the recurring question of how to set filters (single, double, or any multiple) onto H alpha, I have a suggestion. Get a source of hydrogen emission that you have control of... With the spectral source there is no question where you are tuned...

[Joe]


Hi Joe -- excellent suggestions - the discharge tube would be a relatively inexpensive H alpha “accessory.” Perhaps a “discharge tube light box” (using more than one tube?) could be used in optimizing filter tilts and rotations the way a flat-fielding light box is used. This could help to not only get the filter system on-band, but to also help optimize field uniformity as noted by Mark for his triple stacking endeavor.


Tilt tuned systems may have trouble finding a tilt value that doesn't give annoying reflections, pressure tuning is better for stacked systems for that reason. But even pressure tuned systems will show a wavelength variation across the field, such are the woes of filter systems.


Tilt verses pressure tuning is usually more a problem of tuning the primary internal etalon to be on-band. Tilting can introduce “banding” - creating a band of good contrast, and the area outside of it which falls off-band. Tilt-tuned induced banding is more problematic with internal etalons - Andy Lunt stated “I would guess that internal Etalons are about 5-6 times more sensitive to tilt [banding] than external Etalons...” http://luntsolarsyst...lon-performance

Pressure tuning eliminates the problem of banding due to tilt tuning with internal etalons.

Internal etalons are also more sensitive to circular “sweet spots “ from the internal collimating (or telecentric) optics they use. This is due to the difficulty in getting field and instrument angles within the acceptance angle of the etalon for good across-the-filter bandpass performance. Adding “banding” via tilt-tuning exacerbates a sweet spot issue.

On the other hand, ghost reflections are created between multiple etalons (i.e. double stacking). These reflections are relatively easily removed from the field of view with a little extra tilting to the DS etalon. But this again can introduce banding. Pressure tuning would be great for the second etalon, as when it is tilted it goes slightly off-band, and pressure tuning could bring it back exactly on-band, but probably would not do much for reducing "banding." No one is making a pressure tuned front etalon, although at one time Lunt was looking at the possibility. However, one could design a sealed enclosure with pressure tuning to house an existing etalon, using the ERF for the front “window” and optical glass for the second...

Because of the aforementioned sweet spot issues that can occur with internal etalons - and greater susceptibility to banding if tilt-tuned - they seem more likely to exhibit banding and/or sweet spots when double stacked (e.g. the bandwidth is reduced making any sweet spot and banding more evident.)

Using two external (front) etalons seems to minimize the magnitude of both sweet spots and banding, as there are no instrument angles, and the field angles are just those of the sun itself ~ 0.25 degree. Of course it's the most expensive way to go, and the aperture of commercial etalons is currently limited to 100 mm. Tilting the etalons (front to remove ghosts, objective to tune), can produce some banding, which is a more subtle gradient than is seen with internal etalon based systems. However, when properly adjusted, double stacking two front etalons (or evidently even three triple stacked front etalons as shown by Mark's SM40's) can give good contrast uniformity with little evidence of banding or a sweet spot.

DS SM90-SM90/BF30:

Attached Files



#61 Jim Lafferty

Jim Lafferty

    Mercury-Atlas

  • *****
  • Posts: 2640
  • Joined: 30 Jun 2007
  • Loc: Southern California

Posted 03 January 2013 - 01:57 PM

Wow! Excellent discssion!

Jim

#62 marktownley

marktownley

    Cosmos

  • -----
  • Posts: 8603
  • Joined: 19 Aug 2008
  • Loc: West Midlands, UK

Posted 03 January 2013 - 06:08 PM

(or evidently even three triple stacked front etalons as shown by Mark's SM40's)


This was with 2 external double stacked 40mm etalons and a internal 20mm etalon :wron:

#63 BYoesle

BYoesle

    Aurora

  • *****
  • Posts: 4947
  • Joined: 12 Jun 2004
  • Loc: Goldendale, Washington USA

Posted 03 January 2013 - 07:42 PM

Ah-ha. Thanks Mark. This explains your comment related to "sweetspotting issues" being "tricky to get right" - and makes your results even more extraordinary. :waytogo:

#64 Scott Beith

Scott Beith

    SRF

  • *****
  • Administrators
  • Posts: 44515
  • Joined: 26 Nov 2003
  • Loc: Frederick, MD

Posted 04 January 2013 - 02:31 PM

This has been a great thread and the results are pretty darn impressive!

#65 Bill Cowles

Bill Cowles

    MacGyver

  • *****
  • Posts: 2843
  • Joined: 16 Apr 2006
  • Loc: Utah

Posted 04 January 2013 - 04:44 PM

I just hope there is not going to be a test. ;)

Bill

#66 Spectral Joe

Spectral Joe

    Vostok 1

  • -----
  • Posts: 181
  • Joined: 28 Feb 2011
  • Loc: Livermore CA

Posted 04 January 2013 - 07:14 PM

Tuning any combination of bandpass filters is going to be tricky, especially if you aren't sure of what to tune to. If it's of any use to anyone, this is what the core of H alpha looks like at 0.35 Angstrom bandpass:

Posted Image

Taken during a rare moment of good seeing on 4/24/09, 1.1 arcsec per pixel.

The hydrogen lamp looks like this in use:

Posted Image

It's easily mounted in any sort of enclosure you prefer. I use it for alignment and bandwidth checking.

#67 George9

George9

    Messenger

  • *****
  • Posts: 459
  • Joined: 11 Dec 2004

Posted 04 January 2013 - 08:28 PM

Using two external (front) etalons seems to minimize the magnitude of both sweet spots and banding, as there are no instrument angles, and the field angles are just those of the sun itself ~ 0.25 degree. Of course it's the most expensive way to go, and the aperture of commercial etalons is currently limited to 100 mm. Tilting the etalons (front to remove ghosts, objective to tune), can produce some banding, which is a more subtle gradient than is seen with internal etalon based systems. However, when properly adjusted, double stacking two front etalons (or evidently even three triple stacked front etalons as shown by Mark's SM40's) can give good contrast uniformity with little evidence of banding or a sweet spot.


Nice shots, Bob and Joe.

The DS internal pressure-tuned etalons also achieve large sweet spots (by large I mean good; i.e., bigger than the sun) and no banding. The weak point is the reflections, covered on other threads. The SS is pretty close to perfect in terms of sweet spot, banding, and reflections.

My external Coronado ASP-60s were free of sweet spots or banding, just as Bob describes, until I bought a binoviewer. Then I had severe banding. I never quite figured it out. Perhaps the Denkmeier OCS (more or less a Barlow that goes in front of the blocking filter so you can achieve focus) changed the light cone so that the blocking filter induced banding. Tuning the blocking filter did move the band up and down (the blocking filters used to be tilt tuned).

George

#68 marktownley

marktownley

    Cosmos

  • -----
  • Posts: 8603
  • Joined: 19 Aug 2008
  • Loc: West Midlands, UK

Posted 05 January 2013 - 04:06 AM

The DS internal pressure-tuned etalons also achieve large sweet spots (by large I mean good; i.e., bigger than the sun) and no banding.


Is this a result of the etalons being 'oversized' to what you might expect? ie. only the central area of an etalon is being utilised...

#69 BYoesle

BYoesle

    Aurora

  • *****
  • Posts: 4947
  • Joined: 12 Jun 2004
  • Loc: Goldendale, Washington USA

Posted 05 January 2013 - 01:07 PM

Ditto, great close up detail and discharge tube pic Joe. :)


Perhaps the Denkmeier OCS (more or less a Barlow that goes in front of the blocking filter so you can achieve focus) changed the light cone so that the blocking filter induced banding. Tuning the blocking filter did move the band up and down (the blocking filters used to be tilt tuned).


The blocking filter is also a interference filter, therefore the incident angle of the rays it receives will affect it's performance. Per David Lunt (in an old Coronado blog):

"...the focal ratio ideally should be greater than f10 to maintain performance of the blocking filter....but in fact you will see good results at considerably shorter f ratios than this."

My understanding is that a standard barlow will increase the field angles (bad). This could be made worse with tilting as described above, and might indeed produce banding. Using a TeleVue Powermate would be better, since it has "telecentric" properties which will minimize the increase in field angles.


"The DS internal pressure-tuned etalons also achieve large sweet spots (by large I mean good; i.e., bigger than the sun) and no banding."

Is this a result of the etalons being 'oversized' to what you might expect?


I’m in no way an optics expert, but from what I have been able to learn here’s how etalon size relates to sweet spots:

Typically the sweet spot with a collimated etalon is caused by a magnification of the field angles of an extended object. If the field angle exceeds the acceptance angle of the etalon (i.e the angle needed to meet the filter's band pass specification), that portion of the image (in this case the limb of the sun) will begin to fall off-band. A decease in contrast (a loss of chromosphere detail) or prominence visibility will be noted outside the sweet spot.

For the sun’s limb, the field angle starts out being ~ 0.25 degree (i.e 1/2 the sun's angular diameter, or when optically centered the center-to-limb angle). For an etalon mounted on front of the objective, this is the field angle, and amounts to an f 108 optical cone convergence. Generally there is a very wide sweet spot at this angle and configuration.

In a collimator based internal etalon system, the field angle magnification is the ratio of the objective to collimator lens focal lengths. Because of the optical geometry, this is also roughly proportional the ratio of objective diameter to the etalon (working) diameter. Therefore, the larger the internal etalon, the less the magnification of the field angles, and the better the on-band performance.

For example, the Lunt LS80 uses an internal 50 mm etalon, and dividing 80 by 50 = 1.6. Hence the internal field angle of the suns limb can be generally determined to be 0.25 x 1.6 = 0.4 degree. For the LS100, which uses the same 50 mm etalon, the collimated field angle is 0.25 x (100/50) = 0.5 degree, and so on. So we can see that the smaller the etalon is compared to the objective (i.e. the collimator lens FL and geometry) the greater the field angle magnification, hence the smaller the sweet spot of good on-band performance.

Of note for narrower bandwidth filters is that they also have a narrower acceptance angle to remain on-band. This perhaps can explain why a single stacked system with no evidence of a sweet spot may develop a sweet spot (and possibly banding from tilting the second etalon) when double stacked, due to the constraints of the system to have a narrower acceptance angle for proper on-band performance.

Again I’m not an expert, so if I’ve erred, I hope someone will correct me.

#70 Spectral Joe

Spectral Joe

    Vostok 1

  • -----
  • Posts: 181
  • Joined: 28 Feb 2011
  • Loc: Livermore CA

Posted 06 January 2013 - 12:12 AM

I'm trying to figure out what is causing the banding while using the binoviewer. Nothing past the etalon will affect its bandpass. Is it actually a band or a central hot spot? If it a spot I suspect it's a relay imaging issue with the design of the binoviewer.

#71 marktownley

marktownley

    Cosmos

  • -----
  • Posts: 8603
  • Joined: 19 Aug 2008
  • Loc: West Midlands, UK

Posted 06 January 2013 - 04:59 AM

I'm trying to figure out what is causing the banding while using the binoviewer. Nothing past the etalon will affect its bandpass. Is it actually a band or a central hot spot? If it a spot I suspect it's a relay imaging issue with the design of the binoviewer.


Newtons rings maybe?

#72 BYoesle

BYoesle

    Aurora

  • *****
  • Posts: 4947
  • Joined: 12 Jun 2004
  • Loc: Goldendale, Washington USA

Posted 06 January 2013 - 07:27 AM

Perhaps the design of the blocking filter used with the ASP 60 might be part of the banding issue George noted when used with the accessory optics...

The [ASP-60] etalon is in the front element. The diagonal (PROM-15T) contains the blocking/prominence filter. This does not 'tune' the whole filter. The narrow element (etalon) has a temperature coefficient of passband shift of <1 Angstrom/2000 degrees C. However, the blocking filter is somewhat temperature sensitive and needs compensating, - by tilt, - for large differences in ambient temperature. This filter isolates the H-alpha passband in the etalon and eliminates all the others. If it moves with temperature, without being compensated, it would allow one of the other etalon orders to come through and, thereby, lower the contrast.

[Old Coronado blog - emphasis added]




#73 marktownley

marktownley

    Cosmos

  • -----
  • Posts: 8603
  • Joined: 19 Aug 2008
  • Loc: West Midlands, UK

Posted 06 January 2013 - 01:29 PM


I’m in no way an optics expert, but from what I have been able to learn here’s how etalon size relates to sweet spots:

Typically the sweet spot with a collimated etalon is caused by a magnification of the field angles of an extended object. If the field angle exceeds the acceptance angle of the etalon (i.e the angle needed to meet the filter's band pass specification), that portion of the image (in this case the limb of the sun) will begin to fall off-band. A decease in contrast (a loss of chromosphere detail) or prominence visibility will be noted outside the sweet spot.

For the sun’s limb, the field angle starts out being ~ 0.25 degree (i.e 1/2 the sun's angular diameter, or when optically centered the center-to-limb angle). For an etalon mounted on front of the objective, this is the field angle, and amounts to an f 108 optical cone convergence. Generally there is a very wide sweet spot at this angle and configuration.

In a collimator based internal etalon system, the field angle magnification is the ratio of the objective to collimator lens focal lengths. Because of the optical geometry, this is also roughly proportional the ratio of objective diameter to the etalon (working) diameter. Therefore, the larger the internal etalon, the less the magnification of the field angles, and the better the on-band performance.

For example, the Lunt LS80 uses an internal 50 mm etalon, and dividing 80 by 50 = 1.6. Hence the internal field angle of the suns limb can be generally determined to be 0.25 x 1.6 = 0.4 degree. For the LS100, which uses the same 50 mm etalon, the collimated field angle is 0.25 x (100/50) = 0.5 degree, and so on. So we can see that the smaller the etalon is compared to the objective (i.e. the collimator lens FL and geometry) the greater the field angle magnification, hence the smaller the sweet spot of good on-band performance.

Of note for narrower bandwidth filters is that they also have a narrower acceptance angle to remain on-band. This perhaps can explain why a single stacked system with no evidence of a sweet spot may develop a sweet spot (and possibly banding from tilting the second etalon) when double stacked, due to the constraints of the system to have a narrower acceptance angle for proper on-band performance.

Again I’m not an expert, so if I’ve erred, I hope someone will correct me.


You may not be an expert Bob, but you can be on my pub quiz team anyday ;)

Taken me a while to ingest all this and get my head around it, alongside some other stuff about it elsewhere on the web. I've never really given the whole field angles thing much thought, but it really does make a whole lot of sense and gives an interesting viewpoint to look at Ha systems:

Some of the best images i'm seeing to date on the web are coming out of LS80s - the pictures are stunning in terms of nice tight and even bandpass across some quite large chipped CCDs. Maybe our field angles have a lot more to do with it than we think.

I'm curious now as to the etalon diameters in the Coronado equipment... Hmmm :)

#74 BYoesle

BYoesle

    Aurora

  • *****
  • Posts: 4947
  • Joined: 12 Jun 2004
  • Loc: Goldendale, Washington USA

Posted 06 January 2013 - 03:23 PM

Thanks Mark! :rainbow:

I found this article and blog discussion from several years ago by David Lunt be very helpful in understanding how important the issue of angles (among other factors) are to narrow band interference filter performance:

link

#75 George9

George9

    Messenger

  • *****
  • Posts: 459
  • Joined: 11 Dec 2004

Posted 06 January 2013 - 07:17 PM

Wow, this thread just keeps staying interesting.

Joe, it’s a band, not a spot. Maybe a quarter or third FOV high, and it is as wide as the FOV. I can shift the band up or down in the field of view by tuning the blocking filter. Therefore, I don’t think it is actually the binoviewer. If I had the filter, I could just throw an eyepiece in place of the binoviewer with the OCS in it and try it; not sure why I never tried that before. The other thing I just realized is that the band was always horizontal (like the tilt of the blocking filter), so it wasn’t the etalon (whose angle varied depending on how tightly I screwed in the TMax).

Bob, I think your explanation is right on. The OCS is magnifying the field angles. Great job finding that document.

Coronado’s first solar filters basically took a prominence filter, which was an ERF plus a PROM15T (prominence, 15mm, maybe T for tilting?) and inserted an etalon behind the ERF. The prominence filter by itself showed no solar disk details. I remember David Lunt commenting that I should not need to tune the blocking filter with the etalon; just leave it untilted because the temperature effect was not that big. This worked until I got the binoviewer.

Over time I guess they switched to a blocking filter designed for the etalon. I don’t know, but I suspect that the new blocking filters were wider: still narrow enough to avoid other etalon orders but wide enough to avoid having to tune it with temperature variation (and perhaps cheaper to produce).

In your document, David Lunt mentions the dark element in front of the blocking filter to reduce reflections, and how they later brightened the view with a lighter one. I remember he had me send my blocking filter to Isle of Man to swap in the brighter version. Makes we wonder about the DSII and whether an element could be put between the two etalons to reduce the reflections (reflections would pass three times, but the primary would pass only once), but I guess the view would become too dim.

George






Cloudy Nights LLC
Cloudy Nights Sponsor: Astronomics