Refractors and IR light

[Luke]

I have noticed when imaging the Sun in white light (using safety solar filters, of course) that a few of my doublet refractors do not cope very well with IR light. The image is fuzzy until I add a UV/IR cut filter. I just wondered, do triplets or quadruplets cope far better with IR light? Sorry if that's a daft question. I'm trying to understand why there seems to be a variation in how badly affected some refractors are when you don't use a UV/IR cut when doing white light imaging of the Sun.

[ronin]

They might is the answer.

A triplet controls 3 wavelengths and "forces" them to be at the same focal plane, say 650, 520 and 400nm.

The wavelengths between theses cannot really go very far and although not on the focal plane will have to be very close to it.

Outside 650 and 400nm however the focal plane will go off one way or the other and it depends on how fast it goes away for the IR section that will be the problem. I would expect it to deviate quite rapidly but rapid in optical terms is a small change.

No way to find out I suspect as you would need the glass types and refractive indices, the curvatures of each face. the lens edge thickness and the lens seperation to work it out. Feed all that into a lens design package then ask what the focal length is at say 750nm or 800nm.

Easy way is to picture the focal plane as a parabolic curve in the vertical. The bit we use is the turning point area, so picture the turning position itself (Green) and a small band from the bottom up a little. That would be the optical wavelength. Above this band the starts curving upwards more rapidly, that is in effect the IR area (also UV) and the change in focal length for the wavelength is more rapid. That is the situation, however a parabola is typical of a doublet, it is a X² law, a triplet so an X³ law, the change is more rapid.

[Merlin66]

Luke,

Which camera are you using?

The silicon CCD chip doesn't record above 1100A, so any IR beyond that should have no effect....and even then the QE curve is so low out beyond 900A you'd probably need very long exposures.

A typical UV-IR filter will cut-off at 400nm and 700nm. 

None of the typical glass types used in our scopes are designed for IR suppression - this needs a special glass like KG3.....

What ever the issue, it sounds like something is happening between 360nm and 400nm or 700nm and 1100nm.....

[Luke]

Thanks, Ronin and Ken!

Ken, I am using a Grasshopper 3 CCD (GS3-U3-28S4M-C) with the ICX687 chip. The Quantum Efficiency curve is here on page 3:

http://www.ptgrey.com/support/downloads/downloads_admin/Dlds/Grasshopper3ImagingPerformance.pdf

And I use a Baader Solar Continuum, I bumped into this transmission curve for it that Mark Townley posted on solarchat (you've probably seen the curve before?):

http://solarchat.natca.net/media/kunena/attachments/109/Ba-Cont-lin.gif

Mark posted it in this thread:

http://solarchat.natca.net/viewtopic.php?f=4&t=356&start=25

WIsh I could have gone to the IAS show to meet Mark and some of the other solar gang!

Thanks,

Luke

Luke,

Which camera are you using?

The silicon CCD chip doesn't record above 1100A, so any IR beyond that should have no effect....and even then the QE curve is so low out beyond 900A you'd probably need very long exposures.

A typical UV-IR filter will cut-off at 400nm and 700nm. 

None of the typical glass types used in our scopes are designed for IR suppression - this needs a special glass like KG3.....

What ever the issue, it sounds like something is happening between 360nm and 400nm or 700nm and 1100nm.....

[Merlin66]

Luke,

The QE curve looks pretty typical i.e. zero at 1100nm

Do you use just a Herschel wedge/ Continuum/ND3 arrangement or the ND3.8 (or ND5) solar film?

[Luke]

Ken, these are two combos where I have to use a UV/IR cut to avoid a fuzzy/blurry image:

(1) Onyx 80 + Lunt 1.25 inch Herschel Wedge (inc. integrated ND3 filter) + Solar Continuum + UV/IR cut + Grasshopper 3

(2) Tele Vue 85 + Lunt 1.25 inch Herschel Wedge (inc. integrated ND3 filter) + Solar Continuum + UV/IR cut + Grasshopper 3

With both (1) and (2), if I use an eyepiece instead of a camera  and UV/IR cut, the visual image is very good indeed! That was what really confused me: why the eyepiece view was very good, but the video image was poor? There seemed to be little I could do to sharpen up the stacked images from setups (1) and (2) when not using the UV/IR cut. With the UV/IR cut, no problem.

This is my white light imaging setup which seems to work absolutely fine without UV/IR cut:
(3) Equinox 120 + Baader 2 inch Herschel Wedge + ND1.8 (imaging ONLY filter, of course) + Solar Continuum + Grasshopper 3

If I take away the UV/IR cut with setups (1) and (2), I can see on screen straight away that the image is not as contrasty or sharp as with (3). Stick in the UV/IR cut, then I enjoy the lovely view on screen that I get with (3) without the UV/IR cut.

I've only recently got the TV 85. I suppose a good test when I get time would be the Baader and ND1.8 with the TV 85, with and without UV/IR cut? I *think* I can get prime focus with that setup - I couldn't get prime focus with the Baader in the Onyx, there was not enough inwards focuser travel!

Is there anything you'd suggest for me to test? I'd love to find out more about what is going on!

Thanks for your help.

Luke

[D4N]

Since I have an interest in this as well I decided to do a little experiment.

Equipment:

TSED70Q

ASI120MM

Baader Solar Continuum Filter

Baader Dark Blue Filter (for comparison)

TV Remote

Software:

oaCapture

In a dark room I set up the camera on a 30ms exposure, low gain, no filters. Screen looked black.

Pointed it at the computer screen and it showed a very dark grey.

Used the TV remote in front of it and it was completely saturated.

Next tried the same with the solar continuum filter.

Pointing at the computer screen gave little response.

Using the TV remote gave dark grey. The filter is clearly leaking IR but not much.

Now I tried the same with the blue filter.

This showed dark grey looking at the screen and quite bright white (but not fully saturated) when using the remote.

Now I connected the scope.

With no filter the remote showed dark grey on screen.

With the solar continuum filter it showed completely black.

I then increased the exposure to 100ms, still black.

Then cranked it up to 1s, now it showed dark grey when I used the remote. Without the remote on it went black again so it wasn't background light.

So it seems to me that both my scope and the Baader Solar Continuum filter block a lot of IR, together they block a huge amount of IR but not everything.

However on short exposures this isn't going to be seen.

Sent from my iPhone using Tapatalk

[Merlin66]

For info: the IR controllers usually work with 930-950nm near IR........

Luke, sounds like the differences between the Lunt and Baader wedges may be the issue?

[Luke]

Interesting experiment, Dan! I like the idea to test that.

There's a link to someone finding an improvement with UV/IR cut here on cloudynights:

http://www.cloudynights.com/ubbthreads/showflat.php/Cat/0/Number/6463295/Main/6461685
 
Ken, I use the Baader UV/IR cut. The curve is as you expected, I think?:
 

I wondered how much sunlight arriving at the telescope is IR versus visible. According to Wikipedia:

"In terms of energy, sunlight at the earth's surface is around 52 or 55 percent infrared (above 700 nm), 43 or 42 percent visible (400 to 700 nm), and 5 or 3 percent ultraviolet (below 400 nm)."

And I remembered Baader do recommend a UV/IR cut with the Continuum:

"For digital imaging, we also recommend the Continuum filter be combined with our UV/IR Cut Filter (to completely cut the defocused far infrared wavelengths)."

So I think when I have a bit of time to mess around, I will try the TV85 with the Baader wedge, with and without UV/IR cut.

I don't think I can try the Lunt 1.25 in the Equinox 120, they say the Lunt 1.25 is best used with 100mm or less aperture, as far as I remember.

However, I do have a SW 100 ED Pro, with Schott glass, so it might be interesting to try both wedges with that! I don't tend to image with the 100 as it has the same focal length as the 120, so I figure that for higher res stuff I might as well use the 120. But it would be interesting to see what happens!

[Stu]

Hi Luke,

I remember reading about this issue a while back and have found the link for you. Plenty of hopefully useful and relevant info on here.

http://www.schursastrophotography.com/solar/Articles/solararticle3.html

Cheers,

Stu

[Merlin66]

Not sure where we're going with this one....

Chris mentions the DMK as having higher sensitivity towards the red/ IR - this is generally not correct ( the exception being the xx618)

We accept the Continuum has a IR leak.

IMHO the  only part of the rapidly declining IR atmospheric transmission curve which can come into play with CCD sensors is <1100nm

Checking the transmission curves for Schott and other astronomical type glasses show a consistent transmission out into the IR.

I would therefore expect the type of refractor used not to be a variable.....

[Merlin66]

I found the actual data for the FPL-53 used in the ED refractors.....high transmission all the way into the IR.

Also a good Irradiation curve...

[Stu]

If all refractors appear to transmit the ir, is it possible the 120ed brings it to focus better or is ir always unfocused relative to visible light?

The only other variable is, as already mentioned, the wedge so Luke's experiment may prove or disprove this theory!

Stu

Sent from my iPhone using Tapatalk

[Luke]

Some very interesting curves, Ken, great find!

Stu, it seems to me a few experiments are on the cards, to hopefully prove the theory! I think I will test both the TV85 and the SW ED100 when I get the chance as they should both be happy with the Lunt and the Baader wedges. That said, I am hoping the Quark will arrive soon, and that I think will require a fair bit of attention and testing in various scopes!

[michael.h.f.wilkinson]

Two different issues at play: transmission and differential focus. We can simply ignore wavelengths not detected by the sensor (whether the retina or CCD). If the image is sharp on the retina, but not in the CCD then light outside the 400-700nm range of the retina is bound to be the culprit. Slower scopes, and scopes with better chromatic correction are likely to produce better results. I would expect ED scopes, in particular slower ones, to do better than achromats of comparable focal ratio. Fast triplet could actually perform worse. To be on the safe side, and maximise contrast, a UV-IR cut filter to remove the leak outside the 400-700nm, but still within the range of the CCD makes perfect sense.