Abnormal "extra bloom" in ST10-XME

Attila · Nov 30, 2021

Hi Everyone,

I purchased a very good SBIG ST-10XME astro-camera with CFW-10 filter wheel and AO-8 adaptive optics quite a long time ago, in 2011, and I have used them with full satisfaction so far, especially for photometry. They are perfect tools; I needed just to replace the old O-ring in the filter wheel one time so far.

But during the last few months or one year, I realized a big problem on the images of my ST-10XME camera (see attached jpg pictures converted from calibrated fits files). You can see normal blooming spikes above and below the bright star; this is okay since this is a NABG sensor. But you can see a bright band as well above the star only and it goes from the star up to the edge of the picture with constant width and brightness. Just in one direction...

What a hell can it be? This is not internal reflection I think because I have checked the whole optical path (I use this setup on my 160/F8 APO Triplet refractor), but I do not know what the source of this band can be. Maybe sensor aging problem?

Could anyone help me to solve this problem, please?

Image in focus:

Image out of focus:

In details, within the band, smooth filament-structure can be seen:

Thanks in advance,

Attila Mádai
Hungary

Doug · Nov 21, 2021

It's badly out of focus, resulting in a very wide bloom. Look at all the other stars, they look like donuts.

Attila · Nov 21, 2021

First of all, thanks for your reflection, Dough.

At the same time, sorry to say but I am afraid, you do not understand the problem.
The bloom is acceptable/normal. The extra band from the star up to the upper edge of the picture is abnormal.

I have taken two shots: one in sharp focus and another one in out of focus (intentionally).

There are two absolutely different phenomenons on both pictures:
- blooming spikes under and above the bright star (saturated pixels) and
- unknown bar or band only above the bright star (pixels below saturation).

I have taken two pictures just to emphasis and make the unknown phenomenon more visible.
The unknown band is visible but extremely thin on the sharp picture. Therefore I have taken another picture which is in out of sharp focus. The strange bar is much more visible there.

Sorry, my English is not native English; maybe this is why my problem-description is not clear enough to you...

Attila

Doug · Nov 21, 2021

Well, first of all it's always much more helpful to provide FITS files rather than JPEG.

Maybe check whether the shutter is working correctly. You'd expect an error message if it didn't cycle properly, but if somehow it wasn't closed during readout it would cause a vertical trail.

Attila · Nov 21, 2021

Thank you for the tip.

The camera works perfectly except this problem; there is no any kind of error message.

In spite of it, as you proposed, I will try to check the movement of the shutter immediately after exposure, during reading out as soon as I am at home again in my backyard observatory (next weekend).
Additionally, I will upload the original fits version of these files as well then.

Thanks again,
Attila

P.S. I have uploaded compressed JPEG files just because of the large file size of the FITS.

Doug · Nov 21, 2021

Sure, but I can't analyze a JPEG image. It has to be a raw, completely unprocessed FITS image. You should be able to upload to dropbox or similar.

Attila · Nov 29, 2021

I have checked the operation of the shutter visually (through the light-path window) and I could not recognize any kind of obvious malfunction. No error messages.
I tested the shutter using several exposition times on ambient temperature and -20°C as well.
Please find the original raw (fits) versions of the pictures attached here as requested.

Doug · Nov 29, 2021

I think we can rule out sensor "aging". Typical age-related effects are damage to the array due to cosmic ray strikes over time. That will produce more hot pixels and/or columns over time.

The design of a typical CCD sensor has an insulating barrier between the columns, which is why blooming doesn't happen horizontally. During readout the charge has to be clocked out upwards, so there are no insulating barriers vertically. Instead, there are a series of electrodes of alternating voltage that hold the charges in place during the exposure, and then are cycled to move the charge upwards during readout.

When you have a star that is far too bright, the accumulated electrons create a very large negative voltage, which is greater than the electrode voltages used to hold them in place. It overwhelms the electrodes, resulting in the charge leaking up and down the array.

In addition, when you read out the sensor, the electrodes clock in a cycle that pushes the charge upwards. Due to the excess potential the clocking is less effective in the oversaturated area, and charge tends to lag behind during readout. That produces the classic appearance of a longer downwards tail. The upwards bloom all happened during the exposure; a good chunk of the downwards bloom happened during readout.

Both exposures look normal in that regard; the upwards bloom looks shorter than the downwards bloom as usual. The bloom isn't as long on the out-of-focus exposure because the saturation isn't as intense - the star was spread out horizontally among many different columns, resulting in less oversaturation in each column.

So the bright, very oversaturated regions look completely normal.

Both images show a much fainter tail going to the top of the frame. It's about 1000 ADU above background on the in-focus image, and only 200 ADU above background on the out-of-focus image. There is a "braiding" effect in that area on the out-of-focus image - it's not a perfectly vertical streak. This indicates that this effect is NOT blooming. I'm not certain exactly what it is. The braiding effect suggests an optical cause - perhaps that is an effect of seeing motion of the star. It's not not a diffraction spike. It almost looks like what you'd get if the shutter was open during readout, except if that happened it would trail down the array not up.

I'd suggest making sure that the window and filters are clean. Faint smudges can cause odd effects.

The effect is not visible on any other star in the frame. Try taking an exposure that just barely reaches saturation, and see if you can see a faint trail above the star. If not it's going to be irrelevant to your photometry.

Attila · Nov 29, 2021

Doug said: ↑

Both images show a much fainter tail going to the top of the frame. It's about 1000 ADU above background on the in-focus image, and only 200 ADU above background on the out-of-focus image. There is a "braiding" effect in that area on the out-of-focus image - it's not a perfectly vertical streak. This indicates that this effect is NOT blooming. I'm not certain exactly what it is. The braiding effect suggests an optical cause - perhaps that is an effect of seeing motion of the star. It's not not a diffraction spike. It almost looks like what you'd get if the shutter was open during readout, except if that happened it would trail down the array not up.
Click to expand...

Yep, this is the phenomenon which I am speaking about.

And the strangest features of these bands, fringes or "braids" are their constant brightness and the internal pattern. The braid-pattern is obvious on the afocal picture but surely exists in the focus as well I am sure but the pixel size is to big to show these details.

Optical cause seems to be logical conclusion but I do not know any optical defect which can cause this special "one sided banding" just from the star along up to the edge of the frame with constant brightness and width. At the same time, I have already been investigating this potential optical imperfection in the telescope and focuser without any result. Additionally, if I rotate the CCD around the optical axis on the telescope, the orientation of the braiding effect compared to the sensor remains the same (vertical)!!!

So, according to the facts above, I am sure the source of the problem has to be connected to the camera.
It is true that photometry is probably not affected by this problem since the measured stars have to be well below saturation.
But I would like to use a "perfect" camera...

So what do you think?

Doug · Nov 29, 2021

Since it rotates with the camera, I agree the issue is associated with the camera (or attached accessory such as filter wheel). I also agree that the one-sidedness of it makes it hard to explain optically.

Yet it is also hard to explain on the semiconductor side.

Is the "braiding" a consistent pattern or does it change from image to image? Maybe it's some kind of bizarre RBI effect.

Attila · Nov 29, 2021

As amateur astronomer, professionally, I am a mechanical engineer with some electronic and IT background as well but I am at a loss now. This is a mysterious effect...

I know what residual bulk image (RBI) means but I have never seen this kind of "reading-out-RBI".
On the other hand, RBI can be generated in the SECOND (or later) exposures only but never in the very first exposure.

As soon as I have a stable clear and starry night, I will test this phenomenon much more deeply and will report the results for a common brain storming...

P.S. I used the "empty" filter socket of the wheel for these exposures so, filter could not exert any influence to be on the image. The plane-parallel glass plate of the AO-8 is crystal clean.

Doug · Nov 29, 2021

Yep it's a weird one.

If you were driving electrons into the bulk silicon, which you would be when blooming hard, then I could see there being an immediate effect of some kind. That's why I asked if the braided pattern was constant or varied from image to image. If the latter it may be some kind of seeing effect, which would be a clue as to its source. If the former then it's happening inside the silicon.

Attila · Nov 29, 2021

I will test it accordingly.

Just one more idea: is there anyone here who could reproduce this phenomenon with the same camera, please?

Thanks in advance,
Attila

Attila · Dec 14, 2021

Doug said: ↑

Yep it's a weird one.

If you were driving electrons into the bulk silicon, which you would be when blooming hard, then I could see there being an immediate effect of some kind. That's why I asked if the braided pattern was constant or varied from image to image. If the latter it may be some kind of seeing effect, which would be a clue as to its source. If the former then it's happening inside the silicon.
Click to expand...

I have made an animation of ten out of focus images (10 x 0.1s exposure @ -20°C, calibrated, but a bit cloudy sky):
https://drive.google.com/file/d/1Mh7y3MXYiQDBsj-a2wx0pVD0KSeQ3GkG/view?usp=sharing
It is better to download and play it locally in auto-repeat mode.

As you can see, the effect is varied from image to image. But I do not have any tip what kind of seeing effect can cause this phenomenon... I do not think this is an optical problem.
Maybe an old capacitor has lost its value and the reading circuit produces false ADU values? Since this camera is ten years old, IMHO, this can be a much more realistic hardware cause.

What do you think?

Doug · Dec 14, 2021

Honestly that does look like a seeing phenomenon to me. Almost looks like smoke above a candle! The star is likely dancing around a lot due to seeing, and if you were to drag that out vertically it might very well look like that. But the question is, why is it doing that?

The animation shows the tendrils of light running vertically. That cannot happen during sensor readout. We're seeing variations on a pixel-to-pixel scale during horizontal readout, and yet 2,200 pixel cycles and one vertical clock cycle later we're seeing a very high correlation? Not possible. It's something else.

The only thing I can think of on the electronic side is the flush cycle. This is an interline sensor so it flushes the whole pixel array instantaneously by strobing the voltage on the substrate. That doesn't clear the vertical registers, though, so they are cleared by rapidly clocking the sensor vertically. (At the same time it is continuously flushing the horizontal register so charge doesn't build up at the edge of the chip.)

Maybe charge from that intense star is leaking into the vertical registers during flushing? It might be possible if there's a problem with the camera's analog power supplies, so that the bias voltages and/or clock voltages are wrong.

If you send the camera in to Bill he should be able to confirm that possibility. I don't know about actually repairing it, though, because all the parts in that camera are long since obsolete.

Attila · Jan 21, 2022

Based on your tip, I measured the voltage of the output pins of GlobTek HES61-30 power supply (no-load voltage) which was provided by SBIG for ST-10XME camera:
1. pin: +5V/2A --> +5.32V
2. pin: +12V/0.5A --> +12.66V
3. pin: NC
4. pin: -12V/0.1A --> -11.95V
5. pin: GND

Since pin-4 seemed to be too low (especially without any work-load), therefore I replaced each electrolytic capacitor in the power supply to new one (the power supply is more than 10 years old so its capacitors could be degraded) but, unfortunately, nothing has changed. Both voltage values and camera test have remained the same.

Do you think the low voltage on pin-4 can be the problem potentially?

(Bill has sent me back to you since, according to his reply, you could know this problem better then he knows...)

Doug · Jan 21, 2022

I believe those power supplies are regulated, which means the voltage shouldn't change much as the load varies.

If it is a power supply problem, I think it's more likely one of the camera's internal regulators is acting up.

Attila · Jan 21, 2022

Thanks for your prompt reply.

So I will try to increase the negative voltage of pin-4 as first step (e.g. with a resistor-replacement) but, in case of lack of success, I will open the camera cover as well.
I will report the result.

Just one more question: in your opinion, what the probability can be the sensor chip failure?

Doug · Jan 21, 2022

It is extremely unlikely that increasing the input voltage will improve things.

It is also very unlikely that the CCD sensor has failed. It is more likely that it is not being biased or clocked properly due to a failure of an internal camera power supply or the clock drivers.

Attila · Jan 21, 2022

Thanks for the tips.
Is it possible to find (or ask) electric scheme of this camera from Diffraction Limited?
(The problem-searching would be much-much easier having a drawing...)

Log in or Sign up

Abnormal "extra bloom" in ST10-XME

Attila Cyanogen Customer

Doug Staff Member

Attila Cyanogen Customer

Doug Staff Member

Attila Cyanogen Customer

Doug Staff Member

Attila Cyanogen Customer

Attached Files:

In_focus.fit

Out-of-focus.fit

Doug Staff Member

Attila Cyanogen Customer

Doug Staff Member

Attila Cyanogen Customer

Doug Staff Member

Attila Cyanogen Customer

Attila Cyanogen Customer

Doug Staff Member

Attila Cyanogen Customer

Doug Staff Member

Attila Cyanogen Customer

Doug Staff Member

Attila Cyanogen Customer

Share This Page

Log in or Sign up

Abnormal "extra bloom" in ST10-XME

Attila Cyanogen Customer

Doug Staff Member

Attila Cyanogen Customer

Doug Staff Member

Attila Cyanogen Customer

Doug Staff Member

Attila Cyanogen Customer

Attached Files:

In_focus.fit

Out-of-focus.fit

Doug Staff Member

Attila Cyanogen Customer

Doug Staff Member

Attila Cyanogen Customer

Doug Staff Member

Attila Cyanogen Customer

Attila Cyanogen Customer

Doug Staff Member

Attila Cyanogen Customer

Doug Staff Member

Attila Cyanogen Customer

Doug Staff Member

Attila Cyanogen Customer

Share This Page

Useful Searches