QHY9 in-focus smearing problem

[mldee]

I went through a series of problems with a new QHY9 mounted on RC8 and what looked like random "hot pixel noise", After substituting cables etc, it appeared to be USB-induced noise from the pier USB hub-power 5V connection. All good now.

My present problem is that when I focus the QHY9, using Nebulosity's "Fine Focus" function and a reasonably bright star such as Peacock, I can get good focus, but the image on the main screen then shows smears to the right of the star, in effect making the star go very egg-shaped.

It doesn't happen if I use a QHY8 Pro in exactly the same setup, and I tried the QHY9 without the filter wheel and it still did it. I have not yet tried any other software. Clouds.....

Has anyone else ran into this problem and if so, solutions?

[bmitchell82]

Pictures.?

When you say smear to my mind its something like the qhy9 fosted and then has dragged gunk from the PCB and then dried on the cover slip. Its happend to me before with a little bit of good acetone a hair dryer and dessicant :) happy days!

[RickS]

Are you binning? The KAF8300 can generate short horizontal blooms due to the limited depth of the horizontal shift register. Examples here: http://www.pbase.com/wjshaheen/horizontal_blooming_in_the_kaf8300_ ccd

Cheers,
Rick.

[bmitchell82]

Im happy mine doesn't have that issue! phew

[mldee]

Yep, exactly like that, and it would be caused by the fact that Neb by default uses 4x4 binning in the focus mode. I can change that to no binning and will check again. Clouds permitting.....

The blooming actually starts out as short horizontal lines that gradually merge into the "bloom". Interestingly, the little star image in the fine focus window doesn't seem to bloom.

Does this mean that binning is a no-no for RGB in actual imaging, rather than focus mode?

[bmitchell82]

most of my RGB is 2x2

[mill]

Mike, only bin to 2x2 in RGB and bin L in 1x1.
Make sure USB clamping is on in the ascom driver and also no USB during readout is on.

[marki]

Clamping on for the old drivers, clamping off for the new ones. No USB during download for all versions.

Mark

[LightningNZ]

Binning just isn't very effective on the KAF-8300 full stop, but especially not on the colour version I suspect. It'll still work if you have have a faint overall field but you'd be better off just imaging a little longer from what I understand of these chips. For focusing it shouldn't matter much - just reduce the exposure time till you no longer get the smears.

Kodak would have made one the best CCDs just about perfect if they had dedicated accumulation registers rather than regular pixels for horizontal shifting.

Cheers,
Cam

[RickS]

So long as you don't exceed the capacity of the HSR (believed to be about 40Ke-) you won't see this effect. So, as suggested, you might get away with binning 2x2 with colour filters. Narrowband filters should be OK too.

Cheers,
Rick.

[mldee]

Thanks everyone for your comments and suggestions.

I should point out that I use the generic QHY9 driver, not ASCOM, with Nebulosity, it just works. I have no idea if it does the USB off or clamping tricks, I guess I'll investigate and maybe ask Craig Stark.

I should have plenty of time, clouds and rain forecast for the next month.......

[gregbradley]

I can't comment on the QHY9 but I have never seen anything other than perfect stars in any binning with my FLI ML8300. So I think the statement the KAF8300 has horizontal blooms is more of an implementation error by the camera maker.
I think the antiblooming level can be adjusted and its probably more that this has not been set at the right level. I know you can adjust the antiblooming level of the STL11 KAI11002 so perhaps the KAF8300 can be also. Worth asking about.

Also the bulk of those horiztonal bloom images are tracking errors.
If your tracking is good for say 2 minutes and then you get a group of bad PE then the star will shift for a while and you get a 2nd sharper star sticking out the side of the larger star (which got exposed the longest).

I have seen plenty of those but only when there were tracking errors.

Also binning 2x2 may not be 4x the sensitivity as you may expect as commented here on the way the registers are handled but there is a gain.

From personal use estimate something like 2X gain. I have gotten some pretty stunning Ha images this way so don't write it off.

If the seeing is poor you may be better off shooting luminance at 2x2. The KAF8300 has the smallest pixels of the current crop of Kodak sensors so there is a lot of resolution available there. Shooting Ha at 2x2 works well as often Ha only adds colour and contrast rather than detail.

I have been using mine for 3 years.

One thing I have noticed is the small wells on a fast system will overexpose stars that can wreck the brighter stars in an image and even lose some colour data. In that case I found limiting exposures to 5 minutes solved that.

Greg.

[Tandum]

I've never seen horizontal artifacts like that from my qhy9. I've seen a vertical bloom line but only when focusing on a bright star in a sub frame and I've seen black lines running left to right from bright stars due to clamping issues but I'm running the latest drivers which has the clamping option and using maxim to capture.

I also checked LRGB subs from an old image that had RGB binned x2. Measuring a corner sample star, definitely not the brightest star, 600sec L binned x1 = 3288, 300sec RGB binned x 2 = 8407 when totaled. If x2 binning doubled sensitivity I'd expect that RGB total to be close to the L value but it seems closer to triple? I stopped binning the camera when imaging unless I have to.

The ascom drivers allow you to set gain for each binning mode. I run X1 at a gain of 5 but all binning modes are at gain zero. Perhaps driving the camera directly from nebulosity into the old drivers carries the gain setting through to all binning modes?

I been using this one for 2 years now, a change would be nice :)

[mldee]

Greg,

My blooming in 4x4 looks very much like the 2nd and 3rd pics from above, although I also get some eggy tracking images as in the other pics, but no blooming in actual imaging type shots and duration shots, just the 1sec or so repeat focussing pics at 4x4. Makes checking focus a pain.

I will try and grab a screen shot when the the clouds let me. I may also ask Theo or the QHY forum after I investigate a little more.

David, I have the ASCOM drivers and may give them a whirl, it's just that I've had enough on my plate with a new scope, a new camera, the move from OSC to Mono, so have left the new software curve for a bit later.

Oh and it's also time to mount the C8 Hyperstar again and do some OSC widefield on Fornax and Orion.....Time flies when you're havin' fun.

Thanks all for your constructive comments.

[gregbradley]

What does clamping mean here? I have not heard that word before with other cameras.

Does it mean locking in place settings?

Greg.

[Tandum]

I may be wrong, but I believe clamping refers to a diode clamp to stop saturation of the ADC.

Another thought is you can also get artifacts when focusing as you are exposing faster than the shutter can close. If you take flats with the qhy9 you'll see that anything less than 3 seconds will give artifacts from the shutter in the flat. So if you are focusing at a second or so, you will most likely get bloom lines running from a bright star as the shutter isn't completely closed as it's reading out. I see this a lot when focusing on a bright star in a sub frame image in maxim.

[mldee]

Hmm, must admit I hadn't given that damn shutter any thought. I realise it's necessary to have one in this design of CCD, (dunno why though), but it sure is clunky to look and hear that thing chunking away. I'll check it.

The alternative is the QHY8 Pro with a fan that sounds like a Super Hornet on take off!

Regarding Neb and the USB during download, I do note that it totally pauses everything during the few seconds of download, I wonder if that may be their equivalent? I'll ask the author.

[Tandum]

No USB only refers to the camera itself. Here is the answer I got regarding it.

[mldee]

That's basically what appears to happen in Neb, nothing happens on screen during the time the status line reads "downloading" or words to that effect....whether it's actually doing the RAM storage bit first I'll try and find out.
I notice in the Neb Help file that while it doesn't mention the USB thingy, it does have "high speed readout" and "disable high speed readout" options, with disable being the default, which I use.The help file says that is to reduce readout noise, for whatever that's worth.

[mldee]

Well, finally had a couple of hours clear sky, and happy to report that changing Neb frame and focus mode to zero binning stopped the smearing problem. Then the clouds came back......

Also just received my Tak collimator scope, and with some kind help from RickS, I will try my hand at some RC8 collimation whenever the clouds allow.