Information on QHY cameras.

[alpal]

QHY's website is down & so is their repair service.

Does anyone have close up high resolution pictures of the inside of a QHY9m camera
showing the parts that are on all sides of the PCBs - printed circuit boards?

I am looking for any adjustment pots - or potentiometers.
I believe there is one that adjusts the horizontal bleed lines & can turn them from black to white or neutral in between.

There is such a pot on the QHY12 see link:
http://qhyccd.com/cc...p?topic=3523.15 (http://qhyccd.com/ccdbbs/index.php?topic=3523.15)




cheers
Allan

[Maurice]

Hi Allan

If you still need this info, I can PM it to you....

cheers
Maurice

[alpal]

Thanks Maurice - go ahead.


cheers
Allan

[RobF]

I'd be giving Theo a ring first Allan?
Having said that, you've piqued my interest what you're up to (as a fellow QHY9 owner).

[alpal]

OK - I've got the information for anyone who also needs it.


How to remove horizontal bleed or blooming lines from a QHY9 mono camera.


I'm an electronics technician & it looks fairly simple.

Disclaimer:
A camera is a very delicate piece of electronic equipment & it needs
to be opened on an anti-static bench by a qualified electronics technician
who uses an anti-static wrist strap.

After downloading -
The document only shows arrow positions properly when I enable editing in Word.
200% zoom is best for viewing.

The Pot to adjust is R27 but the measuring point is unclear.
It looks like its at the cathode end of a diode as marked.
The 2.2 Volts needs to be measured with an oscilloscope as per the photo.

I have not tried this adjustment yet but I will report back when I do.
It may be prudent to wait until I do the adjustment before anyone else
uses the information provided.
Anyone using this information does so at their own risk.

The document is here:

https://drive.google.com/file/d/0B2VJKGAcwQ0ZQnNQNlZ1RTl0Wms/view



cheers
Allan

[alpal]

I just got the latest document in pdf form.
It has a few minor changes:

https://drive.google.com/file/d/0B2VJKGAcwQ0ZNGFQRE0xZ3FsVEE/view?usp=sharing


I think what I need to do is to create an external potentiometer
that screws into the housing so I can adjust it while under the stars.
It would be too much mucking around to undo & keep adjusting all the time in order to test it.

cheers
Allan


This is what I'm getting now:
The Running Man Nebula - stack of luminance -with the black lines
exaggerated using HDR toning.

[RobF]

That's with clamping on Allan?
Thanks for posting this. Very interesting.

[alpal]

Hi Rob,
I don't know.
I have clamping ticked.
I'll try it both ways.

cheers
Allan

[alpal]

Me
I wonder if it's possible to create an artificial star indoors to test the camera?

It's a pity that I could do an adjustment -
reassemble the entire camera -& then have to wait until I set up
the whole system to test for the results.

Perhaps shining a very dim light from a LED into an optic fiber & then illuminating
a small part of the sensor?
I have an optic fiber cable in my cupboard somewhere.

Any ideas?

cheers
Allan

[Shiraz]

if you don't have a lens to fit Allan, a pinhole will allow you to image indoors. http://www.sxccd.com/handbooks/SXVR-H694%20handbook.pdf shows how to make one (page 6) that should be fine for your cam. Then just put a led torch or similar in the field of view to generate an overloaded condition.

[alpal]

Thanks - I will definitely try that to see if I can re-create the black lines indoors.


cheers
Allan

[AndrewJ]

Gday Allan

Just interested in how you interpreted the 2.2V from the pdf,
as its very light on other details ( ie like where ground is )
A mate wandered round today with a QHY8L that was/is showing problems.
When i stuck a CRO on it as per the instructions, i got a pulsetrain whilst processing but it was only about 0.3V pk-pk.
Adjusting the pot as per instructions appeared to adjust the DC bias only,
ie it stayed as a 0.3V pk-pk waveform, but adjusted its DC bias.
The pdf supplied says the upper edge of the waveform should be 2.2v but it doesnt clearly indicate whats zero.
If i assume a DC bias of zero is what is required for operation, then the waveform should be about 4.4V pk-pk and the pot merely sets the DC bias to zero.
What did you end up seeing?

Andrew

[alpal]

Hi Andrew,
I haven't taken it apart yet.
It worries me that you didn't see what you expected.

I think the first thing to do is to try & re-create the fault indoors by
putting a lens in front of the QHY9 & using a LED as an artificial star.
I doubt I have the adapters to do that.

I'll have to get on to this job - & take pictures of waveforms etc to help others.
It annoys me that I always had this problem from brand new.
It tells me it was never adjusted correctly at the factory.
Also - what if I see spots on the CCD glass cover? - inside it or outside.
I would need Argon gas to replace what's inside to effect a repair.
I don't know what solvent to use to clean the glass.
All these doubts have made me think twice before opening it up.

One solution could be to put an external pot in place of the PCB mounted one
& adjust it while outdoors at night.
I don't know it's value so I would have to take it all apart twice.
It could be better to put a resistor in series each side of the pot -
& another one parallel to the pot to make for finer adjustments.
More research is needed.


cheers
Allan

[AndrewJ]

Gday Allan


We ran the unit with a fully exposed sensor and couldnt get data, let alone an image. I was expecting at least a saturated image, but didnt even get that.


In this case, the unit had been working OK, then slowly died over a few minutes, so i do suspect something else. He was sent the voltage setting pdf to try that first, but what we saw simply doesnt match expected :(


Dunno about your unit, but on the 8L, that pot is TINY and very sensitive.
It is also inside the cooled chamber section and i assume that is more for thermal stability, as it would have been much more sensible to have it external to the sealed chamber otherwise.



Have attached what we got with the unit at first.
Tweaking the pot merely appears to adjust the DC biasing of the trace
( Looking at your linked doco for the QHY9, i see their vertical scale is 2V, ie the waveform is really 4V pk-pk. I cant read the scale in the 8L pdf, but had deduced thats what we were expecting, and it was merely a bias tweak we were doing )

Andrew

[alpal]

Hi Andrew,
So there is another fault in your camera?
It doesn't work at all?


Yes- it looks very tiny & also of poor quality.
It should be a long 20 turn pot if it was that critical.
It is not inside the sealed chamber for the KAF8300 sensor but it is
in the cooled area.


But your unit has another fault.
It could be just a bias pot - I'd have to see what happens when I adjust it.
What is the bandwidth of your CRO?

cheers
Allan

[AndrewJ]

Gday Allan

Its not my camera, i was just helping a mate out, but it does look like it is not outputting data at all, as i expected at least a saturated image with 1 sec grabs.
He got the test procedure from QHY, so we just looked at the signal as per the test docos. I could see that adjusting the pot only appeared to tweak the DC bias of the signal, but its pk-pk value was way off expected. No idea why yet.


My CRO is a 60 MHz rebadged Dick Smith unit (really a Pintek PS-605 )
Run in DC coupled mode for the screen grab posted

Andrew

[alpal]

Hi Andrew,
60 MHz should be enough.
I can't see a waveform on yours that looks like the .pdf document -QHYs notes.
For a start you can see ringing on the top of the waveform in QHYs notes.

I think you'll have to tell your friend to send it back to QHY unless
you can see something obvious like a cracked solder joint.

cheers
Allan

[alpal]

Hi Andrew,
could you tell me what the value of the pot is in the QHY9 camera?

It may just be a lot simpler to buy one & install it.
Of course I'd buy a 10 turn or 20 turn pot & install it externally via a tiny hole or
I may even be able to get a pot that screws on with an airtight seal.
Like this:
http://au.rs-online.com/web/p/trimmer-resistors/2770228/

Just drill a hole in the camera case & install it with fine wire
that is covered with fine heat shrink.

Then I could adjust it while outside under the stars.
It is possible that it's adjustment would depend on temperature
& how bright some stars are in the frame.
therefore it may need to be adjusted for each target.
I just don't know yet how critical it is.
I certainly don't want to be taking the camera apart again & again.
A one shot - one time repair is all I want to do.

cheers
Allan

[AndrewJ]

Gday Allan

I dont have access to the camera anymore and it was a QHY8L, so may well have used a different pot.
As to tweaking the pot based on brightness of target etc, i cant see that being critical as if it was, it would have been set out in the pdf.
All they say is set the waveform to 2.2V, which looks like it is merely centring the waveform at that point to set a DC bias of some sort.
Dunno enough about how sensor reading works to comment properly tho:P

Andrew

[alpal]

Hi Andrew,
A QHY8L - OK so your results are not valid for my QHY9m.

If the pot is used in a voltage ratiometric way then it would be immune from temperature changes.
In other words the top section of the pot would change by equal proportion to the bottom section then the voltage output would be identical regardless of temperature.
However - we don't know that because I don't have the circuit.
As it will work at very low temperatures compared to the 20 degrees room temperature where it is adjusted - the temperature
could change any setting.
I am in the dark here without a circuit.

cheers
Allan

[alpal]

I opened up the camera & made the adjustment.
The value of the Pot R27 is 100 ohms.

The info provided by QHY is misleading.
They say 2.2 Volts but their Oscilloscope documents show 2 + .2 of a division
( because they are on 2 Volts per centimeter ) which is actually 2.4 Volts.
When I checked my QHY9 it was 2.4 Volts but the waveform does not look
identical to their picture.
I used 1 volt per centimeter to get more accuracy.
I can't see any ringing on top of the waveform but maybe my CRO doesn't have the bandwidth..

I adjusted it to 2.2 Volts & will test it at the first opportunity
to see if the black horizontal lines on bright stars have gone.

I bought 2 different types of 100 ohm preset pot so that I could install
one of them at an external location but I will see if the above works first.
I added 2 giant photos on flickr of the 2 PCBs - top & bottom.

see pic:

[AndrewJ]

Gday Allan

At least your signal looks better than ours did :-)
One thing tho, looking at their pdf, if you ignore the overshoots, the basic waveform is close to or just above 4V pk-pk
Yours is only about 3.0v.
I wonder if that is part of the problem???
( we only saw around 0.3V pk-pk, but agin, i have no idea what is generating the signal )
Also, what was yr timebase?
It looks like 0.1us but if so, is slightly different to what i measured

Andrew

[alpal]

Hi Andrew,
Yes - their signal has a more negative going section.
It is different to what I measure.
I was very careful to make sure that my zero volt level was correct
on the display & that I was on DC coupled.
I also checked the calibration using the 1 MHz 2 Vpp facility at the front of the CRO.
From the QHY notes I am hoping that the negative part of the waveform is not important.
They only refer to a measurement from zero volts to 2.2 Volts.
They are dealing with the positive half of the waveform.

cheers
Allan

[coldlegs]

Allan
been watching this thread with interest as a QHY9 dark line sufferer like yourself. Have you succeeded in creating an artificial star yet to test the adjustment? If you do then an adjustable pot on the outside will certainly be on the cards for me. Not sure how I will seal it though with thermal expansion and contraction sucking moisture in. Might araldite the back end of the pot into a solid lump although araldite does get old and brittle. Silastic maybe??
Cheers
Stephen

[alpal]

Hi Stephen,
I haven't made an artificial star yet.
I want to use a real star for my testing first.

Well - you might be lucky & not have to seal anything.
I can see the same problem with your pictures.

QHY are obviously mixed up about the settings for the R27 pot
that controls black horizontal lines.
Is it 2.4 Volts or 2.2 Volts?
The picture & the text disagree.
If that's the case then maybe 100s of these cameras are mal-adjusted from the factory?
Don't quote me on that yet.
Mine was faulty from brand new & I thought it was normal.

I need to check the camera now that it's set to 2.2 Volts & see what happens.

Also - you won't believe this -
around all 3 connectors & mostly on the gold plated threaded connector
there is a huge air space -
the camera is not sealed!

If I have to wire an external pot in place - I can easily run wire-wrap
wires out through the space between that connector & the housing.
Therefore don't worry about Araldite to make a good seal - there is no seal.

The problem will be getting the pot R27 off the PCB.
It would need 2 soldering irons -
one for the 2 pins at one end & the other for the one pin at the other end.
If not - the surface mount pot R27 would possibly rip the tracks off the PCB -
if just soldered one side & you tried to lift the part off.

I was going to do that today but I didn't have 2 soldering irons.
I do have that at work - as I'm an electronics technician so
I may rework this job with an external pot if I have to.

cheers
Allan

[Maurice]

Hi Allan

I think that you'll find that the detector is covered over with a windowed cap that is somewhat 'sealed' against the PCB (I think it has an o-ring underneath). The electronics are not sealed to the outside world..

regards
Maurice

[alpal]

Yes - that's true,
Inside that area - it will have argon gas to stop moisture building up.

In some ways the electronics on the PCBs should be sealed too.
It would be easy enough to do with some Dow Corning 738 white non-acidic silicon.
I may consider that when Ive proven I've fixed the black line problem.

cheers
Allan

[multiweb]

The aluminium chamber comes off. There is a rubber gasket on the PCB,

[alpal]

Thanks - I'm not going to touch that part.
I would need Argon to re-gas it

If you think about it -
the whole camera should be sealed & then filled with Argon.

[alpal]

Update:

I made in an indoor test of the results using an Antares guide scope attached to the QHY9.
Luckily my old Starshoot auto guider had exactly the right threads &
T piece - M42/0.75 - to screw on to the QHY9 so I used that.
It focused on to the target which was only 2 meters away
& I used my trusty tripod to center the target.
For an artificial star I tried many variants but the best ended up
some aluminum foil over a LED lamp.
There were already tiny holes in the old aluminum foil so I didn't need to make any.
I also tried many different exposure lengths.
The bright "stars" all exceed the well depth.
I was unable to see any black tails on my artificial stars
on any of the frames I took no matter what processing I used
& if the camera was warm or cold.
I know that's not very scientific because I didn't test the camera this way
before I adjusted it.

The only true way to test this is under real stars.
Actually the "black tails" are hard to see from RAW frames.
They come out mostly after heavy stretching & processing.

cheers
Allan

[multiweb]

That's true with the old QHY8 but with the QHY9 you'll find that you still can image with the main body of the camera opened to ambient air and the CCD chamber will not fog up. I've used it many times with lenses and they're not sealed. Only recently, after 2 years, did I have to regaz the chamber.

[alpal]

That's interesting,
how did you re-gas it?
Where did you buy the specially dried gas?

[multiweb]

Some info here (http://www.iceinspace.com.au/forum/showthread.php?t=109715&highlight=argon). It's an old thread, but still applicable.

For the QHY9, I popped the chamber open then reassembled it inside a plastic freezer bag filled with argon. So the whole camera was immersed in the gaz.

[alpal]

Thanks for that information.
I might buy a bottle of Argon.
I am thinking that if my adjustment proves successful I'll find out the best way to clean all
of the glass surfaces inside the camera to stop dust doughnuts -
then seal around the 3 connectors with Dow Corning 738 non-acidic silicon -
then gas the whole camera with Argon .
That would certainly be an improvement to the factory setup
& extend the camera's life.

It's great when we post information for other people.
I've tried to contribute valuable information here -
from adjusting the camera to attempting artificial star tests.
Thanks for the Argon gassing tip.

cheers
Allan

[alpal]

New ideas:

I have thought of a way to improve the artificial star test.
I am lucky that I can focus at only 2 meters from the target with that Antares guide scope attached to the QHY9.

What I need is a simple cardboard tube 2 meters long painted black on the inside.
That would make it super dark inside & mimic as close as possible real dark conditions.
I could then paint some Aluminum foil black as well & make tiny holes in it.
I would also need a focus adjustment on the guide scope.
The LED light needs to have a variable power supply to alter the brightness.

I have proven the concept - & I might build the above setup.

cheers
Allan

[multiweb]

This might help. It's a shot of the sensor with the chamber flipped up.
The only thing to look for is the top right corner of the sensor in the photo. You can see two hairline thin wires. That's the TEC. Don't break these. The gasket is on the PCB. It's pretty thin. The black block is aluminium. Btw that's a QHY9 mono.

re: silicon, I have one reply (http://qhyccd.com/ccdbbs/index.php?topic=4911.0)so far on QHYCCD forums.

DOW CORNING
3140 RTV COATING (http://www.dowcorning.com/applications/search/products/default.aspx?R=109EN&country=AUS)

There is one supplier in Adelaide which I still have to contact:

SOUTH AUSTRALIA REYNELLA, ADELAIDE
ELLSWORTH ADHESIVES AUSTRALIA

TEL: +61 8 8321 9346

3/65 MOORE ROAD
REYNELLA, ADELAIDE SOUTH AUSTRALIA 5161

[coldlegs]

Seems to be some on Ebay http://www.ebay.com/sch/i.html?_from=R40&_sacat=0&_nkw=3140+RTV&LH_PrefLoc=2&_sop=15

Cheers
Stephen

[alpal]

Thanks ,
It great to get all this QHY9 information together.
We have already exceeded what's on their QHY website.
The 2 photos I posted on flickr of inside the camera are the largest & most detailed I can find on the web.
Hopefully people can keep adding tips.

One more tip would be useful.
What is the best way to clean the window that covers the CCD & also the window in the nose of the QHY9?


cheers
Allan

[multiweb]

QHY9: I've read that the glass on the CCD itself has an anti-reflection coating. I've never put a finger on it. Best talk to Qiu about cleaning it. It can be damaged apparently.

The glass on the chamber itself, I'd say just wipe it as usual. Wet kim wipes, etc...

QHY8 (the square model) has a thick piece of glass directly on the CCD. I've cleaned it in the past with Qtips. No issues whatsoever. I don't know about the new QHY8s , like the pros or cylindrical 10,11,12 etc...

[alpal]

Thanks - I was thinking of using one of these products:
http://au.rs-online.com/web/c/facilities-cleaning-maintenance/electronics-cleaners-protective-coatings/air-dusters/?cm_mmc=AU-PPC-0914-_-google-_-1_Facilities_Cleaning_%26_Maintenan ce_PSSS_435974-_-PSF_436021_Air%20Dusters_Phrase_Air %20Dusters_P&gclid=CITvlt7Y98MCFRZxvAodapwAtQ

but some have solvents and are not just air.

It is difficult to decide.
I have left my QHY9m well alone for now.

cheers
Allan

[RobF]

Thanks for the interesting reading all.

I usually put some argon in the body of my 9 a few times each year according to the older linked thread here.

Sometimes every 18months or so may polish the outer CCD chamber very gently. Have used Q tips and isopropanol before, but generally its so clean a puff/brush then gentle breath and polish is sufficient. Can tell from reflections of bright room light if there's anything on there before and after.

Love the 2 stage cooling on my 9. :evil:
-25 in the middle of Qld summer (when there are no clouds ;))

The chamber warming circuit generally works very well, as long as you ramp up the cooling over 10-15 mins. Easy to do with modern software while everything else is getting sorted.

[alpal]

Thanks for that info.
Yes - I like my QHY9 m too - it's good value for the price.
Obviously it has a few minor problems but they can all be overcome.

cheers
Allan

[alpal]

Update after the black horizontal line problem:

Using my 8" f6 Newt -
I finally got to do a proper star test after I adjusted the pot in the QHY9 mono.
It's much better - I had to really stretch it with HDR toning to see any black line.
See the pic of Alpha Centauri.

I also did a Ha 20 minute bin 2x2 exposure of the statue of Liberty Nebula
which came out quite well.
I also took a fair bit of LRGB of the same target.
I have a lot of processing to do.

cheers
Allan

[alpal]

Update 29th January 2017.

The QHY9M horizontal bleed line problem is solved.
I took 2 pics & posted them here on the forum:
http://www.iceinspace.com.au/forum/showthread.php?t=152997


The trick was to adjust the pot R27 here:
https://www.flickr.com/photos/24719437@N03/16597181705/in/photostream
to get the correct waveform &
then only use slow download - about 12 seconds.
Don't tick the box for fast download when in capture mode.
( Actually it says that in the book but it should be in giant letters - I missed it.)

Also - the slow download has far less noise than fast download.
The dark noise on fast download is about 4225 bit count of the maximum 65536 bit count
whereas the dark noise is only about 1029 bit count for slow download.
4000 is about the average sky noise so you get a considerable advantage in signal to noise ratio too.

see the attached pic showing the difference using NASA FITS liberator & centering
the slider in the middle of the noise hump of 2 dark frames
depending on the readout speed.
Left hand side is slow readout speed & Right hand side is fast readout speed.


cheers
Allan