• Welcome to Orpington Astronomical Society.
 

News:

New version SMF 2.1.4 installed. You may need to clear cookies and login again...

Main Menu

CMOS Blotch Problem

Started by The Thing, Oct 15, 2018, 09:52:46

Previous topic - Next topic

0 Members and 2 Guests are viewing this topic.

The Thing

The way round it I'm using is not to use any calibration at all. I'm tweaking my scope to give an evenly illuminated field as its supposed to provide one. I'm consequently getting better results. I've tried a long/short alternate exposure plan and that seems to do the trick. I've done a new sigma stack of Orion using these and the 120s sub's I used before and the result is very much better.

RobertM

I'm gutted that you're having all this trouble Duncan.  Do you think you'll be able to get over this issue ?

Robert

The Thing

Well, yes. As mentioned, since calibration causes the issue, ditch the calibration.

Carole

Quoteditch the calibration.
How do you compensate for lack of flats re vignetting and dust bunnies?

Carole

The Thing

Hi Carole,
Never had a dust bunny visible on a stacked image and I've think I've got my Telescope Service scope sorted so there is minimal vignetting. In StarTools the best starting point is an uncalibrated stack. There are very good tools for correcting vignetting and noise. I've spent days trying to get the same sort of results in PixInsight to no avail.

Duncan

The Thing

Just been reading Marks post about diffraction patterns and microlenses.

At least with the colour blotches there are things to try. I have images where they didn't manifest at all so it must be solvable. The next thing I'll try is a stable DC power supply with low/no ripple for the USB3 hub (which powers the camera) and separately for the cooling which currently varies the chip temp by more than 0.5c once its 'reached' temperature. At the moment they are both fed via 'direct' takeouts on the Kendrick dew heater controller, so that switching may be affecting things as well. Voltage instability may be a function of living out in the sticks and/or having an electric water heater for the underfloor heating (off most of the night) and another for hot water heater (on all night for cheap rate juice). I'll keep experimenting.

Also I have had trouble collimating for a while so I went back to first principles. The focuser (held by two tiny grub screws) wasn't in straight! I also re-adjusted the crayford tensions and that noticeably straightened the focuser tube! Collimating with the Concenter tool was a doddle and things finally 'looked' right. If it's now right I will have a lot more signal arriving on the CMOS and this should swamp the blotch causing whatever in the CMOS. Where are those pesky stars when you need em!

The Thing

Hi All,

This may come in useful to someone...

I have replaced shared power supplies (12V laptop variety) with individual ones for each power need. This affects the 'industrial' USB3 hub which powers the cameras and the TEC cooling on the ASI294MC Pro. This seems to have completely solved the blotch problem. I suspect power output ripple and variation caused by items changing their power needs is a factor. I need to do some PI processing and stretching to verify this as I have been using StarTools recently due to PI display issues.

I also now think ambient temperature affects the efficacy and stability of the TEC cooling to a great degree. I might try making an insulating jacket for the camera body so that in warm weather it's not fighting the cooling by absorbing heat from the environment. In cold weather this shouldn't be needed. With an ambient temperature of 5C the TEC only drew 50% power to keep the CMOS at -20C, at 15C it was 100%. The CMOS temperature also flips up and down 0.5C every few seconds, this might be a measurement error or the effect of the solder blobs connecting the sensor to its backplane not being effective conductors.

I have also upped the ADU count for my flats, I was using 20000 which is the default in the APT Flats Aid tool, I have increased it to 32000ADU which is about 1/2 full well depth on this camera and suddenly the flats are correcting properly again!

All good stuff.

Duncan

ApophisAstros

Quote from: The Thing on Dec 18, 2018, 10:34:59
I might try making an insulating jacket for the camera body so that in warm weather it's not fighting the cooling by absorbing heat from the environment. In cold weather this shouldn't be needed. With an ambient temperature of 5C the TEC only drew 50% power to keep the CMOS at -20C, at 15C it was 100%. The CMOS temperature also flips up and down 0.5C every few seconds, this might be a measurement error or the effect of the solder blobs connecting the sensor to its backplane not being effective conductors.
is that just for your colour camera as i don't seem to have a blotch problem with my cmos mono and given the amount you pay for this equipment should you have to do this for efficiency as they should be made to work in hot temps as well as cold.
Roger
RedCat51,QHYCCD183,Atik460EX,EQ6-R.Tri-Band OSC,BaaderSII1,25" 4.5nm,Ha3.5nm,Oiii3.5nm.

Carole

#23
Glad you are getting sorted Duncan, though these cameras should be fit for the purpose they are intended in the first place, and not having to solve mysteries as to why they behave badly.

How are you managing to power 12V power separately and how many have you got?

I currently have 5 to power.
2 Mono cameras, 2 Electronic filterwheels and my dew heaters, all running off the same 12v adapter through 12V splitters.  Luckily mine seems to work fine.

I don't as a rule use a USB hub though but use two separate laptops instead.

Carole


The Thing

Quote from: Apophis on Dec 18, 2018, 12:02:59
is that just for your colour camera as i don't seem to have a blotch problem with my cmos mono and given the amount you pay for this equipment should you have to do this for efficiency as they should be made to work in hot temps as well as cold.
Roger
Yes. Other colour CMOS cameras seem to exhibit the same sort of artefacts but I don't think any mono CMOS users have reported the equivalent grey blothc problem.

The Thing

Thanks Carole. I think it's down to the change of tech. I am sure if you went back to the dawn of CCD there would have been similar problems identified and overcome, hence the use of Darks and Bias frames and the introduction of cooling for instance.

I now have separate 12v supplies for USB3 hub (main camera, guide camera, TemperHum sensor, mount control) 4A, camera cooler 5A, mount 5A, dew heater controller which still shares with the focuser controller (6.5A) and of course laptop (which is 19v). They run through surge protected plug strips and there are ferrite rings at strategic locations - mostly as part of the power bricks' leads.

RobertM

Hi Duncan,

I'm really glad you've sorted the problem out, must admit I was a bit shocked to see where you were powering the camera from.  I've got a few of ultra low ripple (>1mv rms) PSU modules to go into my mk2 power brick, one will be dedicated to the camera another to the USB hub and the last to mount/dew bands and other "dirty" noise sources.

Regarding ADU levels for flats, I go for about 2/3rds full well count (40000 ADU).  The main reason is that the edges, where the signal is lowest, will add more noise if the ADU values are low.  If you are using 32000 peak then the lowest edge level might be just 10000 depending on vignetting - worth considering as the edge is also part of the image and needs just as much s/n ratio as the rest.

Robert

The Thing

Hi Robert,

I've used a PC power supply unit for DSCs etc. for years, obviously OK for a DSLR but not a proper camera. I haven't been using that in the observatory though Have you got a particular make of power supply? I've had a look and there seem to be endless bench supplies and other units/modules/PCBs to choose from but very few give ripple figures. I've got a DC-DC converter coming which does have a resonable ripple figure quoted and I'm going to feed that from a Toshiba laptop power brick (19v) and use it for the hub and by implication, the camera.

And I'll try a higher ADU for the flats...

Carole

Hmm, I have been using between 22,000 and 30,000 ADU for flats, and generally try to aim for around 22,000 - 24,000.  I find they work for me but maybe it is different for a CMOS camera.

Carole

ApophisAstros

I might try two sets on next session to see what works best
Roger
RedCat51,QHYCCD183,Atik460EX,EQ6-R.Tri-Band OSC,BaaderSII1,25" 4.5nm,Ha3.5nm,Oiii3.5nm.