#1  
Old 30-09-2019, 10:30 PM
RyanJones
Registered User

RyanJones is offline
 
Join Date: Jun 2018
Location: Melbourne,Australia
Posts: 1,439
DSLR Temperature Noise Test

Hi All,

Ive been contemplating a bit of an experiment / DIY project to cool my DSLR. Let me start by saying i have no misconceptions that I will ever get anywhere near even the cheaper cooled astro cameras but its something I'd like to try.

The first thing to do to qualify the experiment is to capture some base line noise data to compare to. So for those that are interested, here is the start of my project.

I took a series of 180sec Dark subs @ iso 400 ( because its what I regularly image at ). Firstly at room temperature of approx 14deg. The second with the camera in my fridge at approx 4 degrees. The Camera is an old Astromoddified Canon 350d which is expected to be quite a noisy camera to start off with. It does not have a sensor temp sensor but this doesn't bother me too much. Given the design of the camera and the design of my potential project, I'll be aiming to reduce the overall temp of the camera, not directly cool the sensor. Although not as effective, the plan is to design something noninvasive.

Below are 2 of the subs ( 1 from the end of each set ). Before even looking at them, the difference was apparent in the file size. The cooled subs size varied by only a handfull of kilobytes between the first and last sub. The ones that came from the ambient temp climbed steadily over time and at the end of the set were nearly 150 kilobytes larger. In total the final cooled sub were nearly 250 kilobytes smaller than the ambient sub.

Questions an Critiques invited

Cheers

Ryan
Attached Thumbnails
Click for full-size image (Ambiant 5.jpg)
37.7 KB75 views
Click for full-size image (Fridge 5.jpg)
34.9 KB71 views
Reply With Quote
  #2  
Old 03-10-2019, 09:45 AM
Startrek (Martin)
Registered User

Startrek is offline
 
Join Date: Dec 2017
Location: Sydney and South Coast NSW
Posts: 5,990
Ryan
My 60mm axial Cooling Fan reduces sensor temp by 5 degC . I’ve tried it over several nights at different conditions
I’m only looking to preserve the life of the camera as dark current noise reduction would require at least a 20 deg C reduction
Martin
Attached Thumbnails
Click for full-size image (613AE34D-8120-4C24-80E5-9B76E52B340F.jpeg)
103.0 KB37 views
Click for full-size image (8E0F620E-E9D5-4837-8220-01826729D932.jpeg)
112.1 KB49 views
Reply With Quote
  #3  
Old 03-10-2019, 03:33 PM
ChrisV's Avatar
ChrisV (Chris)
Registered User

ChrisV is offline
 
Join Date: Aug 2015
Location: Sydney
Posts: 1,737
So as far as noise is concerned what's the mean, median and std dev of the two subs?
Reply With Quote
  #4  
Old 03-10-2019, 04:28 PM
casstony
Registered User

casstony is offline
 
Join Date: Feb 2006
Location: Warragul, Vic
Posts: 4,493
Shouldn't a 5 degree drop roughly halve the noise?
Reply With Quote
  #5  
Old 03-10-2019, 06:33 PM
RyanJones
Registered User

RyanJones is offline
 
Join Date: Jun 2018
Location: Melbourne,Australia
Posts: 1,439
Quote:
Originally Posted by ChrisV View Post
So as far as noise is concerned what's the mean, median and std dev of the two subs?
Do be honest I don’t know Chris. I assessed the noise visually. The two images I posted only really show the difference in the hot pixels once shrunk to JPEG but the RAWs from the camera show a really significant difference in noise levels visually. How would I go about finding the mean, median and standard deviations of the subs ?
Reply With Quote
  #6  
Old 03-10-2019, 06:42 PM
RyanJones
Registered User

RyanJones is offline
 
Join Date: Jun 2018
Location: Melbourne,Australia
Posts: 1,439
Quote:
Originally Posted by Startrek View Post
Ryan
My 60mm axial Cooling Fan reduces sensor temp by 5 degC . I’ve tried it over several nights at different conditions
I’m only looking to preserve the life of the camera as dark current noise reduction would require at least a 20 deg C reduction
Martin
My experiment showed a huge difference over 10deg Martin. I’m sure the gains aren’t linear though. A fan like yours is by far the easiest way to go but it requires the ambient to be low enough. I’m not for a minute pretending that I’m going to work miracles with a DSLR but I am going for a more significant drop in temp. If I can maintain a temp of under 5 in an ambient of not much more than 20, I’ll be stoked.
Reply With Quote
  #7  
Old 03-10-2019, 07:56 PM
gb44 (Glenn)
Registered User

gb44 is offline
 
Join Date: Jul 2012
Location: Gold Coast
Posts: 273
Ryan
I like the idea of preserving the camera. The further one gets into this concept the more one just wants to put it all back on a manufacturer. The cooled CMOS cameras are more affordable. Here are some questions...

Can you achieve a stable + repeatable temp? If not then prepare to repeat all the calibration frames every session. Honis cooler boxes didnt use themostats did they? Cycling Peltiers is not recommended is it?

Can your system avoid condensation? My Honis - style cooler boxes always had water in them.

Will you use a closed tube system with the cooled camera? Frost quickly developed on my light pollution filter in my Newtonian telescope and using a dew heater strip was required but not always successful.

Does the cooler system require more power cells?
Does it add difficulties with scope balance and set up time? Do we have time?

Cheers
GlennB
Reply With Quote
  #8  
Old 04-10-2019, 06:58 AM
RyanJones
Registered User

RyanJones is offline
 
Join Date: Jun 2018
Location: Melbourne,Australia
Posts: 1,439
Hi Glenn,

All valid questions you ask. I have been thinking a lot about all of these issues, moisture being the significant one and that’s certainly something I want to avoid. I’ve been thinking of how I can integrate a desiccant tray into the system but it’s going to be a case of think of the best solution and see if it works.

Re: filter fogging up, I already have this issue when using my camera with my Newt and have already addressed it with a lens heater.

The control of it is another thing I’ve considered. You’re right that cycling the Peltier isn’t recommended however it’s temperature could possibly be controlled by varying the heatsink fan speed ? As far as repeatable results, I do fresh calibration frames every session anyway because the ambient is always changing so if I have to continue doing that it will be just less gain out of the project rather than making things worse.

Thank you for your interest. I’m keen to hear what you think about the pitfalls and what you think of my solutions ?

Cheers
Reply With Quote
  #9  
Old 06-10-2019, 05:16 PM
gb44 (Glenn)
Registered User

gb44 is offline
 
Join Date: Jul 2012
Location: Gold Coast
Posts: 273
The idea of a camera where one can say set the temp constantly at say 10 degrees is pretty nice. From session to session that keeps things simple. Everything else is a bit like reinventing the wheel nowadays.

With the cooler box I've tried doubling the fans and speed controller rheostat and got the temps down okay. I suppose I stopped when the water issue got bad and the camera froze one time!

I had thought about a timer where after decent cooling then the whole box is turned off and on but I never tried it. I had considered using a manual xmas light timer for that.

Cheers
GlennB
Reply With Quote
  #10  
Old 06-10-2019, 06:41 PM
RyanJones
Registered User

RyanJones is offline
 
Join Date: Jun 2018
Location: Melbourne,Australia
Posts: 1,439
Thank you again for your input Glenn.

I decided to pull apart the camera on Saturday and see where I could get cool air into the body of the camera and how I could dehumidify it. Every thought I had raised more concerns then I had a moment of what I thought was clarity. I’m here now, why not just pull the rest of the camera apart and put a cold finger in..... Not a job for the feint of heart and if I’m honest I’m not sure I’d do it again. So here’s my progress :
Attached Thumbnails
Click for full-size image (3D72F84A-FC38-49E9-A464-F049C3D72CD6.jpg)
196.6 KB47 views
Reply With Quote
  #11  
Old 07-10-2019, 10:06 AM
gb44 (Glenn)
Registered User

gb44 is offline
 
Join Date: Jul 2012
Location: Gold Coast
Posts: 273
Wow, well done!
Bare, basic, light, functional.

Are there sites that describe controllers for the finger?
Waterproof?

Cheers
GlennB

PS looks like $20 will get the right controller to do the job -
https://jimchungblog.com/2017/01/01/...n-jays-ticket/

Last edited by gb44; 07-10-2019 at 10:25 AM. Reason: answer own question
Reply With Quote
  #12  
Old 07-10-2019, 09:12 PM
RyanJones
Registered User

RyanJones is offline
 
Join Date: Jun 2018
Location: Melbourne,Australia
Posts: 1,439
Thank you Glenn. This little 350d has come a long way lol. It's got some history

Thank you for the link. I’d seen these controllers but I wasn’t sure if they were what I was after. That’s my next move. Sadly I’m going to have to open the old girl up again to put the sensor in. This has turned into a really learning process

Cheers
Reply With Quote
Reply

Bookmarks

Thread Tools
Rate This Thread
Rate This Thread:

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off

Forum Jump


All times are GMT +10. The time is now 03:50 AM.

Powered by vBulletin Version 3.8.7 | Copyright ©2000 - 2024, Jelsoft Enterprises Ltd.
Advertisement
Testar
Advertisement
Bintel
Advertisement