Jump to content

  •  

CNers have asked about a donation box for Cloudy Nights over the years, so here you go. Donation is not required by any means, so please enjoy your stay.

Photo

[DIY] Astro CCD 16-bit Color 6Mpx Camera

astrophotography ccd DIY equipment imaging
  • Please log in to reply
1241 replies to this topic

#76 Relativist

Relativist

    Cosmos

  • *****
  • Posts: 8,152
  • Joined: 11 Oct 2003
  • Loc: OC, CA, USA

Posted 26 April 2015 - 06:49 PM

I find this project very interesting. The main question I have is if it's a viable option for someone who mainly does semi-live viewing (aka opto-electronic astonomy or EAA). I'd probably use it with Astrolive, DSS live or DSS with Astrotoaster.

 

As for the cooling, if someone is able to model the camera in a multi-physics package like Comsol or Ansys we might be able to refine it prior to construction. That said, it might be a good idea to use materials like microporous insulation and radiant barriers to improve the performance.



#77 gregj888

gregj888

    Mercury-Atlas

  • -----
  • Posts: 2,511
  • Joined: 26 Mar 2006
  • Loc: Oregon

Posted 26 April 2015 - 08:30 PM

 

Any mileage in making a water cooled backend option ?

 

Steve

 

 

YES, really the only way to do deep cooling.  It also cuts vibration or should. 

 

There are several papers that cover most of the "trick's I've learned so far.  Here's a good starting point:

http://articles.adsa...000158.000.html

 

This paper only mentions polishing the Cryostat, Etch and electro-polishing is better if you can.  The etch gets rid of the microscopic pockets of oil from the machining operations.  The polish reduces the surface area which reduces water adhesion (from the atmosphere)

 

My next Cyostat will have a copper slug for the backside water jacket... it may even be a copper CPU cooler and reside inside the cryostat (so it can be cooled below dew point without getting condensate).

 

I haven't done the electro-polish.  This paper mentions Viton O-rings, which are not needed for air, butyl is as good.   Instead of Torr-Seal a much cheaper option is Hysol C1...

 

I make my own "vented screws"  on the lathe and can help there if anyone heads down this road.

 

A really bad photo, but here's the photometer prior to painting (outside only)...

Attached Thumbnails

  • Picture1.jpg


#78 gregj888

gregj888

    Mercury-Atlas

  • -----
  • Posts: 2,511
  • Joined: 26 Mar 2006
  • Loc: Oregon

Posted 27 April 2015 - 12:51 AM

Steve, 

 

Another point to bring up is overall noise.  Saw a good presentation over the weekend.  One significant point is the shot noise from the sky glow for most will swamp out any benifit from cooling, certainly the -50c or more being mentioned.  Everyone will have to do the math for themselves, but you'll need a real dark site to justify colling below -20c to -30c.

 

The NIR sensor is different, you get into black body pretty fast there.



#79 macona

macona

    Ranger 4

  • *****
  • Posts: 365
  • Joined: 18 May 2011
  • Loc: Beaverton, OR

Posted 27 April 2015 - 01:49 AM

 

  • Integral heated window/UV-IR filter.

I might be able to coat windows with ITO with copper pads for heating.



#80 steveastrouk

steveastrouk

    Apollo

  • *****
  • Vendors
  • Posts: 1,166
  • Joined: 01 Aug 2013
  • Loc: State College, Pa.

Posted 27 April 2015 - 07:50 AM

 

 

  • Integral heated window/UV-IR filter.

I might be able to coat windows with ITO with copper pads for heating.

 

 

 

The way I had it, I had a washer shaped 1/8" circuit board, with a copper spiral on it, with the filter/window in the centre. The PCB dissipated about 150mW, as I was playing with it - I have no idea what it would really take - and I figured that the heat at the edge would conduct nicely to the centre of the glass.

 

Steve



#81 wasyoungonce

wasyoungonce

    Viking 1

  • -----
  • Posts: 543
  • Joined: 07 Jun 2007
  • Loc: Land Downunder

Posted 28 April 2015 - 12:07 AM

Ok, my fleabay Nikon D70s body arrived today.

 

Works just fine as said, seller threw in extra battery also CF card, another battery and charger...$86 AUD (~$78USD).   Nice pixel size...7.8um I believe, which is a real nice size.

 

I want to play with it for awhile (I've purchased a "D ring" to test it fully, I have no lenses for it in anycase) then when the DIY camera is finalised......it'll be relived of its Sony ICX453AQ sensor! :smirk:   I only purchased the working D70s body as this was selling on ebay and only had an hour to go and not many bids.  So I bid and won (under my max auto bid) at what I consider a low value for a working unit!  The gods do shine occasionally upon me!

 

Now on to learn more about this camera!

 

Brendan


  • Raginar and hippieua like this

#82 steveastrouk

steveastrouk

    Apollo

  • *****
  • Vendors
  • Posts: 1,166
  • Joined: 01 Aug 2013
  • Loc: State College, Pa.

Posted 28 April 2015 - 09:35 AM

Hippie is respinning the board, so its 4" diameter - and my new camera case can accommodate 6 filters and a dark.


  • Raginar likes this

#83 Raginar

Raginar

    Cosmos

  • *****
  • Posts: 9,594
  • Joined: 19 Oct 2010
  • Loc: Pensacola, FL

Posted 28 April 2015 - 11:03 AM

My guess would be no. Most of the fancy cameras use a two stage tec. How do they work? Do you just put a tec on the tec? I think if we could get it down to -10 or 15C during the summer, people would be pretty darn happy.

#84 hippieua

hippieua

    Vostok 1

  • *****
  • topic starter
  • Posts: 192
  • Joined: 12 Dec 2014
  • Loc: Kiev

Posted 28 April 2015 - 11:54 AM

tec on tec is not enough. They have to be different sizes. Coz you cant cool the hot side of one tec with the same(size and power) tec.

 

BTW: is anyone here can offer effective and proven shema for controlling tec by current limiting? (not by PWM)



#85 gregj888

gregj888

    Mercury-Atlas

  • -----
  • Posts: 2,511
  • Joined: 26 Mar 2006
  • Loc: Oregon

Posted 28 April 2015 - 12:25 PM

 Raginar,

 

You can stack your own but they don't work as well and the commercially made multistage.  The photo I posted above is a home brew stack I was testing until the multistage arrived.  As Hippieua says, you need the right cooling at each stage.  I think it’s got a 4 stage now and under vacuum (~3x10**-4 T) gets to -65 (20c water for cooling) with a little more to do.  The commercial stacks don’t have the extra interface between TECs and are optimized stage to stage for cooling. 

 

There are TECs designed that move a lot of heat but not a lot of delta-T and others with good delta-T but fewer watts of cooling.  So the right single stage may be enough.

 

The issues with multistage are-

1)  hard to find anything above a 2 stage in stock, but there are a few

2) Very little cooling (watts) on a 4 or 5 stage even though the delta-T is high < 10w  Heat load goes way up with a difference of -100c

 

Low hanging fruit would be a 2 stage TEC cooler and water cooling on the back side.   Use a copper CPU cooler off EBay for the heat exchanger behind the TEC.  If you need more cooling, cool the water to just above dew-point.  Check out some of the “over clocking” DIY cpu water chillers for ideas.   Note, these take power… not great for battery powered operation.

 

Hippieua, I have one that seems to work...  I think I updated the schematic with the modes.  It's a copy of another on-line controller with some minor tweaks.

Attached Files


Edited by gregj888, 28 April 2015 - 05:37 PM.

  • hippieua likes this

#86 mdavister

mdavister

    Mariner 2

  • -----
  • Posts: 259
  • Joined: 10 Jul 2007
  • Loc: Pennsylvania

Posted 28 April 2015 - 12:48 PM

tec on tec is not enough. They have to be different sizes. Coz you cant cool the hot side of one tec with the same(size and power) tec.

 

BTW: is anyone here can offer effective and proven shema for controlling tec by current limiting? (not by PWM)

 

You are going to want to use PWM in some form to control the current into the TEC.  You could set up the control loop to run in current mode so the feedback signal is the current instead of the voltage.  If one is concerned about electrical noise one option would be to use a very slow PWM and limit the rate of voltage change such that EMI is reduced.  Another thing possible would be to shut down the cooler (or set it full on) during the image download.  This would keep any transitions down to a minimum and help reduce the noise generated.



#87 hippieua

hippieua

    Vostok 1

  • *****
  • topic starter
  • Posts: 192
  • Joined: 12 Dec 2014
  • Loc: Kiev

Posted 28 April 2015 - 01:09 PM

 

tec on tec is not enough. They have to be different sizes. Coz you cant cool the hot side of one tec with the same(size and power) tec.

 

BTW: is anyone here can offer effective and proven shema for controlling tec by current limiting? (not by PWM)

 

You are going to want to use PWM in some form to control the current into the TEC.  You could set up the control loop to run in current mode so the feedback signal is the current instead of the voltage.  If one is concerned about electrical noise one option would be to use a very slow PWM and limit the rate of voltage change such that EMI is reduced.  Another thing possible would be to shut down the cooler (or set it full on) during the image download.  This would keep any transitions down to a minimum and help reduce the noise generated.

 

Its not about the noise. To maintain the direction of the heat flow, DC current is required.

 

More info here


  • wasyoungonce likes this

#88 macona

macona

    Ranger 4

  • *****
  • Posts: 365
  • Joined: 18 May 2011
  • Loc: Beaverton, OR

Posted 28 April 2015 - 01:22 PM

FWIW, in the stuff from SITe my friend got there are a whole bunch of the double stage peltiers he is selling that were used in their cameras.



#89 gregj888

gregj888

    Mercury-Atlas

  • -----
  • Posts: 2,511
  • Joined: 26 Mar 2006
  • Loc: Oregon

Posted 28 April 2015 - 03:56 PM

macona,

 

Is that at Surplus Gizmos?  If so, good to know.

 

One note, the coolers need to be sized for the heat load and cooling desired, then add some extra.  The photometer has 3 stage cooler (just checked)  listed as a max delta-T of about 110c, I get about 85c at the cold finger.  If you just want "some cooling" and temp regulation, then it's easier and a surplus cooler would fill the bill.

 

hippieua, I've only seen analog circuits on the few scientific cameras I've seen.  The coolers will be mostly on anyway...  The driver on the circuit I posted is low resistence and doesn't get hot when cooling my flea3 or Skyris (whole camera).  Regulation is quite good, better than 1c in cruddy service (whole camera) and probably closer to 0.2c  With a little tuning in a camera should be well under 0.1c.  One of the circuits I borrower from claims 0.01c, but need to see that to claim it... :-)   I'll know more when I connect it to the Photometer.

 

Mitch, I agree PWM would be more efficient and the drivers will run cooler, but probably not the issue when counting electrons.  Not saying you couldn't make it work, but the analog works well and is proven.  Peltiers aren't supposed to like PWMs much either, but as you say, some edge shaping maybe some extra caps would help on that front.

 

If you want to lower the power required, the best way is to lower the cooling requirement: delta temp, heat load.  Temp is easy, heat load is conduction (slots, small wires, few wires, vacuum but it takes a good vacuum).

 

As always, YMMV



#90 wasyoungonce

wasyoungonce

    Viking 1

  • -----
  • Posts: 543
  • Joined: 07 Jun 2007
  • Loc: Land Downunder

Posted 28 April 2015 - 08:20 PM

QTY 2 Nikon D70 sensors avail here cheap, ebay $40USD $5 P&P:



#91 macona

macona

    Ranger 4

  • *****
  • Posts: 365
  • Joined: 18 May 2011
  • Loc: Beaverton, OR

Posted 28 April 2015 - 08:55 PM

macona,

 

Is that at Surplus Gizmos?  If so, good to know.

 

 

 

Yup, that's it.

 

-Jerry



#92 steveastrouk

steveastrouk

    Apollo

  • *****
  • Vendors
  • Posts: 1,166
  • Joined: 01 Aug 2013
  • Loc: State College, Pa.

Posted 28 April 2015 - 09:02 PM

 

 

tec on tec is not enough. They have to be different sizes. Coz you cant cool the hot side of one tec with the same(size and power) tec.

 

BTW: is anyone here can offer effective and proven shema for controlling tec by current limiting? (not by PWM)

 

You are going to want to use PWM in some form to control the current into the TEC.  You could set up the control loop to run in current mode so the feedback signal is the current instead of the voltage.  If one is concerned about electrical noise one option would be to use a very slow PWM and limit the rate of voltage change such that EMI is reduced.  Another thing possible would be to shut down the cooler (or set it full on) during the image download.  This would keep any transitions down to a minimum and help reduce the noise generated.

 

Its not about the noise. To maintain the direction of the heat flow, DC current is required.

 

More info here

 

 

 

I was going to say, the best way to do this is with SMPS technologies., but I am leery of putting a SMPS inside something as sensitive to noise as a CCD camera.



#93 wasyoungonce

wasyoungonce

    Viking 1

  • -----
  • Posts: 543
  • Joined: 07 Jun 2007
  • Loc: Land Downunder

Posted 28 April 2015 - 10:07 PM

QTY 2 Nikon D70 sensors avail here cheap, ebay $40USD $5 P&P:

Jeez...some gents on Australian IceInSpace forum snapped them up within minutes......guessing they are joining the project ;)

 

Steve, expect some PMs :lol:

 

Brendan



#94 glend

glend

    Surveyor 1

  • *****
  • Posts: 1,958
  • Joined: 04 Feb 2014

Posted 29 April 2015 - 02:37 AM

I was one. PM sent Steve.

 

Some of the ideas flying around in this thread seem to be inceasing the complexity of the original project design. Honestly a single stage TEC is all that's needed for a delta T of -25C, the fan and heatsink being an important choice. CPU cooler heatsink/fans with copper pipes move alot of heat but add weight (I run one on my Canon 450D cold finger).

 

I want the 2" focuser, that makes sense.

,

Also, the addition of a filter wheel and going mono will increase the cost of this project significantly for those of us who do not own a set of filters, or desire to. So what is in, what is out, and what is being held back for the next version.  I realise you can't please everyone, and  I was happy enough with OSC and single stage TEC.


Edited by glend, 29 April 2015 - 03:31 AM.

  • Redstone2 likes this

#95 mdavister

mdavister

    Mariner 2

  • -----
  • Posts: 259
  • Joined: 10 Jul 2007
  • Loc: Pennsylvania

Posted 29 April 2015 - 06:59 AM

 

I was going to say, the best way to do this is with SMPS technologies., but I am leery of putting a SMPS inside something as sensitive to noise as a CCD camera.

 

 

SMPS are probably in a great deal of CCD cameras.  It's about the only way you will get a camera that will run on 12V single input.  Most CCD's (as I'm sure alot of you know) require multiple bias and drive voltages.  I do SMPS design professionally, it is possible to get them to be low noise, provide the DC that a TEC requires and keep EMI down.  A lot of this requires careful layout and attention to the circuit design. 

 

There may have been a misunderstanding about using a PWM to power a TEC.  It seems the general assumption was that the PWM would be directly driving the TEC, thus what was linked to the article was correct.  I was thinking of a filtered PWM driving the TEC, you get the best of both worlds then. 

 

If clean DC is critical, a lot of applications will use a SMPS for the main bulk power conversion, followed by some sort of linear post regulator plus a filter network to clean up the power.



#96 steveastrouk

steveastrouk

    Apollo

  • *****
  • Vendors
  • Posts: 1,166
  • Joined: 01 Aug 2013
  • Loc: State College, Pa.

Posted 29 April 2015 - 07:33 AM

I was one. PM sent Steve.

 

Some of the ideas flying around in this thread seem to be inceasing the complexity of the original project design. Honestly a single stage TEC is all that's needed for a delta T of -25C, the fan and heatsink being an important choice. CPU cooler heatsink/fans with copper pipes move alot of heat but add weight (I run one on my Canon 450D cold finger).

 

I want the 2" focuser, that makes sense.

,

Also, the addition of a filter wheel and going mono will increase the cost of this project significantly for those of us who do not own a set of filters, or desire to. So what is in, what is out, and what is being held back for the next version.  I realise you can't please everyone, and  I was happy enough with OSC and single stage TEC.

 

 

You don't have to have the filter, or mono, the deal is still about the board and components. My designing a case is a side project.

 

In discussions with other here, I am proposing options like this:


1.) Semiconductors, passives and PCB (full electronic kit)

2.) Semis only

3.) PCB only

 

Then a case design:

 

4.) Cold finger in aluminium

5.) Case with cold finger and window assembly and filter, complete with sealed connectors for electronics

6.) As 5, no filter.

7.) Any components in the case.

 

8.) TEC/Peltier cell to suit.

 

9.) All of 1.) + 5) (or +6)  and 8 (complete camera kit)

 

 

Price for item 1 to be $50..$60, subject to final PCB costs.

 

Steve


  • wasyoungonce likes this

#97 steveastrouk

steveastrouk

    Apollo

  • *****
  • Vendors
  • Posts: 1,166
  • Joined: 01 Aug 2013
  • Loc: State College, Pa.

Posted 29 April 2015 - 07:35 AM

 

 

I was going to say, the best way to do this is with SMPS technologies., but I am leery of putting a SMPS inside something as sensitive to noise as a CCD camera.

 

 

SMPS are probably in a great deal of CCD cameras.  It's about the only way you will get a camera that will run on 12V single input.  Most CCD's (as I'm sure alot of you know) require multiple bias and drive voltages.  I do SMPS design professionally, it is possible to get them to be low noise, provide the DC that a TEC requires and keep EMI down.  A lot of this requires careful layout and attention to the circuit design.

 

There's a switcher in the camera as it is now. I've fought shy of designing them, except simple designs using the National SimpleSwitcher, in anything I've done: I just buy them for integration.

 

Hippie would really appreciate your input for the TEC driver.



#98 1E1HFPPE

1E1HFPPE

    Ranger 4

  • -----
  • Posts: 318
  • Joined: 19 Apr 2009
  • Loc: Washington, DC

Posted 29 April 2015 - 08:17 AM

I would like to sign up for the group buy if it is not too late.

 

Randy



#99 pbunn

pbunn

    Vanguard

  • *****
  • Posts: 2,000
  • Joined: 29 Sep 2010
  • Loc: Spartanburg, SC

Posted 29 April 2015 - 08:19 AM

I got a working D70 this morning on the auction site. Looking forward to the build.

 

Pat



#100 mdavister

mdavister

    Mariner 2

  • -----
  • Posts: 259
  • Joined: 10 Jul 2007
  • Loc: Pennsylvania

Posted 29 April 2015 - 08:28 AM

How much heat does a CCD generate while operating?  I would guess not much, but have never really looked at it.




CNers have asked about a donation box for Cloudy Nights over the years, so here you go. Donation is not required by any means, so please enjoy your stay.


Recent Topics





Also tagged with one or more of these keywords: astrophotography, ccd, DIY, equipment, imaging



Cloudy Nights LLC
Cloudy Nights Sponsor: Astronomics