1252 replies to this topic

[wasyoungonce]

mdavister, on 25 Apr 2015 - 03:33 AM, said:

    To clarify, the kit would not include the CCD sensor, you would still have to salvage one, or procure one yourself.

To clarify, the kit would not include the CCD sensor, you would still have to salvage one, or procure one yourself.

 

 

 

Yep my D70s (body only, supposedly working ok) purchased on fleabay is on its way to me (incl CF card, battery and charger).  Picked it up for $86 AUD...~ $67 USD.

Lets hope it arrives ok.

Edited by wasyoungonce, 24 April 2015 - 07:12 PM.

[steveastrouk]

We're now over 12, and some of the current project members are trying to haul in their friends too, so it looks like we'll reach the better magic number of 25 sets.

 

I've revised my model case, and I'll be putting that up for critical comment later.

 

I am grateful for all the helpful comments I've received by PM from CN members for ways to accomplish this.

 

HippieUA is still enjoying a nice vacation somewhere warm, so I need to seek his comments on the PCB design, before I place an order for the quantity.

[hippieua]

Wow, nice! We have build around 20 cams for 2,5 years. So i wonder, those 12+ people from here are more interested in getting "end user device" for astrophotography or taking part in developement or maybe some other reasons? Please let me know, because I am very excited with so many participants.

[steveastrouk]

Wow, nice! We have build around 20 cams for 2,5 years. So i wonder, those 12+ people from here are more interested in getting "end user device" for astrophotography or taking part in developement or maybe some other reasons? Please let me know, because I am very excited with so many participants.

 

This is a step to building a community of camera builders/users which kind of died out when the "Cookbook CCD camera" chips became way smaller than the latest and greatest CCDs. The Webcam hackers seemed to give up too. One of the problems was the speed that you could practically download a device over a standard serial port - relatively simple USB interfaces have revolutionised the speed issues.

 

Steve

[hippieua]

Ye Steve, i know your motivation as we talked a lot, but what for the others?

[steveastrouk]

Ye Steve, i know your motivation as we talked a lot, but what for the others?

 

...its what ATM'ers DO ?

[Jacques 370]

Hello.
I am also interested in a kit for this exciting project.

[Raginar]

What Steve said. I just think it'd be neat to have a hobbyist camera. I don't do development or anything. I just like tinkering.

[gregj888]

Steve,

 

I'm interested as well.  I have an old SITE thinned Back Illuminated 1k x 1K chip that was going into a SW Cryo camera.  I lost steam on the build and the design has gotten so old I haven't gotten back to it.  At this point a new power supply and controller might bring it back to life.  It's monochrome, so have to deal with that but shouldn't be a big deal.

[steveastrouk]

Well, as I promised, here are a couple of renders of my new version. I reverted to an off the shelf box, with the lid replaced by a custom 1/4" (6.35mm) aluminium plate. Box is roughly 120 x 100mm x 36mm, 5" x 4" x 1.25"

Weight TBA

 

Salient features.

• 2" focusser mounting.
• Integral heated window/UV-IR filter.
• 30mm square TEC cooler, with cold finger clamped to CCD.
• Large (rather pretty I thought....) heatsink, clamps onto hot side of TEC with a "hot finger" with a drill chuck (Janot) taper lock.
• Large screw cover, with 'O' ring seal,  to access filters for changing, also, cover is deep and hollow, to hold a big bag of dessicant, retained with a spring strip.
• 5 position filter wheel, driven by DC motor inside system - can't wait to get the little motor I've found for it.
• Everything is assembled to the 1/4 inch plate - you can design your own housing, if you don't want to use the box I've found. This is to keep the plane of the CCD exactly perpendicular to the axis of the focusser.

Not shown, but thought about: hermetic, or at least IP68 connectors for USB and power, screws and fixings, and the support to hold the CCD assembly.

Attached Thumbnails

• 
• 

Edited by steveastrouk, 25 April 2015 - 08:13 PM.

[wasyoungonce]

Gents if your interested then PM steveastrouk!

 

My first major foray...was to modify my Philips SPC900NC and install a mono CCD sensor and flash it for raw monochrome out.   **** thing works a treat.

 

I'm really not good at development software, just soldering (so I can assist here) and making noise...FWIW.   But my main interest is to achieve a DIY camera and the process & fun of getting there.

 

Brendan

[ahend]

Nice work Hippieua, I appreciate your approach and dedication.
When you talk about the possibility of helping in the development of your project, in what areas would you most likely involve the CN community? I could see beta testing by almost any user but others might offer firmware programming help to assemblers, application development for testing and tuning, design and machining of mounts, heat sinks or casements, even circuit analysis.

Andrew

[steveastrouk]

Nice work Hippieua, I appreciate your approach and dedication.
When you talk about the possibility of helping in the development of your project, in what areas would you most likely involve the CN community? I could see beta testing by almost any user but others might offer firmware programming help to assemblers, application development for testing and tuning, design and machining of mounts, heat sinks or casements, even circuit analysis.

 

One thing he and I have talked about is noise optimising the circuit board - not a field I've had to do a lot of work in, but at this 14-16 bit level, its critical.

[ahend]

I see there is a lot of copper pouring in that regard on the PCB's. Perhaps some physical shielding would also be helpful?

[ahend]

Well, as I promised, here are a couple of renders of my new version. I reverted to an off the shelf box, with the lid replaced by a custom 1/4" (6.35mm) aluminium plate. Box is roughly 120 x 100mm x 36mm, 5" x 4" x 1.25"

Weight TBA

 

Salient features.

• 2" focusser mounting.
• Integral heated window/UV-IR filter.
• 30mm square TEC cooler, with cold finger clamped to CCD.
• Large (rather pretty I thought....) heatsink, clamps onto hot side of TEC with a "hot finger" with a drill chuck (Janot) taper lock.
• Large screw cover, with 'O' ring seal,  to access filters for changing, also, cover is deep and hollow, to hold a big bag of dessicant, retained with a spring strip.
• 5 position filter wheel, driven by DC motor inside system - can't wait to get the little motor I've found for it.
• Everything is assembled to the 1/4 inch plate - you can design your own housing, if you don't want to use the box I've found. This is to keep the plane of the CCD exactly perpendicular to the axis of the focusser.

Not shown, but thought about: hermetic, or at least IP68 connectors for USB and power, screws and fixings, and the support to hold the CCD assembly. 

 

 

 

Great looking case, and yes, heat sink. What would be the footprint for the CCD electronics? 

[steveastrouk]

Great looking case, and yes, heat sink. What would be the footprint for the CCD electronics? 

 

LOL

 

Mmm. As currently drawn, about 4 5/16", (107mm) and 3" wide (77mm). Actual height, I'm not sure. The CCD is the highest part apart from connectors I think. I need to know for the case drawings.

 

Steve

[gregj888]

Steve,

 

Not sure this is the time or the place, but a couple of comments on noise.  One of the easy things to do is to split the power planes so the analog and digital portions have separate regulators and power planes.  This would probably take a 4 layer board to implement. 

 

There are optio-isolators but TI for one makes solid state isolators with better response.  These can also level shift I believe.

 

To add to Ahend's comment.  a few grounded thruholes and break in the solder mask on the ground plane would allow a physical metal shield to go over the CCD pin area and analog amp (1/2 the ADC).   Think of a thin metal box that would cap the pin area with a hole for the cold finger.  

 

You have to be careful, but slots in the PCB isolating the CCD help cut thermal load.  I have some Manganin wire if anyone wants to get crazy... :-)

 

In no way suggesting a change, simply a few comments and FYI.  Great job BTW.

 

My thought is to add a daughter ( Personality) board for the specific CCD.  Simply plug it into the  CCD socket if noise is a problem.  -60 to -70 C probably isn't the target cooling level either...  Please remember though, if you build it, we will hack it <G>

[ahend]

Steve,

 

Not sure this is the time or the place, but a couple of comments on noise.  One of the easy things to do is to split the power planes so the analog and digital portions have separate regulators and power planes.  This would probably take a 4 layer board to implement. 

 

There are optio-isolators but TI for one makes solid state isolators with better response.  These can also level shift I believe.

 

To add to Ahend's comment.  a few grounded thruholes and break in the solder mask on the ground plane would allow a physical metal shield to go over the CCD pin area and analog amp (1/2 the ADC).   Think of a thin metal box that would cap the pin area with a hole for the cold finger.  

 

You have to be careful, but slots in the PCB isolating the CCD help cut thermal load.  I have some Manganin wire if anyone wants to get crazy... :-)

 

In no way suggesting a change, simply a few comments and FYI.  Great job BTW.

 

My thought is to add a daughter ( Personality) board for the specific CCD.  Simply plug it into the  CCD socket if noise is a problem.  -60 to -70 C probably isn't the target cooling level either...  Please remember though, if you build it, we will hack it <G>

 

Greg, I have been tinkering with a Photometrics CCD and was wondering about the slots adjacent to the detector. I figured they were there to equalize the partial vacuum on either side of the PCB but was scratching my head over the size and position of these features. Very interesting, thanks for the insight!

 

Does the Manganin wire require a special flux or solder? I see it's almost 90% copper so I would gather it would behave.

 

Andrew 

[steveastrouk]

Steve,

 

Not sure this is the time or the place, but a couple of comments on noise.  One of the easy things to do is to split the power planes so the analog and digital portions have separate regulators and power planes.  This would probably take a 4 layer board to implement. 

 

There are optio-isolators but TI for one makes solid state isolators with better response.  These can also level shift I believe.

 

To add to Ahend's comment.  a few grounded thruholes and break in the solder mask on the ground plane would allow a physical metal shield to go over the CCD pin area and analog amp (1/2 the ADC).   Think of a thin metal box that would cap the pin area with a hole for the cold finger.  

 

You have to be careful, but slots in the PCB isolating the CCD help cut thermal load.  I have some Manganin wire if anyone wants to get crazy... :-)

 

In no way suggesting a change, simply a few comments and FYI.  Great job BTW.

 

My thought is to add a daughter ( Personality) board for the specific CCD.  Simply plug it into the  CCD socket if noise is a problem.  -60 to -70 C probably isn't the target cooling level either...  Please remember though, if you build it, we will hack it <G>

 

Some interesting ideas there. I hope Hippie is reading them on his beach somewhere...My understanding is after three days of sun and not much else, he's itching to play camera electronics again...

 

Steve

 

Steve

[Raginar]

Would you consider building the filter wheel with a 7 slot? If you're going to image with it in mono, you'll want LRGBHaSiiOiii.

Have you guys read the threads on stripping the Bayer matrix from these chips? It doesn't look terribly difficult and the results are fairly impressive. The 'mistakes' can usually be removed via flats.

[steveastrouk]

Would you consider building the filter wheel with a 7 slot? If you're going to image with it in mono, you'll want LRGBHaSiiOiii.

Have you guys read the threads on stripping the Bayer matrix from these chips? It doesn't look terribly difficult and the results are fairly impressive. The 'mistakes' can usually be removed via flats.

 

I've read it. Until I can source the sensors more readily, just shout and I'll come and watch.

 

Firing up the CAD now to see what 7 slots means to the CAREFULLY DESIGNED AND SELECTED DIMENSIONS I ALREADY HAD.

[steveastrouk]

 

Firing up the CAD now to see what 7 slots means to the CAREFULLY DESIGNED AND SELECTED DIMENSIONS I ALREADY HAD.

 

 

4.17", was 3.42", The only other mechanical changes are the box and front, all the rest are common to the camera core.

 

Steve

Edited by steveastrouk, 26 April 2015 - 10:15 AM.

[gregj888]

Steve,

 

Manganin solders like Cu...  It does need some lead length to be effective so the easiest way would be to wire the socket point to point.  It's all about the numbers of wires and cross section, so a carrier for the socket wired with the minimum numbers and sizes of wire are best thermally.

 

On by NIR photometer I got an extra degree C for every copper wire I swapped with manganin (~ #30).  Since then I've picked up some very thin wire but haven't tried it yet.  The NIR cold finger is at about 208k (-65 C) in a 22 degree ambient with water cooling.  The thin stuff (0.04mm) is off the normal wire charts, and I would only use it if both sides were mounted. 

 

For these temps you really need a vacuum "cryostat" and 4-5 stages of TEC cooling or dry-ice.  I would not suggest it for the general thread of this project.  IMHO, If you try to turn this into a really state of the art camera, it will loose it's more general appeal due to expense and build difficulties.  Comments below are only comments... ;-)

 

If you slot the PCB, add vias on each side of the slot so the traces can be cut and wire jumpers put in.  That's pretty easy and non-evasive.  Then someone can make the change if looking for that extra degree or three of cooling. If the vias are spaced for a header, the personality board is a slam dunk for those that want.  On the ground, you want a "single point ground... usually at the analog input of the A2D.  If you put a ground plane under the socket that is isolated from the main (analog)  ground plane, connect it with a trace as close to the A2D as possible. 

 

Also note, I am a hacker at this stuff!!!  There are likely others on this thread and Hippie is probably one of them, that are way ahead of me. 

 

I made a TEC controller that works pretty well.  Had I used SMT parts I could have made it two channel and wish I had.  What I did do was to add a "proto" area, a bunch of vias with access to the extra unused op-amp, pwr, gnd etc.  Nice to build in some flexibility.  A via on a signal you may want access too can really help later.  2 channels?--  one for the cold finger and one for a thermal shield around the cold finger sensor...  for my NIR photometer.

 

If you have a proto board and want to, I have access to a Spectrum analyzer and near field probes.  Would be easy to do a quick scan of the board and check for radiated EMI and frequencies.  No CCD is needed but it would need to be powered.  last half of the video-   https://www.youtube....h?v=CQgJmmn-eSE

 

Disclosure- I sell Rigol products in the Pacific NW, but that's why I have access... :-) 

[steveastrouk]

If you have a proto board and want to, I have access to a Spectrum analyzer and near field probes.  Would be easy to do a quick scan of the board and check for radiated EMI and frequencies.  No CCD is needed but it would need to be powered.  last half of the video-   https://www.youtube....h?v=CQgJmmn-eSE

 

Disclosure- I sell Rigol products in the Pacific NW, but that's why I have access... :-) 

 

Place I work bought ALL the kit to do their own EMC work, conducted, radiated, emissions, susceptibilty. Used it once....

 

BTW, if folks aren't aware, we are working in exponential realms, small changes in temperature massive affect dark current - so a couple of degrees is well worth getting.

 

Any mileage in making a water cooled backend option ?

 

Steve

Edited by steveastrouk, 26 April 2015 - 03:09 PM.

[wasyoungonce]

Would you consider building the filter wheel with a 7 slot? If you're going to image with it in mono, you'll want LRGBHaSiiOiii.

Have you guys read the threads on stripping the Bayer matrix from these chips? It doesn't look terribly difficult and the results are fairly impressive. The 'mistakes' can usually be removed via flats.

Yes I have...but I suspect it might be easier to get the unit working first then make it mono.  That said...you have to remove the sensor to remove the bayer matrix and this design has the sensor soldered in, obviously to minimise stray capacitance & noise.

 

As said removing the bayer also removes the microlenses and not really sure of the impact on this.

steveastrouk, on 27 Apr 2015 - 06:05 AM, said:

    Any mileage in making a water cooled backend option ?

 

Probably but I see this as a high end option.  Not that I don't see this DIY as "not high end" I see it as achieveable but I see adding water cooling as an unecessary thing.  Maybe it could be added as an option addon?

 

Brendan

Edited by wasyoungonce, 26 April 2015 - 05:45 PM.