1220 replies to this topic

[archimede]

 first light from the cam 84, attached with adhesive tape to the fire of a newton 200mm f5

 

[ahend]

 first light from the cam 84, attached with adhesive tape to the fire of a newton 200mm f5

 

 

 

Great work! I'm glad you stuck to it and made a working camera out of it!

[Toups]

 

 

In particular, I am looking at ImageJ https://imagej.nih.gov/ij/index.html  I had some success using ImageJ version 149 and 150 with the http://www.neutronop....com/ascom.html additions.  However these did not work completely out of the box for me.  First I had to run in XP compatibility mode (probably due to an issue with ASCap.vbs that I haven't bothered to track down yet.)  Also I added the following code block to ASCap.vbs.  This allows me to adjust the cam84s settings dialog prior to exposure.  However since settings aren't remembered I will investigate further mods to ASCap.vbs of ASCap.txt to fix the settings issue.  Also ASCOM Platform 6 => Tools => Profile Explorer will let you make permanent changes to the settings.

 

Found the issue that was requiring the code to run in XP compatability mode.  Around line 150 or so in ASCap.vbs is the line

image.ImageArray = camera.ImageArray

change it to

Dim anImage
anImage = camera.ImageArray
image.ImageArray = anImage

After this change, it should run under Windows7 without requiring compatibility mode.  This change also allows it to work with the new ImageJ2 at http://imagej.net/Welcome

 

 

The version at NeutronOptics has been fixed with respect to this bug.  I sent this issue to the owner and he graciously updated the code.  You will still want the addition that creates a dialog to allow you to update the settings before the exposure.  This latter issue I did not pass back to him since this is a nuance unique to the cam8s drivers creating a settings dialog after the ASCom driver requests a connection.

 

At some point, I may cobble together a ImageJ plugin to access the cam8s ASCom drivers or port the drivers into an ImageJ plugin.  However that is a project for another time.

[axsdenied]

First of all a big thanks to all the people involved in this project for their hard work.

 

I am thinking of starting a build and I am trying to get an overview of the current state of various projects. Please correct me if I am wrong:

Cam84 (square PCB):
As far as I can see there has not been any changes for a while. The last revision gerbers I found is 84_v41.
However, the round version had several updates which are not in the square version, see here and here:
How important are those updates and should/will the be ported into the square version?

Cam84R (round):

The file archive contains gerbers for 84_R3.1 but there is a folder with a single image called 84_R3.21. Is this newer/finished?

 

Cam85:

Is this supposed to be a successor of cam84? What is the status of this build? Is it worth doing this instead of cam84?

Housing:
Square is simple (Gainta G107).

What about the round housing? Is there a file that can be used for CNC routing the housing? I have seen that few people have finished round housings.

 

Thank you

[hippieua]

All round 84 were worse than square one. So if you want to start building right now, take the Cam84_v41 gerbers.

 

Cam85/Cam85m right now is a draft, and its being in development, it give better results that all previous cameras. All info only on my local forum and a little bit on astroccd.

 

Housing ...

I saw only one round housing

[gregj888]

IMHO, the advantage of the round board is that it's easier to make a vacuum rated case for it and easier if you only have a lathe.

 

The rectangular board is easier with the stock case.

 

For those of us who just can't color inside the lines, you can do your own rectangular case as well.  You can also make a rectangular case that's vacuum capable, but this one is not.  This one does have water cooling for the Peltier and will seal pretty well.  There's a pocket for descant to try to keep frost off the sensor.  4 more holes to drill and tap for the plastic cap on the back and some paint, then assembly.

 

Greg

Attached Thumbnails

• 
• 
• 

[karlfoxman]

Wow Greg that's a lovely case! Nice work. 

[gregj888]

Karl,

 

Thanks, hope it works as intended...  The Peltier cooling should be fine, it's the pillars for the other chips that I hope keep the insides cool.  The black panel on the back is replaceable so I can integrate the Peltier controller at a later date.  I have one I designed some time back, but it's bigger than I want and no USB control.  The one I have will work for now, but may be a follow-on project.  Yes, watching yours as well... :-)

[karlfoxman]

I have a strange issue with the usb on my hub at the moment, trying to track the issue down. It appears as if something damaged the d- or d+ on my hub and I'm very sure it's no short to supply. It was only the camera usb that was damaged. The camera itself is still fine on a USB and the tec controller part of my hub is still fine. I didn't put any protection on the usb ports as the ic has a level of protection built in. I hope to find the issue soon. I did manage quite a few bias frames with the entire setup and a few dark frames at delta t of 40 deg c. Pretty impressive and a huge impact on the quality of the images. Good luck with your setup it really looks great.

 

I'll share my hub/tec when it's ready. Credit for the entire tec is with the guy who's link is in the first post it seems to work great! Just my hub needs work, a new hub ic arrives tomorrow and the old one is removed already.

[archimede]

Edited by archimede, 06 June 2016 - 06:07 PM.

[archimede]

now I have to figure out which depends all the vertical bands that are seen in pictures ..... is strange .....

[Rich_B]

are those vertical lines stable ? Same position, same configuration regardless of camera settings? Are they present in the full size image or just in the rescaled downsampled version? Are they present in darks ? In bias frames ?

[Rich_B]

IMHO, the advantage of the round board is that it's easier to make a vacuum rated case for it and easier if you only have a lathe.

 

The rectangular board is easier with the stock case.

 

For those of us who just can't color inside the lines, you can do your own rectangular case as well.  You can also make a rectangular case that's vacuum capable, but this one is not.  This one does have water cooling for the Peltier and will seal pretty well.  There's a pocket for descant to try to keep frost off the sensor.  4 more holes to drill and tap for the plastic cap on the back and some paint, then assembly.

 

Greg

Nice machining . How are the 2 posts touching the chips they're supposed to cool ? It seems pretty hard if not impossible to have the 2 post surfaces , the 2 chips surfaces, the 4 board mounting points , the CCD surface and its Peltier cooling aluminum post all in thermal contact ? These are too many surfaces to be made coplanar and stay coplanar after temp changes , circuit board bow due to mechanical and thermal gradients . Do you have some thermal pads somewhere? How's the CCD held in the socket ? Repeated thermal cycles of the aluminum post behind it could work it out of the socket gradually . The post expands because it's heated by the Peltier hot side, the socket and CCD pins shrink due to being cooled and as a direct result, the CCD back is pressed harder against the Peltier cold side . This causes the CCD pins to slip out of the socket . At least that's my theory     

[gregj888]

Rich,

 

You are full of questions :-)  Unfortunately, they are good ones... :-)  Real answer is I have the same ones.

 

I like options and flexibility, so here's where I am and have planned...

 

I do have some Gap Fill (referenced earlier, thank you)... I've measured and milled to the best of my ability and plan to use the Gapfill on the posts.  If that doesn't work, I'll drill the posts and put in spring loaded plungers with thermal compound, so lots of compliance.  There are actually posts on both top and bottom of the board.  Those on the CCD side contact the board, hoping to pull heat from the copper cladding.  The posts are not opposite each other s hopefully won't stress or crush anything.

 

BTW, the case should be almost iso-thermic with the water cooling, at least that's the expectation.   The camera only dissipates a couple of watts...

 

The board is pretty "springy" so provides some compliance for the CCD even being clamp top and bottom.  The cold finger is milled to a few thousands long and very square (top and bottom parallel).  The socket is quite tight.  With the compliance in the board, there may be some movement in the socket, but that should stop at some point.  Will the CCD still be in contact with the cold finger is a question.  Bottom of the Peltier has been lapped to the case, will also do the cold finger to the Peltier.  Not sure about the CCD and cold finger, doing that has other issues.  I'll use heat sink compound for these connections.

 

If I have trouble with the CCD unplugging itself, I'll put compliant "bumpers" between the CCD and front cover.  That will hurt cooling but I think I have plenty of that (and 2/3 stack or bigger Peltiers on the shelf).  Perfect example for needing through holes on the board to allow clamping the CCD to the cold finger.

 

So yes, and I'll have to try it to see how it does. 

 

Greg

[archimede]

I have to check if those vertical shadows also come with bias, and flat, I can just do the test and do you know thank you !!!

[flolic]

Housing is almost finished. Still to do cold chamber...



Sent from my SM-G920F using Tapatalk

Edited by flolic, 07 June 2016 - 05:37 AM.

[gehelem]

Still investigating why i have no signal on SL jumper :

Didn't find any bad solder....

Any idea ?

When is this signal active ? Acquisition or download or ????

Gilles

[vakulenko_sergiy]

SL = SLOAD pin at ad9826 - used for SPI communication between FT2232H and ADC.

Active signal is Low. Active when configuring ADC (initial configuration when connecting camera, gain-offset initialization before starting exposure).

So if there is a issue with SL - camera driver can't correctly configure ADC.

[gehelem]

SL = SLOAD pin at ad9826 - used for SPI communication between FT2232H and ADC.

Active signal is Low. Active when configuring ADC (initial configuration when connecting camera, gain-offset initialization before starting exposure).

So if there is a issue with SL - camera driver can't correctly configure ADC.

ok, thank you Sergiy

i think i can't see it because i'm triggering on SCK :

Regarding the datasheet of ad9826 , this event is only expected a few times, i'll go deeper tonight.

 

BTW, my ad9826 is on revision AD9826KRSZ, and i didn't find JRSZdatasheet to compare.

 

Gilles.

[vakulenko_sergiy]

AD9826KRSZ is a ordering code, see page 20 of ad9826 datasheet

[axsdenied]

Thank you for clarifying things. The square version it will be.
 

All round 84 were worse than square one. So if you want to start building right now, take the Cam84_v41 gerbers.

 
Are there any changes from the latest round version that are worth incorporating into the Cam84_v41? For example the changes listed in your posts here and here.

The parts "upgrade" was also mentioned here but the BOM is still the original so it may not even be worth doing it???

 

Housing ...
I saw only one round housing

My bad, all photos of the housing being built were from the same person (flolic).

[Toups]

 

 

 

In particular, I am looking at ImageJ https://imagej.nih.gov/ij/index.html  I had some success using ImageJ version 149 and 150 with the http://www.neutronop....com/ascom.html additions.  However these did not work completely out of the box for me.  First I had to run in XP compatibility mode (probably due to an issue with ASCap.vbs that I haven't bothered to track down yet.)  Also I added the following code block to ASCap.vbs.  This allows me to adjust the cam84s settings dialog prior to exposure.  However since settings aren't remembered I will investigate further mods to ASCap.vbs of ASCap.txt to fix the settings issue.  Also ASCOM Platform 6 => Tools => Profile Explorer will let you make permanent changes to the settings.

 

Found the issue that was requiring the code to run in XP compatability mode.  Around line 150 or so in ASCap.vbs is the line

image.ImageArray = camera.ImageArray

change it to

Dim anImage
anImage = camera.ImageArray
image.ImageArray = anImage

After this change, it should run under Windows7 without requiring compatibility mode.  This change also allows it to work with the new ImageJ2 at http://imagej.net/Welcome

 

 

The version at NeutronOptics has been fixed with respect to this bug.  I sent this issue to the owner and he graciously updated the code.  You will still want the addition that creates a dialog to allow you to update the settings before the exposure.  This latter issue I did not pass back to him since this is a nuance unique to the cam8s drivers creating a settings dialog after the ASCom driver requests a connection.

 

At some point, I may cobble together a ImageJ plugin to access the cam8s ASCom drivers or port the drivers into an ImageJ plugin.  However that is a project for another time.

 

 

It appears that adding the line

camera.connected = false

just before

End Sub

near the very end of ASCap.vbs will save the parameters.

 

I have a mod to turn on or off the dialog to allow parameter changes, but that is for another time.

[gehelem]

 

SL = SLOAD pin at ad9826 - used for SPI communication between FT2232H and ADC.

Active signal is Low. Active when configuring ADC (initial configuration when connecting camera, gain-offset initialization before starting exposure).

So if there is a issue with SL - camera driver can't correctly configure ADC.

ok, thank you Sergiy

i think i can't see it because i'm triggering on SCK :

Regarding the datasheet of ad9826 , this event is only expected a few times, i'll go deeper tonight.

 

BTW, my ad9826 is on revision AD9826KRSZ, and i didn't find JRSZdatasheet to compare.

 

Gilles.

 

Ok, i can see a very brief signal on SL, when  connecting

But nothing when starting exposure, sure it's my scope

=> will have to check other signals directly on CCD

I'll  compare with the ones provided on previous posts.

 

(OMG what am i doing ? ;-) )

Edited by gehelem, 09 June 2016 - 07:34 AM.

[gregj888]

Gehelem, sorry, on the road this week.  If you change the gain setting you will probably see a burst on SL as well.  Sounds like It's working as expected.

 

You saw the discussion earlier about the resister value confusion?  It's a few pages back.

 

Greg

[Jose25]

Hi, I'm first in this project.

 

But, I have a really big interest in this project because of my project in laboratory.

 

I trying to follow this topic but, In first page, schematics information of camera and another information is not connected.

 

I think it is expired(?).. So,,, dose anyone have an information of hippieua's project schematics? I really appreciate if anyone upload that information. 

 

Thank you so much.