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
1192 replies to this topic

#751 kalalaa

kalalaa

    Explorer 1

  • *****
  • Posts: 75
  • Joined: 28 Nov 2015

Posted 02 April 2016 - 04:58 AM

If you have good S1 signal goes into IC4 but bad RG comes out i think you need to check the components around IC4. R14 and C37 is the right components? Well soldered? Your RG looks like inverted.

Update:

Scope3_www.kepfeltoltes.hu_.jpg


Edited by kalalaa, 02 April 2016 - 05:27 AM.


#752 ahend

ahend

    Vostok 1

  • *****
  • Posts: 111
  • Joined: 22 Nov 2012
  • Loc: Grand Rapids, MI

Posted 02 April 2016 - 01:11 PM

IC4 S1 seems ok but the reset gate signal from IC4 is amiss. I will be scrutinizing IC4 and its support components.....

 

 

 

Scope1.jpg?raw=1

 

Scope2.jpg?raw=1


Edited by ahend, 02 April 2016 - 05:01 PM.


#753 kalalaa

kalalaa

    Explorer 1

  • *****
  • Posts: 75
  • Joined: 28 Nov 2015

Posted 02 April 2016 - 02:15 PM

IC4 Pin1 RG signal looks similar (but maybe on an analog scope the amlitude is bigger because it is faster). V-s and H-s are definitely the same.

But CCD pin 20 and 17 is really different. Your scope was at DC mode right? Maybe the Vsub positive spikes can cause the RG spikes or the RG cause the VSub. But the RG was ok before C69.

So Vsub comes from cxd1267 with a lot of components. around it. Some diodes (maybe wrong direction) some capacitors and resistors, so much more cold solders or other problems can cause in this area.

I can see on my results the Vsub has a large 10-11V offset and only negative spikes. You have positive and negative spikes and don't know the offset. Maybe a diode (D3 or D4)?

I don't know now.If you need some help (scoping, measuring) you can find me.



#754 ahend

ahend

    Vostok 1

  • *****
  • Posts: 111
  • Joined: 22 Nov 2012
  • Loc: Grand Rapids, MI

Posted 02 April 2016 - 05:04 PM

IC4 Pin1 RG signal looks similar (but maybe on an analog scope the amlitude is bigger because it is faster). V-s and H-s are definitely the same.

But CCD pin 20 and 17 is really different. Your scope was at DC mode right? Maybe the Vsub positive spikes can cause the RG spikes or the RG cause the VSub. But the RG was ok before C69.

So Vsub comes from cxd1267 with a lot of components. around it. Some diodes (maybe wrong direction) some capacitors and resistors, so much more cold solders or other problems can cause in this area.

I can see on my results the Vsub has a large 10-11V offset and only negative spikes. You have positive and negative spikes and don't know the offset. Maybe a diode (D3 or D4)?

I don't know now.If you need some help (scoping, measuring) you can find me.

 

You were right about CCD 20 and 17, I was AC coupled at the time. I had switched over to find the signal and forgot - idiot! I have updated the images above - take a look and 17 and 20 from CCD now.



#755 kalalaa

kalalaa

    Explorer 1

  • *****
  • Posts: 75
  • Joined: 28 Nov 2015

Posted 03 April 2016 - 02:59 AM

Better results but CCD 20 and 17 low signal is too wide. Mine is only 0.1sec wide but yours is 0.6sec. Don't know why. What was your expo times? If same 0.5sec as mine, maybe the VSub is the problem.... Check twice your R15-16-18; C34-39; D5- 6; C34 and diode direction.

I have no idea....



#756 kalalaa

kalalaa

    Explorer 1

  • *****
  • Posts: 75
  • Joined: 28 Nov 2015

Posted 03 April 2016 - 03:28 AM

I checked CXD1267 pin 17 and 18 and no signal just DC voltage. So Vsub on the CCD is not an input?! It's an output?! Can somebody verify this?

If yes we need go back to RG. But RG before C69 looks good so why bad after it?

 

If somebody make a board ask him to scope it out RG and Vsub without CCD.


Edited by kalalaa, 03 April 2016 - 04:08 AM.


#757 ahend

ahend

    Vostok 1

  • *****
  • Posts: 111
  • Joined: 22 Nov 2012
  • Loc: Grand Rapids, MI

Posted 03 April 2016 - 08:14 PM

I have been looking at some of the supporting components in my problem areas and have found at least one cap that is incorrect, C37 of IC4. Mine is measuring 172pf instead of 33pf, although it was ordered correctly. My 1uf caps are more like 20% tol. instead of 10% but I don't believe that should be a huge problem. 

 

I am going through the rest of the board now, component by component,(L, C, R), and desoldering them for a check. Looks like I will be ordering some more components....



#758 Toups

Toups

    Vostok 1

  • -----
  • Posts: 196
  • Joined: 01 Sep 2007
  • Loc: Greater Boston Area, Massachusetts, USA

Posted 03 April 2016 - 10:51 PM

Just a few brief notes, I managed to briefly get back to this project.

 

First the good news:

 

1) A paint stripper containing methylene chloride can be used to release the chip from the carrier.  I soaked it for several days and was able with a gentle pry to remove the chip from carrier.

 

2) Since I had bad luck trying to dead-bug solder U1 after lifting some pads to replace a chip that died due to a shorted image chip.  (This was before I broke down and bought a Weller WES51, with the right tip, maybe I would have had better luck) I bought some "QFN24 0.65mm 0.5mm to 2.54mm DIP Adapter PCB Board"s from China on e-Bay.  Soldered U1 to this board by using a ETDD tip from the backside after tinning the contact pads and tinning the chip "leads" and heat sink.  Also came back and used a ETU to ensure good contacts from the top.  Next cut the board down to about a 7 mm square using a Dremel cutoff wheel.  Stripped the solder mask off using an X-acto knife.  Removed all traces from bottom of board except heat sink solder pad (but did trim slightly) using the Dremel cutoff wheel as a grinder.  Soldered the board to my original board and used wires to repair.  A picture will better show what I did, but no time tonight.  After this, the U1 voltages were good and stable.  

 

3) Good news is that using a socket the CCD turns out to be a very good idea.  (See #4)

 

Now for the bad news:

 

4) Although this CCD is not shorted and drawing excessive current unlike the first I tried, it may not be working.  Since this came from a image board bought on e-Bay, it may be that the CCD was bad before I removed it from the circuit board or I wasn't careful enough removing it.  Time to debug and scope the circuit.  A few quick checks makes me think the board is good. The good news is that since the CCD is socketed it is easy to remove during testing or to replace.  After some additional debugging I'll see what I need to do.

 

Trials and tribulations:

 

5) I am using ImageJ version 149 from https://imagej.nih.gov/ij/index.html with the ASCOM plug ins from http://neutronoptics...com/ascom.html  Had lots of trouble recently with it crashing and don't know if I forgot how I managed to use it or something changed on my system from a recent update, but in order to get it to work I need to run ImageJ.exe in Windows XP SP3 compatibility mode on my Windows 7 system.  I won't mention how much time I wasted before I discovered (or rediscovered) this fact.

 

6) Notice that there is an ImageJ2 fork at http://imagej.net/ImageJ2  However, I couldn't figure out if it is compatible with the original ImageJ ASCOM drivers.  If anyone else knows more, I would be interested in learning more about this.

 

Question:

 

7) My -15 V supply is at -14.5 V and stable.  Is this in spec?

 

I'll report back at some point with pictures of my patch job and the results of my board debugging.


Edited by Toups, 03 April 2016 - 10:53 PM.


#759 Toups

Toups

    Vostok 1

  • -----
  • Posts: 196
  • Joined: 01 Sep 2007
  • Loc: Greater Boston Area, Massachusetts, USA

Posted 04 April 2016 - 10:53 PM

As promised, photos of my board repair: (See my previous post for some more discussion of this fix.)

cut.JPG

Backside after removing unneeded traces.

back.JPG

Initial attempt at attaching wires to new board.

wires.JPG

 

Ended up removing some insulation and replacing some wires for the final repair.  Note the diodes and a capacitor have been shifted from original positions or tilted to enable everything to fit.  In some cases solder mask was scratched off to allow a new placement.

 

Warning, the weight of the insulated wire lifted one of the new pads so I had to solder a wire directly to the pin.  So far it is holding up well.

 

Here it is after soldering (from backside) to original board and making all connections.

repair.JPG


Edited by Toups, 04 April 2016 - 10:54 PM.


#760 Toups

Toups

    Vostok 1

  • -----
  • Posts: 196
  • Joined: 01 Sep 2007
  • Loc: Greater Boston Area, Massachusetts, USA

Posted 05 April 2016 - 06:05 PM

Hope this isn't out of place here, but I noticed a Cam85 at http://astroccd.org/2015/11/cam85/ by a different author.  Is this worthy of a new thread?  (Not that I plan on abandoning this one anytime soon.)



#761 gehelem

gehelem

    Explorer 1

  • -----
  • Posts: 68
  • Joined: 01 Mar 2016
  • Loc: France

Posted 06 April 2016 - 01:02 AM

Yep !
There is also this pretty little Cam10...
Mmmmmmm
I know what i'll do in 2017 ☺

#762 WSG

WSG

    Lift Off

  • -----
  • Posts: 17
  • Joined: 05 Apr 2016
  • Loc: Germany

Posted 06 April 2016 - 10:49 AM

This is really a great project. Now I am eager to build also a cam with the rectangular board. Has someone a PCB of the latest version that he can sell to me?  I live in germany. Therfore something close to me would be nice to keep the shipping cost low.



#763 gehelem

gehelem

    Explorer 1

  • -----
  • Posts: 68
  • Joined: 01 Mar 2016
  • Loc: France

Posted 07 April 2016 - 07:30 AM

Back in the game !

I've restarted a new board  : I've checked all power supplies with my scope, everything looks to be ok except for -8V which is a little bit less (-7.4V), do you think it's acceptable ?
i would also like to try the "slow driver", where can i find it ?
Did anybody try the "test" sofware on Vakulenko's Github ?
https://github.com/v...m8s_view_wdebug

@Toups
I've socketed my CCD like you, thanks for the tip,but i think contacts are not so good as expected, I obtain different results when i try to move the CCD into the sockets

GLM


Edited by gehelem, 07 April 2016 - 07:35 AM.


#764 kalalaa

kalalaa

    Explorer 1

  • *****
  • Posts: 75
  • Joined: 28 Nov 2015

Posted 07 April 2016 - 11:36 AM

gehelem:

Install the latest driver and send me an email adress, and i will send you the two files. You need to replace them and you will have the slower driver.



#765 gehelem

gehelem

    Explorer 1

  • -----
  • Posts: 68
  • Joined: 01 Mar 2016
  • Loc: France

Posted 09 April 2016 - 12:22 PM

Thank you Kalalaa

I've tried with my new board with the slower driver, as i suspected my computer to be too slow.

I'm now testing with nebulosity4

The only thing i get is noise, and it's more clear when i place the camera under light.

But not image at all.

I must now investigate with signals, it will be difficult because my scope has no memory, i can't see anything , it's too fast.

 

Something "funny" also : sometimes, while image is taken, i can ear a little noise, like "biiizzzzz"... strange

GLM


Edited by gehelem, 09 April 2016 - 12:23 PM.


#766 kalalaa

kalalaa

    Explorer 1

  • *****
  • Posts: 75
  • Joined: 28 Nov 2015

Posted 09 April 2016 - 01:11 PM

The noise is normal, i have it too. So the driver works?

Try with a lens, with offset 0 gain 63.



#767 gehelem

gehelem

    Explorer 1

  • -----
  • Posts: 68
  • Joined: 01 Mar 2016
  • Loc: France

Posted 10 April 2016 - 03:44 AM

Yes the driver works.

But no more luck with a lens and offset 0 gain 63

I must check signals like you did with Ahend



#768 gehelem

gehelem

    Explorer 1

  • -----
  • Posts: 68
  • Joined: 01 Mar 2016
  • Loc: France

Posted 10 April 2016 - 11:11 AM

difficult, difficult.

my old scope shows a lot of things, but i must admit i don't know what...

What should i trigger with ? 



#769 ahend

ahend

    Vostok 1

  • *****
  • Posts: 111
  • Joined: 22 Nov 2012
  • Loc: Grand Rapids, MI

Posted 10 April 2016 - 12:19 PM

difficult, difficult.

my old scope shows a lot of things, but i must admit i don't know what...

What should i trigger with ? 

 

This is from Grim on astroccd.org and details CAM84 timing. You can trigger on negative edge of ACK to start.

 

Shown in 4 65uS chunks so you need an appropriately small time period.

 

The entire forum entry can be most helpful. Translate the web page with Google or your favorite interpreter.

http://www.astroclub...p?topic=28929.0

 

GrimTiming.jpg?raw=1

 

 

The analog signal at the end shows reset, reference level "black", and the actual pixel signal.


Edited by ahend, 11 April 2016 - 09:57 PM.


#770 Toups

Toups

    Vostok 1

  • -----
  • Posts: 196
  • Joined: 01 Sep 2007
  • Loc: Greater Boston Area, Massachusetts, USA

Posted 10 April 2016 - 10:35 PM

I am having issues with Q1.  It is configured as a voltage follower.  After removing the CCD from the socket, I tried injecting a signal into pin 19 (CCD Vout) of my socket.  I do not see the corresponding signal on the output of Q1 at R21.  So I replaced Q1 with a spare (IRLML2803TRPBF) but still no signal.  So either Q1 is exceedingly fragile and I destroyed both trying to solder or I have another problem to track down.  Did not find any other problems on an initial check of the related components.

 

Injecting a signal at R21 does allow the A/D to digitize it so the rest of my circuit appears to be working (more or less.)

 

Just a suggestion, those having trouble getting a image might want to check Q1.



#771 gehelem

gehelem

    Explorer 1

  • -----
  • Posts: 68
  • Joined: 01 Mar 2016
  • Loc: France

Posted 11 April 2016 - 03:41 PM

@Toups
What kind of signal do you inject ?

#772 Toups

Toups

    Vostok 1

  • -----
  • Posts: 196
  • Joined: 01 Sep 2007
  • Loc: Greater Boston Area, Massachusetts, USA

Posted 11 April 2016 - 09:42 PM

A couple volt square wave at around a Megahertz or so.  Something's odd, Q1 seems to be working today but probably have bad CCDs.  Might need to check for a cold solder joint around Q1.  Unfortunately, it may be a while before I can get back to this project again.

 

Careful injecting a signal, Q1 could be easily damaged by static, not sure I recommend injecting a signal except as a last resort.

 

May have to go to a known working CCD.  I have a camera with a bad shutter that works occasionally so I know the CCD is good.  (The CCDs I am currently trying came from boards that were sold as working that were removed from camera.)

 

May build board number 2 before going to the known good CCD.  (Was hoping to use it to repair a camera with good shutter but bad CCD.)


Edited by Toups, 11 April 2016 - 09:43 PM.


#773 gregj888

gregj888

    Vanguard

  • -----
  • Posts: 2044
  • Joined: 26 Mar 2006
  • Loc: Oregon

Posted 13 April 2016 - 12:41 AM

Started my build Sunday and did a skillet reflow tonight.  Pretty happy so far but have a lot of checking and verifying to do.  A visual inspection with a microscope showed a couple solder bridges and I found a couple of dry  and overly wet pins.  I didn't use a stencil and that would have helped a lot as would solder paste that hadn't expired :-).

 

Since I wanted to use a reflow technique I'm building the whole board minus the CCD and Q1.  Will add those once everything else checks out.  I didn't want to reflow the board more than once or twice, so a full build seemed the best option (for me).

 

Much more to come, so may split to a separate build blog (there's already 31 pages to wade through, a sort form might be nice).  

Attached Thumbnails

  • pre-reflow.jpg
  • post-skillet-reflow.jpg

  • ahend likes this

#774 ahend

ahend

    Vostok 1

  • *****
  • Posts: 111
  • Joined: 22 Nov 2012
  • Loc: Grand Rapids, MI

Posted 13 April 2016 - 05:45 AM

Can someone please confirm my read back settings of the FT2232H, shown below? 

 

I went back into MProg and did a Read and Parse under the Tools menu. The EEPROM data that is read from the chip will then populate the MProg interface. What comes back is a bit different from the cam84.ept template data.

 

We are disabling the USB serial number so I do not see an issue with the "Di" serial number prefix, or the "Disabled" indication in the serial number string itself, but would that also account for the USB version to show 1.1?

 

 

MPROG-RP.jpg?raw=1



#775 gehelem

gehelem

    Explorer 1

  • -----
  • Posts: 68
  • Joined: 01 Mar 2016
  • Loc: France

Posted 13 April 2016 - 05:48 AM

I think i had the same problem : try to unckeck "disable usb serial number" before writing


Edited by gehelem, 13 April 2016 - 05:59 AM.



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