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

First (viable) light with my new QHY247C ColdMOS camera

  • Please log in to reply
21 replies to this topic

#1 Lord Beowulf

Lord Beowulf

    Surveyor 1

  • *****
  • topic starter
  • Posts: 1850
  • Joined: 13 Oct 2008
  • Loc: Cedar Park, TX

Posted 03 February 2017 - 12:55 AM

So I received my beta test QHY247C ColdMOS camera last Friday and have had a few hoops to jump through to get things to a point where I can remotely automate and get some good results, but tonight I finally managed to pull everything together and get some decent subs to share.  As I'd predicted in December when I ordered, this arrived at the start of Chinese New Year but Dr. Qiu has been checking his e-mails while on the holiday and has been pretty responsive.  I'll be posting pictures of the setup, etc. as I have time, but although I should have been in bed an hour ago, I wanted to at least share the results from this evening.  This is a stack of 9 frames of ten-minute exposures of the Crab Nebula with the gain set to about 3/4 of maximum. The cooler was set for -15C.  No darks, bias, or flats are applied, and only a basic non-linear stretch was performed in Photoshop. I didn't bother framing perfectly since I was just trying to get something and had a good guide star on initial alignment so I went with it.  As you know, the moon is nearing first quarter and at this point was only about two hours (30 degrees) away from the Crab when I took these shots.  Overall I'm pretty pleased with the results now that I'm starting to get the settings dialed in.  I'll have to take new darks/bias/flats after changing the gain, but I want to play with it a bit to find the best settings first.   

 

Click on the photo for the full sized image and then zoom to full size in your browser to see the full resolution and the hot pixels. Note that JPEG artifacts will make those larger than they actually are.

 

4002.jpg link31.png

 

Enjoy!

 

Beo


  • Ricky, ArkabPriorSol, Pauls72 and 4 others like this

#2 Lord Beowulf

Lord Beowulf

    Surveyor 1

  • *****
  • topic starter
  • Posts: 1850
  • Joined: 13 Oct 2008
  • Loc: Cedar Park, TX

Posted 06 February 2017 - 05:55 PM

We've been cloudy since I posted the previous message, but I was able to go get some darks and bias matching this gain setting.  I applied those along with the original "blue sky" morning flats taken at the previous gain settings, stacking in DSS, and for the most part things came out pretty good.  I don't see any evidence of the dust spots, and the hot pixels are all gone.  

 

4044.jpg link31.png

 

Dr. Qiu commented on the slight yellow appearance of the original frames.  I'd scaled the FITS loader in DSS by ~1.54 and 1.56 for B and R to offset the obvious green caste in the original low gain images.  Those numbers came from the sample color balance data listed in the QHY247C manual, since I haven't had an opportunity to do any evaluation on my own.  I went ahead and removed those values for the above stack, and frankly can't tell much difference.  I think after changing the gain it may not matter as much.  Either that, or DSS isn't doing what I thought and hasn't removed the correction I'd applied.  He also suggested that a luminance filter might be needed since the QHY247C does not contain its own IR cut filter.

 

I went ahead and did my best to balance the color as best I could and finish a process on this, since I've never managed to capture a stack of M1 that looked as good as this.  I'd still like to have several more hours of integration on this target, not to mention fixing the color balance, but it's not too bad for a first start.  

 

4045.jpg link31.png

 

I may go back and play with the RGB background calibration in DSS to see what that does as well.  I'm not sure how much the blue (vs. white) flats did to me either.  The other problem I'm encountering with DSS is that it apparently wasn't designed with this sized files in mind and may have a memory leak.  It'll work the first time through, but if I attempt to run a second stack without closing and reloading I get an out of memory error.

 

Beo

 

 

 

 

 



#3 Lord Beowulf

Lord Beowulf

    Surveyor 1

  • *****
  • topic starter
  • Posts: 1850
  • Joined: 13 Oct 2008
  • Loc: Cedar Park, TX

Posted 06 February 2017 - 06:04 PM

As also promised previously, the next few posts will cover my experience with unpacking and setting up what I received.  There are more photos in the gallery than I'll post here, so follow the link icon in the first post to browse the rest of the gallery.  The link icon counts against my image limit per post, so I'll delete them on most of the images.

 

The box arrived a bit worse for wear inside a DHL shipping bag, but everything was packed well enough inside that it survived ok.

 

4005.jpg link31.png

 

4006.jpg

 

Inside the shipping box, we have the box with the camera itself, a 12V 6A (if I recall correctly) power supply, and a bag of cables and adapters.

 

4007.jpg

 

The accessory bag contains a USB 3.0 cable, a 2" nosepiece adapter, a larger clamp-on adapter designed to allow adjusting the face of the sensor to flatten the field, a short power supply extension cable with a locking connector, an automotive outlet power cable, and a tube with dessicant for drying the camera if needed.

 

4008.jpg

 

A nice foam layer surrounds the camera, and the paperwork includes a checkout sheet and a card directing you to the website for drivers and documentation. However, the one thing I would like to have seen would have been a bag with dessicant to keep the camera dust free and dry.

 

4009.jpg

 

Well, I have to go get on a plane here in a few minutes, so I'll have to continue this later!  For those who want to jump ahead, just browse the gallery.

 

Enjoy!

 

Beo


  • Ricky likes this

#4 Lord Beowulf

Lord Beowulf

    Surveyor 1

  • *****
  • topic starter
  • Posts: 1850
  • Joined: 13 Oct 2008
  • Loc: Cedar Park, TX

Posted 07 February 2017 - 10:52 AM

Front, back, and side views of the camera, showing the sensor, mounting flanges, USB and power connections, cooling fins and fan.

 

4010.jpg link31.png

 

4011.jpg

 

4012.jpg

 

For an idea of size, I compared it to my Celestron Nightscape.  It's quite a bit longer (largely due to the better cooling) and slightly smaller in maximum diameter.  The one thing the Nightscape has that I'll miss is the shutter that allows automated collection of darks.

 

4013.jpg

 

Without the shutter and given I didn't have an appropriate adapter to mate the camera's M54 or clamp ring to my Taurus Tracker III OAG, I had to come up with a solution to take some initial test darks.  I chose to use the supplied clamp ring and add a cover to it using some paper and masking tape, since that was all I had handy.  Frankly, with a simple insert in this to protect the sensor window, this would be an ideal way to ship the camera with its own integrated "lens cap". 

 

4019.jpg

 

4021.jpg

 

I had to route a new USB 3.0 extension to my pier, since my embedded solution is all USB 2.0.  Given the 46.2 MB per frame, the USB 3.0 interface is really a must!  Luckily I planned for expansion and have a 4" sewer pipe conduit running under the slab from the pier to the wall. I may rebuild my data ports in the future, but for now this gets me running.

 

4016.jpg

 

Finally, connecting to the PC I was able to get all the drivers set up and capture a set of darks

 

4018.jpg

 

Beo



#5 Lord Beowulf

Lord Beowulf

    Surveyor 1

  • *****
  • topic starter
  • Posts: 1850
  • Joined: 13 Oct 2008
  • Loc: Cedar Park, TX

Posted 07 February 2017 - 12:17 PM

Since I use the Taurus Tracker III off axis guider, and only have T-ring adapters for it, I didn't have a solution for the M54 threads, even with the 2" adapter.  So, it was back home to design and print an adapter that uses the outer clamp ring of the camera and has the mounting flange for the Tracker.  Turned out quite nice, although I forgot to shorten the depth of the camera mount side to make it bottom out on the face of the flange rather than the leveling screws on the face of the camera body as the factory ring is designed to do.  Thus, there is still the potential of light leakage through the serpentine path around the flange as warned in the camera documentation.  There are a couple of other concerns with using PLA for this part.  For one, the surface finish of PLA is quite shiny, so even though it's black, surface reflections may be possible.  Also, PLA is pretty low temperature and tends to flow/relax a bit over time, especially in warm environments.  I'm not sure what will happen to it this summer under the weight of the camera!

 

4022.jpg link31.png

 

4024.jpg

 

4025.jpg

 

Test fitting to my spare Tracker:

 

4027.jpg

 

4029.jpg

 

And the adapter just on the camera.  Note the dust motes already on the window, which explains some of the bad spots in the flats.  Need to clean that already!  I suspect that biggest piece at least is part of the packing foam.

 

4036.jpg

 

And finally, the camera on the telescope.

 

4037.jpg

 

Beo

 

 


  • BenKolt likes this

#6 Lord Beowulf

Lord Beowulf

    Surveyor 1

  • *****
  • topic starter
  • Posts: 1850
  • Joined: 13 Oct 2008
  • Loc: Cedar Park, TX

Posted 07 February 2017 - 12:32 PM

While this probably won't bother most users who set up their system each night, I discovered one design limitation in trying to remotely automate the camera with the rest of my system.  It turns out that the camera uses an internal I/O hub that is USB powered and then the actual camera interface that is powered from the 12V supply, both of which have separate drivers in Windows. Unfortunately that configuration apparently requires a particular startup sequence where the camera must be powered BEFORE connecting it to the USB interface. Since I remotely automate my observatory and power everything up remotely, that was not possible. The USB 3.0 extender and interface would remain powered by the PC and thus never shut down.  I'm hoping that a firmware or driver update will resolve this in the future, but my temporary work-around was to take a USB 3.0 hub I had and add an Arduino Nano controlled relay to disable the power on the USB 3.0 port going to the camera.  This is similar to the way the hubs in the piers work, except in that case the entire 5V supply for the hubs is shut down along with the 12V supply to all the components.

 

4041.jpg link31.png

 

Currently I just have this whole mess tossed up on the top of the pier under the mounting plate.  The cable that came with the hub was flakey, so I had to swap it. I also need to replace all the bright blue LEDs on the board, but for now some masking tape will have to serve to knock down the light.  The good news is that this appears to work so long as I remember the power up sequence!  Just one more thing to keep track of.

 

4043.jpg link31.png

 

Beo


Edited by Lord Beowulf, 07 February 2017 - 12:34 PM.


#7 Lord Beowulf

Lord Beowulf

    Surveyor 1

  • *****
  • topic starter
  • Posts: 1850
  • Joined: 13 Oct 2008
  • Loc: Cedar Park, TX

Posted 12 February 2017 - 11:37 PM

Last night I had a bit of time to take some lunar shots to play with my color balance.  It was windy and humid so I didn't bother trying to focus or anything.  I just wanted a good neutral density image I could use to set my white balance.  Below are the results.  This is considerably different than the numbers I'd tried before, giving me a 0.937 multiplier for red and a 1.912 multiplier for blue (DSS doesn't let you adjust green).  The one problem I see with these settings is that a saturated image comes out with a nonsensical color instead of being white.  Not sure how to fix that.  And while these settings address the color balance for the range of lunar exposures I took, the light from my 10-minute exposures on M1 is so far to the left of the histogram that this scaling doesn't seem to help.  I'm still playing with settings to see if I can figure out the right combination.  

 

4047.jpg link31.png

 

4046.jpg link31.png

 

Beo


  • mattyfatz and artem2 like this

#8 Lord Beowulf

Lord Beowulf

    Surveyor 1

  • *****
  • topic starter
  • Posts: 1850
  • Joined: 13 Oct 2008
  • Loc: Cedar Park, TX

Posted 27 February 2017 - 04:48 PM

I haven't had much opportunity to run further tests on this camera, although I did have one really good night where I THOUGHT I had decent guiding as well.  However, although DSS will detect over 100 stars when manually tested, it only finds a handful during the actual stacking process and refuses to stack more than about five frames out of the 30+ I have for each object.  I haven't had a chance to try anything else yet.  

 

One thing that struck me is that even at ten minute exposures, most of the image data I'm looking at is at the very bottom of the histogram and so I wanted to do a bit of an experiment on color balance with a brighter object.  Thus, over to the Orion Nebula to take a ten minute exposure there and see what happens.  This is the single frame with dark and flat corrections only:

 

4048.jpg link31.png

 

Applying some non-linear stretching of that and doing a bit of masking to bring in the lowest portion without blowing out everything else gives me this:

 

4049.jpg link31.png

 

If instead I do a linear stretch of the lowest 8% of the histogram I get this:

 

4050.jpg link31.png

 

In both cases, the color looks reasonable, but if I look at the histogram, it's apparent the red needs to be boosted a bit.

 

Capture.JPG

 

If I adjust the background calibration to align the histogram I get this, which definitely looks better in the red of the nebula:

 

4051.jpg link31.png

 

I think this was with the per-channel background calibration in DSS and I need to go back and see if the RGB calibration is any better.  I just wish there was a way in Photoshop to apply a linear correction to the entire image that is finer than the 0-255 that it provides in the user interface.  I've always considered it rather silly that my image is 16 bit, but I can only make 8 bit adjustments to it.  

 

Beo

 


  • mattyfatz and BenKolt like this

#9 Lord Beowulf

Lord Beowulf

    Surveyor 1

  • *****
  • topic starter
  • Posts: 1850
  • Joined: 13 Oct 2008
  • Loc: Cedar Park, TX

Posted 04 March 2017 - 05:26 AM

In lieu of the needed IR cut filter, I wanted to attach my skyglow filter, but I wasn't able to attach the 2" adapter directly to the camera.  Thus I needed to design and print a new adapter that included a filter mount that could hold a 2" filter.  It turned out pretty darn good.

 

Here's the filter rim and internal threads on the new adapter:

 

4067.jpg link31.png

 

Front and back views of the threaded ring to hold the filter. Note the holes to allow screwing the ring on and off.

 

4068.jpg

 

4069.jpg

 

A quickly printed installation tool with four pins to mate with the filter ring.

 

4070.jpg

 

All the pieces ready to install the filter in the holder:

 

4071.jpg

 

With the filter installed and the ring tightened, the adapter is ready to go.

 

4072.jpg

 

The whole assembly is then installed on the camera.

 

4073.jpg link31.png

 

Beo


  • mattyfatz and brave_ulysses like this

#10 Lord Beowulf

Lord Beowulf

    Surveyor 1

  • *****
  • topic starter
  • Posts: 1850
  • Joined: 13 Oct 2008
  • Loc: Cedar Park, TX

Posted 04 March 2017 - 05:40 AM

I wasn't able to get DSS to stack all of the frames for this image due to some wobbling of some of the stars, so I downloaded Regim to give it a try.  While being Java based, it's considerably slower than DSS, it did at least actually stack all of the images that DSS refused to stack.  Here's the result with an initial stretch.  The color balance actually looks pretty good.

 

4074.jpg link31.png

 

Beo


Edited by Lord Beowulf, 04 March 2017 - 05:46 PM.

  • mattyfatz, BenKolt and artem2 like this

#11 Lord Beowulf

Lord Beowulf

    Surveyor 1

  • *****
  • topic starter
  • Posts: 1850
  • Joined: 13 Oct 2008
  • Loc: Cedar Park, TX

Posted 05 March 2017 - 03:04 AM

After my success with the filter adapter, I decided to take it a step further.  Out of everything I've designed and printed so far, I think this has to be the most advanced and one of the ones I'm proudest of.  Given the lack of a T-ring adapter on the QHY247C, and that my Geoptik camera lens adapter tended to be too thick to reach proper focus anyway, I decided to create and print my own Nikon to QHY adapter, and while I was at it, I designed a Losmandy rail clamp right into the part and even printed the knobs to go with it. The only parts not printed were the two carriage bolts and the springs for the clamp. I even had room to get a full ORO logo into the print!  All parts are printed as single pieces with no additional support required.

 

Here are all the pieces.  The main adapter body and larger part of the rail are printed together. The clamp assembly includes two guide pins, and springs go around the 1/4-20 carriage bolts used to clamp it. The carriage bolts press fit into the knobs and the clamp ring screws on outside of the body to hold the lens tightly. Not shown is the internal ring for holding the filter, which is taken from the previous filter adapter.

 

4056.jpg link31.png

 

Here are various shots of the fully assembled adapter, including the filter.

 

4057.jpg link31.png

 

4058.jpg

 

 

In this shot you can see the lens bayonet matching the camera and the clamp ring. This is also the bottom face of those two prints, as evidenced by the bit of blue tape still there and the dull finish. The entire adapter assembly is printed from that base with no support.

 

4059.jpg link31.png

 

In this shot (although a bit out of focus) you can see the filter and the all important tab for holding open the aperture of the lens. The lens twists on until the aperture opens fully and then you clamp it in place with the outer screw ring.

 

4060.jpg

 

Note the "chatter" of the printer as it creates the logo, giving a ghosted image.  This also reflects the two different directions the two parts were printed.

 

4061.jpg link31.png

 

Continued next post...

 

Beo

 


  • mattyfatz and BenKolt like this

#12 Lord Beowulf

Lord Beowulf

    Surveyor 1

  • *****
  • topic starter
  • Posts: 1850
  • Joined: 13 Oct 2008
  • Loc: Cedar Park, TX

Posted 05 March 2017 - 03:13 AM

Here's the adapter assembled and installed.  It mounts perfectly to the top Losmandy rail of my 11" Edge HD and then the camera and lens (in this case my Sigma 70-300 mm zoom) attach cleanly. Can you spot my one OOPS! in the adapter design?

 

4062.jpg link31.png

 

4063.jpg link31.png

 

Here's a hint..  Look where the switches are.

 

4064.jpg link31.png

 

Did you see it?  When I added the base, I put it on the wrong side of the adapter, which for the most part is round, except for the bayonet.  Thus, the lens top faces "down" when installed. Actually it's better that way since when the scope is in use I can still see the focus marks. However, if I ever reprint it, I'll change it to put the top of the lens to the side.

 

And finally, a first light shot.  It was incredibly windy, but I had to get it focused and give it a try. Note the blue tape to hold the focus and zoom in place at 70 mm in this case.

 

4065.jpg link31.png

 

4066.jpg link31.png

 

Next up, an image...

 

Beo

 

 



#13 Lord Beowulf

Lord Beowulf

    Surveyor 1

  • *****
  • topic starter
  • Posts: 1850
  • Joined: 13 Oct 2008
  • Loc: Cedar Park, TX

Posted 05 March 2017 - 03:37 AM

I was able to get one frame before closing up the observatory to keep everything from blowing away.  This is with the scope centered on the Horsehead and the lens at 70 mm.  If you look at the full sized image you'll see the stars are all lines due to the wind vibration.  This was stacked in DSS so the colors were all wonky and I had to tweak that as well as manually apply some correction for vignetting.  I'm not sure WHAT the white is on the bottom of the frame.  If I had clouds coming in I didn't notice, but it could have been something in the observatory.

 

4076.jpg link31.png

 

Note I'm also trying to process these images from my laptop in China at the moment (just arrived back in Shanghai after Beijing, Shanghai, and Shenzhen last week) so don't expect things to look great!  Would have loved to visit QHYCCD while I was here.  I was under the impression they're in Shenzhen, but never was able to confirm.  Would have liked to visit DJI was well, but oh well...

 

Beo


Edited by Lord Beowulf, 05 March 2017 - 03:39 AM.

  • mattyfatz and BenKolt like this

#14 Lord Beowulf

Lord Beowulf

    Surveyor 1

  • *****
  • topic starter
  • Posts: 1850
  • Joined: 13 Oct 2008
  • Loc: Cedar Park, TX

Posted 14 March 2017 - 04:59 AM

So I mentioned the 2" nose piece adapter and the M54 mounting ring on the camera.  I've posted a few pictures of the situation on the version I received.   Here's the tilt adjustment ring and 2" adapter:

 

4077.jpg link31.png

 

4078.jpg

 

Installing the adapter on the ring, the outer flange hits the three face adjustment/mounting screws.  Note that this is not the same adapter ring as shown on the QHY347C that's designed to attach to the filter wheel either.

 

4079.jpg

 

4080.jpg

 

Removing the three mounting/tilt adjustment screws allows the 2" adapter to mount flush, but leaves open holes that can let light in. There's also no way to adjust the tilt of the camera, so what' s the point of the adjustment ring? Not that it SHOULD be needed anyway, but still...  Covering the holes with tape is not a viable long term solution.

 

4082.jpg

 

4083.jpg

 

Comparing the threads between the mounting ring and the camera itself, it's apparent that they're the same diameter and thread pitch, so the 2" adapter SHOULD fit directly on the camera.  It doesn't.

 

4085.jpg

 

The reason the 2" adapter won't fit the camera directly, despite the matching threads, is that there's a flange on the inside of the adapter that hits the four pan-head screws holding the window heater PCB.  It looks like it would fit if that extra flange/baffle wasn't there.  I don't really see any reason for it.

 

4088.jpg link31.png

 

Beo

 

 

 

 

 


  • BenKolt and entilza like this

#15 nocturnalguy

nocturnalguy

    Lift Off

  • -----
  • Posts: 5
  • Joined: 03 May 2017

Posted 09 May 2017 - 08:07 AM

Hello thanks for your review!

Could you please tell me the camera body diameter to see if it fits my Hyperstar?



#16 Phil Sherman

Phil Sherman

    Vanguard

  • -----
  • Posts: 2348
  • Joined: 07 Dec 2010
  • Loc: Cleveland, Ohio

Posted 10 May 2017 - 04:43 PM

I had a similar issue with my USB hub at the mount and adopted a different solution. I opened up the USB cable around 12" from the scope end and cut the (red) power wire. The hub is no longer powered from the computer and gets its power from the external power supply for it.



#17 Lord Beowulf

Lord Beowulf

    Surveyor 1

  • *****
  • topic starter
  • Posts: 1850
  • Joined: 13 Oct 2008
  • Loc: Cedar Park, TX

Posted 15 May 2017 - 07:34 AM

Hello thanks for your review!

Could you please tell me the camera body diameter to see if it fits my Hyperstar?

Hi, it's 90 mm according to QHYCCD (don't have it handy to re-measure myself).  You can see a mechanical drawing in the QHY367C manual as they still haven't updated the QHY247C manual to contain the drawing.  Note that they've also added a lot of new adapters for this camera.  If I hadn't already designed and 3D printed my own camera lens mount I'd be interested in the adapters they've made.

 

Beo



#18 Lord Beowulf

Lord Beowulf

    Surveyor 1

  • *****
  • topic starter
  • Posts: 1850
  • Joined: 13 Oct 2008
  • Loc: Cedar Park, TX

Posted 15 May 2017 - 07:41 AM

I had a similar issue with my USB hub at the mount and adopted a different solution. I opened up the USB cable around 12" from the scope end and cut the (red) power wire. The hub is no longer powered from the computer and gets its power from the external power supply for it.

Thanks Phil,

 

I had tried that previously with another cable for resetting my USB connection to the cell phone that provides my internet, but it apparently impacted the integrity of the communication on the cable, or otherwise that cable just wasn't that good to start with.  Given this one's using USB 3.0 that's even a bigger risk.  All of the USB 2.0 hubs in my pier are configured to run off of a separate 5V supply and are disconnected from the incoming 5V on the input connector.  If I decide to upgrade to USB 3.0 in the pier I'll likely do the same.  However, so far I haven't found a USB 3.0 hub that appears completely stable, much less one I could hack to fit in the pier.  And since the 5V comes up with the 12V supply that still might not fix the startup order problem.  Dr. Qiu says they've made a fix to the USB board that will prevent this problem in the future, but they haven't yet sent me my replacement so I can try it out.  I'd really like to do so!

 

Beo



#19 Lord Beowulf

Lord Beowulf

    Surveyor 1

  • *****
  • topic starter
  • Posts: 1850
  • Joined: 13 Oct 2008
  • Loc: Cedar Park, TX

Posted 25 June 2017 - 02:48 PM

So QHYCCD sent me the info on the mod they'd made to make the USB 3.0 interface start up with the camera rather than being out of sync with the camera due to running off the USB power supply.  The mod is a single jumper between two 0402 components, so it's definitely not for the faint of heart.  You need a good SMT rework station with microscope and pinpoint soldering iron.  The USB circuitry on the QHY247C gets its power from the USB port, while the rest of the camera operates off the 12V supply. However, if the USB is connected first, the camera interface doesn't start up correctly. QHYCCD came up with a solution that requires a jumper on the PCB, presumably to hold the USB chip in reset until the 12V supply is powered up. Rather than sending the camera back for modification, they asked me to make the modification myself. Luckily I have the microscope and SMT soldering iron needed to accomplish the task. I still need to test everything completely, but this folder shows the disassembly and re-assembly of the camera.

 

Four screws hold the back plate on the camera. It's interesting to note that the holes in the cover are beveled for flat or oval head countersunk screws.  A ground wire attaches from the USB board to the back panel with a screw that also holds a wire from the power jack. The power jack plugs into the board opposite the USB board.  The body housing is just a sleeve that slips off over the heatsink and PCBs. The power supply and USB boards each have coarsely 3D printed parts that attach them to long standoffs using shorter standoffs for the end cover. The boards just sit in a slot in the bottom 3D part but are screwed to the top one.

 

4228.jpg link31.png

 

With the two short standoffs removed, the USB board is held by two connector-less ribbon cables. I HATE dealing with these flexible printed circuit connectors. The one visible on this side with the white cable uses a connector style that's not too bad, but the back one is a pain to reconnect. Neither gives me total confidence they're seated fully when re-installing.

 

4229.jpg

 

On this top side of the board you can see the large Altera FPGA. To the left is the Cypress Semiconductor USB 3.0 chip, and above is an apparently obsolete Micron Gigabit SDRAM. To the right of the FPGA is a flip style printed ribbon connector.  This also has another view of the 3D printed  mount.

 

4231.jpg

 

On the back are all the connectors (mostly for programming the board or other applications) and various passives under a conformal coating. The other ribbon connector is visible as the black bar above the white housing in the upper left.

 

4232.jpg

 

The USB startup fix consists of adding a jumper between a resistor near the ribbon cable connector and a capacitor at the edge of the board. I'm assuming the capacitor is on the chip power supply coming from the 12V input and the resisotr is going to the reset pin of the USB chip. That would hold the USB chip in reset until the power supply comes up. Normally a mod like this would also include cutting a trace to break the original connection, but hopefully this will work!

 

These are 0402 parts so they had to be done under a microscope with my finest SMT tipped soldering iron and even then it was a challenge. I also had to carefully remove the conformal coating around the area.

 

4234.jpg

 

The solder joint at the resistor doesn't look very pretty, with some residual solder on the isulation and conformal coating. That connection was a real challenge getting the solder to melt at the board/component interface. That part must have a via to the ground plane very close to the connection. I didn't want to apply heat for two long as it would potentially loosen the whole component or damage a trace on the board.

 

4236.jpg

 

Re-assembly is pretty much the reverse process.  Getting the cables reconnected was tricky enough.

 

4240.jpg

 

Even re-connecting the ground wire was a bit of a challenge dealing with the tiny screw and washer that kept wanting to go anywhere but where I needed them.

 

4241.jpg

 

As usual there are more photos in the gallery.  Enjoy!

 

Beo

 


  • BenKolt and Gene3 like this

#20 Gene3

Gene3

    Mariner 2

  • -----
  • Posts: 222
  • Joined: 02 Feb 2016
  • Loc: Del Mar, CA

Posted 16 July 2017 - 07:49 AM

Wow you are one brave dude!!

 

I was considering the QHY247C but now i may wait until they fix this issue



#21 BenKolt

BenKolt

    Viking 1

  • -----
  • Posts: 656
  • Joined: 13 Mar 2013

Posted 16 July 2017 - 09:12 AM

Lord Beowulf:

 

A most impressive review and description of your hard work.  I am very impressed with your skills.  Someday I hope to have the ability to print some of my own parts.

 

Ben

 

P.S.  I tried to rename our dog Grendel after Beowulf's nemesis given the animal's disposition, but she wouldn't let me after I told her the reference.



#22 Lord Beowulf

Lord Beowulf

    Surveyor 1

  • *****
  • topic starter
  • Posts: 1850
  • Joined: 13 Oct 2008
  • Loc: Cedar Park, TX

Posted 21 July 2017 - 06:51 PM

Thanks for the comments.  We finally have had some clear skies with a favorable moon phase and fitting my travel schedule so that I could put things to the test.  After a quick response from Dr. Qiu to resolve a driver problem with the temperature control (something they'd found and fixed by the time I reported it) I've been able to run a couple of nights of imaging with no problems detected so far.  I need to process my images, but I think things are coming together!

 

BTW, I like the name for your dog.  Was that the dog or the wife that wouldn't let you use the name?  smile.gif  While I long ago gave up trying to explain the reference in most fora, I actually came by the handle as a second order attribution.  My favorite author since I was a young teen has been Larry Niven, who makes reference to the Beowulf/Grendel story (and various other classical literature) repeatedly throughout his works.  One of his main characters in most of his future history of the Known Universe is Beowulf Shaeffer, whose name I adapted to Beowulf Schaeffer when I first started using and then hosting my own BBS long before the World Wide Web and before all but a few college geeks had heard of the Internet or Usenet.  It gave me somewhat of a perverse pleasure to sign my posts "B.S." which was quite often a commentary either on the content of my own post, or my opinion of someone else's!  lol.gif  I finally had to give up the full name though when websites came along, not only because no one recognized the reference, but more often than not, because most websites couldn't handle that long of a user name!  At any rate, I've always been amused at the number of Beowulf (original) references littered throughout books, TV, and film (including the Star Trek Voyager episode).

 

Beo   AKA B.S.




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.







Cloudy Nights LLC
Cloudy Nights Sponsor: Astronomics