The data is there, but it obviously needs dozens of hours to squeeze out definition.
And this single 5 min shot shows so much potential of the Rho Ophiuci area:
(Note to Charles: That bright orange star is your "Orange Julius" star.)
Wednesday, February 26, 2014
Flummoxed by the Flux
Here's the original image:
Here's the heavily massaged data turned into a B&W:
I haven't processed any data recently in Deep Sky Stacker (DSS) because I've decided to collect larger amounts of data per night. I've subsequently, skipped taking dark frames which is beginning to catch up to me.
Here's a link to a web page explaining all the other data which must be acquired before you can stack an image:
http://starizona.com/acb/ccd/advimcal.aspx
I have to decide how I'm going to collect dark frames. The easiest way would be to setup the camera on a cloudy night that has temperatures roughly equal to the night of the imaging run. This can be tricky since cloudy nights tend to have warmer temperatures.
Back to the drawing board?
So, I took out the shorter FL rig with the 66mm f/6 scope and the 135mm lens in the hopes that I could have an easier time, a more positive experience. Nope.
I didn't charge the battery from the previous session to see if I could go two nights
of powering the mount on a full charge. I think the answer is a qualified "yes".
I had decided to shoot some part of the flux nebula (IFN) near M81/82. I figured that the proximity to Polaris and the short FL of the system would make it easy to shoot 15 min subexposures.
My plan was to shoot 21 subs at 15 min equaling a little over 5 hours.
After setting everything up, I did a test exposure and there was that awful, familiar vertical trailing. I changed the calibration step duration from 1000 to 400 in PHD2. I turned the autoguiding camera about 30 degrees to the left and I re-calibrated the autoguiding. I watched the PHD2 window and it seemed fine. It was guiding perfectly and there was no obvious error -- expect that every so often, the guider would send a 500+ms command to move the mount. I didn't think much of this at the time.
It didn't help that 2-10 mph winds were pushing the laptop tent around.
The next test 15 min exposure showed less trailing. I decided that the errors were perhaps caused by the wind and that things would calm down eventually. So, I set up Backyard EOS to shoot 21 subexposures at 15 min each. 800 ASA, F/3.2. I also changed the calibration step to 660.
Why change the calibration step again? Who knows.
Went in, took a nap for about an hour and a half, came out and saw trailing in all of the 14 shots that had been taken. >sigh< I did examine the shots and it looked promising as some tweaking with the rudimentary "curves" function in BackyardEOS revealed patches of faint red near M81/M82 and throughout the frame. But still, every frame showed trailing.
If you looked at the successive frames, it was obvious something was wrong as M81/82 starts out in the upper right and then 2 hours later, they are about 1/3 of the way down on the frame. In retrospect, it almost seems like there was no (or possibly very poor) guiding in declination.
I tried other settings in PHD2 and shot another 15 min exposure and I got the same trailing. It just didn't make sense as I wasn't getting any error messages from PHD2.
I decided I needed to pack up. But before I called it quits, I slewed the scope to the Rho Ophiuchi area near Scorpius. I changed the calibration step to 250 (again why?) and got the autoguiding to work fine. And I shot two perfectly fine 5 min shots. I should have tried a 15 min shot, but I was tired and decided to call it a night.
Basically, I had a problem/technical issue and I didn't know how to fix it. Perhaps, I should have used the 50mm finder/guidescope instead? I had been out the previous night and I managed to shoot a several dozen 5 min shots of the Leo Trio. But even on the previous night, there were problems maybe every 8th frame? Perhaps periodic error is to blame and I should run the Eqmod software to deal with it? An interesting note is that the Sirius/Heq5 Pro mount has about a 638 sec worm period. Does that mean that exposures longer than that will always have problems? If so, then periodic error correction should addresss the issue.
Hm...
[Later: I forgot to mention one crucial detail from this session: I didn't polar align. In retrospect, that was probably the biggest mistake and it offers a plausible explanation for my problems. I assumed that since the mount was weighed down by the 50 lb battery that polar alignment would be good enough. But every time I setup for an imaging run -- even though I keep the mount in the same place -- I move the battery around a bit. And the act of removing/adding the OTA/telescope surely adds/releases some stresses in the aluminum structure. I wouldn't be surprised if the mount was off by 2mm in some random direction. The fact that M81/82 seemed to drift over 2-3 hours of subexposures speaks to polar alignment issues. Still I'm faced with the fact that PHD2 didn't fix the drift...which in theory it's supposed to do. I should have shot a 15min exposure of the Rho Ophiuchi area. The bottom line here is that by not polar aligning at the beginning, I added another layer of unknowns to the guiding problems. Lesson: Don't take shortcuts.]
So, I took out the shorter FL rig with the 66mm f/6 scope and the 135mm lens in the hopes that I could have an easier time, a more positive experience. Nope.
I didn't charge the battery from the previous session to see if I could go two nights
of powering the mount on a full charge. I think the answer is a qualified "yes".
I had decided to shoot some part of the flux nebula (IFN) near M81/82. I figured that the proximity to Polaris and the short FL of the system would make it easy to shoot 15 min subexposures.
My plan was to shoot 21 subs at 15 min equaling a little over 5 hours.
After setting everything up, I did a test exposure and there was that awful, familiar vertical trailing. I changed the calibration step duration from 1000 to 400 in PHD2. I turned the autoguiding camera about 30 degrees to the left and I re-calibrated the autoguiding. I watched the PHD2 window and it seemed fine. It was guiding perfectly and there was no obvious error -- expect that every so often, the guider would send a 500+ms command to move the mount. I didn't think much of this at the time.
It didn't help that 2-10 mph winds were pushing the laptop tent around.
The next test 15 min exposure showed less trailing. I decided that the errors were perhaps caused by the wind and that things would calm down eventually. So, I set up Backyard EOS to shoot 21 subexposures at 15 min each. 800 ASA, F/3.2. I also changed the calibration step to 660.
Why change the calibration step again? Who knows.
Went in, took a nap for about an hour and a half, came out and saw trailing in all of the 14 shots that had been taken. >sigh< I did examine the shots and it looked promising as some tweaking with the rudimentary "curves" function in BackyardEOS revealed patches of faint red near M81/M82 and throughout the frame. But still, every frame showed trailing.
If you looked at the successive frames, it was obvious something was wrong as M81/82 starts out in the upper right and then 2 hours later, they are about 1/3 of the way down on the frame. In retrospect, it almost seems like there was no (or possibly very poor) guiding in declination.
I tried other settings in PHD2 and shot another 15 min exposure and I got the same trailing. It just didn't make sense as I wasn't getting any error messages from PHD2.
I decided I needed to pack up. But before I called it quits, I slewed the scope to the Rho Ophiuchi area near Scorpius. I changed the calibration step to 250 (again why?) and got the autoguiding to work fine. And I shot two perfectly fine 5 min shots. I should have tried a 15 min shot, but I was tired and decided to call it a night.
Basically, I had a problem/technical issue and I didn't know how to fix it. Perhaps, I should have used the 50mm finder/guidescope instead? I had been out the previous night and I managed to shoot a several dozen 5 min shots of the Leo Trio. But even on the previous night, there were problems maybe every 8th frame? Perhaps periodic error is to blame and I should run the Eqmod software to deal with it? An interesting note is that the Sirius/Heq5 Pro mount has about a 638 sec worm period. Does that mean that exposures longer than that will always have problems? If so, then periodic error correction should addresss the issue.
Hm...
[Later: I forgot to mention one crucial detail from this session: I didn't polar align. In retrospect, that was probably the biggest mistake and it offers a plausible explanation for my problems. I assumed that since the mount was weighed down by the 50 lb battery that polar alignment would be good enough. But every time I setup for an imaging run -- even though I keep the mount in the same place -- I move the battery around a bit. And the act of removing/adding the OTA/telescope surely adds/releases some stresses in the aluminum structure. I wouldn't be surprised if the mount was off by 2mm in some random direction. The fact that M81/82 seemed to drift over 2-3 hours of subexposures speaks to polar alignment issues. Still I'm faced with the fact that PHD2 didn't fix the drift...which in theory it's supposed to do. I should have shot a 15min exposure of the Rho Ophiuchi area. The bottom line here is that by not polar aligning at the beginning, I added another layer of unknowns to the guiding problems. Lesson: Don't take shortcuts.]
Monday, February 24, 2014
So… I was plagued all evening by weird guiding problems. Though the skies were very good for imaging, I was getting trailing in anything over 4 minutes on the Tak. The trailing was nearly vertical like it was moving in a line from 1 PM to 7 PM. What to do? I moved the weights so to be more east heavy. I checked the cables. I moved the table in case the cables had undo amounts of tension to the computer. I turned off the mount, turned it back on. Re-aligned.
At one point in the evening, I moved the scope in the scope rings because the finderscope was off center. This had no effect.
I decided to shoot the Leo Trio with 4 min sub exposures at 3200 ISO. Not ideal, but I felt I needed to hedge my bet. All my 10 min exposures showed lots of trailing.
I tightened all the screws on the tube holder, the finder mount, the RA and Dec locks.
The problem didn't go away. I kept wondering what had changed. This evening was the first time I used the nFocus controller and it worked well. I wondered if it somehow had caused a change to the driver. It didn't seem likely, but I had to note it since it was the only substantial change to the system.
Earlier in the evening, I did another polar alignment with the "DSLR method." I got the scope as well aligned (if not better) than the previous night. So, I was really wondering.
I just tried turning the auto guider camera, the QHY5L-II about 90 deg and it seems I'm getting good stars…. As I write this, I have a 10 min guided test going on…..
Just got back in. A perfectly fine 10 min exposure of M83.
So, if nothing else, I figured out how to make the Rigel Sys nFocus controller work with BackyardEOS. AND I learned that the QHY5L Autoguider camera can be quite finicky about its orientation in the finderscope. I have a mark that says, "UP" on the camera. I need to remember to keep the UP toward the sides instead.
Next task: Get Eqmod working so I can control the mount with the laptop.
At one point in the evening, I moved the scope in the scope rings because the finderscope was off center. This had no effect.
I decided to shoot the Leo Trio with 4 min sub exposures at 3200 ISO. Not ideal, but I felt I needed to hedge my bet. All my 10 min exposures showed lots of trailing.
I tightened all the screws on the tube holder, the finder mount, the RA and Dec locks.
The problem didn't go away. I kept wondering what had changed. This evening was the first time I used the nFocus controller and it worked well. I wondered if it somehow had caused a change to the driver. It didn't seem likely, but I had to note it since it was the only substantial change to the system.
Earlier in the evening, I did another polar alignment with the "DSLR method." I got the scope as well aligned (if not better) than the previous night. So, I was really wondering.
I just tried turning the auto guider camera, the QHY5L-II about 90 deg and it seems I'm getting good stars…. As I write this, I have a 10 min guided test going on…..
Just got back in. A perfectly fine 10 min exposure of M83.
So, if nothing else, I figured out how to make the Rigel Sys nFocus controller work with BackyardEOS. AND I learned that the QHY5L Autoguider camera can be quite finicky about its orientation in the finderscope. I have a mark that says, "UP" on the camera. I need to remember to keep the UP toward the sides instead.
Next task: Get Eqmod working so I can control the mount with the laptop.
Another single frame from a couple nights ago of the Seagull Nebula (IC 2177) in Monoceros:
It was a 3 min exposure at 1600 ISO. Brought in to PS with levels, curves, color selection with a mask to enhance reds. Then I cheated and used auto color to finish it off. Some noise reduction thrown in as well.
And the usual crop:
Shot with the 135mm lens at F/3.2 through the Canon T3i.
Sunday, February 23, 2014
Planning seems important in astro-imaging. Last night I set up my gear with the Tak on the Sirius mount. I got out earlier than usual to balance the RA and DEC with the scope slightly front heavy. (And the RA slightly weight (bottom) heavy.)
I let everything cool down for two hours, then setup the computer and connected all the cables.
I haven't even gotten the Eqmod and focuser cables figured out and it already seems like there are too many cables wrapped around the scope. Part of planning is to move the scope toward the part of the sky that will be targeted for the night. It's important to do this so to prevent/anticipate any potential cable snags while imaging for 2-3+ hours.
Originally, I was planning on shooting the Rosette Nebula (N2237, etc). But by 9 PM, the Rosette was already at the Meridian. Only an hour after meridian, it began to impinge upon Flagstaff's light pollution dome. Though the light pollution wasn't horrible, it does create gradients in the image that causes problems when post-processing.
There are lots of great nebular targets that lie generally in the plane of the Milky Way. The problem is that during winter, those targets tend to be located in the West. So, I don't have a lot of time to image those targets because they are already setting by the time I get started. There's always this frantic race to get everything setup and ready to go by evening twilight. Of course, this would be partly alleviated by an observatory since setup time would be trivially short. But I digress.
The answer lies obviously in proper planning. The better time to image these nebular targets would be summer into late fall.
So, based on these reflections and the fact that Leo and Virgo are better placed, I had planned to shoot the Leo Trio (M65, M66, and N3628). I decided to get better polar alignment using the "DLSR method" with 190 sec exposures (10 sec stationary, 90 sec with the left arrow on the mount keypad and 90 sec with the right arrow on the mount keypad). I always underestimate how long this will take. I always think it shouldn't take more than 45 minutes. But in actuality, it takes like 1.5 hours if I'm going fast. If I really want to nail polar alignment, I need to plan for 2.5-3 hours of work. (This obviously affects your imaging plan for the evening.)
[Added later: I had used the battery while doing the DLSR polar alignment. And I noticed when I connected to Backyard EOS that my camera temps were in the 30's (Celsius)). Yikes. I switched to the AC plug in and the temps came down into the teens within a few minutes. That battery and screen really push the temps.]
It wasn't until 10:40 that I was ready to image. I slewed to M65 and took a test shot. It was really NOISY. I looked up and immediately saw the cause. High clouds had rolled in and blanketed the sky under a haze. Orion looked fuzzy. I waited an hour and checked the IR weather images. It looked like a small bank of clouds had floated from the West into central AZ. But the northern part of the state remained mostly clear. And that part of the sky looked promising.
So, I went back out and slewed the scope onto M101 in Ursa Major. Took a short test exposure. Then I decided to do an experiment: I shot M101 for 20 minutes at 400 ISO. I had switched the autoguider to take 3 sec exposures (due to M65 being harder to image) and hadn't changed the QHY back to 2 secs (my normal time). The stars looked slightly oval with a small tip on one side. Overall, it looked okay. I anticipate taking longer exposures (30+ minute in H-a) so I wanted to see how 20 would turn out. There were clearly two potential issues/fixes: 1) Taking the QHY camera back to 2 sec exposures; and 2) Rebalancing the scope/weights. The rig was actually west heavy. (It should be East heavy.)
Encouraged by the 20 min shot, I set up Backyard EOS to do a 9 shot series of 10 min. exposures. I picked 9 because the moon was going to be up around 1:30 AM and it was already a few minutes past midnight.
I fell asleep on the couch and didn't get back out until 2:30. It was already done and I glanced at the last two subframes which looked fantastic on the screen. I kept thinking that if I could do a 4 hour run, I would have some real data.
The other thing here to note is that proper planning gives you leeway in case you can't image on a particular night. If your target is somewhere in the East when you begin, then you can still go for it a day or week or month later. But if your target is already in the West, you may not be able to image it in a month.
Bah. Maybe this is all too obvious.
I let everything cool down for two hours, then setup the computer and connected all the cables.
I haven't even gotten the Eqmod and focuser cables figured out and it already seems like there are too many cables wrapped around the scope. Part of planning is to move the scope toward the part of the sky that will be targeted for the night. It's important to do this so to prevent/anticipate any potential cable snags while imaging for 2-3+ hours.
Originally, I was planning on shooting the Rosette Nebula (N2237, etc). But by 9 PM, the Rosette was already at the Meridian. Only an hour after meridian, it began to impinge upon Flagstaff's light pollution dome. Though the light pollution wasn't horrible, it does create gradients in the image that causes problems when post-processing.
There are lots of great nebular targets that lie generally in the plane of the Milky Way. The problem is that during winter, those targets tend to be located in the West. So, I don't have a lot of time to image those targets because they are already setting by the time I get started. There's always this frantic race to get everything setup and ready to go by evening twilight. Of course, this would be partly alleviated by an observatory since setup time would be trivially short. But I digress.
The answer lies obviously in proper planning. The better time to image these nebular targets would be summer into late fall.
So, based on these reflections and the fact that Leo and Virgo are better placed, I had planned to shoot the Leo Trio (M65, M66, and N3628). I decided to get better polar alignment using the "DLSR method" with 190 sec exposures (10 sec stationary, 90 sec with the left arrow on the mount keypad and 90 sec with the right arrow on the mount keypad). I always underestimate how long this will take. I always think it shouldn't take more than 45 minutes. But in actuality, it takes like 1.5 hours if I'm going fast. If I really want to nail polar alignment, I need to plan for 2.5-3 hours of work. (This obviously affects your imaging plan for the evening.)
[Added later: I had used the battery while doing the DLSR polar alignment. And I noticed when I connected to Backyard EOS that my camera temps were in the 30's (Celsius)). Yikes. I switched to the AC plug in and the temps came down into the teens within a few minutes. That battery and screen really push the temps.]
It wasn't until 10:40 that I was ready to image. I slewed to M65 and took a test shot. It was really NOISY. I looked up and immediately saw the cause. High clouds had rolled in and blanketed the sky under a haze. Orion looked fuzzy. I waited an hour and checked the IR weather images. It looked like a small bank of clouds had floated from the West into central AZ. But the northern part of the state remained mostly clear. And that part of the sky looked promising.
So, I went back out and slewed the scope onto M101 in Ursa Major. Took a short test exposure. Then I decided to do an experiment: I shot M101 for 20 minutes at 400 ISO. I had switched the autoguider to take 3 sec exposures (due to M65 being harder to image) and hadn't changed the QHY back to 2 secs (my normal time). The stars looked slightly oval with a small tip on one side. Overall, it looked okay. I anticipate taking longer exposures (30+ minute in H-a) so I wanted to see how 20 would turn out. There were clearly two potential issues/fixes: 1) Taking the QHY camera back to 2 sec exposures; and 2) Rebalancing the scope/weights. The rig was actually west heavy. (It should be East heavy.)
Encouraged by the 20 min shot, I set up Backyard EOS to do a 9 shot series of 10 min. exposures. I picked 9 because the moon was going to be up around 1:30 AM and it was already a few minutes past midnight.
I fell asleep on the couch and didn't get back out until 2:30. It was already done and I glanced at the last two subframes which looked fantastic on the screen. I kept thinking that if I could do a 4 hour run, I would have some real data.
The other thing here to note is that proper planning gives you leeway in case you can't image on a particular night. If your target is somewhere in the East when you begin, then you can still go for it a day or week or month later. But if your target is already in the West, you may not be able to image it in a month.
Bah. Maybe this is all too obvious.
Tuesday, February 18, 2014
Modifying My Canon T3i... Some Notes
Modifying my Canon T3i was nerve-wracking. I put it off for about 6 months as I knew that the mod could destroy the camera. When I finally decided to go for it, I consulted and studied Gary Honis' website where he gives very good step-by-step instructions on the whole process.
http://dslrmodifications.com/rebelmod450d1.html
Why the mod? There are dozens of nebulous targets for the small to medium sized telescope that emit light in the H-alpha frequency. H-alpha lies at 656.28 nm and it's basically a "deeper" red color that is mostly blocked by filters that sit in front of the camera sensor. It can be picked up by a regular digital DSLR, but it is only let through at 20-30%. By replacing the stock filter with one that lets through 90-95% of H-alpha means that you'll be able to detect and image a lot of nebula more efficiently with less exposure time.
In terms of replacement filters, there are two major "brands" and two kinds of replacement options.
As far as I know only Baader and Astrodon produce a UV/IR filter that can simply be swapped for the stock filter. These special filters are similar to the stock filter except that they let more of the deeper red/near infrared onto the sensor. But they still block the infrared and UV. The other type of filter is a clear glass filter that doesn't really block anything. The reason why an aftermarket UV/IR filter is desirable is that stars will begin to bloat in optical systems that involve refractive elements. Refractor telescopes and camera lenses are always compromises in terms of performance and price. Lenses that focus UV light to near IR light cost a lot of money because they require more expensive and exotic glass elements (and designs) to perform well.
I sent my first DSLR away a few years ago to Hap Griffin to have him do a full-spectrum mod which basically meant he removed the stock filter and left it as is. It made the camera extremely sensitive to H-a. But it wouldn't focus with my camera lenses (with some exceptions) and the color balance was really screwy. Another problem was that the sensor became a literal dust magnet. Before each session, I had to use a blower which seemed only to move the dust.
Thus, in an effort to gain the advantage of better H-alpha sensitivity and adding some dust protection, I went with the Astrodon UV/IR replacement filter. I figured I'd save $200 bucks by doing the mod myself. In actuality, since I had to buy $60-70 worth of special tools, it wasn't much of a savings at all.
(DISCLAIMER: I AM NOT CONDONING THIS MOD. YOU WILL LOSE YOUR CAMERA WARRANTY IF YOU DO IT.)
I didn't document the process comprehensively, but I did snap pictures along the way:
Gary Honis' instructions include making a makeshift tool from a tweezer and rubber band segments:
Here I'm trying to use the tool to pull some delicate cables from their connectors:
All the major cables are disconnected:
Here is the sensor assembly:
I've removed the stock filter assembly from the sensor assembly:
The mod requires that you remove the stock UV/IR filter front the plastic support assembly. It is glued in with something that resembles silicone. Though I used an exacto knife to remove a bunch of the glue, it wouldn't budge. I applied some pressure to separate it and it cracked in half! BUT the most alarming part was that I SAW AND FELT glass particles (and a fine mist) shoot toward my face. I felt a piece hit my forehead. I immediately dipped my head down, then slowly walked to the shower and took a very awkward shower to remove any tiny glass particles from my face, hair and eye area.
The Astrodon replacement filter glued in place. I don't show it, but I used a 8x magnifying glass to inspect the replacement filter. There were lots of pieces of dust that took a while to remove:
I waited two days for the silicone to cure before I reattached the filter assembly to the sensor assembly:
Everything going back to together. There is a very congested area in the upper right of the sensor board here where 3 cables sit on top of each other. After screwing this board back in place, I found I couldn't completely reseat two cables into their connectors. So, I had to disconnect all the cables again and unscrew the board and THEN connect the TRICKY stacked cables FIRST then screw down the board.
Here is that tricky area:
http://dslrmodifications.com/rebelmod450d1.html
Why the mod? There are dozens of nebulous targets for the small to medium sized telescope that emit light in the H-alpha frequency. H-alpha lies at 656.28 nm and it's basically a "deeper" red color that is mostly blocked by filters that sit in front of the camera sensor. It can be picked up by a regular digital DSLR, but it is only let through at 20-30%. By replacing the stock filter with one that lets through 90-95% of H-alpha means that you'll be able to detect and image a lot of nebula more efficiently with less exposure time.
In terms of replacement filters, there are two major "brands" and two kinds of replacement options.
As far as I know only Baader and Astrodon produce a UV/IR filter that can simply be swapped for the stock filter. These special filters are similar to the stock filter except that they let more of the deeper red/near infrared onto the sensor. But they still block the infrared and UV. The other type of filter is a clear glass filter that doesn't really block anything. The reason why an aftermarket UV/IR filter is desirable is that stars will begin to bloat in optical systems that involve refractive elements. Refractor telescopes and camera lenses are always compromises in terms of performance and price. Lenses that focus UV light to near IR light cost a lot of money because they require more expensive and exotic glass elements (and designs) to perform well.
I sent my first DSLR away a few years ago to Hap Griffin to have him do a full-spectrum mod which basically meant he removed the stock filter and left it as is. It made the camera extremely sensitive to H-a. But it wouldn't focus with my camera lenses (with some exceptions) and the color balance was really screwy. Another problem was that the sensor became a literal dust magnet. Before each session, I had to use a blower which seemed only to move the dust.
Thus, in an effort to gain the advantage of better H-alpha sensitivity and adding some dust protection, I went with the Astrodon UV/IR replacement filter. I figured I'd save $200 bucks by doing the mod myself. In actuality, since I had to buy $60-70 worth of special tools, it wasn't much of a savings at all.
(DISCLAIMER: I AM NOT CONDONING THIS MOD. YOU WILL LOSE YOUR CAMERA WARRANTY IF YOU DO IT.)
I didn't document the process comprehensively, but I did snap pictures along the way:
Trying to keep track of all the little parts as I disassembled:
Uh oh, what have I done?
Gary Honis' instructions include making a makeshift tool from a tweezer and rubber band segments:
All the major cables are disconnected:
Here is the sensor assembly:
I've removed the stock filter assembly from the sensor assembly:
The mod requires that you remove the stock UV/IR filter front the plastic support assembly. It is glued in with something that resembles silicone. Though I used an exacto knife to remove a bunch of the glue, it wouldn't budge. I applied some pressure to separate it and it cracked in half! BUT the most alarming part was that I SAW AND FELT glass particles (and a fine mist) shoot toward my face. I felt a piece hit my forehead. I immediately dipped my head down, then slowly walked to the shower and took a very awkward shower to remove any tiny glass particles from my face, hair and eye area.
Here is the assembly with a vinyl-type of gasket that holds the filter in place. The gasket was so stretched that I could't re-use it. But everything worked fine without the gasket.
The Astrodon replacement filter glued in place. I don't show it, but I used a 8x magnifying glass to inspect the replacement filter. There were lots of pieces of dust that took a while to remove:
I waited two days for the silicone to cure before I reattached the filter assembly to the sensor assembly:
[Addendum to above image: On the left side of the image, you can see detailed notes I took along with a drawing of a torx screw. Perhaps the single most crucial aspect of the mod involves carefully noting the exact position of 3 screws that determine the "orthogonality" of the sensor to the expected plane of the image. If one of these screws is out of whack, then the sensor plane will not be flat to the body.]
Everything going back to together. There is a very congested area in the upper right of the sensor board here where 3 cables sit on top of each other. After screwing this board back in place, I found I couldn't completely reseat two cables into their connectors. So, I had to disconnect all the cables again and unscrew the board and THEN connect the TRICKY stacked cables FIRST then screw down the board.
Here is that tricky area:
Since I was so freaked out and annoyed, I neglected to document the last re-assembly steps. Let's just say that I was extremely pleased and relieved when the camera started up again.
Friday, February 14, 2014
Working the bugs out on the Sirius with the Tak102 on top.
I just came in from a night of testing and practicing with the Sirius mount carrying the Tak 102. Attached were the Canon T3i and the QHY5L-II autoguider on a Stellarvue 50mm guide/finderscope.
The whole mess was balanced by two 11 lb weights. One weight was approx. 1/3 of the way from the top and the other weight was about 2/3 of the way down.
The Stellarvue 50 has a helical focuser and the QHY5L focuses fine without the need for an extender or a diagonal. The QHY5L does stick out a lot and the tension screw bites the end of the camera.
One drawback of the helical focuser is that there's no way to lock it down. So, inevitably when slewing from one target to another the tension on the USB or autoguider cable causes focus issues.
I was surprised by how much the focus goes out of whack between slews. I taped the helical focuser to stay in one position, but I found it still changed focus when slewing to a different part of the sky.
The big issue for the evening was PHD2. After polar alignment with the polar scope, I slewed
to Orion and started PHD2. Everything worked well and I made my way up to 10 minutes autoguided with round stars. The temperature from Backyard EOS reported pretty high
temps (like 25-35 C) from the camera. I don't know what was up.
The stars were a little bigger than I'd hoped. But I found this is just the nature of the TV .8 flattener and reducer.
After Orion, I slewed to M82 and decided to autoguide again. No love. I turned the autoguider off,
reset everything, turned it back on and tried to calibrate PHD. But I kept getting an error:
Star did not move enough.
It was really odd. I searched online and tried changing settings in PHD2. I increased calibration step duration from the default of 750ms to 2000. It didn't do much if anything. I kept getting the error. I tried pushing the arrow buttons on the handset controller for the mount. Nothing. After a
frustrating hour of re-attaching cables, I tried jiggling the USB cable connecting the camera to the latptop, and lo and behold it started to calibrate "West" like it should have. I also moved the focuser controller from the area of the USB. There may be a chance that "noise" is interfering with the cable that connects the laptop to the autoguider.
I noticed that you can hear the autoguide corrections as a periodic high pitched hum corresponding to some directional info on the bottom the PHD2 window.
But even with a successful calibration, the mount wouldn't guide for more than 60 seconds before there would be a blinking red screen and PHD2 acted like it couldn't guide correctly. Frustrating.
I was getting trailed stars in 60 second exposures.
I tried polar aligning again using the DSLR technique. It looked like I already had pretty good polar alignment. I rebalanced the scope/camera gear and tried PHD2 again. This time it worked.
I set the calibration duration to 1200ms. I got round stars in a 12 min. test exposure.
One obvious thing I learned was that PHD2 calibration should be done locally. That is, it should be done AFTER slewing to your target. If you rely upon the PHD2 calibration from a previous object, it will try to make massive corrections and eventually, it will fail the autoguide.
[Update: The mount made unusual cyclical motor noises when the scope was WEST heavy. This may have contributed to the calibration issues. The fix was straight forward -- moving the weights to balance everything or making the setup slightly EAST heavy.]
The whole mess was balanced by two 11 lb weights. One weight was approx. 1/3 of the way from the top and the other weight was about 2/3 of the way down.
The Stellarvue 50 has a helical focuser and the QHY5L focuses fine without the need for an extender or a diagonal. The QHY5L does stick out a lot and the tension screw bites the end of the camera.
One drawback of the helical focuser is that there's no way to lock it down. So, inevitably when slewing from one target to another the tension on the USB or autoguider cable causes focus issues.
I was surprised by how much the focus goes out of whack between slews. I taped the helical focuser to stay in one position, but I found it still changed focus when slewing to a different part of the sky.
The big issue for the evening was PHD2. After polar alignment with the polar scope, I slewed
to Orion and started PHD2. Everything worked well and I made my way up to 10 minutes autoguided with round stars. The temperature from Backyard EOS reported pretty high
temps (like 25-35 C) from the camera. I don't know what was up.
The stars were a little bigger than I'd hoped. But I found this is just the nature of the TV .8 flattener and reducer.
After Orion, I slewed to M82 and decided to autoguide again. No love. I turned the autoguider off,
reset everything, turned it back on and tried to calibrate PHD. But I kept getting an error:
Star did not move enough.
It was really odd. I searched online and tried changing settings in PHD2. I increased calibration step duration from the default of 750ms to 2000. It didn't do much if anything. I kept getting the error. I tried pushing the arrow buttons on the handset controller for the mount. Nothing. After a
frustrating hour of re-attaching cables, I tried jiggling the USB cable connecting the camera to the latptop, and lo and behold it started to calibrate "West" like it should have. I also moved the focuser controller from the area of the USB. There may be a chance that "noise" is interfering with the cable that connects the laptop to the autoguider.
I noticed that you can hear the autoguide corrections as a periodic high pitched hum corresponding to some directional info on the bottom the PHD2 window.
But even with a successful calibration, the mount wouldn't guide for more than 60 seconds before there would be a blinking red screen and PHD2 acted like it couldn't guide correctly. Frustrating.
I was getting trailed stars in 60 second exposures.
I tried polar aligning again using the DSLR technique. It looked like I already had pretty good polar alignment. I rebalanced the scope/camera gear and tried PHD2 again. This time it worked.
I set the calibration duration to 1200ms. I got round stars in a 12 min. test exposure.
One obvious thing I learned was that PHD2 calibration should be done locally. That is, it should be done AFTER slewing to your target. If you rely upon the PHD2 calibration from a previous object, it will try to make massive corrections and eventually, it will fail the autoguide.
[Update: The mount made unusual cyclical motor noises when the scope was WEST heavy. This may have contributed to the calibration issues. The fix was straight forward -- moving the weights to balance everything or making the setup slightly EAST heavy.]
Here's a shot of the setup from last night. Note the fancy tripod spreader and battery caddy that's holding the 50 lb battery. It really helps keep the mount/telescope stable. Before the battery caddy,
I could tip the tripod off one of the legs with my finger (guessing 10-15 lbs of force). Now, I have
to push considerably harder (maybe 30-40 lbs of force?) to tip up a leg. The battery itself barely
clears the bottom of the stock spreader bolt (which attaches the mount to the tripod).
I wish the TPI spreader was designed to sit about 6-8 inches lower. But I'm sure that would
double or triple the price.
Here's where I ordered it: http://tpiastro.com/index.htm
Here's a 10 minute shot at 100 ISO. Shot with modded Canon T3i. Autoguided with Stellarvue 50mm finder with a QHY5-II. No processing except bringing it into PS, then exporting as JPG.
Lower right corner. Definite elongation. It seems slightly more pronounced in this corner than in the three other corners. I'm a little disappointed, but I can live with it.
This is a 12 minute autoguided shot of the M82 area. I brought it into PS and did an auto levels to see
what sort of vignetting I was getting. Obviously, the reducer/corrector is not centered. But flats will take care of most of this issue. Also, focus is off as I had not refocused since taking the Orion shots which was 4-5 hours earlier.
Closeup of M82 with a red line indicating the supernova.
Subscribe to:
Posts (Atom)