Thursday, November 07, 2013

Astro-Tech 8" Imaging Newtonian, Part V

I am still seeing comatic stars in the AT8IN in spite of my best efforts.  Before adjusting the laser collimator, I have decided to stiffen up the Newtonian tube some more, using a much larger aluminum plate.  Actually I wanted to use a steel plate that went a full 360 degrees on the inside of the AT8IN tube (it would also match the coefficient of expansion of the existing rolled steel tube) but I don't have such steel and I don't have the roller.  But I did have a good supply of aluminum sheet.

So I decided to use an entire pre-cut sheet. Here's the before (small reinforcing plate):



And after (much larger plate, bolted with the four M4 focuser base bolts, four additional M3 bolts, and two M4 bolts of the finder base).



The previous reinforcing plate was rather small, and held in place by the four M4 bolts of the focuser base itself. I had made it small to avoid interference with the spider vane bolts and the two bolts that hold the finder bracket.  The new plate is much larger - I used an entire pre-cut sheet, didn't bother cutting it.  Due to its size, it interferes with one of the spider vane bolts and with the finder bracket bolts, so I added holes for those.

I removed the spider, secondary, and holder, then hand-bent the aluminum sheet and fit it inside the tube. The rear end of the sheet is right up against the first baffle.  I then marked the four holes for the focuser base, and the circular hole for the focuser itself, then removed the sheet and drilled the four M4 holes.  I then put the sheet back in and bolted it in place.


With the aluminum sheet held in place by the four bolts of the focuser base, I drilled additional holes for the finder bracket.  I also added four holes at the corners of the aluminum sheet, by marking the exterior of the Newtonian tube and drilling through the tube and through the sheet behind it.  This is the only destructive and non-reversible part of this modification.


Because I was too lazy to remove the aluminum sheet after it had been bolted down (and I didn't want to use the "hundreds of holes" technique to cut the large hole for the focuser drawtube) I simply took a 76mm bimetal hole saw to the aluminum sheet.  The hole saw cut through the aluminum in less than a minute, and also gouged the internal diameter of the focuser base. Ugly, but not visible once the focuser was mounted.  I painted the gouge with some automotive touch-up paint even so.


So now the focuser base should be much more secure, there should not be any shift in collimation when pointing to different areas of the sky.  Of course, the Taiwanese focuser itself may be the next suspect.  I am somewhat peeved because between the cost of the AT8IN, the Keller reducer, and a Feathertouch (if I add one), this scope would cost the same as a used Takahashi Epsilon 160 being sold on Astromart right now (7 November).  Of course this one would be faster (f/3 versus f/3.3) and have a larger aperture, but it's not as solid and the optics are almost certainly inferior. And it's not a Takahashi.


Wednesday, October 09, 2013

Astro-Tech 8" Imaging Newtonian, Part IV

Found out that collimation is still shifting even with the relocated locking knobs, so I decided to do the next step and replace the collimation springs with stiffer ones.

Looked through RS Online and found this "Spring, compression, music wire, 11.25 dia x 29.5mm" (part number 121-157). These springs are about the same diameter as the stock Astro-Tech springs, and are 29.5mm long. Spring rate is 4.51 N/mm and maximum force (when compressed to 10.8mm) is 85.42 N which is quite substantial.


These springs cost S$ 0.60 each (minimum purchase is a pack of 10) and have nice flats ground into both ends. They were pretty hard to place into the mirror cell, but I resisted the urge to cut them shorter with a cut-off wheel and eventually managed to get them in place.


Side by side with the stock collimation springs, it's rather obvious that the stock springs are woefully inadequate.


Now when in place and the primary mirror collimated, the springs are compressed to a length of about 12mm. This translates to a holding force of (29.5mm - 12mm) x 4.51N or about 79N per spring.  That is very substantial, the equivalent of about 24kg across the three springs.  It is still possible to shift the mirror when pushing it by hand, but I am hoping that with these new springs there will no longer be any collimation shift with position.

Monday, September 30, 2013

Astro-Tech 8" Imaging Newtonian, Part III

After last night's funky stars, I decided to make some modifications to the Astro-Tech 8" Imaging Newtonian in an attempt to improve the optical performance.

From this very good blog post I was able to get some ideas on how bad the focuser sag is on this telescope (check out the video in the link).  I had a look at my focuser and it's also the same CPC MR7W linear bearing, a not-bad focuser (I contrast-enhanced the area of the linear bearing where the part number is printed). Maybe I can hold off on buying a Moonlite, as my gadget budget has been reduced to zero.



First item of business was to remove the focuser.  To do this, you must first remove the two grub screws that hold the focuser body to the curved plate (one of the grub screws is visible in the above picture).  Then remove the four screws that hold the curved plate to the telescope tube.  These are M4 bolts with a nut on the other side; ensure that the nuts don't fall into the innards of the telescope.


Find a sheet of cardboard (I used some spam postcard from the local Omega dealership, which was extolling the America's Cup Omega Team New Zealand). Put the cardboard inside the tube and up against the focuser hole and bolt holes; and trace the outline onto the cardboard.


Transfer the pattern to an appropriate sheet of aluminum.. Make sure that the aluminum sheet is small enough so that it clears the bolts for the finder bracket, and the baffles behind the focuser. Approximately a 5" x 5" aluminum sheet is the correct size.


Then cut out the focuser hole and drill the bolt holes, by whatever means necessary.  I ended up drilling dozens of tiny holes along the focuser hole line, then whacked the remaining aluminum bits with a chisel until the center fell out. A very laborious task.  If you have an appropriately sized bimetal hole saw, you'll produce a much nicer hole, much faster.

It's also necessary to fit the aluminum to the curve of the tube.  Resist the temptation to use the telescope tube as a form, you might end up denting the tube.  I found a round piece of wood and hammered the sheet against it until a rough curve was formed.


The aluminum plate actually goes up inside the tube;  then re-affix the focuser base plate.  The original bolts won't do because they are too short. You will need M4 x 20mm bolts.  The original nuts can be used.


The focuser assembly feels much more rigid with the aluminum plate inside. Note that I didn't bother to paint the plate; although it is shiny, I'm gambling that the baffles would block any stray reflections from the plate.

The second step is to relocate the collimation locking bolts. As can be seen in this photo, the collimation bolts (black) are spaced 120 degrees apart, and the locking bolts (white) are also 120 degrees apart equidistant between the collimation bolts.

I quickly discovered that this arrangement is no good; after achieving a good primary collimation with the black bolts, when you torque down the white bolts, the collimation shifts.



The solution is to drill new holes as close as possible to the black collimation bolts. Use a small drill bit (like a 3mm) and go slowly.  I did not remove the mirror from the cell so had to proceed very slowly. If the drill bit punches through the metal, it should hit the metal of the mirror holder; but you could slip and ding the mirror. So better safe than sorry. If you drill the holes with the mirror in place, make sure to vacuum up all the swarf and metal shavings afterward; you don't want that stuff getting on the mirror coating.

After the three holes have been drilled, put in a 4mm, then 5mm drill bit and enlarge the holes.  Once the holes are 5mm in diameter, you can tap them with an M6 tap.  The locking bolts are M6 thread. This is the result.  Now collimation only shifts slightly when the locking bolts are torqued down. More importantly, because the locking bolts are right next to the collimation bolts, the collimation shift is very controllable.


I also replaced the white collimation locking bolts with some M6 leg levelers with round plastic feet that cost $0.50 apiece.  This isn't strictly necessary, but it's useful to be able to distinguish the collimation and locking bolts by feel while peering through the focuser.


And here's the result: seven, 5-minute sub-exposures, unguided, of M27.  Depth of the 5-minute sub-exposures is about the same as 20-minute subs with the AT90EDT.  However stars aren't nearly as round. Hopefully better collimation with rectify that..



Saturday, September 28, 2013

Astro-Tech 8" Imaging Newtonian, Part II

There's reasons why the Takahashi Epsilon costs over $5000. One of which is that the Epsilon can cover a 44mm image circle. Another is that the Epsilon is very sturdily built, so collimation doesn't shift.

Here's first light from my Franken-scope:  note the very heavy vignetting on the APS-C sensor:


And the stars aren't round across the sensor.  In fact they are all kinds of shapes (these are the four corners):


I will have to collimate the scope more properly.  Also learned that the tube beneath the focuser needs to be reinforced to prevent flexure, something like this (not my picture).


Well, these are things to do on a Saturday..

Friday, September 27, 2013

Astro-Tech 8" Imaging Newtonian, Part I

I've been struggling with the very strong light pollution in Singapore, which forces me to image in narrowband.  I have resorted to 20-minute long sub-exposures using our Astro-Tech AT90EDT, which has been overall a good experience thanks to the Mach1, but at f/6.7 the depth of the images leaves something to be desired.

Obviously the option is to go faster (in terms of focal ratio).  However the Takahashi FSQ-106ED with reducer (f/3.6) and the Borg 125SD with Super Reducer (also f/3.6) are well beyond my wallet's capacity.  I also thought about the Starizona Hyperstar system which is really fast at f/2.0, but our C9.25 is not Hyperstar-capable, so I'd have to find a compatible SCT, on top of buying the adapter itself which is rather pricey as well.

Another option was the Takahashi Epsilon, and the new Epsilon 130D (to be introduced this November 2013) is in the realm of affordability (i.e. half the cost of the Epsilon 180ED). But the problem with the Epsilon 130D is that it is fairly short at 430mm - that's too wide-field for me!

I had been thinking about the Boren-Simon Powernewt, but it also is a pricey solution.  However an Astro-Systeme Austria 2" Keller reducer suddenly popped up on Astromart a few weeks ago.  It was definitely not cheap - in fact it is the most expensive astronomical item I've purchased in dollars per kilogram of weight.  These reducers come up used very rarely and are quickly snapped up, so I moved fast to get it.

Having obtained the Keller reducer, which has a 0.73X reduction ratio and also corrects for coma, it was time to find a Newtonian.  I decided on an 8" f/4 because I didn't feel too comfortable with the size and weight of the 10" - even though the Mach1 in theory should not break a sweat with the 10" model.  Luckily one of my Cloudy Nights forum buddies had an Astro-Tech AT8IN 8" f/4 Imaging Newtonian for sale. I got it at a good price (a stroke of good luck, as it's been back-ordered at Astronomics for many months now). But as usual shipping was expensive.


The QHY8 camera I've been using for the past few years had a nosepiece that provides exactly 55mm of back-focus.  This is for the Baader MPCC coma corrector (and all the generic refractor flatteners and reducers out there).  I'd been using the QHY8 with an Orion non-reducing refractor flattener, and this worked well with round stars to the corners.



I had already received the Keller reducer prior to the arrival of the AT8IN, but I had not factored in the 65mm required back focus of this reducer. Mismatching the back focus would completely undo the coma-correcting features of this corrector that I'd paid so much for. I don't have any extension tubes and  I didn't want to wait a long time for them.

Luckily I had this spare 2" to T-thread female adapter lying around from my misbegotten on-axis guider experiments.  The female T-thread allows the Keller reducer to screw into it, and the 2" eyepiece holder allows it to grip the QHY8 camera's nosepiece.  I used the pipe clamp as a crude parfocalizing ring, so that every time I remove the 2" to T-thread adapter, I can put it back and still get the 65mm (more or less) back focus without whipping out the digital caliper every time.


And voila:


Sunday, July 21, 2013

BeagleBone Black - First Impressions and Astronomical CCD's

Bought a BeagleBone Black from element14 last week, cost was S$ 66.70.  Since then element14 has lowered the price to S$ 56.95.  The BBB is an Arduino-sized Linux machine with a 1GHz Texas Instruments Sitara SoC, 512MB of RAM, 2GB of onboard eMMC NAND flash with the Angstrom Linux distribution, a micro-SD card slot, 10/100 Ethernet, a regular-sized USB host port, a mini-USB slave port, and a micro-HDMI port for video output.


So far I haven't used it with the micro-HDMI port for lack of the proper adapters, I've wired it up to my access point (which has a bunch of Ethernet ports) and powered it off an iPad charger with a USB cable.  I have tried both the latest version of Angstrom Linux, and Ubuntu 12.04 LTS for the BB.   Angstrom is a fairly strange distribution, as it's based on some OpenWRT tools. Ubuntu is much more predictable to a desktop Linux user; however it is missing Chromium and the OpenCV libraries, which are both bundled with Angstrom.

I got a 16GB Transcend micro-SD card (about S$ 17). The onboard eMMC is too small to install a full OS. Currently there is a bug or feature limitation on the BBB where, if you boot off the eMMC, the micro-SD card is not visible; so the only way to use the micro-SD is to put both the OS and the separate file system on it. To get a full OS (e.g. I installed KStars on Ubuntu) you really need a micro-SD card.  16GB seems to be the sweet spot in terms of price/storage.

I've had some ambitions of using the BBB as a stand-alone astronomical imaging controller, e.g. when connected to a telescope and two cameras, it would perform telescope control, auto-guiding, and image capture.  I have a Meade DSI lying around, and with some Makefile kung fu I managed to get the Meade Deep Sky Imager code for Linux to compile on Angstrom.  I need to get my hands on the Meade DSI firmware, and see if I can get things working.

Supporting my QHY8 astronomical camera would be easier on Ubuntu (because the CCD software requires a Qt development environment) but I'll get to that later; I'm not even sure if my QHY8 is supported, since the CCD software specifies the QHY8L.  Tried changing the VID's in CCD but the QHY8 doesn't work.

Tuesday, June 25, 2013

1-dimensional Kalman Filter, Arduino version

Converted the Processing code (which was a conversion of Adrian Boeing's C++ code) to Arduino.

kalman.ino
// simple Kalman filter
// adapted from C code by Adrian Boeing, www.adrianboeing.com 

KalmanFilter::KalmanFilter(float estimate, float initQ, float initR)
{
  Q = initQ;
  R = initR;

  // initial values for the kalman filter
  x_est_last = 0;
  P_last = 0;

  // initialize with a measurement
  x_est_last = estimate;
}


// add a new measurement, return the Kalman-filtered value
float KalmanFilter::step(float z_measured)
{
  // do a prediction
  x_temp_est = x_est_last;
  P_temp = P_last + R*Q;

  // calculate the Kalman gain
  K = P_temp * (1.0/(P_temp + R));

  // correct
  x_est = x_temp_est + K * (z_measured - x_temp_est); 
  P = (1- K) * P_temp;

  // update our last's
  P_last = P;
  x_est_last = x_est;

  return (x_est);
}

kalman.h
class KalmanFilter
{
public:
  KalmanFilter(float estimate, float initQ, float initR);
  float step(float measurement);
private:
  // initial values for the kalman filter
  float x_est_last;
  float P_last;

  // the noise in the system
  float Q;
  float R;

  float K;    // Kalman gain
  float P;
  float P_temp;
  float x_temp_est;
  float x_est;
};

and how to use:

// simplistic Kalman filter for encoder readings
KalmanFilter kf(0, 0.01, 1.0);
float avg_err = kf.step(track_err);

Monday, June 10, 2013

Kalman Filter for Dummies

I found a simple 1-dimensional Kalman filter online.  It doesn't really look much better than an IIR filter, to be honest - no control input, no way of factoring in the "ideal" value into the estimate.  But as the math of Kalman filters eludes me, this will have to do for now.

I've ported it to Processing.  Should be easy to convert to Arduino or similar.

// simple Kalman filter
// adapted from C code by Adrian Boeing, www.adrianboeing.com
class KalmanFilter
{
  // initial values for the kalman filter
  float x_est_last = 0;
  float P_last = 0;
  // the noise in the system
  float Q;
  float R;
  float K;    // Kalman gain
  float P;
  float P_temp;
  float x_temp_est;
  float x_est;
  // constructor - initialize with first estimate
  KalmanFilter(float estimate)
  {
    Q = 0.022;
    R = 0.617;
    // initialize with a measurement
    x_est_last = estimate;
  }
  KalmanFilter(float estimate, float initQ, float initR)
  {
    Q = initQ;
    R = initR;
    // initialize with a measurement
    x_est_last = estimate;
  }
  // add a new measurement, return the Kalman-filtered value
  public float step(float z_measured)
  {
    // do a prediction
    x_temp_est = x_est_last;
    P_temp = P_last + R*Q;
    // calculate the Kalman gain
    K = P_temp * (1.0/(P_temp + R));
    // correct
    x_est = x_temp_est + K * (z_measured - x_temp_est);
    P = (1- K) * P_temp;
    // update our last's
    P_last = P;
    x_est_last = x_est;
    return (x_est);
  }
}
And to use it..

  // initial estimate, measurement noise, process noise
  KalmanFilter kf = new KalmanFilter(0, 0.05, 1);
  // update with a new measurement
  float data = kf.step(measurement);

Monday, May 06, 2013

Womarts 320x240 TFT LCD Shield with Touch Screen

I bought one of these Womarts 320x240 TFT LCD Touch Screen boards on ebay because they are cheaper than the Sainsmart equivalents, and have both the TFT and carrier board in a single package, unlike the separate TFT board and carrier board that Sainsmart has. Heck, these are even cheaper than the 128x64 I2C LCD boards being sold out there.  The downside of the Womarts board is that it consumes a heck of a lot of pins. You'll be hard pressed to do anything else with one of these boards riding your Arduino.

There is a version of the Womarts board for Mega-type Arduino, and another for the Uno form factor. This post is about the Mega version. There has been some discussion on the Arduino forums about how good (or bad) these $25 Womarts boards are, so I had rather low expectations.

The main problem is that one set of pins on the board doesn't match up with my (also off-brand) China Arduino Mega. But the problem is with the board, not the Mega.

Here we see that the board consumes all of the extra pins on the bottom of the Mega. Oops.. there go the SPI pins.. I'm not sure if the SPI pins are actually used (I hope not!)
On the top side of the Arduino (with the USB port on the right):

The bummer: the bottom set of pins don't fit:


The solution is to bend the pins inward. Not the best reliability, but a necessary evil.



And the best part.. the Henning Karlsen UTFT library works out of the box (just uncomment the line near the top of the example program for the Mega).