Changes for the shutter

The shutter is still having trouble closing. Opening is always fine, but sometimes the close operation fails. The lower shutter gets power, but never moves. The Arduino goes into reboot mode, as though power was re-applied to it.

My theory is that firing up the motors maybe causes a momentary power surge / voltage drop, so the Arduino thinks power has been cycled. Perhaps the shutter in the closed position requires more power/torque from the motor to get started, due to the angle of the activator pulling the shutter closed.

Connecting power directly to the shutter (bypassing the “brush/ring” mechanism) works.

Things to try:

  • Perhaps the copper tape on the track, used to distribute power to the Arduino, is torn or worn. I replaced the tape in the park position area. The brush/tape system now has a resistance of around 5-10 ohms.
  • I notice that the shutter failure occurs when the park position does not place the dome in the exact correct spot. The Foster is often off by 6-12 inches when it parks the dome. In fact, I am seeing potential problem here – sometimes the position is 3 feet off! Oh, boy, another Foster problem to debug.

On the other hand, the home sensor is fairly accurate since it works by detecting the sensor magnet. It will be a couple of inches off depending on which direction the dome turned to find home, but it is consistent. So, I moved the home sensor so that home is in the same position as park. Foster version 3.2.3 has a checkbox that indicates Home is the same as park, but Parking does not use the magnet. It just goes to where it thinks the home sensor should be; this position can be off several inches.

Incidentally, the inaccurate positioning started happening with version 3.2.2. Updated software/firmware to 3.2.3, same problem.

I re-synced the telescope so the rotation is consistent with the new home sensor position.

  • I changed the Arduino code to ramp the pulse width modulation (PWM) from 512 (doesn’t move the motor) up to 1023 (full speed) in steps of 64 every 50 msec. I thought maybe this would help avoid the power drop. I can hear a slight change in the motors starting up.
  • If these things don’t work, my last idea is to install 2 large capacitors on the the Sparkfun MonsterMotor board. They leave spots on the board for these capacitors, but I can’t get any information from them on what they are for or how one would size them. Maybe these capacitors would avoid the power drop. If one even exists:)
Advertisements

Accessing my internal system using brew.my-sky.com

ACP provides a neat feature where you can get a no-ip address which allows you to access your observatory using a unique name (mine is brew.my-sky.com).

I have the problem of not being able to access my observatory using my-sky.com from within my home network. Access works fine from outside the home, but within your LAN you have to address the observatory as 192.168.x.x. This leads to awkward coding of web pages/twiddlers, bookmarks that don’t always work, etc.

Solutions
1) Supposedly, buying a fancier router that supports NAT loopback can resolve this issue. I was preparing to do this; the model I was looking at costs about $130. I did not test this out since solution 2 worked.

2) Change the etc/hosts file on your home computer. This file resolves name addresses before the system goes out to the Internet for DNS resolution.

In my case, the observatory computer 192.168.1.10 hosts the ACP server. My home computer is where I mostly work from. I want my twiddlers, and my blog entries, to use brew.my-sky.com to access a couple of web pages I have created. These do things like show current weather, forecasts, and operate the power controller in the observatory.

The hosts file is located in c:\Windows\system32\drivers\etc. I added the line
<code>
192.168.1.10 brew.my-sky.com
</code>
There should be no leading spaces. A single space or tab separates the IP address and the name.

Caveats:
1) When you edit the hosts file, Windows will not let you save it back in the etc directory. Save it into your Documents directory, remove the .txt extension Notepad insists on adding, then COPY (don’t move) the file to etc.

2) In theory the system immediately starts using the new hosts file. To make sure, use a cmd.exe window and enter “ipconfig /flushdns” to clear the cached entries.

3) test by pinging your domain name.    “ping brew.my-sky.com”
a) If the host file is not active, you will see your remote IP address (like 75.67.23.110).
b) If it works, you will see the local address 192.168.1.10.

4) You may need to restart your browser? I did not need to, but several net posts indicated this was necessary. Firefox in particular may cache DNS entries.

5) One system continued to have problems (a laptop with Windows 8.1). I found that I needed to add permissions for Users to the hosts file (Properties/Security on file hosts). It would only let me add a particular setting, but that worked fine.

More Darkness Readings

Went out before the moon came up and got some readings in different areas nearby. The idea was to verify whether the map showing light levels is correct or accurate. In addition, to check the calibration of the SQM-L meter versus the iPhone app.

Light Measurements at three locations near Casa Grande.

Light Measurements at three locations near Casa Grande.

The lower point (6.1 Bortle) is just into an Indian reservation. These readings were taken between 9:00-9:30PM, 25 degrees C.

The upper point (5.8 Bortle) is on the other side of some mountains nearby; I think to the north of this location is another reservation. Readings taken between 10:00-10:30 PM.

The readings at the middle point are my backyard, between 10:30-10:45. The moon rose shortly after.

Conclusions

1) The map’s darker areas are in fact measurably slightly darker. The 6.1 site in particular was noticeably darker.

2) My area just has too many neighbors with outside lights on!

3) The iPhone app clearly has much more variability than the SQM-L. Perhaps if I take about 10-15 readings and average things work out OK. It also looks like I need to add about .15 to the iPhone value, although this difference changes with the different reading values; a greater difference is shown at the 6.1 value than the 5.65.

Maxim hangs STF8300

I have been having an irregular problem where ACP imaging would fail when Maxim gets hung on reading the camera. I start an imaging run and it fails after several images.

Symptoms:
– The cooling on the camera reports as being off. When the camera is restarted, it often gives a Filter Wheel err 7, or Maxim crashes.
– The exposing tab says “Reading”. Guiding appears to be continuing.
– Sometimes Maxim gets a filter wheel error 7.

I set up some tests, using only Maxim and the camera (no FocusMax, ACP, SkyX) running sets of 40-50 exposures and found that
a) the problem occurs if I guide on the camera simulator rather than STi.
b) the problem occurs if I have no guiding.
c) the problem does NOT occur if the exposures are 5 min or less. The original problem occurred with 30 minute exposures; it still occurs with 10 minutes.

Found that the camera drivers are out of date again via SbigDriverChecker. I updated the windows drivers, and also updated the camera firmware.

Initial test indicates that the problem has gone away; will need to keep an eye on this.

I also found a scary web post indicating there is a setting in the Power Settings tab which causes Windows to turn off power to a USB port if it hasn’t been used for a few minutes (unclear how many). I will turn this setting off, by default it is on.

Update: turned off the power setting, started to run a series of Darks. Keeps failing, similar to the previous problems. However, slight difference is it indicates “300 of 300” in the progress bar rather than “Reading”. So, will have to keep looking into this.:(

Update: well, it appears that the problem is due to a slightly loose USB cable. I went out and re-plugged in the cable on both ends, and it might have been slightly loose on the computer end. After that, a couple of tests worked OK. Crossing my fingers…

FocusMax resolved?

For months, I have been having trouble with FocusMax / ACP interaction problems.

The Problem

FocusMax (3.8.0.15) keeps crashing with the typical “Object not set” message. This only occurs when ACP creates a FocuserControl object to adjust the focuser position when a filter gets changed. The crash is intermittent and occurs at different times in the processing. I.e., ACP creates the focuser object, issues perhaps 8-10 commands against it, then releases the object. The crash may occur after any one of the commands.

The ACP AutoFocus.vbs would often fail; the script FilterOffsets.vbs would always fail after 3-5 filters. Typical ACP imaging scripts do regular filter changes for pointing and focusing. These scripts would typically fail after an hour or two. Problems occur with either Maxim5 or Maxim 6, with real cameras or using simulators.

I tried a number of things; eventually I rebuilt the entire observatory computer. I restored the original system from the backup partition, then reloaded all of the software. It continued to fail; now I was also seeing a problem where the Optec was failing to report position to FM.

We (Steve Brady, Bob Denny, myself) were leaning toward some type of hardware issue. So, I went out to map out the USB cables so I could

1) understand how everything was connected,

2) ensure that appropriate USB 1.0 versus 2.0 hubs are being used. I sometimes get a message “This device can run faster” when the STF8300 is powered up.

The map is posted elsewhere. I found a very nice utility USBview, a microsoft tool which unfortunately is not included in the base OS. It draws a useful tree structure showing where evrything is attached and what USB version it uses.

First, I discovered that when I rebuilt the system I installed an old version of the Optec driver (mislabeled folder). Installed the correct version, not the Optec problem is gone.

Second, two weird things happened when I was tracing the USB cables. I was unplugging and re-plugging in cables to make sure I had the right cable/device pairing.

a) twice, when I plugged the cable back into the computer the OS installed the device drivers. Weird, the device was already installed and operating. The system should not need to re-install drivers, they are already there.

b) When I plugged in the Edgeport cable, I got a Blue Screen of Death. The Edgeport supplies 4 RS-232 ports off a single USB cable; I use 2 of these to operate the two focusers Optec and LazyFocus. When the system was rebooted the Edgeport showed up but none of the ports were there. I uninstalled/reinstalled the Edgeport driver, rebooted, and the COM ports were now there.

After weirdness (b), everything is now working. Somehow the USB drivers were “messed up”, but got themselves straightened out. I can now runs AtoFocus.vbs and FilterOffsets.vbs with no problems, using eithere Simulators, the STF8300, or the ST2000M. Everything looks peachy now:)

I moved to FocusMax version 4, the new (non-free) version. Problems continued, with a new wrinkle

Paramount is Here!

The MX showed up today, and I am off to the observatory to install it.

OK, here we go with issues:

1) The plate I had made to receive the mount on my pier doesn’t fit. I hd specified within 0.0001 inches tolerance; the bolt holes are about 1/16 inch off.

To get it to work, I put in 1/4-20 screws. Oops, I don’t have washers. Off to the store for washers. The mount now looks nice and level on the bubble mount. The bottom washers barely fit – they are up against the pier collar.

2) rotated the Versa plate. Odd that you can only use 4 of the bolts to remount it. Unclear what I am supposed to do with the wiring – I just ran it outside the head to the instrument panel. They provide a bunch of wrenches, but not one of a size to fit the Varsa plate.

3) No serial numbers? The manual says the serial numbers for the software are on the packing slip. No numbers are on the slip. There is a serial for the mount itself on the outside of the main box. Maybe that is supposed to get the numbers from the Bisque site or something?

Called OPT, they say call Bisque. Bisque is closed:(

4) through the mount cabling – you have to be kidding. I think one might be able to run raw wires through there, but all my cables have connectors on them. There is no way a connector is going through that tiny area. Looks like my wires will hang off the mount like they always have off the G11.

5) This is a problem – it turns out the two counterweights, all the way down the shaft, do not counterbalance my OTA. I have an 11″ Edge and 2 small refractors, with ST2000XM, Canon EOS, and Starfish cameras. These apparently weigh too much – I will need to either a) figure out how to add more weights, b) but a $250 counterweight, or c) buy the shaft extension $150. Irritating, and is going to hold off doing anything until I can get the extra weight on there.