Jump to content


Photo

10Micron

  • Please log in to reply
301 replies to this topic

#251 Tonk

Tonk

    Fly Me to the Moon

  • *****
  • Posts: 7148
  • Joined: 19 Aug 2004
  • Loc: Leeds, UK, 54N

Posted 22 January 2014 - 08:47 AM

As to the earlier sneering about "going manual" to set up the the mount (quite a few pages back now) - I will reiterate that is an extremely easy mount to set up to a high pointing model accuracy and PA accuracy (both better than 20" error) in just 20 to 30 minutes - ideally done during twilight - and doesn't need anything more than an astrometric eyepiece to accomplish.

So anyone looking for a totally standalone mount (meaning no extra laptop /software needed) that goes unguided up to at least 10 minutes, that can be rapidly setup in the field then the 1000HPS is the mount that can do this very very well. In this respect it is set apart from the 2000, 3000, 4000HPS which are really observatory mounts.

The 1000HPS can be likened to the smaller Losmandy mounts with the Gemini I/II controller (another standalone system) but on steroids performance wise! It is set well apart from the Losmandy mounts in this respect but otherwise the functionality and operation of the controllers is very similar.

#252 HowardK

HowardK

    Apollo

  • *****
  • Posts: 1091
  • Joined: 20 Oct 2010

Posted 27 January 2014 - 10:19 AM

Hi Tonk.

Are you saying here that the 2000 needs Per's modeling software to perform.

Or will an accurate say 3+5 alignment and accurare polar alignment done thru the hand controller suffice for 10 mins unguided with really accurate goto's.

Am considering the 2000 btw to replace my CGE PRO sometime this year.
But as i do not use a computer nor want to bother with maxim dl and model building all this talk of Per's software has me worried how the 2000 performs out of its box so to speak.

#253 EFT

EFT

    Vendor - Deep Space Products

  • *****
  • Vendors
  • Posts: 2566
  • Joined: 07 May 2007
  • Loc: Phoenix, AZ

Posted 27 January 2014 - 10:29 AM

Hi Tonk.

Are you saying here that the 2000 needs Per's modeling software to perform.

Or will an accurate say 3+5 alignment and accurare polar alignment done thru the hand controller suffice for 10 mins unguided with really accurate goto's.

Am considering the 2000 btw to replace my CGE PRO sometime this year.
But as i do not use a computer nor want to bother with maxim dl and model building all this talk of Per's software has me worried how the 2000 performs out of its box so to speak.


What Per's software does is completely automate the process for those people who want to go hands-off. It also leaves the calibration alignment process up to the computer which can arguably be even more precise than the human operator. It is not a requirement, but instead augments an excellent process that is already built into the mount and the mount's external software (when used). In general, it is recommended to get at least 20-25 alignment stars to start unguided imaging. Gotos are highly accurate with only a few stars, but for dual axis tracking and correction, the larger model is better.

#254 Tonk

Tonk

    Fly Me to the Moon

  • *****
  • Posts: 7148
  • Joined: 19 Aug 2004
  • Loc: Leeds, UK, 54N

Posted 27 January 2014 - 10:56 AM

But as i do not use a computer nor want to bother with maxim dl and model building all this talk of Per's software has me worried how the 2000 performs out of its box so to speak.


I don't use Per's software. Its not a requirement as I pointed out above - its a tool for the lazy ;) See my more detailed answer to your question on the other thread ("10Micron club - I'm in").

Just to make it clear the 1000HPS and 2000HPS use the same controller and handpad - so whats said for one applies to the other - differences are mechanical - imaging load, max slew rate etc

#255 HowardK

HowardK

    Apollo

  • *****
  • Posts: 1091
  • Joined: 20 Oct 2010

Posted 27 January 2014 - 11:57 AM

Thanku Ed and Tonk.

Sounds like 2 or 3 iterations at alignment with 10-15 stars used each time (depending on how lazy or impatient i feel at the time) will get things working real well, real quick.

#256 GIR

GIR

    Vostok 1

  • -----
  • Posts: 138
  • Joined: 02 Jan 2010
  • Loc: Finland

Posted 27 January 2014 - 12:25 PM

In general, it is recommended to get at least 20-25 alignment stars to start unguided imaging. Gotos are highly accurate with only a few stars, but for dual axis tracking and correction, the larger model is better.


Mounts with absolute encoders use the model for both purposes, gotos and tracking. Gotos are not a problem with a rather small model but if you want to do tracking properly you should make a rather large model to cover the whole sky ...or build a local model which is possible (as far as I know)only with one mount manufacturer's software at the moment.

I have a permanent setup and am using a 120 point model for the unguided imaging. I also add a 5-10 point local model on top of that sometime when the target is located in low alt position.

#257 Tom Polakis

Tom Polakis

    Surveyor 1

  • *****
  • Posts: 1668
  • Joined: 20 Dec 2004
  • Loc: Tempe, Arizona

Posted 27 January 2014 - 03:52 PM

I have a permanent setup and am using a 120 point model for the unguided imaging. I also add a 5-10 point local model on top of that sometime when the target is located in low alt position.



Are the 5 to 10 points for low altitude modeled to deal with refraction? One thing that has bugged me about the concept of unguided imaging is that refraction is significant for large-scale imaging at moderately high altitudes, like 30 degrees. Even if you have perfect polar alignment and the mount can eliminate periodic error, how do these closed-loop mounts deal with refraction? Thanks.

Tom

#258 Tonk

Tonk

    Fly Me to the Moon

  • *****
  • Posts: 7148
  • Joined: 19 Aug 2004
  • Loc: Leeds, UK, 54N

Posted 27 January 2014 - 04:02 PM

Even if you have perfect polar alignment and the mount can eliminate periodic error, how do these closed-loop mounts deal with refraction? Thanks.


For the 10Micron controller you input site altitude (or use a GPS receiver) and/or the current barometric pressure. So you need a good barometer for best accuracy!

These inputs are for the refraction part of the tracking equation. If you don't add the measured pressure but just the site altitude it assumes a standard sea level pressure

#259 HowardK

HowardK

    Apollo

  • *****
  • Posts: 1091
  • Joined: 20 Oct 2010

Posted 27 January 2014 - 06:05 PM

Can anyone explain the firmware and software updating process for the 2000 HPS.

The 10 micron forum will not let me onto the firmware thread.

#260 EFT

EFT

    Vendor - Deep Space Products

  • *****
  • Vendors
  • Posts: 2566
  • Joined: 07 May 2007
  • Loc: Phoenix, AZ

Posted 27 January 2014 - 06:45 PM

Can anyone explain the firmware and software updating process for the 2000 HPS.

The 10 micron forum will not let me onto the firmware thread.


I'm not sure why you can't get into the thread, but it may be limited to registered owners I suppose. The firmware is updated through a firmware updater program that is run with a connection to the mount computer. I don't know if the connection can be either wired or wireless, but my general recommendation would always be to use a wired connection when updating firmware. Program updates are installed through .exe executable files.

#261 GIR

GIR

    Vostok 1

  • -----
  • Posts: 138
  • Joined: 02 Jan 2010
  • Loc: Finland

Posted 28 January 2014 - 01:45 AM

Are the 5 to 10 points for low altitude modeled to deal with refraction? One thing that has bugged me about the concept of unguided imaging is that refraction is significant for large-scale imaging at moderately high altitudes, like 30 degrees. Even if you have perfect polar alignment and the mount can eliminate periodic error, how do these closed-loop mounts deal with refraction? Thanks.

Tom


Tom

Yes, the main reason for those extra 5-10 plate solves (for low altitudes) is getting rid of the remaining refractions, flexures, plate solving errors etc. It also shrinks the RA and DE tracking errors (which will always be present even in a large model) to practically 0. So it’s very handy to use even in connection with large models and especially with mobile setups. The beauty of the local model is that it takes plate solves from the path your imaging. So it’s VERY accurate and fast to do build (takes 2-3 minutes).

Playing with barometric readings, altitudes etc. is waste of time and not very accurate IMO. Especially when having all kinds of weather/atmosphere issues (temperature changes, reflections from snow, different seeing conditions etc.) Using plate solves from the path your imaging is far more accurate.

Another important factor is the time and place. Place is usually not a problem because you can get/input coordinates in many different ways (needed in any mount, not just mounts with encoders). Having an accurate time (sub-seconds accuracy) is bit more complicated because even the computer time is not very accurate. However, with a computer (and internet connection) you can use any simple (free) software to sync the computer time to some reliable time source (e.g. every 10 minutes). Or use a GPS unit to handle both the time and place automatically and accurately. I’m using a small 40 euro GPS device to do the job and haven’t had any problems even during the cold winter months.

#262 Per Frejvall

Per Frejvall

    Messenger

  • *****
  • Posts: 428
  • Joined: 28 Sep 2012
  • Loc: Saltsjöbaden, Sweden

Posted 28 January 2014 - 02:41 AM

Firmware is updated, as Ed correctly states, with an Updater application that is installed on your computer. It has the firmware embedded, so there are no separate firmware files. You install the updater of the version you want and simply run it.

Connection can be anything; wired RS232, wired Ethernet or WLAN (if you have the WLAN option).

The updater application is also used to add/remove/edit data in the comet and satellite databases in the mount. Once you have transferred your TLE files with the orbital parameters, they are stored within the mount (actually box) and are available without a computer.

As for refraction or not, I beg to disagree, GIR. Refraction is a well researched subject with a very reliable formula taking the temperature and ambient pressure as inputs. Science and research have determined that it is the pressure and temperature at the actual site that matters most.

10Micron handles the refraction calculations in-mount. This means that when you build a model (in any way you want), the refraction parameters need to be correct. Prior to entering the synced coordinates into the model, the mount firmware will nullify the refraction so that the refraction at the time of actual use of the model can be applied. Therefore it is imperative to have the correct parameters.

Should one choose the plate solve "force" method for low altitudes, the net result would be poor tracking when refraction does change to something other than what was present at the time of model building. I am quite sure that ASA also nullifies refraction prior to entering it in the model.

Some final notes on refraction are:

1) The formula most generally used is accurate to about 0.07', or 4.2"

2) Ambient pressure changes refraction by 1% for every 0.9 kPa

3) Ambient temperature changes refraction by about 1% for every 3°K

4) Increased pressure decreases refraction, increased temperature increases refraction

5) At 10° alt, the refraction is about 5', 5° 10', 30° 2'

As per point 5, you can see that even at 30° alt you are still looking at a good 120" of light bending. Use a temperature error of 10°C and you add or subtract 12" to that. You clearly HAVE TO use refraction in this game if you want to be able to use your model on more than one night (or two hour span if a high pressure system is passing by).

Time, then...

We all know that 1 second off in time yields an error in Ra of about 15", which is a lot when we expect sub 5" RMS performance. A GPS communicates at 4800 bps, normally with one start bit and one stop bit. Net performance is thus 480 characters per second, or 2.1 ms per character. Below is the spec from IEC 61162-1 specifying the NMEA 0183 protocol, and more specifically the sentence used to send the GPS time (the shortest one, not necessarily the one your GPS uses). For your GPS it would be like this:

$GPGLL,5916.25,N,01818.18,E,192831.22,A,D*3F

That constitutes 46 characters (including line endings) and takes 96 milliseconds to transfer to your computer. Before that, the GPS has to calculate the position fix with its hardware maths thingy (correlators), assemble the sentence and get it out the door, adding a few milliseconds to the game. Then your computer has to set the clock, a fast process, but nonetheless. In the end, you are at least 1.5" away at that time, most likely much more. Add to that the fact that the time is only specified with 10 ms of precision...

Using a good NTP client (Meinberg is good) via the internet can get your PC clock running at sub millisecond precision, so that is a lot better. Other factors that may or may not drown the errors here have to do with sidereal time of the plate solve. I haven't researched that yet :)

Given this time trouble, I wrote my time setting software so that it waits for a 2 ms slot at an even second, then sends it to the mount and records that time. When the mount responds to the time set request, it is reasonable to assume that the TCP/IP communication is negligible and that the time of response arrival closely matches when the mount set the time. The software reports this and you can reissue the command until you are satisfied with the precision. I usually settle for 30 ms or so. Of course, this is much better handled in ASA's software as it runs entirely in the PC and doesn't have to communicate the time to the mount.


Enough of a morning rant :cool:

All the best,

Per

Attached Files



#263 Hilmi

Hilmi

    Soyuz

  • *****
  • Posts: 3834
  • Joined: 07 Mar 2010
  • Loc: Muscat, Sultanate of Oman

Posted 28 January 2014 - 02:49 AM

So how does the mount handle permanent observatory mounting. Do you need to keep on constantly updating the temperature and pressure readings? Sounds counter intuitive.

#264 Skunky

Skunky

    Ranger 4

  • -----
  • Posts: 357
  • Joined: 16 Sep 2013

Posted 28 January 2014 - 04:18 AM

Any modeled mount will have that issue. But i bet someone will write a program to allow a sensor to update the mount automatically.

#265 Tonk

Tonk

    Fly Me to the Moon

  • *****
  • Posts: 7148
  • Joined: 19 Aug 2004
  • Loc: Leeds, UK, 54N

Posted 28 January 2014 - 05:09 AM

Playing with barometric readings, altitudes etc. is waste of time and not very accurate IMO.


Per has succinctly answered the RI issue re the math.

If you look into it the refraction problem is not that sensitive over the pressure change rates likely to be experienced during a imaging session. The reality is you just need a reasonable fix as to the pressure/temp at the time you start imaging.


In my practical experience inputting barometric pressure/temp works well enough for the real world. I still get ROUND stars in long unguided exposures at 20 degrees alt.

Now I'll point out that rapid pressure changes are usually a symptom of a front passing over which invariably is accompanied by clouds and rain, making this whole scenario rather moot.


#266 famax

famax

    Explorer 1

  • -----
  • Posts: 55
  • Joined: 01 Jul 2007

Posted 28 January 2014 - 05:25 AM

I have a question about temperature and refraction :
We input temperature usually measured at 1m from the ground.
But temperature profile change a lot with altitude
and so should the refraction index...no ?
Sometime in winter they are temperature inversion between hign and low altitude for example , so the refarction parameters seems a bit too simple regarding reality...


I do agree with Gir that a part from direct drive techno
the local tracking thing is a real value for ASA mount, thanks to Mr Keller

#267 GIR

GIR

    Vostok 1

  • -----
  • Posts: 138
  • Joined: 02 Jan 2010
  • Loc: Finland

Posted 28 January 2014 - 05:29 AM


As for refraction or not, I beg to disagree, GIR. Refraction is a well researched subject with a very reliable formula taking the temperature and ambient pressure as inputs. Science and research have determined that it is the pressure and temperature at the actual site that matters most.

10Micron handles the refraction calculations in-mount. This means that when you build a model (in any way you want), the refraction parameters need to be correct. Prior to entering the synced coordinates into the model, the mount firmware will nullify the refraction so that the refraction at the time of actual use of the model can be applied. Therefore it is imperative to have the correct parameters.

Should one choose the plate solve "force" method for low altitudes, the net result would be poor tracking when refraction does change to something other than what was present at the time of model building. I am quite sure that ASA also nullifies refraction prior to entering it in the model.


Per,

If you really think that making the refraction calculation based on some formula and manually inputted parameters on e.g. temperature and ambient pressure …and trying constantly updating those parameters, is a good way to go, we really do disagree completely.

I’m using a good all sky model, which can be supplemented with a local model (done at the time you’re actually imaging), and using that combination to solve some of the remaining issues included when using only the sky model. There is no need play with any refraction parameters and the tracking accuracy is simply astonishing.

If you assume that the refraction parameters will change so much during the imaging session, you can always make another local model (takes only couple of minutes and it’s fully automated). However, I have never had a need to do that. Neither have I ever heard that someone would start updating and inputting new refraction parameters manually when imaging ;)

#268 GIR

GIR

    Vostok 1

  • -----
  • Posts: 138
  • Joined: 02 Jan 2010
  • Loc: Finland

Posted 28 January 2014 - 05:30 AM

I do agree with Gir that a part from direct drive techno
the local tracking thing is a real value for ASA mount, thanks to Mr Keller


Thanks famax...

#269 Tonk

Tonk

    Fly Me to the Moon

  • *****
  • Posts: 7148
  • Joined: 19 Aug 2004
  • Loc: Leeds, UK, 54N

Posted 28 January 2014 - 05:31 AM

Sometime in winter they are temperature inversion between hign and low altitude for example , so the refarction parameters seems a bit too simple regarding reality...


Do you really want to image stuff below 20 degrees altitude?

Where refractive index issues are significant is largely in the "no-go" imaging area of the sky around the horizon so why are we worrying?

#270 Tonk

Tonk

    Fly Me to the Moon

  • *****
  • Posts: 7148
  • Joined: 19 Aug 2004
  • Loc: Leeds, UK, 54N

Posted 28 January 2014 - 05:36 AM

If you really think that making the refraction calculation based on some formula and manually inputted parameters on e.g. temperature and ambient pressure …and trying constantly updating those parameters, is a good way to go, we really do disagree completely.


GIR - with your method do you re-run the local plate solve between each image? I.e. how do you constantly update your parameters? - Oh you say you never have had the need. That says that the problem isn't a sensitive one for your imaging durations. So ...

…and trying constantly updating those parameters


Well if you don't have to .... we don't have to. So why are you raising this unlikely scenario?????




I largely think this plate-solving vs math refraction modeling argument is blown up out of reallity as my experience of using the barometric method gives excellent results - i.e. round stars. Surely thats the goal?

Per is quite right the mathematical modeling of atmospheric refration is accurate and practical enough not to be a problem. Frankly this is just a gearhead argument without recourse to testing reality and too full of "IMO" (actually MYGIBTYG "my gear is better than your gear") comments ;).

GIR - your ASA mount is wonderful and probably has the best and most accurate methods for everything under the night sky. However unless you can really tell me that you don't get accurate enough tracking with a 10Micron mount through practical hands on experience - then your "IMO"s are just guesses - right?

#271 GIR

GIR

    Vostok 1

  • -----
  • Posts: 138
  • Joined: 02 Jan 2010
  • Loc: Finland

Posted 28 January 2014 - 06:12 AM


Well if you don't have to .... we don't have to. So why are you raising this unlikely scenario?????


Well, actually Per raised that scenario...

"Should one choose the plate solve "force" method for low altitudes, the net result would be poor tracking when refraction does change to something other than what was present at the time of model building".

#272 Hilmi

Hilmi

    Soyuz

  • *****
  • Posts: 3834
  • Joined: 07 Mar 2010
  • Loc: Muscat, Sultanate of Oman

Posted 28 January 2014 - 06:40 AM

I would still like some clarification, does one need to enter the temperature and pressure reading with every session in a permanently mounted set-up? If not, then this would tell me that the pressure reading is not that critical to the correct operation of the mount and can be safely ignored. If it makes a difference, then is there a possibly more streamlined solution of doing this?

I can imagine having to update the temperature and pressure readings then rebuilding the model would start getting old very fast in a permanent setup.

Do not take this as product bashing, this is a genuine question from a person truly interested in the product.

#273 Per Frejvall

Per Frejvall

    Messenger

  • *****
  • Posts: 428
  • Joined: 28 Sep 2012
  • Loc: Saltsjöbaden, Sweden

Posted 28 January 2014 - 06:52 AM

OK, folks. Calm down. A few points:

Yes, the math of refraction is well understood. It is based on local temp and press.

If you tell your mount that it is +10° and 1030 hPa when it really is 0° and 997 hPa, and then run a model, that model is going to be really wrong. Period. It will only be good when you have +10° and 1030 hPa.

If you instead input the correct values, the model will work regardless of ambient temp and press if you input the values.

Slew to a star at 30° alt and then change the temperature by 10°C and watch it move. ;) Naturally, refraction affects pointing more than tracking, but the further north you are the more the tracking is affected. And yes, ASA mounts can always do a local model, most likely valid unless the temperature or pressure changes significantly during the run. Autoslew tells you how much of refraction there is and displays coordinates as refraction compensated ones.

For my part, I let ACP send temperature and pressure to the mount at the start of every plan (1 hr). I am used to the temperature changing 10°C or more over a night's run so it matters to me.

So, there is nothing to it. The question was whether it matters, and the answer is that it does. You can also ask yourself why software packages like ACP has a check box that determines whether it or the mount does the refraction compensation...

Tonk, your variant is sufficient. You do not pay much attention to it, but you are using refraction and giving it data to work on. Good enough!

/per

#274 GIR

GIR

    Vostok 1

  • -----
  • Posts: 138
  • Joined: 02 Jan 2010
  • Loc: Finland

Posted 28 January 2014 - 06:59 AM

Time, then...

We all know that 1 second off in time yields an error in Ra of about 15", which is a lot when we expect sub 5" RMS performance. A GPS communicates at 4800 bps, normally with one start bit and one stop bit. Net performance is thus 480 characters per second, or 2.1 ms per character. Below is the spec from IEC 61162-1 specifying the NMEA 0183 protocol, and more specifically the sentence used to send the GPS time (the shortest one, not necessarily the one your GPS uses). For your GPS it would be like this:

$GPGLL,5916.25,N,01818.18,E,192831.22,A,D*3F

That constitutes 46 characters (including line endings) and takes 96 milliseconds to transfer to your computer. Before that, the GPS has to calculate the position fix with its hardware maths thingy (correlators), assemble the sentence and get it out the door, adding a few milliseconds to the game. Then your computer has to set the clock, a fast process, but nonetheless. In the end, you are at least 1.5" away at that time, most likely much more. Add to that the fact that the time is only specified with 10 ms of precision...

Using a good NTP client (Meinberg is good) via the internet can get your PC clock running at sub millisecond precision, so that is a lot better. Other factors that may or may not drown the errors here have to do with sidereal time of the plate solve. I haven't researched that yet :)

Given this time trouble, I wrote my time setting software so that it waits for a 2 ms slot at an even second, then sends it to the mount and records that time. When the mount responds to the time set request, it is reasonable to assume that the TCP/IP communication is negligible and that the time of response arrival closely matches when the mount set the time. The software reports this and you can reissue the command until you are satisfied with the precision. I usually settle for 30 ms or so. Of course, this is much better handled in ASA's software as it runs entirely in the PC and doesn't have to communicate the time to the mount.



Almost missed this one…

Just to be a bit more specific, I’m using the GPS module for determining the location and time for the computer. The GPS module doesn’t communicate directly with the mount. I would guess that the software is smart enough to understand that there is a certain lack with the GPS signal and is able to handle it. However, have never tested if there is some difference with that method or letting some software synchronize the computer clock to a good time source. Have to try that some day…

#275 Per Frejvall

Per Frejvall

    Messenger

  • *****
  • Posts: 428
  • Joined: 28 Sep 2012
  • Loc: Saltsjöbaden, Sweden

Posted 28 January 2014 - 07:26 AM

Meinberg NTP - good stuff!

/per






Cloudy Nights LLC
Cloudy Nights Sponsor: Astronomics