Re: Powering Remote Imaging Systems


Steve Reilly
 

Thanks Dale,

 

NUCs are a bit new to me and have always built my own custom systems based on the observatory needs such as serial and USB ports wanting to have as many "native" as possible which is generally fine for universities and other installments where it is being manned. My experience at SRO even though there is staff there to assist when needed is different in that they need to be my eyes and then I need to ask the right questions and think without the aid of visual assistance. I asked for a picture of the back panel on the PC to see the LAN connections and it was then that I realized the CP4 is connected to the switch and not the 2nd port on the computer but more importantly I saw how dirty the back side and especially the fan section was alerting me to the fact that it had a dire need to be blown out. No telling how much dust/dirt is on the heat sync/fan.

 

From what little I've read on these NUC units I should be able to get most of what's needed and then use a StarTech ICUSB23241 for my needed serial ports. My issue with remote is the handicap it presents when trying to visualize the needs in resolving issues. I’ll need to physically prepare and load the necessary programs on the NUC here at home and then have it ready to be connected to the equipment at SRO. Of course once it’s online I can make adjustments but the more ready it is ahead of time the less down time there should be. The next question this brings to mind is location of the NUC. I’m guessing the unit can be mounted to the pier or other very close location making cabling maybe a bit better in terms of possibly un-needed extensions? Would long duct straps be an acceptable method of securing this unit to say the pier?

 

-Steve

 

 

 

 

 

-----Original Message-----
From: main@ap-gto.groups.io <main@ap-gto.groups.io> On Behalf Of Dale Ghent
Sent: Tuesday, April 20, 2021 8:30 AM
To: main@ap-gto.groups.io
Subject: Re: [ap-gto] Powering Remote Imaging Systems

 

 

Hi Steve,

 

I use units from OnLogic, specifically the ML100G-51 model. They have a variety of system in industrial and rugged designs. I would keep to i3 or better systems. Celerons, especially N-series Celeron CPUs, don't really have the core count or clock speed when it comes to Windows. OnLogic systems are made in the US, and support is out of the US as well.

 

In the US and UK, there is SimplyNUC and their Porcoolpine line of passively-cooled units, but a recent update saw an emphasis on Thunderbolt instead of classic USB ports. They're still fine if you need only 1 or 2 USB connections to the box (ie, you have a USB hub).

 

There are of course a lot of options on Amazon and AliExpress, such as the Kingdel units that Luca mentions, as well as ones from Topton.

 

 

> On Apr 18, 2021, at 11:19, Steve Reilly <sreilly24590@...> wrote:

>

> Thanks for the reply Dale. I probably should replace at least the SRO computer. Any particular brand or vendor you can recommend for getting these? Hopefully they can be well configured which is what I think kept me from using these early on.

>

> -Steve

>

>

> -----Original Message-----

> From: main@ap-gto.groups.io <main@ap-gto.groups.io> On Behalf Of Dale Ghent

> Sent: Sunday, April 18, 2021 11:08 AM

> To: main@ap-gto.groups.io

> Subject: Re: [ap-gto] Powering Remote Imaging Systems

>

>

> I wouldn't worry about computers needing a "rest" unless you have power availability considerations that would necessitate a complete shutdown of everything during idle periods. Computers are pretty resilient things. You can get fanless systems with SSDs and be completely free of any moving parts that could wear out from long periods of running. It's far better to have a fanless system that is designed to be fanless, than one that relies on a fan for cooling and is in trouble if that fan breaks or gets clogged up with the kind of dust and gunk that often floats around inside observatories.

>

> I don't know what your current in-observatory system is, but I would just opt for a low power system instead of trying to come up with some power management/synchronization machination for a single remote PC. Something NUC-ish is suitable for almost any observatory's control and acquisition system. You can find small systems with 15W TDP Intel i3 or i5 CPUs that sip power are completely adequate for the task.

>

> You can have it do stuff during the daytime, too. Run a weather station, monitor inside temperature/humidity sensors, or an all-sky or security cameras, for example.

>

>

>> On Apr 18, 2021, at 08:17, Steve Reilly <sreilly24590@...> wrote:

>>

>> While not directly mount related it does concern powering systems. For a system I run at SRO and one at home I have Web Power switches for the scripting the powering of equipment on/off through ACP Expert which works great but now years after just leaving the computers on 24/7 forever I wonder about the intelligence of that practice. I do have the computer’s BIOS set to restore to previous state should there be a power loss so that it turns back on and in turn opens the programs I have in the Start Menu which is fine for emergencies but I suspect I wouldn’t want to force stop/start it like that on a regular basis if not necessary. I’ve read a bit about Wake On LAN but that requires a computer that is on that network to access the computer you want woken which in this case is doable even with SRO as they have a VPN that sets the remote computer as part of your local network. I can Remote Desktop into that system using the IP address from home when the VPN is connected.

>>

>> But of course I’m looking for more fool proofing such as not relaying on memory to connect and turn on each system every day  as that will most certainly be forgotten from time to time and likely on the very best of nights. So the real question hidden deep in this post is if anyone knows of a mostly foolproof method to have remote computers turn on and off at predetermined times. I would expect you’d want to calculate the earliest time for On in your area considering DST and off by the latest morning time. Say you may want 4PM start for the winter when you could be imaging by 5:30pm in some areas and off at 9am in the morning which would/should allow time for sky flats should you take them both evening and morning. This at least gives the computer a rest of 15+ hours a day and in my case where I use a Starlight Xpress UltraStar guider t to would be powered down as it’s always on when connected to the computer’s USB.

>>

>> Just think there should be a more reasonable way to run these system that would be easier on the computers as well as being more efficient. Anything obvious I’m missing? Suggestions?

>>

>

>

>

>

>

>

>

>

>

>

>

 

 

 

 

Join main@ap-gto.groups.io to automatically receive all group messages.