GOES 16 GRB downlink vs GVAR

Got a quote for the entire computer by a very reputable company. The cost is about what I'd expect for a computer of this magnitude.
* The processor is a AMD Ryzen Threadripper 1920X
12-Core / 24 Threads 3.5 GHz.
* ASRock X399 Taichi motherboard
* 32 Gb ram, Can be expanded to 64 GB
* Main drive for OS 240 GB solid state drive
* Data storage drive 1TB
* Upright computer case
* 550W power supply
* CPU cooler :Arctic Freezer 33TR
* Low end VGA PCI card for admin purposes
* DVD drive
* Keyboard/mouse

The quote came in at 2560.00 for the machine.
With a 1 year parts and labor warranty.
I'm told and agree that this is a very high end gaming computer.
I am also told that this processor will work very well with CentOS linux.
The hardware chosen has high reliability and costs a bit more, it would be silly to choose cheap unreliable hardware for this.
However the company that quoted me this is selling a computer for 1,800 that's about a 1/3 mabie 1/2 of what this one does.
I'm impressed with the advancement of the computer world as 10 years ago the price for this would have been out of sight and almost impossible to get.
The parts are available and can be built within a couple of weeks.
I will be setting on this for a few months at least as I want to see the bird in normal Ops before committing to this machine.
I will add as a final note that this computer costs less then what I paid for my GVAR machine.
Which only has a 1 GHz processor in it.
And just have to add it's super cool. :glasses
AMD_Ryzen_Threadripper.jpg

X399 Taichi motherboard.png
 
The TBS8922 PCIE tuner has been spec'ed for BPSK in several of the TBS listings. I would verify this with TBS, but a $90 card might be compatible and provide an inexpensive hardware solution.

There is a confusing mix of formats sent down by the GOES satellites. I read that GRB uses an outer packet format of DVB-S2/QPSK and the inner format is CCSDS space packet format. Its my understanding that the high spec computer is to decode the CCSDS space packets at 15.5 Mbps per channel (and there are two channels). And the NOVRA S300 is for the DVB-S2/QPSK generic stream. So what we need is an alternative to the NOVRA S300 that does DVB-S2/QPSK generic stream. If the TBS8922 does generic stream QPSK, it would be great.

GVAR, LRIT, and HRIT are plain BPSK with no DVB-S2 involved. The BPSK modulation used by LRIT and HRIT can demodulated in software or hardware. I'm not sure if GVAR requires hardware for BPSK demodulation, or if there is a software solution available?

And there is no getting around the expensive high-end PC for CCSDS space packet.

...
* The processor is a AMD Ryzen Threadripper 1920X
12-Core / 24 Threads 3.5 GHz.
I would be inclined to stick with Intel because there could be subtle variations in AMD that affect compatibility with the Univ. of Wisconsin GRB software. Or maybe check with Univ. of Wisconsin to make sure it has been tested with AMD.

...
I will be setting on this for a few months at least as I want to see the bird in normal Ops before committing to this machine.
I will add as a final note that this computer costs less then what I paid for my GVAR machine....
Good idea. Something could happen to the satellite on its way east. Plus you also might want to verify that the signal is still strong enough after it moves farther away from you.

... As for me, I completely recompiled the Open Satellite Project from scratch and the 'GOES Dump' program is still reporting network errors. There are three separate programs, "Decoder", "Demodulator", and "GOES Dump". I have been running all 3 on the same machine. Next I'm going to try "GOES Dump" on a different system.
 
Ti,
I'm glad you found those as they May work for GRB DVB-S2 reception.
Other options are: TBS5927 Professional DVB-S2 TV Tuner USB, This would be the best option as it should be able to decode the GRB DVB-S2 outer layer.
2nd: TBS 6925 PCI-E DVB-S2 TV Tuner Card, this may also work for the GRB DVB-S2 receiver.
I have as well a: Techno Trend TT Budget S2-1600 that might work but no OS that I can use it with as all my computers that have PCI cards are WIN 2000. :crying
I bought it for NOAAPORT.
But as N6BY said they won't work for GVAR reception. GVAR uses I believe it to be called "legacy" products.
An example of the GVAR demodulator is the Commtech EF data CDM-570/L
Sometimes you can find those on ebay.
It would have been nice if GVAR used DVBS as it could have cut the costs of earth stations, But NOAA wanted to stay with a format similar to what the early spin stabilized spacecraft (GOES 5-6) used when they went to the GVAR format. The only big change in the GVAR format was several years ago when they did some channel changes in the "N-O-P" series spacecraft.
I remember it well, I had to pay for the upgrade to get GOES 13 GVAR and it wasn't cheap. Had to turn to GOES West when 12 went down like many others that didn't have the upgrade at the time.

There is a confusing mix of formats sent down by the GOES satellites. I read that GRB uses an outer packet format of DVB-S2/QPSK and the inner format is CCSDS space packet format.

N6BY,
The DVB-S2 is an outer format as you state.
The CCSDS space packets are what is generated from NOAA, it's their format of sending this type of data.
The GRB stream is made of 3 layers. The DVB-S2 receiver strips of the outer layer sending the CDAU AOS packets to the computer where it decodes those and exposes the space packets that have the NOAA's info inside of them.
After the DVB-S2 outer layer the computer is decoding additional 3 layers and putting all the fragments and different parts of the files back together.
See page 5 of the Pug GRB vol 4.pdf and read on for more info.

And there is no getting around the expensive high-end PC for CCSDS space packet.

There's a lot happening inside of that computer, it may
be possible to use less of a machine but as stated in the manual the possibility of dropping/losing data gets much higher.

I would be inclined to stick with Intel because there could be subtle variations in AMD that affect compatibility with the Univ. of Wisconsin GRB software. Or maybe check with Univ. of Wisconsin to make sure it has been tested with AMD.

There was no recommendation on the processor in the manual, just that is 12 core 2.4 GHz at 64 Bit.
I suggested the Intel unaware that AMD made one as well.
But it might be a good idea. Though, since it will work with the OS well it should be fine with the software.

Something could happen to the satellite on its way east. Plus you also might want to verify that the signal is still strong enough after it moves farther away from you.

It's not so much about that above as it's more I want to see it in ops mode for a period of time. There have been some issues with sudden data dropouts on the GRB side and no details about it.

On the open satellite project: Did you ask what machine/OS they did this on so you can replicate it? Simple question but I have to ask it.

I am reuploading the first file I posted in post #1 as I see somehow the file was compressed and is very difficult to read. See the link for the full size file.
Goes R data types
 

Attachments

  • GOES-R_Data_Types.jpg
    GOES-R_Data_Types.jpg
    132.5 KB · Views: 169
...
On the open satellite project: Did you ask what machine/OS they did this on so you can replicate it? Simple question but I have to ask it...
I compiled it for Windows 32-bit. There were Microsoft Visual Studio project files for most of it. But they didn't include the dependent libraries (there were several) which used different compilers.

It can also be compiled for Ubuntu 16.04. I don't know the minimum system requirements and don't have any email addresses of people to ask.
 
I recommend that anybody wants to view of the GOES R Data types graphic use the link above not the image, as again it's compressed so much that it's tough to read.

But they didn't include the dependent libraries (there were several) which used different compilers.

You mentioned before that the libraries needed libraries.
Nice....

don't have any email addresses of people to ask.

So I guess they expect you (the downloader) to figure it all out. This makes it very hard to build a program.
mann... You got a big task set out in front of you. Keep trying.
 
If the TBS8922 does generic stream QPSK, it would be great.

Back to what Ti suggested.Thanks for the info Ti.

The TBS8922 PCIE tuner has the capability of high speed data download. Their manual has little info besides installing it in the computer.
If this card can do the data stream it would be a great cheap option to the Novra S300 as N6BY mentioned.
But.. I will stress these things.
1. I have used these type of tuners and they do work but make shure that the driver is compatible with your OS. I know what they are saying on the web, I bought one (not a TBS) that said it would work on Win 2000 and it didn't. The driver wasn't made for it.

2. If you use the PCIE version of the tuner keep in mind that (from my experience) that you most likely need to install the DirectX 9 as the viewer is an automatic install with the software. If the video viewer is separate then you got some good software in my opinion.
The tuner running data or video will draw the computer's resources down. If you're trying to process the data at the same time on that computer it can be an issue. Exp. when we are talking data speeds as such.
That's also why I prefer the standalone version (S300) as it don't draw from the computer's resources. The processing is all in it's self. TBS's USB version I'm not shure if it's that way as well.
If this tuner will work for GRB it will work for NOAAPORT.

I'm going to put up a link to the data book for the current generation of GOES satellites (N-P) not GOES R.

GOES N Data book

Since there is interest in the GVAR data stream, this info will be useful + you can learn about all the neat things on the GOES spacecraft.

Section 2-5 P. 34 on the PDF viewer shows the exploded view of the main satellite parts.

Section 15-11 Talks about the GVAR format P. 176 on the PDF viewer.
 
...
So I guess they expect you (the downloader) to figure it all out. This makes it very hard to build a program.
mann... You got a big task set out in front of you. Keep trying.
I've now got the three Open Satellite programs properly compiled and running. I wasn't able to get any images because of low signal quality (its overcast, windy and raining now).

Open Satellite project also has a GRBDump program that you might be interested in.

Also, Amazon is now offering GOES-R datasets: GOES on AWS I think its free for one year of use, but I'm not positive.
 
  • Like
Reactions: hwboy4
I've now got the three Open Satellite programs properly compiled and running

What are the details that you used to make it happen.
Windows OS, libraries, compiler. Glad you got it to work finally. If I can replicate it I might try it.

Open Satellite project also has a GRBDump program that you might be interested in.

That's what Lucas Teske said, They were going to do something with GRB in the software.

Any suggestions for ready to use software to view the data?

You asked this question awhile back and I didn't answer it. So here goes.
I may have a free program that can view the data sets in mind but it will not autoload them.
I need to get my hands on an actual GOES 16 Netcdf-4 file to see if it will work. I know it works with the GOES images on NOAAPORT.

Amazon is now offering GOES-R datasets

That's fine for those who want run off the internet or can't put up a big enough antenna. But you will have to have a connection faster then 3MB speeds (min) as the GOES R data sets are very large.
Amazon uses the PDA server. You have to register to use it. Most likely you get registered through your Amazon account.
I rather put up an earth station for reception. Good info though.
 
What are the details that you used to make it happen.
Windows OS, libraries, compiler. Glad you got it to work finally. If I can replicate it I might try it....
I used Visual Studio 2017 running on my Windows 10 laptop. Visual Studio includes all the necessary compilers. I have used Visual Studio since the late 90's in my day job (I am the author of Natural Scene Designer for Windows and Mac, and DR Processor for Windows). If you haven't used Visual Studio before, I would definitely recommend starting with something easier than the Open Satellite project.

But if you don't want to modify one of the programs or step through the code in a debugger to see how it works then I would recommend using the pre-compiled versions.

Here is the link to the precompiled demodulator and decoder: Releases · opensatelliteproject/xritdemod · GitHub

And here is the link to 'goesdump': https://github.com/opensatelliteproject/goesdump/releases

You will need an AirSpy SDR and 'SpyServer' which you can download from airspy.com. Then you have to edit the configuration file to give it the URL of SpyServer which I would recommend running on another PC.

Also, I had to adjust the gain in the configuration file from 5 to 21 (I have a long cable run from my dish).

... Here is one of the first images I got out of the Open Satellite project. Its the southern hemisphere as seen from 135W.


1510977133-GOES 15-Southern hemisphere-FLSCLR.jpg
 
  • Like
Reactions: GOES 12 GVAR user
I used Visual Studio 2017 running on my Windows 10 laptop.
I would definitely recommend starting with something easier than the Open Satellite project.

I don't have Win. 10 here OR V.S. That's good advice.

link to the precompiled demodulator and decoder

Thanks for the info

You will need an AirSpy SDR and 'SpyServer

Don't have and not buying another SDR, good advice for anybody reading this though.

Your images are looking good. :thumbup :thumbup Isn't that neat as you see the sun light fading away, Wait till you see the moon in one of the images. ;)


Back to the way to open the netcdf files... for GRB
As I said you have to load them manually.
This software is free and can do much much more.
This is the native resolution of the GRB CMI images CH 1-3, These are from CONUS projections. And what I'd expect from a direct feed system.
I can zoom even more then what you see.
I noticed that NOAA just updated it to support the fixed grid projection that GOES R uses and..
ABI L2 CH1.jpg

CH1
ABI L2 CH2.jpg

CH2
ABI L2 CH3.jpg

CH3
I will upload more channels soon with a satellite loop.

This should also be able to open them off the Amazon PDA server. Every release is getting better.
I got my data of of NOAA's CLASS website. You have to register to use CLASS.

Software display kit: NOAA's Weather and Climate Toolkit
 
...
Amazon uses the PDA server. You have to register to use it. Most likely you get registered through your Amazon account.
I rather put up an earth station for reception. Good info though.
I feel the same way -- I'd rather build my own antenna system and software than download images from Amazon or other online sources.

I plan to help the guys working on the Open Satellite project.
 
  • Like
Reactions: hwboy4
ABI CH8, 10 Water vapor at different levels CONUS view. This is a very powerful forecasting tool.
ABI CH8 CONUS.jpg

CH8
ABI CH10 CONUS.jpg

CH10
Depending on what mode NOAA is running you can expect these 2 images (view) along with all the other channels 3 times every 5min.
That's the difference to LRIT and GRB.
 
CH8 Water vapor loop 11/15/2017 around 7 AM.
Won't run here on the forum so have to go to here:

WV Loop

Careful where you click as it may take you some where else, make shure you're downloading the file.
Click on download this file. When you get to the download location

WV Conus Loop 2017_11_15.gif
 
Last edited:
In continuing to stick with the topic of this forum, GVAR data has some similar characteristics.
in terms of channels:
GVAR Imager
CH um wavelength
1. 0.65
2. 3.9
3. 6.5
4. 10.7
6. 13.3

GRB LHCP
CH um wavelength
1. 0.65
7. 3.9
8. 6.185
10. 7.34
14. 11.2
15. 12.3
16. 13.3

Also addition to a couple of channels that are not on GOES N-P series the resolution is slightly higher on GRB.
In simple terms, It looks a little more "grainy" on the images above with GVAR.
 
N6BY,
I got an update through email that I thought you might like to hear. NOAA has the GOES S satellite on schedule to replace GOES West in the fall of next year. GOES S is from the GOES R series spacecraft lineup. So you should in about a year have the GRB signal near that area.

Also the file dropper link is no good anymore as it only lasts for 4 days. So the file of the animation is no longer there. No need to click on it.
Not much will happen here as GOES R will begin moving shortly to it new position of 75 degrees West. In early Dec. it should arrive at the new position and take over as GOES East. Providing weather and other issues don't hamper testing, I will hopefully have some new HDSDR pics up with it on location at 75W. Meanwhile I wait.
 
One year is a long time for me to wait for the new GOES West. So I made a septum feed for 1686.6 MHz last night, which I plan to try out before the move from 89.5 to 75W. The septum feed is quite large - approx. 18"x4.5"x4.5". Here is a link to a septum feed design calculator made by Lucas Teske: Septum Feed Calculator Will see how strong the GRB signal is here before and after the move.

On the OSP chat group ( OpenSatelliteProject ) they have reported major progress with live GRB reception and decoding. On Oct 23rd they captured their first live GRB image using an Ayecka DVB-S2 receiver with custom firmware supporting generic stream. The generic stream support is mostly working, but they are still getting the bugs out. Its not yet available for sale. Ayecka is going to charge more $$ for their receiver with the generic stream support.

So the OSP group is now working with a TBS6903 to keep the cost down. Under a special agreement with TBS, they were able to get some source code -- for 'TBSRecorder' I think. They'll need to rewrite it to stream UDP packets instead of recording to disk. I offered to help, but they are not allowed to share the TBS source code.

Anyway, I am quite impressed with the progress that OSP has made receiving GRB -- and making it very affordable. It sounds we will be able to stream GRB with a TBS card very soon!
 
Last edited:
.... So I made a septum feed for 1686.6 MHz last night, which I plan to try out before the move from 89.5 to 75W. ...
I tried the circular polarity septum feed last night for GRB, aimed at 89.5W, and unfortunately I didn't even see the slightest trace of a signal at 1686.6MHz. :( However, it did pick up a good signal at 1694.1, which is HRIT. So I know that the dish is correctly aimed at 89.5W.

For another test with a circular polarity L-Band signal, I moved my dish over to Inmarsat 4F3 and tuned to the Outernet frequency of 1539.8725 MHz. Was quite surprised to see Outernet coming in with an SNR of 35 db !!! That signal strength completely blows away my previous attempts a year ago, discussed in this thread: https://www.satelliteguys.us/xen/threads/experimenting-with-l-band-antennas.363533/ The best I did back then was about 21 dB with the same dish, but with a helical coil and also a patch feed.

So maybe my new septum feed is not properly tuned for 1686, or they shut off GRB ahead of the scheduled move on the 30th, or ??? Will try again in a few weeks.