GOES 16 GRB downlink vs GVAR

N6BY,
Sounds good. I got some paper tube on order myself to make the prototype I want to make. It will be slow shipping though to the store.

But I will have to put my 10 foot Unimesh back up.

I had a felling that you would have to, 8' here in the US even if it could cut it would be so close that any signal change in path or other would start causing large data loss. From the impression of what I have read that if the receiver loses sync with the file is decoding it discards that file.
A couple of dB margin is well worth all this effort and time.

I have already asked Paul W1GHZ that exact question and his answer was:
Hi Tim

You should be able to receive both polarization's simultaneously, with about 20 dB of isolation between them.
 
N6BY,
Sounds good. I got some paper tube on order myself to make the prototype I want to make. It will be slow shipping though to the store.
I have read that a cylindrical waveguide will more evenly illuminate the dish (vs. a square one). Are you basing your design on N2UO's? http://ok1dfc.com/EME/technic/septum/N2UO%20opt.pdf

... Only 11 more days until GOES-16 is supposed to be operational at its new location. My septum feed is close to done, but I still lack a DVB-S2 receiver that can send a generic stream over IP. If Lucas Teske (Open Satellite Project) does not have any software that works with a TBS card, then it will be time for me to make an attempt at it.
 
Are you basing your design on N2UO's?

Yes, and on the primary design of the Super VE4MA. But mine will not have the taper portion as it's not needed. I will use the choke ring design such as the VE4MA feed. Like this:
13cm feed choke ring.jpg

Hope ON7UN don't mind me using his pic.
And yes a square feed does create a different pattern that's not optimum for a round dish that I researched. But if used with the choke ring it will help that and the choke ring is needed for a dish F/D.4 or in that area.
See here: OK1KIR Gallery: 1296 MHz, Our New System, 1296 MHz feed for CP
 
I like the round design. I'm wondering how he made those inverted cone shaped probes.

... I have been doing lots of research on generic streams and BB Headers. I was able to record a generic stream last night with my TBS6983 (not GRB, but similar).

I am now confident that I can write a program for my TBS card to send a generic stream via UDP, record it to a file, or process it within the program. This will be for Windows only. I am pretty sure that its not currently possible with Linux due to device driver issues.
 
Last edited:
I'm wondering how he made those inverted cone shaped probes.

Most likely on a lathe, That should not be needed on our feed as those cones most likely help the matching over wide bandwidth. In our case bandwidth is not an big issue here as we are at 1 frequency.
The probe dia. (wider) allows a wider band feed.

I was able to record a generic stream last night with my TBS6983 (not GRB, but similar).

Well you should be able to begin Thu. as beginning 16:15 UTC the GRB is to be back on line. GOES 16 is about to become GOES East.
So I guess no news for OSP on their work.

GOES 16 Status

Everyone I have talked to says that Windows will not like running this type of process and may crash.
Since you messed with Windows you should know that anyway. Great if Windows will work with it, with no issues.
On the latter in Linux I.D.K.
I have been doing some research on DVB-S2 Generic Stream (GSE) receivers/demodulators here during the meantime. I have found several rackmounts that would do the job just waiting on quotes. And ya, I figure that the prices are far higher then the ones from TBS.
Even wanting a quote from Ayecka with their updated version.
 
...
So I guess no news for OSP on their work....
Correct, no news from OSP. The developer (Lucas Teske) has not been online in several days.
...Everyone I have talked to says that Windows will not like running this type of process and may crash.
Since you messed with Windows you should know that anyway. Great if Windows will work with it, with no issues.
On the latter in Linux I.D.K...
Not sure what they mean by "Windows will not like running this type of process and may crash". Makes no sense to me. I verified last night that it does work on a Windows PC with a generic stream. I have working source code for my TBS6983 and its not from TBS nor the Open Satellite Project. I just have to add UDP streaming which is not that difficult. Or I may decide to merge the DVB generic stream reading code into the same program that decodes GRB -- for maximum efficiency
...I have been doing some research on DVB-S2 Generic Stream (GSE) receivers/demodulators here during the meantime. I have found several rackmounts that would do the job just waiting on quotes. And ya, I figure that the prices are far higher then the ones from TBS.
Even wanting a quote from Ayecka with their updated version.
I'll be interested in knowing their price. When I get my code working for TBS cards, I'll sell it to you for 1/2 of the lowest quote you get from Ayecka and the others. :)
 
Last edited:
It's been a week so I thought I would give an update on my progress trying to build a GRB downlink station.

I completed construction of the septum feed. It tested much better than my previous attempt, but the signal is not quite strong enough on my 8.5' BirdView. I am in the process of setting it up on the 10' Unimesh. I built the septum mount today and now have what I need to mount it to the Unimesh at the focal point. I hope it will pull in 10 dB of SNR.

On the software side, I have made a lot of progress with the program that will stream from my TBS6983. Now I just need a GRB signal to test it with.

Brett
 
  • Like
Reactions: hwboy4
Slow here. The raw materials are in for the building of the prototype septum feed so next month I will be able to start building that.
I did another signal test Thur. The signal level is the same as previous tests with the satellite at 85W. So all good there.
Brett,
I again figured you would have to go to the 10' dish as 8.5' (mabie if you're down along the equator) is just not enough in Cal. You should be able to get similar results of what I have got.


I have made a lot of progress with the program that will stream from my TBS6983.

Sounds good. Be interested on how that comes out.

I got at least one quote from a receiver Mfgr. so far.
MainData
Optisat S2/S receiver, BBF to UDP

This is a rackmount version, quote 1,700, 1,900 USD.
Much less than the Quorum receiver.

He also has a version that is:
PCIe board, which is able to receive DVB-S2 generic stream and get BBF frames.
But:
I would have to write my own application to output via UDP.
Quote of 1,300 USD.

As expected the cost is far more then the TBS units but these are also commercial grade.

Not sure what they mean by "Windows will not like running this type of process and may crash".

The software written for this is usually done for Linux or in Java script and run on a Linux platform. From what I have been told Windows has a hard time managing the large amount or resources required to process and handle these datasets. It also probably stems from that Linux has a lot of open source code. Now I was told this when GVAR was still the used dataset. OS's have changed and mabie the new Windows OS's do a better job at running high demand applications now. Also these ingest stations were designed on servers for that use.
I'm not saying that it can't do it, I'm just passing along what I have been told and why. If it will work, use it.
But I WILL make this point. The computer doing GRB should be dedicated for THIS and this only.

Found this in my searches.
Here is a link to the CCSDS standard. This is the format the data files are sent inside of the DVB-S2 stream or the base band format I believe.
Link: CCSDS Standard
 
Slow here. The raw materials are in for the building of the prototype septum feed so next month I will be able to start building that....
I look forward to hearing how well your cylindrical septum feed works. I'm curious to know what diameter of tubing you are using? Mine is square and a photo of it is at the bottom of the following thread: Experimenting with L-Band antennas

My biggest mistake was that I ordered 5 inch wide (O.D.), 1/4" thick aluminum square tubing. It is heavy! My septum feed is 12 pounds. Hopefully the dish struts won't sag with all that weight. I could have ordered much thinner tubing. But I needed the inside to be approximately 4.5 inches according to the septum calculator.
...
I got at least one quote from a receiver Mfgr. so far.

Optisat S2/S receiver, BBF to UDP

This is a rackmount version, quote 1,700, 1,900 USD.
Much less than the Quorum receiver.

He also has a version that is:
PCIe board, which is able to receive DVB-S2 generic stream and get BBF frames.
But: I would have to write my own application to output via UDP.
Quote of 1,300 USD.
...
That is a lot for those receivers. But some of their customers (govt., industry, etc) probably wouldn't blink at those prices.

BTW, I was kidding when I said you could have my TBS generic streamer program for 1/2 of the lowest bid. It will be freeware.

...
Found this in my searches.
Here is a link to the CCSDS standard. This is the format the data files are sent inside of the DVB-S2 stream or the base band format I believe.
Link: CCSDS Standard
@weather01089 from the Open Satellite Project was nice enough to send me a 400 MB file containing the GRB stream today. I examined it and found the data to be as expected. The frames are 7274 bytes each (including the 10 byte BB header). Here is an excerpt from the CCSDS standard that you linked:

bbheader.png


Using the BBAnalyzer program, the 400 MB file had one synchronization near the beginning and no CRC errors! So 1 faulty frame out of nearly 55,000. That's encouraging!

With the sample file I got today, I will be able to complete development of the TBS program to stream GRB, even if I can't get a stong enough signal here in California.
 
Well, my septum feed prototype uses cardboard or paper so it's size can be changed easily. You got me sold on that Brett, It was a good idea. The base tube used will be a 6" light duty concrete form. Because it's made of paperboard I can cut a section out of it and make the Dia. smaller. A bit of Elmer's, hot glue and paper will go a long way. The same goes for the choke ring but I will need to add to that tube as the Dia. will be too small from the store.
The form I had to special order is only 12' long so only needing 2' of it I will have plenty left if I error. With a price around $17 it was a no brainier decision to get.
Ya, Your waveguide is a bit heavy. Looks good and, Well, you don't have to worry about much going to damage it.
If you're lucky enough to have tube to be the size needed otherwise....
That's why most use sheeting and form it to the dimensions needed. Dish struts are pretty strong. It's the way you mount it in those struts that will be important.

I was kidding when I said you could have my TBS generic streamer program for 1/2 of the lowest bid. It will be freeware.

I appreciate that AND all the time you and others have spent on this subject. :biggrin

The frames are 7274 bytes each (including the 10 byte BB header).

1 byte = 8 bits therefore 7274 = 58192 the BBFRAME size stated in the pug GRB Vol 4. Pg. 10.
That sounds great.
By what you're telling me that you should have .nc files somewhere on your computer.
With a 10' dish you should be able to get the GRB stream there. I'm seeing no reason why not.
Looking forward to hear about the final version of your streamer.
 
I got my septum feed mounted on the 10 foot dish and aimed at GOES-East. GRB comes in about 11-12 dB with an SDR. But my TBS card sees nothing at 1686. It does, however, see strong signals at around 1972 and 2125MHz , which I think are terrestrial.

Going to add another LNA inline and see if it brings the signal level high enough for my TBS6983.
 
But my TBS card sees nothing at 1686.

Not sure why it don't see something. Even if you were on the threshold you would see the receiver try to lock. Though you shouldn't be.
I would recommend that you try with just one of the polarization's at a time. I assume you're using both.
Make sure that the settings are correct, Again seems silly but make sure.
Otherwise I recommend asking OSP since they have used these cards or similar.

I doubt that another LNA will solve the issue.
Almost sounds like a polarization issue, or setting issue.

BTW, All the tests I have done here are with one polarization, I wouldn't know what it would be on both as I have never tried that.
 
I had only a 20 dB LNA. I added a 32 dB LNA in series and was able to see a 2.3 dB signal in EBS Pro, but not near enough to get a good lock.

I am just using one side of the septum feed, so just one polarization.

I ordered a 64 dB LNA from Hong Kong, so until it arrives I am out of things to try.
 
I just got my program to send a generic stream from a TBS card to GRBDump working!

The stream I used for testing was from a satellite sending a generic stream (but it wasn't GOES-16 since I can't get it yet). The streaming program runs on Windows only, but you can stream to another computer running the Linux version of GRBDump.

I still need to add more controls to the program's interface and test it to see if it can recover gracefully after loss of signal.

After that, I will need to send it to @weather01089 for testing with the actual GOES-16 GRB stream. (He is the only one I know that has a dish which can receive GRB).
 
  • Like
Reactions: Titanium
@weather01089 has tested my new 'GRBStreamer' program and it works well enough to produce some images in GRBDump. (One is below). :) But about 1 in 200,000 Baseband frames it receives from the TBS card have synchronization errors -- before the data gets to my program.

So I tested TBS Recorder to see if it had the same problem and it does. :( I used it to record a generic stream to a file, and then analyzed it with a program called BBAnalyzer. I saw the same kind of errors in the file as GRBStreamer encountered.

The bottom line is that there is nothing I can do to fix the problem, except hope that some day TBS fixes it or another vendor releases an affordable card that works for this purpose.
GRBStreamer.jpg


conus.png
 
Thanks Brett,
1 in 200,000 don't seem to be bad at all in my opinion. You might want to bring that up to TBS's attn. about that issue. That's a VIS CONUS image there ,nice. How many files did it generate in the 200,000 frames?

Well, over the holiday weekend I finally built the prototype septum feed that I wanted to build.
The foil is on the outside of the cardboard tube for the waveguide. It does have a bit of an egg shape at the probes end. That was a result of the install method of the septum. The entire feed is about 2' long.
I haven't tested it and don't expect to anytime real soon as winter is in full swing here in the Midwest.
I'm not dragging my gear out in single digit temps and snow.
Here are the pics of it. This is the prototype not the final version.
Choke ring.jpg Inside septum.jpg Looking down at probes.jpg Outside of feed.jpg Back of Feed.jpg

Along the same line GOES 13 has been put is storage mode so GVAR on Goes East is done, Unless something happens to GOES 16 it's only GRB now. GVAR is sill available on Goes West.
 
  • Like
Reactions: N6BY
I don't know how many files @weather01089 was able to get out of GRBStreamer and GRBDump. All he said was that it worked OK, but did not get any of the full disk images -- probably because the chances of a single error occurring in a large file are much higher.

I didn't report it to TBS because I think its unlikely that they would issue a fix.

Also, @weather01089 has ordered a DD Cine S2 V7A card to test. They reportedly have issued new firmware which will send a generic stream. Another guy that already has one of those cards is going to send me a stream capture file to analyze for errors. We should know in a week or two if the DD Cine card works for our purpose.

Meanwhile here I am trying various combinations of amplifiers and filters to try to get a decent signal. The 64 dB LNA and the Lorch 1675 filter combo didn't work. But it could be a cable or connection problem and I have not given up yet.

Edit: Forgot to ask about your prototype septum. Does the septum make electrical contact with the tube body? According to some of the articles I read, it needs to make contact with the tube all along the top and bottom.
 
I didn't report it to TBS because I think its unlikely that they would issue a fix.

TBS has other cards out there now. I wonder if they have the same problem.

Wondering... Did OSP have the same problem with their TBS6903 card they were using?

I will be interested to hear of the next try with the DD Cine card. I have herd no more from the DVB-S2 receiver Mfgs here so nothing new with that.

Keep working on the RF side, You should be able to get the GRB signal there with that 10' dish.

Does the septum make electrical contact with the tube body? According to some of the articles I read, it needs to make contact with the tube all along the top and bottom.

Yes, when putting a Volt/ohm meter to the feed it shows a short (continuity) at all points. If you look at the back of the feed picture you will see that I used the shorted end to make contact to all other surfaces. I also used plastic 2-56 screws for the connectors so that wont detune the feed.
And by the way it's waveguide is 5" dia. or very near that. Choke ring is 13" dia.

Brett, will your GRB streamer have a readout of the frames it's processed and errors?
 
Interesting, wonder if it's in the firmware of the receiver cards.

GRBStreamer was updated several days ago to report both valid packets and errors encountered.

Great. Thanks.

I'm going to keep looking as well on this end for hardware, There are several choices but not cheap ones at that.
Although the Ayecka SR1 receiver isn't priced real bad in terms. Ebay has it at $570, I payed more for that for EMWIN N. Although I would prefer to purchase it through the mfgr not Ebay. Though, I sill haven't got a quote from them.