GOES 16 GRB downlink vs GVAR

[GOES 12 GVAR user]

Well it 3 AM CST and I figure lets bring up a thread on the new GOES 16 GRB info or GVAR as well. First off I have talked on here about EMWIN several times so some of you may know me already. For those of you who don't I will give you some info. I receive 3 weather data downlinks here. 1. EMWIN N ,2. GVAR ,3. NOAAPORT. EMWIN on an 8' dish, GVAR on a 12' dish and NOAAPORT on a 10' dish. I see most discussions on EMWIN or a few on NOAAPORT so I wanted to bring up one on GVAR or GRB. Okay, GVAR this is a 4 MHZ wide BPSK signal transmitting the GOES East or GOES West processed sensor data for the imager and sounder. The data rate is 2.11 MBPS. The format is a variable gridded style that most don't have the software to plot or display. Since the signal comes down on the L band at 1685.7 Mhz the hardware is not at all common for reception for this data. What does all the above mean, expensive $$$ to build a ground station. Mine is a commercial one and it wasn't cheap though it could have been even more $$$ then what I payed. It also requires no less then a 10' dish for adequate reception 12' if you want a couple of DB of margin. I'm shure the neighborhood covenants allow this size of dish in the yard....No. Since NOAAPORT was converted to a DVBS-2 format the price for these stations have dropped in a major way. Here's why NOAAPORT used to be transmitted on the C BAND (and still is) with a BPSK modulation and it had 4 channels, so you had to have to have 4 receivers to get all of what it had to offer. I had an old style NOAAPORT station quoted and is was 100k + for one, Wayy out of most's price range. When DVBS-2 came along the hardware costs went way down and most could build one. Along with the hardware costs coming down software that was affordable came on the market to decode the frames coming from the DVBS-2 receiver. This software I currently use (weather message). As with the same for EMWIN, since the coming along with SDR's the hardware costs are plummeting for a ground station. Where am I going with all this. I am making the attempt to build a GRB downlink with commercial available hardware and software that is available on the internet to decode the GRB frames. The great thing about GRB is that it's a DVBS-2 format so the receiver is readily available on the market. Next the software is available to decode the frames and turn the packets into NETCDF-4 files. The not so great part is that since the ABI is transmitting much more data it requires a wider bandwidth, how wide you ask? how about 10 Mhz wide. With the wider band a reduction in transmitter power may be an issue. The current GOES 16 antenna sold is a 4.5 meter dish, Pretty much a 15' dish, you turn heads and cars slow down with a dish that size in your yard. I have been told that it possible to get it on a 12' dish here in the midwest and that is what I want to find out. Since the L band is within the DVBS-2 receiver range no downconverter is needed however the receiver must be mounted close to the antenna as the 1686.6 Mhz signal is very lossy on any type of reasonable transmission line but waveguide. Once the signal is at the receiver it turned into ethernet packets and sent to the processing computer where the software does the rest. This in terms is a very simple setup compared to what is currently. But if I can't get enough signal then the rest is mute. GRB is a circular polarized signal as there are bolth horizontal and vertical coming from the transmitter. I am interested only in the vertical side as it has the imager data I want. Perhaps receiving the one polarization will give a little extra to make the 12' dish work. Anyway that's my attempt. Some of the data above may be inaccurate as I'm doing all this from memory tonight. Through in your 2 cents and mabie I'll find out something that I didn't know before. Also I am including an image of the two moreso the GRB data info.

 

[GOES 12 GVAR user]

Some more info I want to add since I posted the last bit of info. The GRB bandwidth can be up to12 Mhz wide per polarization (as stated in the PUG Volume 4 GRB). It also stated that a 12.5' dish is assumed to be adequate for reception for most CONUS locations. I will include some more links on the GRB downlink documents.
Some reading material if you want more info yet.
Basic GRB signal downlink info
GOES-R Product Definition and Users' Guide (PUG) Volume 4 (GRB)
GOES R (16) is currently located at or very near 89.5 degrees West longitude, It will begin it drift to GOES east position Nov 30 and end Dec 11.
It's planned that GOES R will take over GOES east position, GOES 13 will go into storage.
More info: GOES-16 Drift and Transition Plan

 

[N6BY]

Thanks for all the information and links. I wanted to see how well my dish could receive GRB, but I got stuck on the meaning of "dual circular polarization".

I have successfully received circular polarized L-Band with my dish (Outernet and another service which tracks airplanes over the oceans).

Does 'dual circular polarization' mean that you have to receive both right and left hand polarization at the same time and combine the signals (in phase with each other), or does it mean that some of the data streams use LHCP and others use RHCP? I searched all over and could not find the answer to that.

Before GOES-16 moves from 89.5 to 72W, I would like to see how much of a signal I can get on my 8.5' and 10' dishes. Since I'm in California, it will be weaker for me when its at 72W. I do still have a setup to receive RHCP on L-Band. If that fails to receive GRB (which it probably will) , I will build a linear L-Band LNBF for GVAR using a tin can as a wave guide.

Below is a clip from the specification where they mention 'dual circular'.

 

[GOES 12 GVAR user]

Does 'dual circular polarization' mean that you have to receive both right and left hand polarization at the same time and combine the signals (in phase with each other), or does it mean that some of the data streams use LHCP and others use RHCP? I searched all over and could not find the answer to that.

Hi, Thanks for replying to my thread. To answer your question: The latter. There are 2 data streams coming from the GRB transmitter. See page 4 of the PUG GRB Vol 4 (table 3.0-2) to see the detail of the two. The reason they do this is to double the bandwidth of the transponder without actually increasing the bandwidth. It's pretty much the same C band does to their TV transmission, Some of the channels are on one polarization and some are on the other. The circular polarization pattern used keeps isolation between the two data streams. What is required to receive BOLTH is 2 DVBS-2 receivers (stated by the software). Unlike the TV satellites that you can select circular polarization in the settings, this has 2 separate streams that are ingested at the same time. A simple coffee can feed will work with this as all you need is 2 probes 90 degrees to each other. See image. That is why I am only interested in the LHCP stream as it has the data I want + I only need 1 DVBS-2 receiver to get it. It also lets me use hardware (feed horns) that I already have for 1690 Mhz. Your 10' dish is likely your best bet for GRB as 8.5 will fall short, but you can give it a try. Now a 10' dish on GVAR will work but it's not DVBS-2 format as GRB. You would most likely have to point at GOES West for the GVAR as GOES East could be harder to get out there and a 10' is on the edge to get the GVAR signal (if your in an area of a higher noise floor all bets are off). Make shure everything is peaked. Your 8.5 dish will work fine for HRIT on GOES 16 and that is at 1694.1 Details on this stream can be found here EMWIN Stakeholder Meeting Presentation. Good luck and please keep me informed. I am including some images of the 2 polarizations and of the feed that I plan use to receive GRB (single probe version).

 

[N6BY]

I'm glad to hear that GRB does not require both polarities at the same time. One year ago I tested two different types of L-band feeds on my Birdview. I got equal performance with either a 3-turn helical antenna or a patch antenna. Here is a link to that thread: Experimenting with L-Band antennas

Now I just need to figure out where in my cluttered garage I stashed those antennas...

 

[GOES 12 GVAR user]

You have a very good info there, nice. After reading and thinking about the polarization issue I have found out that I am right and wrong. I am right about that you don't need to receive both streams at the same time,in other words you can receive RHCP OR LHCP. I am wrong about my horn antenna as I have it designed to receive only linear polarization. Again see 1690 cir pol horn.jpg. You would have to use both of the probes to receive one of the streams. If only one were used it would be 3dB down every 180 degrees.The solution should be simple to fix it as all that is needed is to combine the two probes in to one coax. You can purchase a commercial combiner or make your own. The DVBS-2 receiver can select the type of circular polarization needed. (What I state here I am not shure it would work, To receive the other polarization would require 2 more probes combined as the first. Again not shure about this) And after the signals are combined use a preamp before it get to the receiver. The helical antenna is not the best match for a dish ( My thought) as it may not illuminate the reflector very well. The patch antenna on the other hand will work as my commercial GVAR feed (linear) uses this type of antenna, works pretty good. I was told though the horn has the best match for dish use and I would agree. There are however many details that can very and a big one is the F/D and size of the dish as it will have a great affect on gain. The better your feed illuminates the reflector without overspill/reception excess noise the higher efficiency = higher gain. Providing that the gain don't go into sidelobes. I have never dealt with circular polarized antennas so this is a new one on my part. But I get the idea of how they work.Though I had to think about it.

 

[GOES 12 GVAR user]

Have been doing some research lately about this. My horn idea will require a 90 degree hybrid splitter/combiner (Mini circuits ZX10Q-2-25-S) to work. So I am on the right path. Also during my research I found that Lucas Teske has made that same thing and put his results on twitter where he does have some of the GRB data on there. He must have got something to work as he has received the GRB data since he built that but there is no detail otherwise.He did mention that the design didn't get the 27 dB of isolation required between the polarization's but only 15 dB. I am also including some patch antennas that may work for this however they are not tuned for the frequencies used here. They are set up more for the GPS frequencies. They are: digi key part #'s RHCP 931-1109-ND LHCP 931-1362-ND These have no active components and are very affordable. All they need is a ground plane of some sort and a short coax and connector.

 

[GOES 12 GVAR user]

I have a 10' dish up for testing. I should soon have some pics of the downlink spectrum of GOES 16 here soon, In next couple weeks. From info I have learned on the internet a 10' dish (may) be a winner for this. I will be using a waveguide horn (diagram above) and LNA for the setup. I'll use HDSDR for satellite spotting, early testing but go to my spectrum analyzer for the final details and S/N info.

 

[GOES 12 GVAR user]

Well, It has been an interesting last 5 days trying to acquire the GOES satellites. It put up quite a fit trying to look a the 1.6 Ghz signal directly. I have under estimated how much gain is on the system with the downconverter. Looking at it with only the preamp of 27dB wasn't enough to see the signal on the analyzer. It took the extra 20 dB to do the job and it took me setting up my 4' dish to find the real issue. Anyhow here are the details.
My system used:
10' .38-.39 F/d prime focus mesh dish (standard C band dish)
Wilmanco block downconverter
DEMI UTRA Phemp LNA
Home made feed horn, 2 styles used in the test.
Bolth horns have vertical polarization.
W/ Choke ring in the final test.
System has a min of 47 Db of gain.

Let's start off with the current GOES East satellite for apples to apples comparison. In the first picture it shows GOES 13 seen on my HP141T analyzer. I label the signals of interest. These signals were seen on a 12' prime focus dish.

The next pictures show the dish and first feed I was testing to get everything working as I wanted it. Tin foil is used to extend the horn to get it closer to the focus without changing the dish's hardware. This wouldn't be left this way. It did work pretty good for the beginning of the tests I wanted to try.

These pictures show what the final horn will resemble as this is what I want. The waveguide is long so unwanted modes will be suppressed, It has a choke ring as this helps the get better illumination of the dish's reflector. It has the 2 probes already installed so all that I will need to add is the 90 degree combiner. This feed did well mabie a couple a dB better then first. I feel with more adjustment I can get 3 or so more dB out of it which is a big deal.

Now the best part: The 3 pictures show GOES R spectrum and the signals of interest.

This is GRB, it's the replacement to GVAR and as you compare the 2 pictures you can see what I said earlier about the bandwidth of it. The span you see is 12 Mhz wide. GVAR is only about 4 Mhz wide. S/N is around 12.2 dBm

The third pic is of the replacement to EMWIN called HRIT. This is the signal of EMWIN and LRIT combined. S/N is around 13.9 dBm

All these signals were seen on the 70 Mhz band as the downconverter was needed to achieve the gain. The final setup will have to have a hotter 1.6 Ghz preamp(s) as that is required to feed directly in the DVBS-2 receiver. No data was got from this off the streams as I don't have the receive hardware.
Testing will continue but the weather here is not going to be the best for that as of now. Cold and windy don't work to test satellite dish performance.

 

[N6BY]

Nice work and write-up about it! On C and KU bands, DVB-S2 reception usually requires a S/N of at least 10 dB. So your results are very promising. But don't you need circular polarization for optimum GRB reception? I have an Outernet circular patch antenna and LNA that I have been meaning to hook up to one of my dishes here. The Outernet frequency is probably close enough to GRB. I have read that GPS antennas are also a possibility, but you have to remove the notch filter.

I need to try the above before November 30 when they move the satellite!

 

[GOES 12 GVAR user]

Thanks.

But don't you need circular polarization for optimum GRB reception?

Yes, you need circular polarization for GRB. As mentioned before I don't have the 90 degree combiner to do that yet. I was just getting some results to see what I have to start with. HRIT is linear polarized unlike GRB that I'm aware of.

 

[N6BY]

Thanks.

Yes, you need circular polarization for GRB. As mentioned before I don't have the 90 degree combiner to do that yet. I was just getting some results to see what I have to start with. HRIT is linear polarized unlike GRB that I'm aware of.

Couldn't you just make one of the leads connecting the dipoles 1/4 wavelength longer than the other?

 

[GOES 12 GVAR user]

Yes that can be done but... The calculation and cutting of the transmission line is very important as a small error will cause loss and other degradation to the signal.
I have no problem doing that a VHF/UHF frequencies but L band is much more delicate. Considering that I'm squeezing about all I can out of this dish, that type of error would defeat the purpose. For the modest cost $29 minicircuits makes what I need matched and peaked and that can be installed as a block in the system.
With that block all that I will need to do is cut bolth feed lines the same length, the way they connect to the combiner determines the circular polarization received.
If this works the feed that will supersede this one will not need that combiner as the feed itself will do that job and can receive bolth circular polarization's at the same time. But that's another issue.
My big concern is the gain required to make a usable 1.6 Ghz signal to the receiver.
To give everybody an idea of the problem faced, here are some numbers. Running some tests last night I determined a range of my actual signal level coming from the feed. The picture of HRIT above shows a signal level of around -48 dBm. As I stated before with the gain numbers from my system is in the area of 47 dB.
27 db from the phemt preamp and 20 dB min from the down converter. Adding the 47 dB to the level of -48 puts it around -94 dBm. But the range could be somewhere in the area -81 to -107 dBm. Noise floor is near -60 dBm
Max signal is 21 dB below noise at the feed, after that the hole gets deeper in noise.
---------
NOTE I want to make it clear that I am not a tech and do this as a hobby. I have taught myself the best way to figure these things. But it may not be the right way. Wanted to make shure I got that out there.
---------
I'm seeing the reason such a large dish is recommend. I have some other info in the next post on a past subject that I want to commit on.

 

[GOES 12 GVAR user]

Also during my research I found that Lucas Teske has made that same thing and put his results on twitter where he does have some of the GRB data on there.

From what I have learned doing my own tests and seen on his twitter I think he is receiving HRIT. The frequency he shows is for HRIT not GRB though he does have CMI (cloud moisture imagery) on there it looks like LRIT not the full resolution GRB image. For the antenna he shows , I think there no way he is getting GRB with an antenna that size.

 

[N6BY]

...
27 db from the phemt preamp and 20 dB min from the down converter...

DVB-S2 receivers expect an input frequency between 950 MHz and 2150 MHz, and the GRB frequency of 1686 MHz is well within that range. So there should be no need for a down converter.

I'm too tired now, but I will try it tomorrow with my circular patch antenna.

 

[N6BY]

GOES 12 GVAR, please disregard part of my message above. I reread your post #1 and you already acknowledged in it that there is no down converter needed. (Its just part of your calculation to estimate the signal strength from your test). But I'm not sure I understand the calculation.

Also, while the GRB data is sent using DVB-S2 modulation, the data payload is not MPEG video as a typical receiver would expect. Its not clear to me how you would get the receiver to send the raw data over ethernet as described in post #1 ?

 

[GOES 12 GVAR user]

Yep, It was part of the calculation to make it visible on my analyzer.
The units used to measure radio frequency amplitude is dBmecibel to milliwatt.
See wikipedia for more details.
The term dB is added or subtracted from a known level.
So if I start with a -60 dBm noise floor add 27 dB (units) to that number I get -87 dBm. Since were dealing in negitive numbers adding to it makes the amplitude lower.
Likewise if I subtract 27 dB from -60 Dbm I get -33 dBm, this is gain.
It seems backwards but this is how I do it. There may be a average to the gain numbers, but I add or subtract it from my known level.
If we get to the positive levels dBm for receive the process reverses.. we don't need to talk much about that involving satellite dishes for antennas. I explain this to anybody reading it as the information it provides is very important related to the tests or understanding actual signal levels, instead of a meter or display.

I have learned several things today with these next tests I did.
1. I found out that my analyzer is not very good at seeing weak signals in the GHz range. It still doesn't see anything with the dish peaked up. Needs the down converter.
2. I found out that my SDR is much more sensitive than I thought. This was brought to my attn by a (real) radio technician. Who is still unhappy with me.
3. Stupid flies are still biting in Oct.

Now the above numbers will come in play with this next bout of information. The SDR made my day.
Below is the SDR display with no signal there. I do this to show the other garbage that is not related to the signals that SDR sees. Also to show the noise floor of the SDR.

The SDR noise floor is shown about -143 dBm. This is 83 dB below my analyzer's noise floor at those frequencies.
Very well more representative of a receiver.
The noise floor here is lower then my calculations so my signal should be above the noise here and....

GRB is coming through great even though it's 3 dB down every 90 degrees due the to polarization mismatch.
So far GRB looks as strong here as GVAR. That's a very good thing. Also note the the SDR can't see the entire GRB spectrum as it's too wide. It shows 1685.3 - 1687.7 Mhz, 2.5 Mhz wide. That's only a fraction of GRB's signal. With this info I shall continue to build the circular polarized horn.
I also show HRIT on the SDR, it's very strong as the antenna is waayy oversized by 2.5 times or more.

The HRIT signal is lower then what my calculations suggested (means it is weaker then I thought) as it shows a -116.9 dBm level with the preamp.
So again doing some numbers -116.9 add 27 dB = 143.9 dBm signal coming out of the wave guide. That's a whopping 0.0000000000000041 of a milliwatt. Providing that HDSDR is close with it's values.
Last, the line seen going across the signal is a file header. I stated in another post on here [smallest EMWIN antenna] if you can see this you have and can decode HRIT.
I use HRIT as a reference as it's the strongest data signal there.
And finally to answer your question N6BY.

Also, while the GRB data is sent using DVB-S2 modulation, the data payload is not MPEG video as a typical receiver would expect. Its not clear to me how you would get the receiver to send the raw data over ethernet as described in post #1 ?

DVBS allows not only MEPG video but also IP packet data streams. That's what NOAAPORT does and would expect GRB to do the same.

 

[GOES 12 GVAR user]

The units used to measure radio frequency amplitude is dBmecibel to milliwatt.

Should read..The units used to measure radio frequency amplitude is dBm: Decibel to milliwatt.
Wrong character put a smiley there.

 

[hwboy4]

There is an open source effort (OpenSatelliteProject) underway to produce software which will handle GRB from GOES-16. Currently the software is working with LRIT and HRIT. Beta firmware for a commercial receiver is being tested in addition to a DVB-S2 board for a PC.

I've been running the LRIT stuff for the past year under Linux using a grid antenna and an AirSpy and will start HRIT when GOES-16 moves to replace GOES-E.

Lucas Teske is the main developer and he has a RocketChat board at OpenSatelliteProject
Many good people there working on GRB but also currently receiving LRIT/HRIT.

All code is on GitHub.

Stop by, check it out, and best of luck with your efforts. I'm waiting until a definitive word on what dish size is required before taking the GRB plunge.

Chris N2CR

 

[N6BY]

There is an open source effort (OpenSatelliteProject) underway to produce software which will handle GRB from GOES-16. Currently the software is working with LRIT and HRIT. Beta firmware for a commercial receiver is being tested in addition to a DVB-S2 board for a PC....

Great stuff, thanks. I registered at RocketChat and searched the OpenSatellite project for the DVB-S2 firmware without finding it. I'm curious which receiver and PC board is the firmware designed for? i have a TBS6983 board, and a few stand-alone DVB-S2 receivers.

... Also, some connectors that I need to test my patch antenna on my satellite dish are supposed to arrive today. Will see how strong the signal is from 89.5W.