GOES 16 GRB downlink vs GVAR

[N6BY]

Here is a long running thread for big dish hunters: Kijiji, eBay, and Craigslist dishes for you to take a look at

Another name for that thread might be "Big Dish Adoption Facilitators"

 

[Micanopy Mike]

RE: "Big Dish Locator" Thanks for posting that!!
On another topic (SDRs for GRB downlinking), a friend is the director of a team of RF design engineers who are tasked with incorporating bleeding edge technology SDR designs for receiving and processing the RF emissions generated by the latest new product developments of MRI medical imaging machines. The company he works for commands around 45% of the global MRI market. The team is working on completing a project to incorporate an SDR with MIND-BLOWING performance in their new MRI machines. This SDR is a product of XILINX Corporation and a .pdf paper describing it can be downloaded here: http://tinyurl.com/ydelvpp2 The full URL is:
https://www.xilinx.com/support/docu...ta_converter/v2_0/pg269-rf-data-converter.pdf

This SDR is overkill for receiving and processing the GRB signal but it shows where the technology is headed and much less expensive equivalents will be available within the next year or two. Mike Baker, Micanopy, FL

 

[GOES 12 GVAR user]

Oh-- when a PM "conversation" response is made, where does it show up? I presume that some sort of notice shows up that a PM "conversation" response has been sent that some sort of notice is made?
It will show up in the little envelope at the top of the page by your user name. Move your mouse over it and a popup box will show the conversation.

There you go.

 

[N6BY]

I am happy to report that significant progress has been made. In a nutshell, I now have a relatively low cost GOES-EAST GRB Earth Station which is receiving and processing both RHCP and LHCP simultaneously!

1) I have made some improvements to Lucas Teske's 'GRBDump' to make it much more efficient. If anyone is interested, please let me know and I will send you a download link which includes my modified source code and the compiled program for Windows. It needs to be run in 64-bit mode on a system with 8 GB of RAM.

2) I built a computer to use as my new software development workstation. With the improvements to GRBDump the new computer is able to process all the incoming data from both LHCP and RHCP and keep up with the streams. It has a liquid cooled Intel i7-8700 CPU, 16 GB of 3000 MHz DDR4 RAM, a 500 GB Solid State Drive, and a 2 TB hybrid drive. As a test, I ran it for 20 hours straight without a problem.

3) I ordered a pair of cheap bias tees from China to power the NooElec Sawbird's. I power them from the computer's USB ports. They also eliminate the need for a separate DC blocker since they have a capacitor which serves the purpose. Here is the link of where I purchased them: New 10MHz-3000MHz 3GHz Radio Frequency Transformer Bias Feeder Bias Tee 0.4DB 699998805737 | eBay

I am now in the process of writing a program to combine the red, blue, and NIR (near-infrared) channels into natural color images. Also there is a lot additional optimization that can be done to GRBDump. But I want to finish the color image generation program first.

... In case some of you haven't heard, there is some bad news for GOES-17 (the new satellite intended to replace the current GOES-WEST). 13 of the 16 bands don't work at night (bands 4 through 16). Those are the infrared bands. The problem is a failure of the cooling system when the satellite is facing the sun (night time here). But fortunately the 3 visible bands don't need cooling and are unaffected. All the other sensors are OK, including the lightning mapper and sun imager.

 

[GOES 12 GVAR user]

Hi all,
I haven't been on too much lately due to getting things up and going here. I finally got my GRB station going with the help of Brett and others. We found out the TBS5927 didn't work for this so by luck I was able to get a TBS6903 in an open box for the exchange of 5927.
The results have been major as my station is now working. I will make this short as I will post more later.
Brett sent me a feed to test and found that the feed wasn't the issue so that's when we suspected the receiver.
I had to build a computer from an almost nonworking one I had laying around here, luck would have it it had Windows 7 on it.
So once I got the computer up and going and got the new receiver card installed I had data. It took a seconded time as the GRB Dump program was not working well with my computer's networking.
Thanks to Brett he (as posted above) redid parts of the Dump program and it runs really good now. He made for me a 32 bit version as my system don't have a 64 bit OS installed.
So here's a screenshot of the computer after more then 3 hours ingesting one stream of the GRB data.

I started ingesting shortly after 6:00 PM.
This is also using my feed that posted on this forum and pictured several pages back.
I will include a link for some of the CONUS images I got as there are a lot.There are two of each there and one is the uncorrected and the other is the corrected version. They are on my google sites page under 4 different directories.
Here for the images
I have also uploaded files that no one else has here and these are the GRB info packets that are sent on the stream every 5 min.
These have good info on the satellite systems in them.
They can be found here
That's it for now, stay tuned as there is much more to come as the software side of things are getting very interesting.

 

[KWX]

Hi all,
I haven't been on too much lately due to getting things up and going here. I finally got my GRB station going with the help of Brett and others. We found out the TBS5927 didn't work for this so by luck I was able to get a TBS6903 in an open box for the exchange of 5927.
The results have been major as my station is now working. I will make this short as I will post more later.
Brett sent me a feed to test and found that the feed wasn't the issue so that's when we suspected the receiver.
I had to build a computer from an almost nonworking one I had laying around here, luck would have it it had Windows 7 on it.
So once I got the computer up and going and got the new receiver card installed I had data. It took a seconded time as the GRB Dump program was not working well with my computer's networking.
Thanks to Brett he (as posted above) redid parts of the Dump program and it runs really good now. He made for me a 32 bit version as my system don't have a 64 bit OS installed.
So here's a screenshot of the computer after more then 3 hours ingesting one stream of the GRB data.
View attachment 133428
I started ingesting shortly after 6:00 PM.
This is also using my feed that posted on this forum and pictured several pages back.
I will include a link for some of the CONUS images I got as there are a lot.There are two of each there and one is the uncorrected and the other is the corrected version. They are on my google sites page under 4 different directories.
Here for the images
I have also uploaded files that no one else has here and these are the GRB info packets that are sent on the stream every 5 min.
These have good info on the satellite systems in them.
They can be found here
That's it for now, stay tuned as there is much more to come as the software side of things are getting very interesting.

Great news, glad you finally got a receiver that worked out. I've been busy with work and finalizing touches on a LINUX version of the GRBStreamer, so couldn't post much... but fun days ahead for sure of more innovation.

 

[GOES 12 GVAR user]

Great news, glad you finally got a receiver that worked out.

Thanks, it's has been a long process to get this far.
Now I can focus on the other aspects of the project.

I set up again and tweaked my dish's focal length for the best C/N.
As I would have expected the 10' dish is giving me a 8 dB average C/N. With a quality level of 68-69%.
As for a previous question that was asked many pages ago by KWX. The signal level varies by .3 dB C/N here. I have seen no more then that over the 5 hours I have had the station up and going.

I have several movies I made with the images I got from my system on my sites page. They can be found here.
Hurricane Aletta Meso sector

Severe storms over the Upper Midwest meso sector

So now with the info from N6BY I will again look into the septum feed. Since Brett had such good results as well as Weather01089 and KWX with theirs I want to approach this again.
My first round one didn't work out, so I will build a square one with a choke ring. The results should be as good as Brett's. I'm figuring at min. 9 dB C/N could be had.
I can also focus on the final version of the Amplifier block that will be out at the dish. I will have more parts to test for here including a couple of ebay bought preamps.
Ok, now I'm going to put out some theories on here, these are my opinion not tested.
One of the questions that has been going through my mind is why is the septum feed working about 2 dB better then the 2 probe feed...?
Here's what I think.
1. The septum feed gives most likely better isolation.
Why? With the 2 probe feed you can lock to either RH or LH polarization's with out changing a thing, also the 2 polarization's are the same amplitude when either one locked.
This is unexpected, I would have thought that the feed lines going to the probes would have to be reversed so to get the other stream. This goes with the 90 degree phase shift in the combiner. Since that both streams can be locked with no changes I figure that the isolation is not as good as it could be.
2.This may have a strong association with the above:
The circular polarization may not be as clean as it could be. Even though the signal comes in with no errors.
The big thing with the septum feed is the purity of the the circular polarization, and that had a definite effect on the signal level and quality.
3. More connections? insertion loss?
Possible, When ever a combiner or splitter is used there is insertion losses. The losses seem very small but add up quickly before the first preamp.
The septum feed may not have as much as an issue with this as it uses the waveguide it self as the splitter and has no other circuits that the signal needs to go through before the first preamp. This totals in a low loss path to the first preamp.
Summery...
If one is to want RGB or both streams then the septum feed is the way to go. The 2 probe cantenna is a good working feed if built properly but is best used on a single stream. The septum feed though more expensive is well worth the extra price for the performance you get with it.
Can the cantenna be made to work better in the future?
Maybe, but the question then boils down to.. what will it cost? and does that cost out weigh the price of the septum feed?
This is far beyond my knowledge base, though there are some on here that may be able to figure that out. In my case the septum feed (I think) can be made at a reasonable price with a bit of elbow grease and soldering.

 

[N6BY]

Hurricane on June 13th made by GRBStreamer, GRBDump, and my new natural color program

I have been working on a new program to create natural color images from GRB Channels 1-3. It uses a 256x256x256 lookup table to compute the missing green band (using blue, red, and near infrared). Also, the luminance values from GOES-EAST have to be converted logarithmically to simulate human visual perception. It took a lot of trial and error to get this far!

The image below was one of the 'trials'. One of the color bands is upside down.

Finally, here is an image of the sun. GOES-EAST captures the sun at several wavelengths. I found an interesting way to combine three of the wavelengths.

 

[N6BY]

Here's a full globe image I made at 10 AM EDT today.

 

[GOES 12 GVAR user]

RGB CONUS, Impressive, It looks great.

I predict that after we've all got good setups here in a few weeks or so we'll be combining channels and making color images.

Took a little longer then a few weeks...Hi

I am sure when I have a working setup for daytime reception my images will be worthy.

Brett, I say you have defiantly made that goal.

It took a lot of trial and error to get this far!

Man no kidding, It has taken a lot.
Then.. you take that info and apply it to the sun images, wow just wow!

The RGB full disk looks great.
So Brett, what's the next rabbit you going to pull from that hat?

 

[N6BY]

RGB CONUS, Impressive, It looks great.
Took a little longer then a few weeks...Hi

Thanks. Yes, a few months ago I thought it would be a few weeks, lol. I had given up on the possibility of using the septum feed during the day. It wasn't until I gave it another try in May that it somehow started working.

The RGB full disk looks great.
So Brett, what's the next rabbit you going to pull from that hat?

Good question. There are several things I'd like to do:
1) Figure out how to eliminate the need for a cable reset. For me, it often takes multiple attempts. Very puzzling.
2) The new color image generation code has artifacts in low light areas. The images that were used to create the green lookup table did not have dimly lit areas, so I need to adapt the program to deal with it better.
3) GRBDump could use a lot of work -- possibly a complete rewrite in C/C++. Its mostly C# right now. I don't like the way it names and organizes files. Also it leaves lots of temp files to waste disk space. Most of all, it needs a GUI and NetCDF output.
4) Build another septum feed by bending sheet metal, to make it lighter and much cheaper to build. Also i have not spent any time trying to optimize the current design. I think it needs shorter probes.

But some or all of the above may have to wait because I have a lot of non-GRB related projects on my plate right now.

 

[GOES 12 GVAR user]

Yes, a few months ago I thought it would be a few weeks

I wasn't pointing out so much the timeline as what you said in terms of the RGB ability.

It wasn't until I gave it another try in May that it somehow started working.

I believe that you did some electronics changes and shielded better as well and you took away the parts that weren't working with your setup.
Regardless you had a working setup far earlier the me.
Congrads.

Figure out how to eliminate the need for a cable reset.

This is where the info files may come into play, If the stream don't see the GRB Info files in a certain period of time then some how the receiver gets reset.
Or it detects if it has lock. But the actual reset is an issue. On my setup I have a SMA to F connector. The SMA connector makes it easy to reset.
Just unscrew the SMA connector until the signal drops off (though the connector is still connected via some threads) with the play in the connector body I can pull it back and push it back together to get the reset.

The images that were used to create the green lookup table did not have dimly lit areas, so I need to adapt the program to deal with it better.

A few bugs, considering what you have done that shouldn't be too much of a issue.

I don't like the way it names and organizes files

Most of all, it needs a GUI and NetCDF output.

The names are not much to look at nor tell you much, there is a standard for the file naming in GRB. As for the GUI and NetCDF I can't agree more. I will contact you about something that concerns this.

I think it needs shorter probes.

Your feed works well the way it's built. I don't know if you are going to get much (again my opinion) by messing with the probes. I will have soon a price for mine as I am just waiting for my fab guy to get back to me.

a lot of non-GRB related projects on my plate right now.

Yes, like here. The yard, normal jobs around the house. Just plain me time. Fix those unexpected items that break and cost a bunch of money...

 

[N6BY]

GRBStreamer is now a stand-alone program that reads from a DVB card and outputs TIFF images. Instead of streaming UDP packets to another program for processing, it does all the processing itself to create images. Since the program doesn't stream anymore, I need to come up with a new name for it.

I wrote all the processing code in C and C++, and the result is an extremely efficient program that can process all images from both RHCP and LHCP with an average CPU load of about 5%. (This is on a 6 core Intel i7-8700).

Currently the program outputs grayscale images. The next step is to incorporate the color combiner program I wrote a couple of weeks ago so it can output color images. When its all done and I do some more fine tuning, I will release the program.

Here is an image of the earth I made at about 4 PM Pacific time. Its a bit greenish, so some additional work is needed on my color combiner algorithm.

 

[GOES 12 GVAR user]

1. The septum feed gives most likely better isolation.
Why? With the 2 probe feed you can lock to either RH or LH polarization's with out changing a thing, also the 2 polarization's are the same amplitude when either one locked.
This is unexpected, I would have thought that the feed lines going to the probes would have to be reversed so to get the other stream. This goes with the 90 degree phase shift in the combiner. Since that both streams can be locked with no changes I figure that the isolation is not as good as it could be.

Update:
After talking with Brett about the polarization issue it was found out that it does take a cable change to receive the other polarization. The software has no control over what polarization is received. That is determined by the the combiner or what side of the septum you are receiving from, as it should.
I did determine that my feed is working better on the one polarization then the other by almost a dB C/N (this associated with #2).

#2 The circular polarization may not be as clean as it could be. Even though the signal comes in with no errors.

I have changed my cables from my probes to see if the signal increase is seen on the other side. I have no results as of yet as the test site I am using had been flooded and now the heat and (biting) bugs are not very tolerable here.
I have also got a septum feed in the works, it's currently is being made by my fabricator and should have the parts here very soon. I hope to have a step by step on how I made mine with plans, costs and materials.

Otherwise with Brett's software update I am currently waiting on how that continues to play out.

 

[N6BY]

The difference in signal on RHCP vs LHCP is probably at the source. Others on the chat group have reported signal strength differences between the two polarities, and I am seeing it too.

Looking forward to seeing the completed septum.

... I got tired of resetting the coax every time I lock GRB (and many others are too), so I have set out on a determined effort to figure out why its necessary and hopefully fix it.

My strategy is to look at all of the STV6120 tuner and STV0910 demodulator registers both before and after resetting the coax, and see which registers change. It took me over a week to get the STV register reading code to work, but I can now read registers in both chips. (25 in the STV6120 and 975 in the STV0910).

The verdict is partially in. None of the STV6120 tuner registers change values before vs. after the coax reset, but some of the STV0910 demod registers do indeed change. Some changes are to be expected as a normal part of the lock process. I haven't tested all 975 of the demod registers yet because I haven't completed entering all 975 register addresses into my code. I expect to finish that later today.

Hopefully the demod registers that change will provide a clue of what changed in order to lock. But there is little to no documentation available for the STV0910 registers, other than what can be learned from the open source Linux driver.

... I have set aside my work on the standalone GRB program to solve the coax rest. Hopefully late this week or next I will release a beta of the standalone program, planned to be named 'GRBImager'.

 

[GOES 12 GVAR user]

The difference in signal on RHCP vs LHCP is probably at the source. Others on the chat group have reported signal strength differences between the two polarities, and I am seeing it too.

Interesting, I haven't heard of that. Good to know.

Looking forward to seeing the completed septum.

Thanks. It should be done or very very soon. I'll have to go pick up the parts.

I got tired of resetting the coax every time I lock GRB (and many others are too), so I have set out on a determined effort to figure out why its necessary and hopefully fix it.

Wow, 975 addresses. That's a lot to go through. Please keep us informed on the progress.
Be looking forward to the release of the new software.

 

[N6BY]

I'm taking a break from trying to understand and fix the locking problem. Back to work on GRBImager. Below is an image it made at Noon EDT today. Will probably take me a couple more weeks before I can release a beta version.

 

[N6BY]

GRB Imager made this at 3:15 PM today showing the smoke over California. The wind is blowing northeast and today the air is clearer than it has been in a while.

I am getting close to releasing the first beta version of GRB Imager. Mostly I need to work on the user interface.

 

[GOES 12 GVAR user]

I have changed my cables from my probes to see if the signal increase is seen on the other side. I have no results as of yet as the test site I am using had been flooded and now the heat and (biting) bugs are not very tolerable here.

I did gain a 1/2 dB with the two probe feed after the cable change. Bringing my signal C/N to 8.5 average.
This will be as good as I can get with the 2 probe feed.

I have also got a septum feed in the works, it's currently is being made by my fabricator and should have the parts here very soon. I hope to have a step by step on how I made mine with plans, costs and materials.

The septum feed is done and the results are great. I got a 10.6 dB C/N signal from it. This goes with what I brought up in my questions above about the septum feed's performance. It's doing far better then the 2 probe feed. Though both are not giving any errors, they both work fine. But the septum feed has got much better efficiency with it so the dish's gain is better used.
I will put up the drawings and plans I used to build it here, soon. I have been very busy here as many changes have been happening so my time is limited.

Here are some notes that I noticed when using the septum feed.
It goes back to KWX's issues with signal fluxes.
I did see the signal level flux by a 1/2 a dB C/N maybe a little more. This occurred when the sun's rays were able to enter the mouth of the feed, this happened 2 times.
The first time was when the sun was at the dishes edge to my West so that the rays could just barely enter the feeds mouth, this was just enough to increase the noise level and hence the slight loss in signal strength. Once the sun got past this point the signal went right back
to 10.6.
The next short period of signal reduction was when the sun was getting behind the dish again allowing a little bit of the sun's ray to enter the feed's mouth. And again once this period passed by the signal went right back to 10.6.
This was with the septum feed only as Brett has set up. I'm thinking that the choke/ matching ring may help this little problem as it may cause more of a shadow over the feed's mouth.
Part I haven't built yet.

So there's my update here. I'm hoping to have those plans up here in a couple of weeks.

GRB Imager made this at 3:15 PM today showing the smoke over California. The wind is blowing northeast and today the air is clearer than it has been in a while.

Looks great Brett, GOES is seeing that smoke. How does the IR channels look withGRB Imager? Could you put a couple of samples up here please.

 

[N6BY]

...
Looks great Brett, GOES is seeing that smoke. How does the IR channels look withGRB Imager? Could you put a couple of samples up here please.

I thought I would have GRB Imager done yesterday, but got stuck on trying to get solar color images looking like I want. Anyway, here is a Band 11 Full Disc with about 1/2 of the Earth in sunlight, but you can't really tell from this image. You can see the rest of the bands when I release GRB Imager. It should be very soon.