GOES 16 GRB downlink vs GVAR

[N6BY]

...
Been too cold outside lately to do much.

71 degrees and sunny here yesterday. Today we're expected to hit 72!

I'll be at my dish re-installing the Quaker cantenna. Until I get the Minicircuits combiner, I'll try it first with a 1/4 wave difference in feed line length.

Edit: Just saw your previous post. Thanks for giving me the heads up on the inverted grayscale. Looks a lot better when its correct.

 

[GOES 12 GVAR user]

71 degrees and sunny here yesterday. Today we're expected to hit 72!

It got to freezing here today, a little warmer here tomorrow then more snow and cold. It's winter here.

I'll be at my dish re-installing the Quaker cantenna. Until I get the Minicircuits combiner, I'll try it first with a 1/4 wave difference in feed line length.

I think that will work better for you, though the septum feed I don't think is performing real bad. But you need every dB you can get for out there.

Since you guys like that, I played with the Brightness, contrast and gama of the images
Brett, yours came out great see below..

Played with the water vapor one and enhanced the vapor detail...

KWX I enhanced the land on yours...

This is one of the great benefits of this data is you can change it to see what you want to see.
Did all this is a photo edit program nothing fancy.
By the way Brett, The sample data that was available for the ABI had the same issues. So it seems the software is seeing the data as it should but not how we see it. Just flip the colors and presto.

 

[N6BY]

Wow, you made the images come alive! Now they actually look real. 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.

...
I think that will work better for you, though the septum feed I don't think is performing real bad. But you need every dB you can get for out there.,,,

The septum is OK, but if I did it again I would have made it much lighter. Didn't realize how heavy that 1/4" aluminum alloy tube would be when I ordered it. It makes the dish struts bend slightly, and the wood frame mount I made for it is not strong enough to keep the septum pointed at the center of the dish. Its so heavy that its tilted up in the wooden mount -- not aimed at the dish center.

 

[KWX]

Thanks for letting me know, I'll order the same (plus a terminator). I assume these just contain passive coils and no input power required, correct?

Congratulations of getting your first GRB images.

My GRB signal error rate here in California is way too high right now for image generation. It worked once in the middle of the night, but most of the time its just frustrating.

I'm going back to the Quaker cantenna as soon as my ZX10Q-2-25 arrives. Just have to add another probe at 90 degrees. The Quaker cantenna is about 98% lighter than the septum feed, which is so heavy it sags on the dish struts.

Looking at the specs of the ZX10Q-2-25, the cross polarity isolation is 20 to 25 dB -- very good!

Thankfully GOES-S (17) will be up after March 1st (knock on wood), and with the calibration data already collected from GOES-16, then I would dare to assume that the commissioning of GOES-S will be a lot quicker which will give you the CNs you need to take care of business in no time.

 

[weather01089]

I'm glad to see weather01089 on here.

Weather01089 what Linux system are you running there with hardware details if you would detail please?
I know of no one that's using CSPP-GEO software other then Gov, or Universities.
I posted a while back the hardware needed to make such a required computer and it wasn't cheap. High end gaming computer.
I am most curious to know what it's running on as the sys. requirements are quite impressive.

2 dB C/N difference from there to here. Interesting.
Thanks for all your info and coming on here, welcome.




CSPP-GEO is a very nice piece of software.
Again Thanks for the info.





So is their customer service.

I am using an HP DL380G7 server with dual 6 core xeon processors to test CSPP-GEO. Its fine it appears until you try to create any
images with the data. Then it starts missing things with both feeds running. Im going to try one polarity at a time next.
Hopefully Lucas fixes the issues with GRBdump, its far superior. Interested in the ZX10Q-2-25 and can antenna! That certainly
brings the cost down for most. The DD card though thus far has been really solid, Im impressed. The other receiver they want to sell 2
of at a high price is sitting on the side for the moment Only will do one feed at a time anyway. I use it to test signal for go/no go quick with the green "lock" light on it LOL.

 

[weather01089]

Thanks for reminding me about the velocity factor. I had completely forgotten about it. Just looked it up and its about 80% for RG6. So the wavelength is 0.8 * wavelength in space.

Also, as you probably already know, the wavelength in the wave guide is longer than in free space.

By dull I meant that there is no visible earth surface and dark, grey images.

Also, I haven't gotten a full disc image because the Windows version of GRBDump crashes when it tries to save them.

My goal is full disc color images in daylight. I know it will require combining channels.

Dull? The Veggie channel (3) shows up land really well during the day. If you use it as the G in RGB though, its overdone for green.
With 1 min apart images on mesos, its anything but dull when there is any kind of weather!. Also, there is are sensors called GLM, for Global Lightning Mapper.
They detect lightning over the entire visible area of the satellite. Im ingesting them into my AWIPSII and displaying, comparing them to the current fee services.
They are at least as good, better in most cases, and over more area. The batches come down every 20 seconds.!

 

[weather01089]

Thanks!!!. As expected from a 6 to 6.5 signal, I'm getting some sync errors. ~2% with BBFrames selected and when CADU is selected then is ~0.5%.

There's an interesting behavior though, on the same machine I'm testing the TBS5925 on, which also has a "Hauppauge 723x BDA ATSC/QAM Tuner" PCIe card installed. GRBStreamer won't lock using the TBS5925 selection from the list box, but instead it will lock if I choose the "Hauppauge 723x BDA ATSC/QAM Tuner"... crazy. Of course, this won't be the intended "server" I'll be running it from but interesting behavior.

Tomorrow, I'll try a 3m dish I have for my AWIPSII setup which uses the NOAAPORT feed and see how it goes when I point it GOES16. I guess worst case scenario, in the event that I can't squeeze any more CN out of the 2.4m to just have them swap roles.

I have a 14.5 ft dish here for NOAAPORT. I don't like the goes16 images on them, better resolution on GRB since they don't reproject them to Lambert. They don't send the GLM lightning from GOES16 on there either since it competes with the pay services the weather service uses. I send the files over the the EDEX server for AWIPS and display them on the workstation. Amazing to see both hemispheres.

 

[weather01089]

As to my Ayecka, they sent me one to test on this feed that receives GS, allegedly. They say only one of the 2 tuners can be used, have to buy 2 for both feeds.
It is highly annoying to use, takes sometimes 1-3 hrs to lock on to the signal. Once it does, its fine though. But EXPENSIVE! They have no interest in fixing that issue, or
making both tuners work since one box doesn't make it worth it they told me. Therefore, I have avoided recommending it to anyone. They did log on to my system and run some
diagnostics, and had no idea why it wouldn't lock fast with an 11db CN. The TBS card reports that signal as well.

Ray

 

[weather01089]

This is a current GLM plot, none in the US at moment though

 

[N6BY]

Dull? The Veggie channel (3) shows up land really well during the day. If you use it as the G in RGB though, its overdone for green.
With 1 min apart images on mesos, its anything but dull when there is any kind of weather!. Also, there is are sensors called GLM, for Global Lightning Mapper.
They detect lightning over the entire visible area of the satellite. Im ingesting them into my AWIPSII and displaying, comparing them to the current fee services.
They are at least as good, better in most cases, and over more area. The batches come down every 20 seconds.!

Dull because they were at night. I am sure when I have a working setup for daytime reception my images will be worthy.

 

[KWX]

I have a 14.5 ft dish here for NOAAPORT. I don't like the goes16 images on them, better resolution on GRB since they don't reproject them to Lambert. They don't send the GLM lightning from GOES16 on there either since it competes with the pay services the weather service uses. I send the files over the the EDEX server for AWIPS and display them on the workstation. Amazing to see both hemispheres.

Looking into finalizing the CentOS 6 install and configuration before loading CSPP-GEO so I can get the GLM feed into AWIPS. I agree, I prefer the GRB's resolution (heck even the HRIT) to the re-projected version AWIPS works with. I understand you want to maximize bandwidth/resource usage on workstation and may be local network but at least give the guy an option to opt in/out of higher resolution imagery LOL!!!... I won't complain much since its way better than having nothing.

 

[GOES 12 GVAR user]

The septum is OK, but if I did it again I would have made it much lighter. Didn't realize how heavy that 1/4" aluminum alloy tube would be when I ordered it. It makes the dish struts bend slightly, and the wood frame mount I made for it is not strong enough to keep the septum pointed at the center of the dish. Its so heavy that its tilted up in the wooden mount -- not aimed at the dish center.

Lessons learned, there are calculators for the weight of metal. But as you said that was the only size you could find close.
The latter makes your antenna not a good as it could be by not being in the center. One thing you don't need.
But what most tend to do is make waveguide out of sheet bent to size. I had a couple of feeds made out of near 1/16" thick aluminum sheet. The fabricator rolled the metal to make the tube needed and spot welded the seam and back cap on the tube for a semi reasonable price. All I needed to do was install the probe.
I would recommend the you have a local metal fabricator make the tube, square or round in the future as the is the easiest way to do it and maybe the cheapest. Get it quoted and see how to go from there.
Square is pretty easy to make as all that is needed is the material bent on a brake.
Round may be slightly more pricey as the forming is a bit more involved.
I use copper flashing/trim sheet found in rolls in the home store. The roll runs about $100.00. But you can make several feeds from a roll. Cutting bending and soldering is required to make the feeds. I suppose you could rivet it together but that's something I haven't tried.
They sell aluminum rolls as well but you would have to fabricate the feed as you would with copper.

 

[N6BY]

...
I would recommend the you have a local metal fabricator make the tube, square or round in the future as the is the easiest way to do it and maybe the cheapest....

I am reusing the Quaker cantenna from 3 months ago. Today I modified it so that there are 2 probes at 90 degrees to each other. 2 coax lines (one is 1/4 wavelength longer) get combined and then go to the Nooelec SAW/LNA. (I did take the 0.80 velocity factor into account with the coax lines -- one is about 35 mm longer)

The resulting GRB signal looked good on my SDR, but the TBS card saw only about 3.5 dB of SNR -- far too low to lock. So I will have to wait until my MiniCircuits 90 degree combiner and terminator arrive.

Here is a link to the cantenna calculator I used: Cantenna Calculator I just had to enter the frequency and the diameter of my cantenna.

 

[GOES 12 GVAR user]

weather01089 - I am using an HP DL380G7 server with dual 6 core xeon processors to test CSPP-GEO. Its fine it appears until you try to create any
images with the data. Then it starts missing things with both feeds running.

Thanks for the response Ray,
I was wondering if you were going to tell me.
That computer you have there, are you using it for the display machine as well or just the data processor?
If you're using it for both then that thing is loaded down.
How much RAM do you have installed?

weather01089 - Hopefully Lucas fixes the issues with GRBdump, its far superior.

Has anybody contacted Lucas and asked what he was running to get the GRBDump to work correctly?
See if it can be reproduced on another machine.
Seems to be a strange issue since KWX looked at the details of the hardware usage due it the software's lag time.
I think it would be nice if Lucas would make his software spit out the Netcdf files OR the images.
Kinda like weathermessage does with the NOAAPORT images.
Seems since it already does a portion of that now, why not output the Netcdf files, OR binary files.
But it's nice he did as much as he did.
Regardless there is a lot of data coming through those links.

weather01089 - Interested in the ZX10Q-2-25 and can antenna! That certainly
brings the cost down for most.

I'm impressed that you like the combined dual probe feed, since you are the only one here that has a well working septum feed. Those feeds (septum) are tricky to get working well. I'll play more with mine later as I get things working here. And it warms up more.

weather01089 - The other receiver they want to sell 2
of at a high price is sitting on the side for the moment/
use it to test signal for go/no go quick with the green "lock" light on it LOL.

Ya, Ayecka. I'm not impressed with them at all, too many unknowns with their hardware/software and lousy C.S. I figured I'll take my chances with the TBS5927 receiver. Ayecka don't seem to care much about our needs. (My opinion)
Shame to use a costly piece of hardware for signal lock tests. Too bad because they could have a really nice receiver. But oh well.

They say only one of the 2 tuners can be used, have to buy 2 for both feeds.

Yep, that was something Ayecka told me when I was in touch with them.

It is highly annoying to use, takes sometimes 1-3 hrs to lock on to the signal. Once it does, its fine though. But EXPENSIVE! They have no interest in fixing that issue, or
making both tuners work since one box doesn't make it worth it they told me

That's plain ridiculous that it takes that time to lock.
+ they (Ayecka) said they weren't seeing issues like that when I was talking with them.
Yep, They don't care to mess with it. Sooo why are they making it?
Again, Oh well.


Nice view of the GLM.
Ray what software suite are you using to display the files?

 

[GOES 12 GVAR user]

I am reusing the Quaker cantenna from 3 months ago. Today I modified it so that there are 2 probes at 90 degrees to each other. 2 coax lines (one is 1/4 wavelength longer) get combined and then go to the Nooelec SAW/LNA. (I did take the 0.80 velocity factor into account with the coax lines -- one is about 35 mm longer)

Quick question for you Brett, Are you using a "TEE"
to combine the coax cables?
If so..Did you calculate the TEE's length into the combiner?
Yep I, knew that your were reusing that but I made some suggestions for everyone on here.
Again lessons I've learned.

 

[N6BY]

Quick question for you Brett, Are you using a "TEE"
to combine the coax cables?
If so..Did you calculate the TEE's length into the combiner?
Yep I, knew that your were reusing that but I made some suggestions for everyone on here.
Again lessons I've learned.

The coax leads are joined together inside a TV splitter that I gutted (replaced the contents with two equal length wires). So all I had to do was make one of the coax cables 35 mm longer.

But there is an impedance mismatch in the splitter with two 50 ohm inputs merging into one. That makes 25 ohms -- not good. The whole point of what I did today was to swap out the septum feed with the cantenna in preparation for the arrival of the MiniCircuits ZX10Q-2-25. It will be fed with 2 equal length pieces of coax since it does the 90 degree phase adjustment internally.

 

[N6BY]

Ok, so after building/tuning the new can-antenna for 1690MHz with combiner, I got the following for GRB:

2.4m CN = 7.7dB to 8dB.
3m CN = 8.3dB to 8.5dB (if no trees then likely get more and may be a bit of feed leg tuning)...

Did you use the same can-antenna dimensions and probe locations that you get from the can-antenna calculator I referenced in post #253? Your's works very well. Also, what diameter is yours? Thanks.

 

[KWX]

Did you use the same can-antenna dimensions and probe locations that you get from the can-antenna calculator I referenced in post #253? Your's works very well. Also, what diameter is yours? Thanks.

Almost the same dimensions. I used a can of a favorite little snack which has a diameter of 5" and a length of 6 1/2"... so 1/2" less than what GVAR recommended (as a minimum). In regards the probes, I used the dimensions specified in the calculator, but I did trimmed them a bit as I was tuning them with a spectrum analyzer/noise/tracking generator I have.

 

[N6BY]

Which snack can has that dimension? I looked all over the supermarket and the best I could do was Quaker Oats. But I would prefer a metal can.

 

[GOES 12 GVAR user]

Dull because they were at night. I am sure when I have a working setup for daytime reception my images will be worthy.

I think the main comment on "dull" was the grayscale was inverted and it didn't look right that way.
The IR channels at night give quite a lot of detail from the ABI.
From what I have seen channel 14 has pretty good resolution for night images.
But the visible images are the eye catchers.

Brett, Got a question for you.
Since we are both using Unimesh dishes 10' in size, what is the distance from the center to the rim of the feed?
I have been running a 45 - 45.5" distance on mine.
As I get up to 46" the gain drops a little.
I ask just to compare notes.

The coax leads are joined together inside a TV splitter that I gutted (replaced the contents with two equal length wires). So all I had to do was make one of the coax cables 35 mm longer.
But there is an impedance mismatch in the splitter with two 50 ohm inputs merging into one. That makes 25 ohms -- not good.

Brett you probably know this but I put it up for anyone else.
The coax for both has to be ODD and EVEN multiples of 1/4 waves. It can't be of any random length.

Such as Leg one 1/4 wave (has to be odd multiples 3,5 exc.)
Leg two 1/2 wave (has to be even multiples 2,4,6 exc)

The distance is the phase offset. Being a 1/4 wave apart will maintain a 90 degree phase shift.
Use 75 ohm leads at the probes and a 50 ohm for the feed line.
You can use a 100 ohm resistor (optional) across the probe outputs to get an exact 50 ohms but it can't be of the inductive type.
Picture... Wilkinson splitter/combiner

For anyone who wants to try making one.