Date
1 - 20 of 40
Howto monitor/graph Tellicast/receiver device in Linux?
|
Christian Peters
Ernst,
yes....it's really a nice and good looking tool you found with eLuna! Maybe the new Eumetcast Linux monitoring standard...!? ;-) Yes...maybe I really first stay on the driver and I even thought about what happened if I do a apt-get update/upgrade!? As Ubuntu sometimes changes the kernel and the modules are build for the current then I really have to recompile the TBS driver every time...?! :-O It goes against the grain for me not to update from time to time...but I think I should stay some time...! Would be great if your mod while merge into the driver. It's just raining dog and cats here and S/NR goes down...! :-O So switching MODCODS would be maybe sometimes really useful. But first I will try to increase S/NR by readjusting and installing another LNB....! Regards, Christian Am 22.02.16 um 21:19 schrieb ernst.lobsiger@... [MSG-1]:
[Non-text portions of this message have been removed] |
|
|
|
Ernst Lobsiger
Christian
That's really nice to see eLuna + RRDtool work on another GNU/Linux Eumetcast receiver system! <cite> Interesting...maybe you can contact with TBS to add this into their linux drivers ! So testing you modified stv-driver means recompiling the driver and install over the present one!? Phew.....not sure whether I didn't break my running system then....!? :-D </cite> Yes my driver means recompiling. But if you still have the TBS-tree you just backup and replace the stv090x.c source. Make will just recompile and link my new stv090x.c source, You could even copy a single module then. ... Be aware that if you update/upgrade your GNU/Linux system you might get a new kernel and the full new module tree of your distribution. This will mean recompiling and re-installing all the TBS-Drivers again too. So I propose you let your system run and graph for a while ... I'll contact TBS again regarding MODCODS. Cheers, Ernst |
|
|
|
Christian Peters
Ernst,
yes...you are right with the missing 0x...but it's running now and eLuna makes nice graphs. You can have a look at http://eluna.hcpeters.de:81/ Thank you very much! :-) I really would like to test and add HVS receiption but not sure howto set up the second tellicast instance. Even I think maybe I will get problems with my ramdisk set on /run/shm as brd doesn't work on Ubuntu... Interesting...maybe you can contact with TBS to add this into their linux drivers ! So testing you modified stv-driver means recompiling the driver and install over the present one!? Phew.....not sure whether I didn't break my running system then....!? :-D Regards, Christian Am 21.02.16 um 17:57 schrieb ernst.lobsiger@... [MSG-1]:
[Non-text portions of this message have been removed] |
|
|
|
Ernst Lobsiger
Andrew
I used an SR1 as a pilot user back in 2014. Even this SR1 did just calculate the link margin. There is no way to get a figure that is actually measured by the TBS-6983 demodulator chip. These are mostly theoretical values: For HVS you need SNR (EsNo) > 9.3 dB For BS you need SNR (EsNo) > 5.9 dB The BS limit applies only when MODCOD 16APSK 2/3 is switched off (not possible in GNU/Linux so far) !! The TBS-6983 works well down to these theoretical/experimental limits. So you can just calculate: Link margun HVS = SNR - 9.3 Link margin BS = SNR - 5.9 With RRDtool used by Cacti you can do these calculations when setting up the plot command. There is no need to have an RRD that saves SNR + two link margins. So you can go on with your existing RRD. The bash script I use for eLuna to graph is below (I am sure it can be adapted for Cacti that also uses rrdtool to graph): # Link margins are just calculated, that's how the SR1 does it too! $GRAPH_TITLES{'snr'} = "{#server#} - TBS-6983 V12 Signal to Noise"; $GRAPH_CMDS{'snr'} = <<"SNR_GRAPH_CMD"; --title "{#server#} - TBS-6983 V12 Signal to Noise" --vertical-label="Signal to Noise [dB]" --lower-limit 0 --upper-limit 20 DEF:snr={#path#}snr.rrd:snr:AVERAGE CDEF:lm8psk=snr,5.9,-,0,MAX CDEF:lm16apsk=snr,9.3,-,0,MAX AREA:snr{#color5#}:"Signal to Noise " GPRINT:snr:LAST:"Current\\: %4.1lf " GPRINT:snr:AVERAGE:"Average\\: %4.1lf " GPRINT:snr:MAX:"Maximum\\: %4.1lf \\j" AREA:lm8psk{#color4#}:"Link Margin 8PSK 3/5 " GPRINT:lm8psk:LAST:"Current\\: %4.1lf " GPRINT:lm8psk:AVERAGE:"Average\\: %4.1lf " GPRINT:lm8psk:MAX:"Maximum\\: %4.1lf \\j" AREA:lm16apsk{#color3#}:"Link Margin 16APSK 2/3" GPRINT:lm16apsk:LAST:"Current\\: %4.1lf " GPRINT:lm16apsk:AVERAGE:"Average\\: %4.1lf " GPRINT:lm16apsk:MAX:"Maximum\\: %4.1lf \\j" LINE1:snr{#linecolor#} SNR_GRAPH_CMD 1; # Return true Cheers, Ernst ---In MSG-1@..., <fox91fox@...> wrote : Thanks, In the coming days I will correct dish pointing. I will use `femon` in SSH on my smartphone to see the SNR value in realtime. @David, there is no problem to add my station on your site. The only thing I need to know is how to extract the data of Link Margin from my TBS6983. Regards, Andrew |
|
|
|
David and Andrew
There is no such thing as an easy Ayecka console or BDADataEx for a TBS-6983 under GNU/Linux. But allong the lines how you get the Es/No values for RRDtool or MRTG I wrote two bash scripts that give you a running console BAR display with Signal Strength and Es/No and even uses the Beep command increasing the frequency for better values. One script is for local use (or over an ssh connection). Script two is for use on a LapTop using GNU/Linux to be able to also run the Beep locally. Script one should run on a Android? Phone with an simple ssh client as descibed here (DISCLAIMER: I still have no mobile Phone, so don't ask me ...) : http://www.techrepublic.com/article/easily-make-an-ssh-connection-from-your-android-device/ My best practice to point an EumetCast dish: 1) Use the sun for pointing to get a first strong Signal 2) Improve the Signal Strength while locked with your Receiver 3) Finally use the Es/No of your Receiver for optimum fine Tuning @Andrew: You can go right to point 3). The method you use to have local figures when you point your antenna depends on how far away (from the Receiver or a local network plug) and how accessible (roof?) your dish is. Cheers, Ernst ============================== Thanks, Ernst. I see at least two good, free SSH clients for Android (and there are some for iOS) so seeing a graph or text from your antenna location should be straight forward. https://play.google.com/store/apps/details?id=com.sonelli.juicessh&hl=en_GB https://play.google.com/store/apps/details?id=org.connectbot&hl=en_GB To Andrew, remember to adjust first azimuth and elevation of the dish, then skew, and then focus (move the LNB in and out if possible). Iterate over the last two adjustments to get the best results. Cheers, David -- SatSignal Software - Quality software written to your requirements Web: http://www.satsignal.eu Email: david-taylor@... Twitter: @gm8arv |
|
|
|
Thanks,
In the coming days I will correct dish pointing. I will use `femon` in SSH on my smartphone to see the SNR value in realtime. @David, there is no problem to add my station on your site. The only thing I need to know is how to extract the data of Link Margin from my TBS6983. Regards, Andrew ============================= Andrew, Thanks! That will be appreciated as EUMETSAT also refer to these pages. Under Windows and TBS devices, I use and recommend the BDADataEx program, with an add-on for monitoring. See: http://www.satsignal.eu/wxsat/dvb-s2/sr1-mrtg.html#bda My own monitoring is here: http://www.satsignal.eu/mrtg/performance_stamsund_bdadataex.php You can calculate the Link Margin knowing the Es/No and the signal parameters - I think Ernst has the numbers for BS and HVS. Cheers, David -- SatSignal Software - Quality software written to your requirements Web: http://www.satsignal.eu Email: david-taylor@... Twitter: @gm8arv |
|
|
|
Ernst Lobsiger
David and Andrew
There is no such thing as an easy Ayecka console or BDADataEx for a TBS-6983 under GNU/Linux. But allong the lines how you get the Es/No values for RRDtool or MRTG I wrote two bash scripts that give you a running console BAR display with Signal Strength and Es/No and even uses the Beep command increasing the frequency for better values. One script is for local use (or over an ssh connection). Script two is for use on a LapTop using GNU/Linux to be able to also run the Beep locally. Script one should run on a Android? Phone with an simple ssh client as descibed here (DISCLAIMER: I still have no mobile Phone, so don't ask me ...) : http://www.techrepublic.com/article/easily-make-an-ssh-connection-from-your-android-device/ http://www.techrepublic.com/article/easily-make-an-ssh-connection-from-your-android-device/ My best practice to point an EumetCast dish: 1) Use the sun for pointing to get a first strong Signal 2) Improve the Signal Strength while locked with your Receiver 3) Finally use the Es/No of your Receiver for optimum fine Tuning @Andrew: You can go right to point 3). The method you use to have local figures when you point your antenna depends on how far away (from the Receiver or a local network plug) and how accessible (roof?) your dish is. Cheers, Ernst |
|
|
|
fox91fox@...
Thanks,
In the coming days I will correct dish pointing. I will use `femon` in SSH on my smartphone to see the SNR value in realtime. @David, there is no problem to add my station on your site. The only thing I need to know is how to extract the data of Link Margin from my TBS6983. Regards, Andrew |
|
|
|
Andrew
[] And yes the figures seem low for a 150cm dish in Italy. There was a discussion last year about the satellite moving in smal "figure-8" on the sky and this can produce dayly variations of Es/No. See espacially message Nr. 20783 on this list or search for "figure-8". I think your dish is slightly missaligned. It's hard to get it right with such a monster dish. You must make very small adjustments and have an Es/No indication on a LapTop or some other devive beside you. There is also plenty of discussions about how to do this best on this list. Cheers, Ernst ============================== Folks, I've found the easiest way to do this is to use TeamViewer. You can share a programs showing SNR (such as the Ayecka console) and have that display on your table or phone. No need for a laptop. More recently I've bought a low-cost Windows-10 tablet (GBP 120 ish) and you can run the Ayecka Console directly on that. Perhaps that would work on an Android phone? Can you run Java programs under Android? Unsure what monitoring program you might use for the TBS under Linux - under Windows I would use the BDADataEx program and share it via TeamViewer. As you say, a 1.50 m dish requires very careful adjustment, and it /is/ bulky to handle. Cheers, David -- SatSignal Software - Quality software written to your requirements Web: http://www.satsignal.eu Email: david-taylor@... Twitter: @gm8arv |
|
|
|
Ernst,
I've deleted my signal and quality graphs. I've also added Es/No graph: http://imageshack.com/a/img923/8917/TwoB6Y.png I think my Es/No are low (given that I have a satellite dish 150cm in northern Italy). Thanks for your help. Andrew ======================== Andrew, I look forward to adding your graphs to those for Europe: http://www.satsignal.eu/mrtg/performance_eumetcast-europe_link_margin.php You will be able to compare your installation with similarly located stations, and make adjustments for dish size. Link Margin is a more sensitive indicator of small changes than Es/No, and I've recently found it useful to see how the HVS data stops when the link margin becomes too small. The "year" graphs give a long-term measure so that any gradual change in the system can be more easily detected. The station of Daniel Hurtmans in Belgium: http://cpm-ws4.ulb.ac.be/reception/ also uses RRDtool, I believe, and was able to format the graphs so that were comparable to those from MRTG, making comparison between stations much easier. Perhaps he can suggest the configuration to use? I can put you in touch if you like. Ernst, it would good to have your data on the page for comparisons, too! We're relatively short of stations in Europe. Cheers, David -- SatSignal Software - Quality software written to your requirements Web: http://www.satsignal.eu Email: david-taylor@... Twitter: @gm8arv |
|
|
|
Ernst Lobsiger
Andrew
You should change the headline of your graphs. I think you diplay Es/No in dB not dBx10 (as given back by the TBS-6983 driver under GNU/Linux). And yes the figures seem low for a 150cm dish in Italy. There was a discussion last year about the satellite moving in smal "figure-8" on the sky and this can produce dayly variations of Es/No. See espacially message Nr. 20783 on this list or search for "figure-8". I think your dish is slightly missaligned. It's hard to get it right with such a monster dish. You must make very small adjustments and have an Es/No indication on a LapTop or some other devive beside you. There is also plenty of discussions about how to do this best on this list. Cheers, Ernst ---In MSG-1@..., <fox91fox@...> wrote : Ernst, I've deleted my signal and quality graphs. I've also added Es/No graph: http://imageshack.com/a/img923/8917/TwoB6Y.png http://imageshack.com/a/img923/8917/TwoB6Y.png I think my Es/No are low (given that I have a satellite dish 150cm in northern Italy). Thanks for your help. Andrew |
|
|
|
fox91fox@...
Ernst,
I've deleted my signal and quality graphs. I've also added Es/No graph: http://imageshack.com/a/img923/8917/TwoB6Y.png I think my Es/No are low (given that I have a satellite dish 150cm in northern Italy). Thanks for your help. Andrew |
|
|
|
Cheers,
Ernst P.S. I am sure there are more elegant ways to retreive the S/N values (David might call it a "bodge"). But my lines also work for drivers that do not indicate the FE: line or when status is empty because you have no lock ... =============================== Some of my most useful programs are screen-scrapers! Another "bodge". David -- SatSignal Software - Quality software written to your requirements Web: http://www.satsignal.eu Email: david-taylor@... Twitter: @gm8arv |
|
|
|
Ernst Lobsiger
Christian
My first line is: upd=0x`femon ..` You did not type the 0x there and had to add it later. So your two-liner should work as well. If you have a 1m dish you should get better S/N values and have a good chance for HVS. I modified a stv090x.c driver source that takes an unsigned long integer to set MODCODs on the fly. I have two bash scripts to set and get MODCODs pretty much like TBS-IP-Tool under WINDOWS. It worked for my PCIe TBS-6925 card back 1 year. It should work for the TBS-5925 as well. If you want to test it at your own risk ? Cheers, Ernst P.S. I am sure there are more elegant ways to retreive the S/N values (David might call it a "bodge"). But my lines also work for drivers that do not indicate the FE: line or when status is empty because you have no lock ... ---In MSG-1@..., <hctpeters@...> wrote : Ernst, yes...strange!?! This worked here on Ubuntu: upd=`femon -a0 -c1 2>/dev/null | tail -n1 | cut -d'|' -f3 |cut -d' ' -f3` upd=N:`echo $((0x${upd})) | awk '{print ( $1/10.0 ) }'` Maybe a shell problem here? Yes...I read something about and you mentioned before that's not easy to switch off HVS on Linux!? At the moment I have a 1m dish and got around EsN0=11.0-11.5dB. Should be a little bit higher if I read your values!? Maybe I should test another LNB (at the moment a Rocket LNB, but have an Inverto Black already here) and should try to do some more adjustments on the dish to get more headroom? ... |
|
|
|
Christian Peters
Ernst,
yes...strange!?! This worked here on Ubuntu: upd=`femon -a0 -c1 2>/dev/null | tail -n1 | cut -d'|' -f3 |cut -d' ' -f3` upd=N:`echo $((0x${upd})) | awk '{print ( $1/10.0 ) }'` Maybe a shell problem here? Yes...I read something about and you mentioned before that's not easy to switch off HVS on Linux!? At the moment I have a 1m dish and got around EsN0=11.0-11.5dB. Should be a little bit higher if I read your values!? Maybe I should test another LNB (at the moment a Rocket LNB, but have an Inverto Black already here) and should try to do some more adjustments on the dish to get more headroom? But I think it's interesting to get even HVS if it is possible with one TBS USB Box!? Can I stay on 2.4.3a to receive HVS additional or should/have I to switch to the new Client 2.12_1 you tested already? But I think I then have to register first for HVS at Eumetsat!? Would be hard to switch now and do all the configuration again!? :-D But maybe even fun.....:-) I will have a look to that link to hardcode switch off 16APSK 2/3....! Maybe a good solution to get a really stable BS. Regards, Christian Am 21.02.16 um 16:40 schrieb ernst.lobsiger@... [MSG-1]:
[Non-text portions of this message have been removed] |
|
|
|
Ernst Lobsiger
Christian
I checked my code to update S/N (EsNo). I do not see what should be changed: I take the HEX number from femon and add 0x in front. Conversion is then done by awk when printing and dividing by 10 at the same time. With a 90cm dish on a clear day you should get around EsN0 = 11.5 dB. A 1m dish should give 12 - 13dB. One more word about BS versus HVS. Your TBS-5925 takes both and when the link margin for 16APSK 2/3 goes to 0 this will harm BS even if you do not take HVS data with a second instance of tc-recv. So if you only intend to take BS and nothing else you should switch off HVS in the demodulator. This is easy under Windows but there is no MODCOD selection tool under GNU/Linux yet. The TBS-5925 is special because the stv090x.c demodulator code is open source. One easy way to switch of HVS is by initializing the demodulator in DUAL MODE (this in no typo). CrazyCat shows how this is done hardcoded when compiling your driver. See here: http://www.tbsdtv.com/forum/viewtopic.php?f=26&t=9243 http://www.tbsdtv.com/forum/viewtopic.php?f=26&t=9243 In dual mode your BS only receiver will work error free down to 3.5dB lower S/N values than before. Cheers, Ernst ---In MSG-1@..., <hctpeters@...> wrote : Ernst, all is working like a charm! Thanks for you work and the files. Very interesting to read the code and learn. I had to slightly change the code of hex-decimal-conversion in the SNR update file but now it's working and looking great! :-) Even data is coming in again...I had to wipe the dish as it was covered by snow....so investing in the snow shield for the LNB was good, but didn't solved the snow problem at all. Maybe I need a dish heater...!? :-D Regards, Christian |
|
|
|
Christian Peters
Ernst,
toggle quoted message
Show quoted text
all is working like a charm! Thanks for you work and the files. Very interesting to read the code and learn. I had to slightly change the code of hex-decimal-conversion in the SNR update file but now it's working and looking great! :-) Even data is coming in again...I had to wipe the dish as it was covered by snow....so investing in the snow shield for the LNB was good, but didn't solved the snow problem at all. Maybe I need a dish heater...!? :-D Regards, Christian Am 20.02.16 um 09:12 schrieb ernst.lobsiger@... [MSG-1]:
|
|
|
|
Christian Peters
Simon,
toggle quoted message
Show quoted text
thank you for that information and the .zip. I already set up eLuna (many thanks to Ernst) and I'm quite happy with it. :-) Regards, Christian Am 20.02.16 um 10:08 schrieb Simon Proud simon.proud@... [MSG-1]:
|
|
|
|
Ernst Lobsiger
Simon
There is no problem with MRTG except for it's the ancestor of RRDtool and very limited. I used it for traffic graphs of my Cisco Switches and if you want to display your receivers on David's page you are probably better off with MRTG. Most of the MRTG scripts can certainly be used for RRD updates too, your offer is very useful. With MRTG you will never be able to things like those: http://oss.oetiker.ch/rrdtool/gallery/index.en.html http://oss.oetiker.ch/rrdtool/gallery/index.en.html or the graphs I do when monitoring my GNU/Linux Eumetcast receivers. So it's also a matter of taste. Cheers, Ernst I've missed most of this conversation but what's the problem with just using MRTG for this? I got that working fine on Ubuntu 14.04, took an hour or so of work but there's plenty of guides on the internet. Francis Breame also provided some useful scripts/tips. My scripts are here, if they're useful: 130.226.13.99/mrtg/mrtg.zip Simon |
|
|
|
Simon Proud <simon.proud@...>
I've missed most of this conversation but what's the problem with just
using MRTG for this? I got that working fine on Ubuntu 14.04, took an hour or so of work but there's plenty of guides on the internet. Francis Breame also provided some useful scripts/tips. My scripts are here, if they're useful: 130.226.13.99/mrtg/mrtg.zip Simon On 20 February 2016 at 08:45, Christian Peters hctpeters@... [MSG-1] < MSG-1@...> wrote: Ernst, |
|
|