Moderator Control Panel ]

FE_READ_BER and MER

FE_READ_BER and MER

Postby IPMAN » Thu Oct 25, 2012 12:30 am

Dear Support,

I have a problem with the following cards:
- DVB-T2 TurboSight TBS 62x0 DVBT/T2 frontend
- DVB-S2 TurboSight TBS 6984 DVBS/S2 frontend

Here are the problems:
a) On DVBT2 I get a BER measurement something like 31 (a two digit number, sometimes even 3 digits, up to 253). This number obviously means nothing as it is supposed to be ratio output, something like 1E-6, for example. I can always get a measurement on the DVBT cards, but the value is meaningless to me.

The command issued is from the DVB API which is FE_READ_BER
How do I interpret the "31" (or 2/3 digit number) or output of the FE_READ_BER command to a proper Bit Error Ratio, seeing I am using DVB API ?
OR is this command not supported by the TBS drivers?
(Note: This page: http://www.linuxtv.org/pipermail/linux- ... 25300.html from Linux.org seems to suggest it is directly related to drivers)

b) Oon the DVBS2 card, the BER measurement is 0. Although this number is ALWAYS 0 (regardless of how strong/good the signal is), 0 means nothing but it could also mean its not reading any value?
It should also be a number in format 1E-6, for example. It seems that no matter which carrier I lock the card to, I Always get 0 value for BER on DVB-S/S2.
How do I interpret the "0" or outpuf of the FE_READ_BER command to a proper Bit Error Ratio, seeing I am using DVB API ? Or is there some kind of setting in modprobe I have to enable/add?
OR is this command not supported by the TBS drivers?
(Note: This page: http://www.linuxtv.org/pipermail/linux- ... 25300.html from Linux.org seems to suggest it is directly related to drivers)

c) I am trying to also get an MER measurement from the cards, however there is no FE_READ_MER command in the DVB API.
The question is, do the TBS cards support MER measurement? and if they do, what is the command/script to add/use in order to capture the MER value?

The information you want to see regarding the dmesg and lspci are attached.

The 'uname -a' command returns the following:
Linux dvbx-test 3.0.0-26-generic #43-Ubuntu SMP Tue Sep 25 17:20:50 UTC 2012 i686 athlon i386 GNU/Linux

Hopefully this can be resolved as it is very important.

P.S. By the way, I've tried to upload the dmesg log and the lspci log in .txt, .doc and .pdf formats and I always get the error "The extension xxx is not allowed.". Could you please tell me which extension this forum allows to upload files?

Thank You.
IPMAN
IPMAN
 
Posts: 11
Joined: Sun Oct 21, 2012 10:02 am

Re: FE_READ_BER and MER

Postby IPMAN » Mon Oct 29, 2012 2:37 pm

Bump... Surely someone knows about this ???
IPMAN
 
Posts: 11
Joined: Sun Oct 21, 2012 10:02 am

Re: FE_READ_BER and MER

Postby aturkin » Tue Oct 30, 2012 9:23 pm

I second this question. I need to interpret BER and signal strength readings from TBS 6280 card; from that I understand, the card reports values in 0-4095 range for signal strength readings, and in 0-1048575 range for BER (for DVB-T). I need to map signal strength to dBuV and BER values to, well, bit error rate.

In fact, if you (TBS guys) can provide some docs on CXD2820R demodulator, that would be great. I understand you might be limited in what you can provide because of NDAs, but any information would be greatly appreciated.

Thanks
aturkin
 
Posts: 2
Joined: Mon Oct 29, 2012 10:33 pm

Re: FE_READ_BER and MER

Postby IPMAN » Wed Nov 07, 2012 2:03 pm

Does anyone in support actually care to reply?
IPMAN
 
Posts: 11
Joined: Sun Oct 21, 2012 10:02 am

Re: FE_READ_BER and MER

Postby updatelee » Wed Nov 07, 2012 9:36 pm

the signal strength to convert to dbuv or dbm just needs a lookup table. Here is an example of how its done in the stv090x driver

Code: Select All Code
/* RF level C/N lookup table */
static const struct stv090x_tab stv090x_rf_tab[] = {
        {  -5, 0xcaa1 }, /*  -5dBm */
        { -10, 0xc229 }, /* -10dBm */
        { -15, 0xbb08 }, /* -15dBm */
        { -20, 0xb4bc }, /* -20dBm */
        { -25, 0xad5a }, /* -25dBm */
        { -30, 0xa298 }, /* -30dBm */
        { -35, 0x98a8 }, /* -35dBm */
        { -40, 0x8389 }, /* -40dBm */
        { -45, 0x59be }, /* -45dBm */
        { -50, 0x3a14 }, /* -50dBm */
        { -55, 0x2d11 }, /* -55dBm */
        { -60, 0x210d }, /* -60dBm */
        { -65, 0xa14f }, /* -65dBm */
        { -70, 0x07aa }  /* -70dBm */
};


now the issue is, those values arent correct. They are close in some cases, but defn not calibrated. The root of the problem is multiple hardware devices use the same demodulator therfore the same demodulator driver. Differences in hardware is going to mean differences in readings on the same signal level. Windows and linux are both known for this issue, its not going to change any time soon.

Not sure what you want to happen with the BER. BER = bit error rate, doesnt need to be converted. its already in the correct format. Maybe your thinking signal quality (Snr) ?

UDL
TBS6925/5980, Prof 7301/7500/8000, Genpix Skywalker-1, Skystar 2 Express HD
Hauppauge 950Q, Kworld 330U/435v3/445v3
Dreambox 800

I use Linux and support open source projects.
User avatar
updatelee
 
Posts: 374
Joined: Wed Jul 25, 2012 11:48 am
Location: CFB Edmonton

Re: FE_READ_BER and MER

Postby IPMAN » Thu Nov 08, 2012 5:57 am

Updatelee, BER does Not = bit error rate, it is a Ratio measurement, not a rate measurement. Thus it is Bit Error Ratio.
Also, the output that it provides is something like "31" or "251" etc... I've never seen a BER measurement that looks like that and I've been in the satellite game for 20 years. Same goes for the MER measurement I'm looking for, Modulation Error Ratio.

BER output, in this context, is always expressed as: "The number of bit errors out of the number of received bits of a data stream over a communication channel that have been altered due to noise, interference, distortion or bit synchronization errors and so on". How can one determine this if there is only 1 number? Answer is you cannot.

The output that I require, should be looking something like this:
BER = 1E-6
or
BER = 5E-4
and so on... depending on the actual numbers of errors present in the data stream out of x number of total bits in data stream.

However the problem is even bigger because the output on the DVB-T2 card gives a flat number like 31... how does one obtain a ratio from a single number?

The second part of the problem is that the DVB-S2 cards ALWAYS give a "0", no matter what type of signal you have. How can it be that the BER measurement is "0"... its obviously an incorrect output, even though I have lock on the carrier.

If you can answer these and provide me with a method of giving an output in BER and MER in the correct format, I will take my hat off to you and you will become my master :D

IPMAN
IPMAN
 
Posts: 11
Joined: Sun Oct 21, 2012 10:02 am

Re: FE_READ_BER and MER

Postby IPMAN » Mon Nov 19, 2012 12:09 pm

Surely by now someone from support would have replied with an answer?
Or does no reply mean that support doesn't know about their own product? It can't be... surely it can't be.
IPMAN
 
Posts: 11
Joined: Sun Oct 21, 2012 10:02 am

Re: FE_READ_BER and MER

Postby SergioDanielG » Wed Nov 21, 2012 10:10 am

Hi IPMAN.
I'm not sure if you are asking about this http://www.linuxtv.org/downloads/v4l-dv ... E_READ_BER , but maybe this page is util for you.

Best regards
SergioDanielG
 
Posts: 356
Joined: Tue May 31, 2011 3:53 am
Location: Tostado (SF) Argentina

Re: FE_READ_BER and MER

Postby IPMAN » Sun Nov 25, 2012 11:45 am

Hi Sergio,

Actually I know about this page already, its only the function calls for various info. However this isn't the information that I require.
If you want, you can try it yourself, you can issue commands FE_READ_BER and you'll see the output... it will be "0" or a 3 digit number like 123 or something. That to me, is NOT a BER output and isn't conforming with standards either. I am guessing its some other reference value that needs to be converted, but the problem is, no one is providing this information, its not in the readme in the driver files and its not on the internet Anywhere.

In the README, there is a little section at the bottom of the driver doc and its got some information regarding the modprobe file settings but even if you try those, they don't work either, meaning they don't output a value in a format that conforms to standards or at the very least, that one would be expecting.

So my question remains still unanswered, it seems crazy that such a simple question is taking literally months to answer when even in the TBS driver documentation it is referenced, yet the actual values do not output in a manner that one can interpret.

I've purchased now over 20 TBS cards, all high quality, no complaints there, but this problem relating to the output of such simple values, should have been addressed but hasn't been after many months. I just hope an ADMIN can see this post and reply to it with the correct information.

Thanks
IPMAN
IPMAN
 
Posts: 11
Joined: Sun Oct 21, 2012 10:02 am

Re: FE_READ_BER and MER

Postby SergioDanielG » Sun Nov 25, 2012 11:46 pm

Hi IPMAN
While we wait some admin answer, if yuo whant, we can see something...

BER output as is, isn't 3 digit, is 8 charactes, isn't decimal digits, its hexadecimal (or some like this)
Look femon output for my TBS6925 (and similar to TBS6920 and TBS6928)

In this case with -H parameter. H of Human redeable. In this case no ber.
Code: Select All Code
$ femon -H -a 2
FE: STV090x Multistandard (DVBS)
status SCVYL | signal  57% | snr  58% | ber 0 | unc 0 | FE_HAS_LOCK
status SCVYL | signal  57% | snr  58% | ber 0 | unc 0 | FE_HAS_LOCK
status SCVYL | signal  57% | snr  58% | ber 0 | unc 0 | FE_HAS_LOCK


Other case with ber and -H parameter.
Code: Select All Code
$ femon -H -a 2
FE: STV090x Multistandard (DVBS)
status SCVYL | signal  58% | snr  75% | ber 1580 | unc 0 | FE_HAS_LOCK
status SCVYL | signal  58% | snr  75% | ber 539 | unc 0 | FE_HAS_LOCK
status SCVYL | signal  58% | snr  75% | ber 537 | unc 0 | FE_HAS_LOCK


The same but without -H parameter.
Code: Select All Code
$ femon -a 2
FE: STV090x Multistandard (DVBS)
status SCVYL | signal 947a | snr c1f7 | ber 0000fda7 | unc 00000000 | FE_HAS_LOCK
status SCVYL | signal 947a | snr c19b | ber 0000021b | unc 00000000 | FE_HAS_LOCK


Maybe you can check femon source ( http://linuxtv.org/hg/dvb-apps/file/3fc ... on/femon.c ) and obtain utilful information.

Best regards.
SergioDanielG
 
Posts: 356
Joined: Tue May 31, 2011 3:53 am
Location: Tostado (SF) Argentina

Next

Return to Linux

Who is online

Users browsing this forum: No registered users and 2 guests