47 replies to this topic

[Erik Slagter]

Well, wouldn't that be consistent with percentage being linear and dB's logarithmic?

The SNR actual value IS logarithmic, that's why it's represented in a logarithmic scale (dB) which makes the value lineair, indeed comparable to percents. So compare dB and % = OK, compare raw SNR and % = NOT OK.

[Erik Slagter]

Except in the extreme low range that is almost never used... (were do you see FECs of 1/2 nowdays???)

FEC of 1/2 is afaik valid only for DVB-S2 and indeed I've never seen it. It means 50% overhead, it may be better to step down to QPSK in that case (if using 8PSK) probably.

And even the mearement equipment used above is no rocket science.... and may give incorrect values in these ranges....

How about consumer grade tuners. I think any non-integer fraction of the dB value is pure guessing and should be discarded anyway (unless for using on the same tuner).

[littlesat]

But al least in general the dBs are somehow comparable between tuners.. by approximation.....
The %-ges are really not comparable at all.

And again the only true good information is using the bit error rates on the different FEC as it is on now.....

[Huevos]

Except in the extreme low range that is almost
never used... (were do you see FECs of 1/2 nowdays???)

Well, 120W, 110W, 107W, 105W, 99W, 91W, 89W, 87W, 85W, 58W, 53W, 50W, 43W, 31W, 27W, 24W, 22W, 20W, 15W, 11W, 8W, 5W, 5E, 7E, 16E, 21E, 23E, 36E, 42E, 45E, 49E, 50E, 53E, 55E, 57E, 64E, 68E, 70E, 74E, 75E, 76E, 78E, 91E, 96E, 100E, 105E, 108E, 115E, 134E, 138E, 144E, 150E, 156E, 166E, and 180E all carry 1/2 FEC transponders (which you'll see if you check OpenPLi's own satellites.xml file). The reason I chose a low FEC transponder is because it is the only way to get an accurate reading for low dB values without the transponder unlocking.

FEC of 1/2 is afaik valid only for DVB-S2 and indeed I've never seen it. It means 50% overhead, it may be better to step down to QPSK in that case (if using 8PSK) probably.

That's not true. It is both DVB-S and DVB-S2 (fec_inner = 1). And also 2/3 is the next DVB-S FEC setting and locks into the low 5dB range right where the algorithm goes skewy. And there are literally thousands of those transponders.

And even the mearement equipment used above is no rocket science.... and may give incorrect values in these ranges....

My 2 pro-level meters agree with each other, and also the values correspond with those in the Tandberg manual.

Edited by Huevos, 5 October 2012 - 17:14.

[kocus37]

Openpli 3.0 Hispasat packaged Meo channel SIC, on DM500HD 11,9 db and on VU+ uno 9,7 db.

[Erik Slagter]

huevos you've made clear that you know what you're talking about

Can you please give me some reference values for the VU+Duo?

[littlesat]

The best is to have a calibration table for each different tuner... but hen you have another issue... each VU+ Duo != another VU+ Diuo and the same for ET boxes and DMM boxes.... especially in the lower range ..... it is still an approximation of +/- 0.5 dB........

[Erik Slagter]

For the moment I focus on VU+ because it didn't have mapping at all.

[littlesat]

And likely the VU+Duo have a different mapping.... ... Wo we need a table to calculate a new formula

[Erik Slagter]

And likely the VU+Duo have a different mapping.... ... Wo we need a table to calculate a new formula

That's why I asked huevos for some data...

[Sotos65]

Hi, maybe a naive question, but why the maximum SNR in dB is 16.2dB? In various programs for DVB cards as well my DR.HD F15 the same time my VU+ Uno shows 16.2dB (or SNR 100%) they may give 17, 18, 20 dB or even more (the DVB cards).

Edited by Sotos65, 6 October 2012 - 19:41.

[littlesat]

The big question is what is realistic....?

[Sotos65]

Yes, that is a good question, but on the other hand why 17 or 18 dB wouldn't be realistic? In any case if this value is the absolute maximum, it should not be reported so easily by the receiver...

[Huevos]

Yes, that is a good question, but on the other hand why 17 or 18 dB wouldn't be realistic? In any case if this value is the absolute maximum, it should not be reported so easily by the receiver...

Another approach might just be an out of range type of output, for example display "< 6dB" or "> 16dB".

I asked huevos for some data...

I'll post some later comparing the Ultimo to the Duo.

Edited by Huevos, 7 October 2012 - 11:40.

[Erik Slagter]

Thx.

[Sotos65]

Hi, this a preliminary comparison I did between VU+ Uno, a Technotrend TT S2-1600 DVB card, a TBS 6925 DVB card and a DR.HD F15 receiver. I'm living in west Crete (Chania) and my dish is a 1.10m Tecatel with a Smart Oxid LNB.

Attached Files

•  Comparison_dB_test1.jpg   113.51KB   8 downloads

[Erik Slagter]

If you want the VU+Solo added, you need to give the output of cat /proc/bus/nim_sockets as well.

[Huevos]

Here's a start. All from 28E comparing the Duo and Ultimo. Most noteable thing is the Ultimo tuner seems to become less sensitive above 12400MHz.

Attached Files

•  Duo-v-Ultimo-dB.txt   517bytes   19 downloads

[VACI_]

@Huevos
take in your consideration frequencies over 11590, cause it is same frequency received by tuner, take in mind that LNB has different local oscilators to receive frequencies below/over 11700MHz.
Frequency received by tuner is 1800MHz, so maybe your LNB is reason for lower sensetivity.
12440-10600=1840MHz high band, active L.O. 10,6GHz
11590-9750=1840MHz low band, active L.O. 9,75GHz

[Huevos]

maybe your LNB is reason for lower sensetivity.

No, because 1) this is a known issue (which may be driver related as it doesn't happen in lo-band), and, 2) I used the same aerial for both tests.

Edited by Huevos, 8 October 2012 - 12:50.