TBS 6903 tested by DXer

[sato]

I and Peon (alias Onacila) are known to our colleagues as hard core DXers who test their satellite hardware to the limits. We both like quality build products that we can rely on and that provides excellent results while doing DXing. This is why there are multiple cards in Peon`s computer, mostly manufactured by Tenow (TBS).

It has been 2 years when Tenow launched its state of the art PC card TBS 6983 Peon and many, many users worldwide can enjoy its great capability But technological progress can not be stopped so Tenow has come with improved TBS 6903 to substitute the old model. I don`t wanna talk about hardware improvements as this is mentioned on official web site.
I`d like to focus on something, that makes state of the card wanted. It is its capabilities in everyday use.
Peon has tested TBS 6903 for some time now and can provide info that you may find useful.

What we know so far:

- no software or driver issues even multiple cards are installed and in use at the same time (TBS 6903, TBS 6983, TBS 5925, TBS 6522 and Prof 7301)
- the lowest SR received so far is SR 128
- the highest SR received so far is SR 65441
- flawless reception of Multistreams, T2-MI or mixed transponders that contain various transmission standards (Multistream + T2-MI)
- the same excellent sensitivity as TBS 6983
- no bitrate issues even @ 133.7 Mb/s
- flawless reception of 16 APSK (TV station)

- we could lock Kasikorn TV on Thaicom 78.5E freq. 4091 V, but TBS 6983 failed


There will be more results if we find something interesting to share.
These results will be uploaded one by one as Peon compares this card to his TBS 6983, 6522, 5925 and Prof 7301 that are installed on his PC he uses for DXing. Peon also searches for suitable signals in C, Ku and Ka band.

Peon`s workshop where all TBS cards are tested and used on daily basis

Device list as installed on Peon`s PC

The lowest SR 128 so far

Multistream + T2-MI, SR 60000 and bitrate @ 133.7 Mb/s

Multistream + T2-MI, SR 60000 and bitrate @ 133.7 Mb/s

78.5E 4091 V Kasikorn TV (TBS 6903)

16 APSK reception

Multistream reception

The highest SR 65441 - band scan (freq. 11752)

The highest SR 65441 - freq. 11752 details

[DVB]

Can I ask You about some test?

Please run BlindScan2 mode at Hot Bird between 11,7 and 11,9 GHz and please check do You have 11,766 GHz, pol. V and 11,804 GHz, pol. V found in results.

For some time I'm fighting with 6903 and it is constantly missing transmissions with FEC 1/2 and 2/3, but only if they are in DVB-S. Also because of that a lot of misses at 68,5°E, because many frequencies are there in DVB-S with FEC 2/3.

The funniest thing is that:
- TBS BlindScan is able to find DVB-S FEC 1/2 and 2/3, only CrazyScan/EBS Pro don't
- I can receive these transmissions without problems, I got problem only with blindscan and only with CrazyScan/EBS Pro.

In BlindScan mode I'm missing most of transmissions, I got problem to lock even strong signals (despite fact, that I can receive them in any DVB player).

[Peon]

Can I ask You about some test?

Please run BlindScan2 mode at Hot Bird between 11,7 and 11,9 GHz and please check do You have 11,766 GHz, pol. V and 11,804 GHz, pol. V found in results.

For some time I'm fighting with 6903 and it is constantly missing transmissions with FEC 1/2 and 2/3, but only if they are in DVB-S. Also because of that a lot of misses at 68,5°E, because many frequencies are there in DVB-S with FEC 2/3.

The funniest thing is that:
- TBS BlindScan is able to find DVB-S FEC 1/2 and 2/3, only CrazyScan/EBS Pro don't
- I can receive these transmissions without problems, I got problem only with blindscan and only with CrazyScan/EBS Pro.

In BlindScan mode I'm missing most of transmissions, I got problem to lock even strong signals (despite fact, that I can receive them in any DVB player).

Hi.
The problem is in StreamReader. TBS Blindscan uses a different algorithm searches. It is necessary to contact the author StreamReader Mr CrazyCat. (Sincere thanks for his work)
TBS6983 also has a problem with high SR, the latest versions StreamReader. For TBS6903 best StreamReader is in version 1.2.3.1348, but the minimum 5MHz step. A smaller step, omitting some TP with SR 20000 and above.
For TBS6983 best StreamReader is version 1.2.2.1248
With EBSpro this problem is not found.
HotBird scan 11700-12000 step 500kHz H/V

[DVB]

Thanks for reply

I did some tests with StreamReader 1.2.3.1348, but anyhow in BLScan2 mode I'm always missing 11,766 and 11,804 from Hot Bird and generally any DVB-S transponder with FEC 1/2 or 2/3. I had already contacted CrazyCat, but looks like we are out of ideas



@steven - Did TBS changed anything between 6983 and 6903 related to TBS BlindScan - some methods or results/values? Maybe just simply TBS algorithm for 6903 is reporting DVB-S with FEC 1/2 and 2/3 in some other way and thus SteamReader.dll cannot interpret correctly answer from TBS drivers or something like that?

But BLScan2 problem is one thing. In BLscan mode I got a lot of missing frequencies, but I had always tested with 1 MHz step. I switched now to 5 MHz, and I see much better results, but with 5 MHz step I will for sure miss a lot of transmissions with low SR. Step in 6903 is also setting for search range?

In 6983 I didn't have such problem.

PS: At EBSPro website donation is no longer possible - EBS Pro cannot be registered now?

[crazycat]

Please run BlindScan2 mode at Hot Bird between 11,7 and 11,9 GHz and please check do You have 11,766 GHz, pol. V and 11,804 GHz, pol. V found in results.

For some time I'm fighting with 6903 and it is constantly missing transmissions with FEC 1/2 and 2/3, but only if they are in DVB-S. Also because of that a lot of misses at 68,5°E, because many frequencies are there in DVB-S with FEC 2/3.

Now checked 13E High Band with my samples 6983 and 5927. All ok, both described transponder found in BLScan2 mode.

13.0E High band scan with TBS 6983

13.0E High band scan with TBS 5927

[crazycat]

TBS6983 also has a problem with high SR, the latest versions StreamReader. For TBS6903 best StreamReader is in version 1.2.3.1348, but the minimum 5MHz step. A smaller step, omitting some TP with SR 20000 and above.
For TBS6983 best StreamReader is version 1.2.2.1248

I not see any critical changes in my STV0910-related code after 1248 or 1348

[DVB]

Thanks for reply

Now checked 13E High Band with my samples 6983 and 5927. All ok, both described transponder found in BLScan2 mode.

At 6983 I also got both transponders - only on 6903 they are missing (despite that both can be found with TBS BlindScan and BLScan2 should have just the same results). That's why my theory that TBS changed something related to their BlindScan results on 6903 (in drivers?) that is messing with StreamReader

[Peon]

Thanks for reply

I did some tests with StreamReader 1.2.3.1348, but anyhow in BLScan2 mode I'm always missing 11,766 and 11,804 from Hot Bird and generally any DVB-S transponder with FEC 1/2 or 2/3. I had already contacted CrazyCat, but looks like we are out of ideas



@steven - Did TBS changed anything between 6983 and 6903 related to TBS BlindScan - some methods or results/values? Maybe just simply TBS algorithm for 6903 is reporting DVB-S with FEC 1/2 and 2/3 in some other way and thus SteamReader.dll cannot interpret correctly answer from TBS drivers or something like that?

But BLScan2 problem is one thing. In BLscan mode I got a lot of missing frequencies, but I had always tested with 1 MHz step. I switched now to 5 MHz, and I see much better results, but with 5 MHz step I will for sure miss a lot of transmissions with low SR. Step in 6903 is also setting for search range?

In 6983 I didn't have such problem.


PS: At EBSPro website donation is no longer possible - EBS Pro cannot be registered now?

Try to EBSpro Blind Scan. (Or OTFS) Not Blind Scan 2. I think that BS2 is made for terrestrial TV . It must not operate correctly. CrazyCat can explain.

[DVB]

Thanks for replies After a break I'm back on this issue with my final findings. Be aware that everything is confirmed on two 6903 cards with different PC configuration and compared to 6983 (which don't have these problems).

Problem 1: In BLScan mode 6903 got effectiveness degradation for high SR transmissions with step 1 MHz. Setting step to 2 MHz improves ability to find high SR transmissions, but of course bigger step = easier to omit low SR transmissions. If step and search range are set to 1 MHz often 6903 got problem to lock even strong 27500 MCPC transmission, often need to set frequency exact to carrier center frequency. 6983 don't have such problem.

I would say that it looks like for 6983 step and search range values were not the same and despite step 6983 was able to search carrier in some inteligent way, 6903 cannot do that. In IQ Scan window (CrazyScan) setting search range to 10 MHz gives 100% lock for high SR. 1 MHz mostly gives no lock if I would try to lock frequency with even small offset which shouldn't happen for high SR, where carrier width is >20 MHz.

Problem 2: BLScan2 mode for 6903 is skipping everything with DVB-S and FEC 1/2 or 2/3, always. But only DVB-S. DVB-S2 is OK. Maybe exactly related to QPSK modulation...

Examples - BlindScan with 6903 at 12,5 - 12,75 GHz on 68,5°E.
1. What results should look like - there should be 20 active frequencies (at time when I did scan attempts)

Code: Select All Code

0=20
1=12522,V,27495,34,1
2=12553,V,4769,23,3
3=12561,V,2865,12,2
4=12562,H,27500,34,1
5=12575,V,9698,12,1
6=12591,H,1832,34,4
7=12602,V,26650,23,1
8=12611,H,6709,12,2
9=12625,H,1157,34,4
10=12628,H,1399,12,2
11=12629,V,8498,34,4
12=12640,V,4999,56,2
13=12643,H,2299,23,3
14=12648,V,2643,23,2
15=12648,H,2225,34,4
16=12657,V,4882,34,1
17=12682,V,29990,23,1
18=12682,H,26650,23,1
19=12722,V,26650,12,1
20=12722,H,26650,23,1

2. Results for BLScan with step 1 MHz - a lot of misses

Code: Select All Code

0=14
1=12553,V,4769,89,3 ;-44 dBm, 14.1 dB
2=12561,V,2865,12,2 ;-44 dBm, 13.9 dB
3=12575,V,9698,12,1 ;-43 dBm, 13.4 dB
4=12591,H,1832,34,4 ;-44 dBm, 21.4 dB
5=12611,H,6709,12,2 ;-45 dBm, 9.6 dB
6=12625,H,1157,34,4 ;-47 dBm, 12.5 dB
7=12628,H,1399,12,2 ;-47 dBm, 8.5 dB
8=12629,V,8498,34,4 ;-45 dBm, 9.5 dB
9=12640,V,4999,56,2 ;-45 dBm, 13.7 dB
10=12643,H,2299,12,2 ;-47 dBm, 11.6 dB
11=12648,H,2225,34,4 ;-47 dBm, 9.5 dB
12=12657,V,4882,34,1 ;-45 dBm, 16.7 dB
13=12682,H,26650,23,1 ;-43 dBm, 15.3 dB
14=12722,V,26650,12,1 ;-43 dBm, 14.3 dB


Overally 6 frequencies skipped. Every attempt gives different results, but always many skipped frequencies.

3. Results for BLScan with step 2 MHz - nearly perfect, but bigger step will result in easy skipping low SR transmissions <1 MHz.

Code: Select All Code

0=19
1=12522,V,27495,34,1 ;-41 dBm, 14.5 dB
2=12553,V,4769,23,3 ;-44 dBm, 14.2 dB
3=12561,V,2865,12,2 ;-44 dBm, 13.5 dB
4=12575,V,9698,12,1 ;-43 dBm, 13.1 dB
5=12591,H,1832,34,4 ;-43 dBm, 21.9 dB
6=12602,V,26650,23,1 ;-42 dBm, 15.2 dB
7=12611,H,6709,12,2 ;-46 dBm, 9.5 dB
8=12625,H,1157,34,4 ;-47 dBm, 12.4 dB
9=12628,H,1399,12,2 ;-47 dBm, 9.4 dB
10=12629,V,8498,34,4 ;-45 dBm, 9.2 dB
11=12640,V,4999,56,2 ;-46 dBm, 12.3 dB
12=12643,H,2299,23,3 ;-47 dBm, 11.5 dB
13=12648,V,2643,23,2 ;-45 dBm, 8.1 dB
14=12648,H,2225,34,4 ;-47 dBm, 8.8 dB
15=12657,V,4882,34,1 ;-45 dBm, 16.5 dB
16=12682,V,29990,23,1 ;-43 dBm, 12 dB
17=12682,H,26650,23,1 ;-43 dBm, 15.3 dB
18=12722,V,26650,12,1 ;-43 dBm, 14.3 dB
19=12722,H,26650,23,1 ;-43 dBm, 14.1 dB

Nearly 100% OK, for some reason 12,562 GHz, pol. H is always missing, but everything else is present and this is much better that in step 1 MHz (in 6983 always smaller step gave better results).

4. Results for BLScan2 - DVB-S FEC 1/2 and 2/3 are missing

Code: Select All Code

0=14
1=12522,V,27495,34,1 ;-41 dBm, 14.3 dB
2=12553,V,4769,23,3 ;-43 dBm, 13.8 dB
3=12561,V,2865,12,2 ;-44 dBm, 13.2 dB
4=12562,H,27500,34,1 ;-44 dBm, 15.4 dB
5=12591,H,1832,34,4 ;-43 dBm, 20.7 dB
6=12611,H,6709,12,2 ;-45 dBm, 8.4 dB
7=12625,H,1157,34,4 ;-47 dBm, 8.6 dB
8=12628,H,1399,12,2 ;-47 dBm, 6.3 dB
9=12629,V,8498,34,4 ;-43 dBm, 9.2 dB
10=12640,V,4999,56,2 ;-45 dBm, 13.6 dB
11=12643,H,2299,23,3 ;-47 dBm, 11.3 dB
12=12648,V,2643,23,2 ;-45 dBm, 8.3 dB
13=12648,H,2225,34,4 ;-47 dBm, 7.3 dB
14=12657,V,4882,34,1 ;-45 dBm, 16.7 dB

Check that BLScan2 found everything except these (from BLScan):

Code: Select All Code

4=12575,V,9698,12,1 ;-43 dBm, 13.1 dB
6=12602,V,26650,23,1 ;-42 dBm, 15.2 dB
16=12682,V,29990,23,1 ;-43 dBm, 12 dB
17=12682,H,26650,23,1 ;-43 dBm, 15.3 dB
18=12722,V,26650,12,1 ;-43 dBm, 14.3 dB
19=12722,H,26650,23,1 ;-43 dBm, 14.1 dB

All these got FEC 1/2 or 2/3 and DVB-S QPSK mode - 100% precisely skipped all.

Because I see the same results on both 6903 cards - I believe that really something is not correct here. Currently because in both modes in StreamReaderEx-BDA and in TBS BlindScan app (in other way, about which I had wrote in earlier post) something is wrong - 6903 cannot be used to reliable blinscan, because with every way there is high possibility of skipping some transponders for different reasons.

PS: BLScan and BLScan2 are both for satellite (AirScan is for DVB-T/C), BLScan is crazycat engine, BLScan2 is wrapper for TBS interface.
EBS Pro is no longer available to buy, anyhow I believe that EBS Pro shouldn't get different results as this is wrapper for crazycat's StreamReaderEx-BDA library, so it would be very weird if crazycats own apps would be less effective in using his library than somebodys other apps

PS2: 68,5°E is also good for test TBS BlindScan app issue, about which I mentioned earlier (this issue is present for all DVB cards). That's because at 68,5°E there are present transponders with same frequency, but opposite polarisation. Look at this:

When You count list items - there are 16, not 19 (and that's the bug), but when You will save .ini, You get 19. In window if frequency is the same - it is filtered out even if polarisation is opposite (which is incorrectly).

Code: Select All Code

0=19
1=12521,V,27495
2=12553,V,4769
3=12561,V,2865
4=12561,H,27495
5=12574,V,9698
6=12591,H,1832
7=12601,V,26653
8=12610,H,6709
9=12624,H,1157
10=12627,H,1399
11=12628,V,8498
12=12640,V,4999
13=12643,H,2299
14=12648,H,2225
15=12656,V,4882
16=12681,V,29994
17=12681,H,26652
18=12721,V,26652
19=12721,H,26652

(missed 12,648 GHz, pol. V, but all DVB-S FEC 1/2 and 2/3 are present in results)

If anybody really got DX on heart - please test it with me carefully and help to find solution, because 6903 got great hardware, but really some (I believe) software shortcomings are putting 6903 effectiveness behind 6983 (or maybe that's some other HW revision of 6903 which need to different treatment?). This really need to be looked at...

First easy test can be done in CrazyScan in IQScan window (that one with constellation plot). Try to lock strong MCPC signal eg. at Hot Bird with search range set to 1 MHz. For example 11,449 GHz, pol. H. And set frequency for example 11,442 GHz. Check 6903, then check othert DVB card. For me 6903 got no lock until I change search range to higher value or put exact frequency, at any other card, also 6983 - lock without problem.

I believe that 6903 itself is not damaged, because finally with different combinations of software probably everything can be found, but all those combinations were not needed for 6983 and that's the clue.

Also it would be nice if somebody would check this exactly as I wrote at 6908 as I believe that it should have the same problems.

[crazycat]

BLScan2 is wrapper for TBS interface - only when disable direct hardware control in streamreader.ini