Installed new TBS6985 DVB-S2 tuner card and DiSEqC switch won't work, + other issues (2 Viewers)

[clarkebelt]

After the experience with the stupid Prof 7301 tuners that would not allow you to have more than one installed in a system at the same time, we purchased a TBS6985 quad input DVB-S2 tuner card and after installing it today (and completely removing the Prof cards and drivers) initially had no luck with it at all. After two or three hours of trying various things I finally decided to try connecting LNB's directly to the tuner, bypassing the DiSEqC switch. Suddenly I was able to receive signals, although I had some other issues I will mention in a moment. But I had planned to use a DiSEqC switch on each tuner and for the moment it appears that plan is shot. We even tried a totally different brand of DiSEqC switch and it would not work either. These DiSEqC switches (4 port DiSEqC 2.0) worked great with the Prof 7301's. So we are wondering if you have ever encountered this issue before and if you know of a fix for it. We're currently running MediaPortal 1.7.1.0 (that is what the TV Server reports), would updating to 1.8.0 possibly have any impact on that issue?

In the meantime to get some functionality we simply connected one LNB output to each of the four tuner inputs. Specifically, we have:

A dual C-Band LNB (each output of which is totally independent of the other) connected to tuner inputs A & B, and the channels associated with that tuner mapped to both inputs.

A different C-Band LNB connected to tuner input C, and the channels associated with that tuner mapped to that input only.

And yet another C-Band LNB, plus a Ku-band LNB, connected through a 22 kHz tone switch to tuner input D, and the channels associated with either of those LNB's mapped to that tuner.

The C-band LNB's all work great, however for some reason the one Ku-band LNB isn't working at all. Again this is something that worked fine with the single Prof tuner using the method mm1352000 explained to us to trigger the tone switch. I'm not sure if we somehow missed a step or what, but in all other cases simply remapping the previously scanned channels to the correct tuners seemed to work.

Also, there is an issue with playing Live TV, where attempting to start a new stream from one computer usually (but not always) causes any other Live TV streams playing on other computers to freeze up. This apparently does NOT happen with recordings, as far as I can tell, because a short time ago I recorded three programs from three different satellites simultaneously and they all appear to have recorded just fine. This is probably not going to be a major issue for us because we hardly ever watch Live TV off the satellites, but still it would be nice if it didn't happen.

The #1 issue for us right now is getting the DiSEqC switches working so if you have any thoughts as to how to fix that, they would be very much appreciated! I just don't understand why we are having a problem with them because I could connect a DiSEqC switch to the Prof 7301, and that card controlled the DiSEqC switch with no issues at all.

 

[mm1352000]

Hello again

For DiSEqC support you need to install the patch from here:
https://forum.team-mediaportal.com/threads/tbs-ci-cam-support-and-other-improvements.103276/

To save you hunting, the version for MP 1.6 and newer is here:
https://forum.team-mediaportal.com/...ther-improvements.103276/page-51#post-1051247

...however for some reason the one Ku-band LNB isn't working at all.

Log files are required to troubleshoot this.

Also, there is an issue with playing Live TV, where attempting to start a new stream from one computer usually (but not always) causes any other Live TV streams playing on other computers to freeze up.

Enable preloading to fix this:
http://wiki.team-mediaportal.com/1_...nfiguration/02_TV_Servers#Edit_Tuner_Settings

mm

 

[clarkebelt]

Thank you VERY much mm1352000, I will give this a try tomorrow and if we still can't resolve the 22 kHz tone switch issue I will send you log files. I especially appreciate you pointing out the correct post in that long thread, because finding that would have been like finding a needle in a haystack, and even if I had found that thread I would still have probably thought it that might not be the correct file because of the 1.6 designation.

 

[mm1352000]

No problem.

finding that would have been like finding a needle in a haystack

I've now linked it in the first post, with a comment.

 

[clarkebelt]

Hello again

For DiSEqC support you need to install the patch from here:
https://forum.team-mediaportal.com/threads/tbs-ci-cam-support-and-other-improvements.103276/

To save you hunting, the version for MP 1.6 and newer is here:
https://forum.team-mediaportal.com/...ther-improvements.103276/page-51#post-1051247

https://forum.team-mediaportal.com/...ther-improvements.103276/page-51#post-1051247

These patches worked great - I now have both DiSEqC switches connected and the card is controlling them. Thanks very much!

...however for some reason the one Ku-band LNB isn't working at all.

Log files are required to troubleshoot this.

Unfortunately this problem still exists so I have sent you log files. If you look right at the end (starting around 1:38 AM on July 10) you will see several attempts to tune Ku channels, from both XBMC and MediaPortal, all of which were unsuccessful.

Also, there is an issue with playing Live TV, where attempting to start a new stream from one computer usually (but not always) causes any other Live TV streams playing on other computers to freeze up.

Enable preloading to fix this:
http://wiki.team-mediaportal.com/1_...nfiguration/02_TV_Servers#Edit_Tuner_Settings

This also worked great, but now I am experiencing a different issue (maybe). If you look in the log files I sent you, at 23:59 (11:59 PM) on July 9 I had scheduled a one hour long recording to start, and that worked. Then at 00:29 (12:29 AM) on July 10, one half hour later, I scheduled a half hour recording which came from the same satellite but a different transponder. For whatever reason, the second recording started, but only recorded about 10 to 15 seconds of program, whereupon the file abruptly ended. Yet MediaPortal seemed to think it was still recording that channel for the entire duration, as evidenced by a message displayed in XBMC that indicated the recording was still in progress. I just wonder if you can see any reason that might have happened in the log file. I'm not discounting the idea that this might be a tuner issue or some other problem, but I was hoping perhaps you might see something that would give you some indication of what happened to cause that recording to die after such a short time.

Anyway, if you can determine anything about what is causing either or both of these issues and how to fix them, I'd be most appreciative. And thank you again for the fixes above!

 

[mm1352000]

Unfortunately this problem still exists so I have sent you log files. If you look right at the end (starting around 1:38 AM on July 10) you will see several attempts to tune Ku channels, from both XBMC and MediaPortal, all of which were unsuccessful.

Hmmmm, the tuner driver doesn't like something about those tune requests. I suspect this is related to the LNB settings [again!].
Please try to change your LNB frequency override settings to:
low = 5150
high = 10750
switch = 11000

I'm just guessing that the driver doesn't like the switch frequency to be less than the high band LOF frequency...

This also worked great, but now I am experiencing a different issue (maybe).
...

I can see 3 continuity errors in the TsWriter log file almost exactly 15 seconds after the recording starts. They almost certainly mark the true end of the recording as you indicated. I speculate that the tuners simply stopped streaming for some reason. There isn't much that TV Server can do about that. The cause... I have no idea. TV Server is not actively interacting with the tuners at that time.

 

[clarkebelt]

Thank you for looking at that, mm1352000. I don't get why that one channel sometimes freezes after 15 seconds if you tune to it while another tuner is on a different channel on that same satellite. I have seen that happen a couple times now in testing with live streams, and also a couple times where if one tuner was in use on that satellite, and I attempted to tune to another channel on that same satellite, it would just act as if there were no signal present, yet if I immediately tried again it would tune it. I don't really see that effect if tuning two signals simultaneously from other satellites, even if they are two different transponders on the same satellite. And I don't see it happen every time I repeat the actions that cause it - it appears to be a hit-or-miss thing. It's just that one particular satellite that mainly seems to have an issue, although we haven't had this card long enough to say it never happens with other satellites or channels. I understand that you have no idea what causes it and that it may not be a problem related to TV Server at all, but what's most frustrating about it is that it's an intermittent issue, so I can't say that if I try to tune these two channels at the same time, one will always fail, nor can I be certain that the channels I have noticed this issue with are the only ones affected.

It would be nice if MediaPortal could detect when those continuity errors occur and when the stream has stopped, and retry the signal. It seems as if once you get past that 15 second mark the signal becomes stable, although because this card is so new to us I cannot say that with absolute certainty. We always try to start recordings one minute prior to the actual program start anyway.

The Ku problem is stranger. I tried changing the switch frequency to 11000 as you suggested and it didn't help; I got the same errors. It's not that the card is incapable of controlling a 22 kHz tone switch - if I use the TBS BlindScan program it finds the transponders on that satellite with no problem:

And as I mentioned, MediaPortal's 22 kHz switch control worked fine with the Prof 7301, so I'm not sure why it's not controlling the TBS card. Does it seem like this is a problem that might be fixed, or do you think that it is something I will just have to live with for now?

 

[mm1352000]

And as I mentioned, MediaPortal's 22 kHz switch control worked fine with the Prof 7301, so I'm not sure why it's not controlling the TBS card. Does it seem like this is a problem that might be fixed, or do you think that it is something I will just have to live with for now?

DiSEqC and tone control almost always require specific/proprietary code for each vendor... and sometimes different code for different tuners from the same vendor. In other words: unfortunately just because something works for a Prof tuner doesn't mean it will work for a TBS tuner, and vice versa.

I wrote the original TBS patch. It was tested using a TBS 5980 with 22k and DiSEqC combinations. Everything worked fine. I know that with TBS the USB tuner handling is different to the PCIe and PCI tuner handling. So, I speculate there is something wrong with the PCIe/PCI tuner handling in the patch. In particular, I speculate that the TBS driver is trying to use the C band LNB LOF to tune the Ku band transponders.

TBS Blind Scan could help us to confirm what is happening.
You've selected:

• LNB type = Ku Circular
• LOF 1 = 10750
• band/22k = high/on

This is different to the settings that MediaPortal would be applying.
Please see what happens if you try:

• LNB type = universal [sorry, I don't know the exact name TBS have used - any setting that allows you to enter 3 LNB frequencies would be fine]
  
• LOF 1 = 5150
• LOF 2 = 10750
• switch = 11000
  
• band/22k = high/on

My hypothesis is that these settings would fail, just like MediaPortal.

Then try changing band/22k to auto.

My hypothesis is that change would work.

Please confirm...

 

[clarkebelt]

Well here is where things take a turn into crazytown...

In looking at the screenshot I posted above I realized that something is wrong. To explain, please understand that at present I am only using the first three ports on the DiSEqC switches - port D or BB (depending on how the software designates it) is unconnected and unused - and of the three that are connected, only the third port (port C or BA) has a tone switch on it, and that tone switch is the only thing that has a Ku-band LNB connected to it. The first two inputs have C-band LNB's directly connected, as does the "off" side of the tone switch. And just in case you start to think I may have them physically miswired, I will just mention that the DiSEqC switches are inside the house, while the 22 kHz tone switches are out at one of the poles. It would therefore be physically impossible to accidentally connect a 22 kHz tone switch between the tuner and the DiSEqC switch.

Yet when I scanned DiSEqC A above, it found the Ku-band LNB and showed the signals from it. Same thing happens if I select DiSEqC B or C (the correct one) - it shows the same signals on roughly the same frequencies. Here is a comparison of what it shows on DiSEqC A (which has NO Ku LNB or tone switch) and DiSEqC C (note that some signals apparently dropped between the two scans, which would be expected on this satellite because it is used for news uplinks and other wild feeds, but you can see the similarities):

But the next weird part is that while these signals also show up on DiSEqC B, on DiSEqC D (the unconnected one) this does NOT happen:

Now I wondered if a similar bleedover between DiSEqC's happens with C-band signals, and it does not. Scanning C-band shows only the signals that should be there on each DiSEqC switch position. I want to emphasize that there is no way the Ku band signals can actually be appearing on all three DiSEqC inputs, so I am at a loss to explain the above. Also it happens on both of the in-use tuners which are connected to separate DiSEqC switches, so I doubt it's a DiSEqC issue (also we never saw anything like this with an actual satellite receiver connected). I am totally at a loss to explain why the TBS scan program would report a Ku satellite on a DiSEqC position other than the one it's actually connected to, but be that as it may, here is the information you asked for. This is with the tone switch on, and it did NOT fail as you expected:

There is no "auto" setting for the tone switch, there is HIGH/ON, LOW/OFF, and ALL. Setting it to ALL showed the next screwy thing - even though I had set the lower frequency to 11700, it found both C-band AND Ku-band signals (the two screenshots below are from the same scan but with the slider moved):

At this point I got to thinking, this stupid scan program is going to find that Ku LNB no matter what I set it to. So as a test, I set it back to circular but turned the tone switch off. Take a wild guess what happened:

My theory is that either this scan program has a bug in it that is not allowing it to work correctly with this particular tuner, or there is something going on in the driver that is allowing the Ku signals from one DiSEqC port to appear on others. Or something like that. Actually this is so weird to me that I can't think of any logical explanation. If the DiSEqC switch were faulty and were allowing signals to bleed over between ports then port D should also be affected, and C-band should also be affected, and neither of those is true (and also I find it hard to believe that two different DiSEqC's would be similarly affected, and that I never saw this issue when it was connected to a receiver). As I noted, it would be physically impossible for me to have miswired it so that the tone switch is between the tuner and the DiSEqC switch. And nothing explains why this program still sees the signals from the Ku LNB when the tone switch is set to off.

So anyway, there is the information you asked for, but I don't know if it can be trusted at all, since this scan program apparently wants to find that Ku LNB no matter what I set it to, as long as it's not C-band.

 

[mm1352000]

Hello again

I haven't got time to go through the information in detail right now. Just wanted to say that a tuner driver can't tell the difference between a Ku and C band signal in the cable. They're both heterodyned down to the same intermediate frequency range (= 950 to 2150 MHz) in the cable. Blind scan results will be interpreted and displayed based on the information you supply. So, if you supply Ku band parameters the results will appear to be Ku band; if you supply C band the results will appear to be C band.

In other words, what you're seeing as Ku band signals on DiSEqC A and B are your C band signals, just interpreted as Ku.

Example:
Port A 11727V (as shown, first result in your first screenshot).
11727 - 10750 = intermediate frequency in cable = 977 MHz
C band equivalent = 5150 - 977 = 4173 MHz
Do you have a transponder at that frequency on the satellite from port A?

On port C you should be able to see differences in the results depending on whether you scan high or low band.

Got to run.

mm