DVB S/S2 cards on legacy PCI interface. (1 Viewer)

Kotik

MP Donator
  • Premium Supporter
  • March 19, 2009
    699
    485
    Athens
    Home Country
    Greece Greece
    Ok so I think I am getting closer.

    The issue seems like a serious bug.

    I went for a clean MP installation and dropped my DB just to make sure we are starting from zero.

    Changed the LNB and checked signal with ProgDVB, it reports that all the sats I am interested in are OK.

    Step 1: I went ahead and stopped TV Service, then i run ProgDVB and move the dish via USALS to sat 19.2E, close ProgDVB and start MP TV Service go to DiseqC Motor menu and select 19.2E and hit the "store position" button after that i move to the scanning page and scan for some channels. MP is able to pick up all the channels and they are working fine.

    Step 2: I stop TV Service and again run ProgDVB this time i switch the dish to 13.0E and i can confirm that it is turning since i am doing all of this over TeamViewer at my rooftop, after that i repeat step one and scan for channels @ 13.0E

    It is all working 100% fine with 0 issues, if i go to DiseqC motor tab and move my dish to different positions without tuning to a channel it never fails. It is reliable 100%, now lets move on to the tricky part.

    I then go into manual control and play with channels from there, it works fine no issues there either UNLESS i get into a channel that either has bad signal or gives me a PMT Time Out, if that happens it is all downhill from that point, i discovered that when MP is unable to tune to the specific channel it will somehow mark all channels on this sat as "bad" and will even somehow move the GoToX for all the channels on the current sat, or at least this is what i think it does, cause after this point all the channels on this sat will stop working, even the GoToX from the DiseqC tab will not send the dish to the right sat, instead it will send it to a different position this is really frustrating. In order to fix it i have to turn off TV Service and run ProgDVB, move the dish to the broken sat, close ProgDVB and start TV Service, from there store one more time the position for the sat and rescan, after that it will work again until the next random No Signal or Time Out PMT error. The channels that are located on different satelites remain unaffected and if i tune to them they are always working unless ofc i get a PMT error on the other sat aswell.

    The issue is really annoying and i am lost as to how to proceed.
    --- Post updated ---
    EDIT: I just now retested and can confirm that MP is somehow confusing GoToX commands, cause a once working stored sat position suddenly will always move the dish to a totally other sat position.
    In other words prior to NO Signal/Time Out for PMT the dish will always move to the right satelites, BUT after the error the dish will move to wrong positions, in my case 13E will move to 9.0E and 9.0E will remain unaffected, how does that happen i cannot still figure out but it does happen and this is why @mm1352000 i see full signal and yet cannot detect any channel, cause after the error it will move to the wrong sat and scanning for channels that don't belong to this sat will result in full signal but no channels found issue. This can explain that problem but it still doesn't explain why suddenly GoToX positions are getting thrown left and right without me even touching the DiseqC tab after successfully storing them initially.

    FYI this was not an issue with my previous SkyStarHD2 card but it still might be a compatibility issue between my motor and TBS or TBS and MP, the way i see it the issue might be narrowed into two possible causes: 1. GoToX are somehow lost/shifted within MP (something that i cannot confirm since i not familiar with MySQL) or 2. in the motor's memory. Since i cannot troubleshoot it over ProgDVB with GoToX commands (using USALS there) i am now stuck and need help.
     

    Kotik

    MP Donator
  • Premium Supporter
  • March 19, 2009
    699
    485
    Athens
    Home Country
    Greece Greece
    @mm1352000 "eurika!!!!!!" Or better "eurika......." it is what i would call a TBS bug. Damn.

    After keeping a list of the channels that give me either no signal or pmt not found i was finally able to reproduce the issue in ProgDVB even with USALS, mind u i am using a brand new LNB today so that was not the issue.

    The issue is that the tuner is unable to get PMT when u tune from specific frequency to another specific one. This explains everything, like why i was not able to scan in MP although had full signal, it is cause the tuner cannot transit between specific frequencies, so if it happens that my previously scanned/watched frequency was a "problematic" one then i wont be able to scan or watch another one that is not compatible.

    To answer ur question @mm1352000 when i move to the problematic frequency i will get a time out, if i retry the same frequency i will get another time out, if i try a different channel but on the SAME frequency i will still get a time out. But if i try another channel that is NOT on the same frequency but is on the same satelite then everything works.

    After that i ca tune again to the problematic frequency and it will work.

    Mind u the problematic frequency becomes a problematic one only if the previous frequency is a specific one.

    The distance the motor has to cover has no influence since i can tune apart as much as 15 degrees and still get everything working, yet i can replicate the issue in small distances like from 19.2 to 13.0 which is half the distance of a working path.

    I will post later the frequencies that give me issues.

    In other words i think something is wrong with the card or the drivers, ofc it could still be MP/ProgDVB issue but i doubt it unless u suspect something.

    I did try an older version of TBS drivers but the issue remains.

    If u tell me that there is nothing to be done i will mail Steven from TBS with my findings and i really hope he has an idea or even better a solution.

    Thanx in advance.

    Sent from my GT-I9205 using Tapatalk
     

    mm1352000

    Retired Team Member
  • Premium Supporter
  • September 1, 2008
    21,577
    8,224
    Home Country
    New Zealand New Zealand
    Sorry, I'm struggling to follow your explanation. Especially statements like this:
    Mind u the problematic frequency becomes a problematic one only if the previous frequency is a specific one.

    I will post later the frequencies that give me issues.
    This could help me to understand.
    I don't think I should say anything more until you do this.

    If u tell me that there is nothing to be done i will mail Steven from TBS with my findings and i really hope he has an idea or even better a solution.
    I'm not going to say that there's nothing to be done when I don't understand the situation.
     

    Kotik

    MP Donator
  • Premium Supporter
  • March 19, 2009
    699
    485
    Athens
    Home Country
    Greece Greece
    Sorry, I'm struggling to follow your explanation.
    Me too:)

    So here lets hope a better explanation.

    We got 5 channels:

    1. DMAX
    19.2E
    12480 V DVB-S
    MPEG-2 27500
    3/4

    2. THT
    13.0E
    11034 V DVB-S
    MPEG-2 27500
    3/4

    3. CTC
    13.0E
    11034 V DVB-S
    MPEG-2 27500
    3/4

    4. 1TVRU
    13.0E
    12597 V DVB-S
    MPEG-4 27500
    3/4

    5. TVCI
    4.9E
    12380 H DVB-S
    MPEG-2 27500
    3/4

    We start watching channel No.1 (working fine), we then tune to channel No.2 (not working, we will get a no signal or no pmt), we switch to channel No.1 again (working fine), we then switch to channel No.5 (working), we switch then to channel No.2 (working), we then switch to channel No.1 (working), we switch to channel No.2 (not working), we switch to cannel No.3 (not working), we switch to channel No.4 (working), we switch back to channel No.1 (working).

    As u can see from 1 ----> 2 or 3 (both located under same transponder) always fails, any other combination will succeed.

    I did add more channels from 19.2 to see if all of them have issues, and some had some didn't, as an example if I try to tune to channel No.2 from:
    19.2E
    11347 V DVB-S2
    8PSK
    MPEG-4
    HD 22000
    2/3
    it will fail but if I switch to channel No.2 from:
    19.2E
    12422 H DVB-S
    MPEG-2 27500
    3/4
    then it will succeed.

    I hope this makes the situation more clear.
     

    mm1352000

    Retired Team Member
  • Premium Supporter
  • September 1, 2008
    21,577
    8,224
    Home Country
    New Zealand New Zealand
    As u can see from 1 ----> 2 or 3 (both located under same transponder) always fails, any other combination will succeed.
    Okay, let's focus on this part.

    Do you think the problem is with 1 (since you can switch to 2 or 3 from other channels), or do you think it is with 2 and/or 3?
    Can the failure be either "failed to lock in on signal" or "PMT not found", or is it always one of them?
    In physical terms, is the failure explained by incorrect positioning of the motor/dish, or is it something else [unknown]?
     

    Kotik

    MP Donator
  • Premium Supporter
  • March 19, 2009
    699
    485
    Athens
    Home Country
    Greece Greece
    I dont think the problem is connected to the channels themselfs but rather to a property these channels have, something in the line of from V polarisation to H or something like that, this was actually what i just tested to make sure it is not a switch between satelites during which switch we also change polarisation, apparently it is not since both 1 and 2 are V and i added a channel on 19.2 that is also V and was able to switch to 1 or 2 from it. So it is something else.

    Regarding the physical aspect i will have to check.

    I will visit my roof in 5 minutes and try to switch from 1 to 2 and see if the dish is moving to 13.0E.

    Sent from my GT-I9205 using Tapatalk
    --- Post updated ---
    EDIT:
    Regarding the physical status:
    In ProgDVB the dish ias always moving to the right position, in MP the dish will move to the right position only the 1st time after that the GoToX position for 13.0E is totally messed and each time u switch to a channel that is on 13.0E MP will send the dish towards 4.8E to be more precise I would say 4.7E.

    I know it makes no sense but this is exactly what is happening, I then have to store again the position for 13.0E from the diseqc TAB if I want to tune to channel that are on 13.0E.

    In MP the error will be a PMT error.
     

    Kotik

    MP Donator
  • Premium Supporter
  • March 19, 2009
    699
    485
    Athens
    Home Country
    Greece Greece
    EDIT2:
    Regarding the physical position, sometimes in MP it will randomly go to a wrong direction and report No Signal, so when I see no signal that means it didn't move to the right direction, but when I see No PMT it means it moved to the right direction but could not find PMT.
    FYI No Signal is rare and hard to reproduce and it is random, has nothing to do with specific channels, or at least I didn't notice smething like that, while no PMT well is something I can reproduce 100%, at least for now.

    EDIT3:
    In case of a No Signal error I always have to move the dish to a different satellite and then move it back to the right position and after that it will work, regarding the No PMT I don't have to move the dish to another satellite, all I have to do is tune to a different channel on another transponder but on the same satellite and it will work.
     

    mm1352000

    Retired Team Member
  • Premium Supporter
  • September 1, 2008
    21,577
    8,224
    Home Country
    New Zealand New Zealand
    In ProgDVB the dish ias always moving to the right position, in MP the dish will move to the right position only the 1st time after that the GoToX position for 13.0E is totally messed and each time u switch to a channel that is on 13.0E MP will send the dish towards 4.8E to be more precise I would say 4.7E.

    I know it makes no sense but this is exactly what is happening, I then have to store again the position for 13.0E from the diseqc TAB if I want to tune to channel that are on 13.0E.
    In my opinion understanding and solving this part of the problem should be the first priority. If the dish is in the wrong position, other problems will follow.

    So, on this subject...
    First, I'm a little confused. You say here that the "GoToX position for 13.0E is totally messed...". This seems to contradict your earlier reply when you said that tuning from 1 -> 2/3 could fail, but then immediately tuning 1 -> 5 -> 2/3 would succeed. How could this work if the position is messed up? ...or did you leave out the re-save/re-scan part of this test (between failure with 2/3 and succeeding with 1 -> ...)?

    The explanations I can think of for a motor moving to the wrong position are:
    1. TV Server is sending the wrong command due to wrong position stored in the database.
    2. TV Server is sending the wrong command due to a bug in the DiSEqC handling code.
    3. The commands received by the motor are corrupted due to insufficient delay between the motor command and the continuous 22 kHz tone.
    4. The motor's stored positions are somehow getting corrupted.

    We can check #1:
    1. Delete all channels.
    2. Use your normal procedure to create a minimal set of test channels - ideally only 2 transponders.
    3. If possible, confirm that the channels are all working correctly without messing up the stored positions.
    4. Use the import/export section of TV Server config to export the TV channels.
    5. Do whatever you have to do to mess up the stored positions.
    6. Use the import/export section of TV Server config to export the TV channels again.
    7. Zip and attach the 2 exports, and tell me which channel(s)/positions you think are messed up. I can check if there are any differences before and after.
    #2 is harder to check... but since you didn't have this problem with the SkyStar HD2, I doubt that #2 is involved.

    There's no way to check #3 without an oscilloscope or equivalent to monitor the commands on the wire. However if this corruption were happening, I'd expect you'd find that moving the motor/dish by switching to low band channels (transponder frequency less than 11700 MHz) would be more reliable. That doesn't seem to be the case in practice. When you try to go from 1 -> 2/3, that is switching both motor position (19.2E = pos 1 --> 13E = pos 2) as well as high band (22 kHz on) to low band (22 kHz off). I'd have expected this to work more reliably. Therefore I also doubt that #3 is involved.

    Again, I don't think there's any way to test #4. I can only suggest that maybe the motor gets confused with the mixed use of USALs and GoToX commands.
     

    Kotik

    MP Donator
  • Premium Supporter
  • March 19, 2009
    699
    485
    Athens
    Home Country
    Greece Greece
    @mm1352000 thanx for the suggestions I will start working on them and let u know.

    I will experiment and apply ur suggestions and let u know. For the moment I would like to express my gratitude towards ur support, even though the problem still remains at least I feel I am not fighting it alone:)
     

    Kotik

    MP Donator
  • Premium Supporter
  • March 19, 2009
    699
    485
    Athens
    Home Country
    Greece Greece
    So back with info.

    Situation is as following, there are currently 2 issues and they are not tied with each other:

    No Signal ----> Always results from a wrong dish movement.

    No PMT Found ----> This is a tricky one but read my findings and maybe u will understand what is going on.

    Let me 1st focus on the No PMT Found issue, I spent pretty much half the day at my roof top with my TeamViewer, I did notice that 90% of the time moving from a channel (not all channels but many) that is located on 19.2E to 13.0E will result into a No PMT Found error. But there was a 10% chance that it will succeed, in order to troubleshoot it I decided to do a manual dish movement from the MP DiseqC tab, so I tune a channel on 19.2E from manual control tab, then I stop timeshifting and I go to the DiseqC tab, over there I manually send the dish to 13.0E and then revisit the manual control tab and start timeshifting a channel on 13.0E. It works 90%, note that the 10% failure is the No Signal bug (read on regarding that one).

    The results of these tests made me thinking, then I remembered that 2 days ago I actually thought that I fixed the issue when I replaced the LNB and for a brief time I was able to switch between 19.2E and 13.0E with a higher success rate (around 50%), I tried to remember what could have changed and why my success rate went again back to 10% from 50%, the reason is that I adjusted my LNB after placing it to get higher signal quality and after that adjustment my success rate dropped again, so how could higher signal quality affect the issue? At that point I immediately got an idea, what if the issue is that MP/TBS is trying to find PMT while the dish is still moving? What if there is a Lock On happening before the dish arrives into its proper location?

    To test it out I used my hands:) I moved the dish to 19.2E tuned to a channel and then switched to a channel that is located @ 13.0E BUT when the dish started moving I covered the LNB with my palm and waited out the whole movement, when the dish parked I removed my palm and lord and behold, NO error, crystal clear signal with no issues. Sounded too good to be true so I did it a couple more times. 100% success ratio. I then combined this finding with the accidental 50% increased chance of success due to LNB adjustment and it made sense.

    To make sure I am not hallucinating I tried out my theory without the help of my palm:) I changed the LNB skew in a way that the signal quality is low, yes LOW, my theory at that point was the following: we are sending a command to move the dish but at the same time send the tune command, as a result if during the movement there are other frequencies found the tuner will lock on them and MP will start searching for PMT, but there wont be any cause the dish is still moving. By blocking the LNB with my hand I prevent the tuner from locking into any signal while the dish is still moving, this will also explain why manually moving the dish from the DiseqC tab is working, cause during that move we are not trying to tune to any channel.

    So back to the part of changing the LNB skew, the reason I did it was to see if I can get as low signal as possible and prevent the card from locking into anything while moving, cause this way the signal quality of non relevant frequencies will be too low to lock into and the only frequency that will work will be the one that the motor will stop on. The results are positive, by adjusting the skew I got a signal quality of 50 out of 100 for a specific channel that I wanted to test, after that switching from 19.2E to this specific channel always worked.

    Now I am not sure whether the tuner is locking into irrelevant frequencies while the dish is moving or it is locking into the right frequency but too early, but I am pretty sure it is a timing issue, it is an issue of how we/TBS tune into a channel.

    I don't know exactly the logic behind the motor support inside MP, so I got no idea how to work around that. At this point I am not even sure if my theory is correct but it sure does explain lots of issues that I am having.

    Of course sitting at my roof top and blocking my LNB each time I want to switch a channel is not a solution, neither is purposely lowering the signal quality, as this prevents me from finding channels on other satellites that already got low signal at my region.

    Regarding the No Signal error, this one is random, and I mean totally random, and it has a 10% - 20% of happening, for some strange reason the dish will either stop half way or will move to the opposite direction. I got no idea or theory as to why it is happening but @mm1352000 I can confirm that it is not the GoToX positions that are getting corrupted. Moving to a different sat after the error will result in proper movement and then I can move again to the previous sat and it will work, so this means the GoToX positions are not corrupt but it means that the command itself is either sometimes randomly wrong or it is confusing the motor somehow.

    So @mm1352000 does any of this make any sense to you?
     

    Users who are viewing this thread

    Top Bottom