1. This site uses cookies. By continuing to use this site, you are agreeing to our use of cookies. Learn More.

How To Diplex OTA with the AT9 dish & MPEG-4

Discussion in 'DIRECTV Installation/MDU Discussion' started by Cap'n Preshoot, Aug 19, 2006.

  1. Smooth Jazzer

    Smooth Jazzer Legend

    183
    0
    Sep 4, 2007
    Several members have placed the BBC(s) and the splitter(s) indoors. That said the BBC does NOT have to be six inches away from the MSW. The rule to follow is TO place the BBC before the diplexer. In other words the BBC pigtail has to be connected to the diplexer SAT input. The RG6 cable coming from the MSW is connected to the BBC input. Connect the cable coming from the OTA cable or splitter to the diplexer ANT input. The remaining diplexer connector gets the RG6 cable coming from the H20. The other end of that RG6 cable connect it to the diplexer with two six inch RG6 coax jumpers which in turn are connected to the H20. SAT to SAT of the H20. ANT to ANT of the H20. That is the way I have it. It works like a charm and I have both the diplexer and the BBC away from rain and snow.
    Thanks
     
  2. Tom Robertson

    Tom Robertson Lifetime Achiever DBSTalk Club

    21,331
    247
    Nov 15, 2005
    JustRob,
    Splitters
    Splitters can only be used between a multiswitch (normally the one on the dish) and two (or more) multiswitches in parallel. The big keys are power passing on at least one leg of the splitter (all, if passive switches are used), and make absolutely certain the downstream switches are all aligned on the switches. The 13V/no tone inputs for all switches MUST go to the same splitter, likewise with the 18V/no tone, 13V/22KHz, and 18V/22KHz so that all the switches get the same feed from the dish that they both want.

    (The power passing splitters have protective diodes so the voltages don't loop back. Yes, the multiple 22KHz signaling can be an issue in theory, but doesn't seem to be a problem in practice.)

    SWM
    SWM is Single Wire Multiswitch. Its whole purpose in life is to do very fancy frequency shifting on individual transponders as requested by the receivers rather than whole band frequency shifting that the LNB does. Since only individual transponders are on the coax, DIRECTV can put 8 tuners worth of transponders on a single coax (plus one channel for other stuff) so up to 4 DVRs or 8 single tuner receivers can be driven off a single wire.

    And the frequencies the SWM uses are outside the OTA band, so OTA can easily be diplexed in and out of that same single wire. Very nice!

    The only sad part of the SWM tale is that it is just now being deployed in a a few locations. DIRECTV would normally be happy to give you one (or two if needed). I hope existing DIY'ers will be able to order them, perhaps around the first of the year?

    OTA with regular Switches
    That can be done as this thread is all about such. BBC, indoors, before the OTA diplex in.

    Cheers,
    Tom
     
  3. NR4P

    NR4P Dad

    6,382
    280
    Jan 15, 2007
    Sunny Florida
    I posted a detailed diagram on diplexing the Slimline dish and HR20, but H20 will be identical, just with one SAT input.

    It's shown here.

    http://www.dbstalk.com/showthread.php?t=103790

    And I hope Tom's point on the SWM is true. I would like to change my system to a SWM8 when they become available.
     
  4. veryoldschool

    veryoldschool Lifetime Achiever Staff Member Super Moderator DBSTalk Club

    42,684
    349
    Dec 9, 2006
    JustRob,
    Oh goodie, Can I pick on a double E this morning? I see I'm late to the party this morning. :lol:

    Both the WB68 and the SWM have four inputs from the dish. These are the 13/18 volt tone/no tone combinations needed to power the dish. Since these are the same, they can be split to the same input on another switch with power [DC] passing splitters. Since the SWM is powered if the splitter is only power passing on one leg and this is connected to the passing port, the LNB stays powered even though the WB68 leg isn't driving the LNB.
    SWM [single wire multi-switch] can by design split it's output to several receivers that are SWM capable. The HR-20 in "SWM mode" only needs one SAT connection.
    The SWM tunes to the transponder required by each tuner and outputs just that one the SWM channel fro each receiver/tuner.
    SWM5 works with 5 and SWM8 works with 8.
    All of the SWM output channels are above the OTA band [SWM converts frequencies along with tuning to SAT transponders].
     
  5. jefbal99

    jefbal99 Hall Of Fame

    2,838
    0
    Sep 7, 2007
    Just a note to add, the SWM8 has an OTA input on the switch, while the SWM5 requires diplexers on both ends to combine, then split out the signals.
     
  6. Nov 2, 2007 #206 of 323
    JustRob

    JustRob Cool Member

    23
    0
    Aug 7, 2007
    Thanks for the replies, everyone. My inquisitive, "what-if" side has some unresolved questions though. Perhaps once these Q's get answered, they should be stickied as some kind of massive FAQ. :grin: If I have time, I can do this if there is enough interest, and most definitely if D* can get me an SWM8 :icon_hroc

    Indeed, I see how this will work; I had overlooked the obvious implication (below). My 5x8 multiswitch (Terk) actually shows the 4 signals as you show above, to the satellite. The important thing here is to make sure to use a splitter with or-ing (steering) diodes to prevent back-powering a device powered off, and to only combine the same signal from both MS's on this splitter. I.e., don't combine MS1 13V/22kHz and MS2 18V/no-tone with a splitter to the dish.

    I suspect that there shouldn't be a problem, so long as splitters with steering diodes are used. However, I suspect if you tried to three-way split, there could be a problem with the modulated signals (22kHz) due to the additive signal wander (assuming the 2-way split). It's hard to say without knowing the LNB discrimination circuitry. After analyzing this with Octave assuming two signals with or-ing diodes in the splitter/combiner, it's clear that even with a 100ppm oscillator at opposite ends of the tolerance, a 22kHz signal will still reach the satellite (assuming 1Vp 22kHz oscillator and 13V DC signal).

    I knew about the multiple channels but I didn't realize until now the implication here; up to 8 channels multiplexed (frequency-up/down-shifted is probably more accurate). So that makes me wonder the bandwidth of each band, and if there are any repercussions regarding the tilting (frequency response) of all RG6/RG59 cable. Does that mean, you could get a weak signal that would then be using the upper-end of the usable coax band, and then it would be even weaker if it got the penalty of having to use the more attenuative frequency band of the coax. Also, because of the frequency shifting, you must also filter the out-of-that-band harmonics, which means possibly more insertion loss or distortion, especially if the components used are cheap (low-Q).

    Since the SWM8 has two outputs (4 channels each), the signal must be split, so I wonder if the OTA would be better off using a separate feed with power distribution amp using an LNA.

    Yeah I wish I could have gotten in on the "beta testing" that was available until a few months ago. I have the slimline installed, and I explicitly stated on the work order to bring an SWM8 when installing the dish. The installer knew nothing about it, neither did his bosses, but DirecTV knew that the "beta testing period is over". Great, but now we're all in this limbo awaiting for the other foot to drop. I would gladly be a tester, as I did all my own wiring, crimping, and have a firm grasp of requirements and limitations. Heck, I use some of this stuff at work :)

    ======================================

    Yes, that's what I would call diplexer H*LL. Every time you add a diplexer, you're reducing the signal by approx 1-3dB (especially the crap ones sold in retail chains). The highest quality diplexers I have seen have 0.25-0.5dB insertion loss, >50dB rejection, and they're made by Soontai (Taiwanese company). We use them at work for for MoCA and CATV feeds. Some people can tolerate the loss, but I have an attic antenna and don't have enough SNR margin for more loss which will affect the ATSC 8VSB signal quality.

    With the SWM8, you'll need a diplexer at each receiver to receive the OTA 8VSB ATSC signal. The HR20 automatically (after power cycling or cable removal/insertion) determines it is connected to an SWM.

    ======================================
    Sure, I'm game. I've got a thick skin, you know us E^2s are always trying to have the last word :lol:

    So, I think we can figure out how to get the original, 3 LNB (MPEG2) satellite reception. You must introduce some new algorithm for getting the other two (MPEG4) transponders (LNBs), right? At worst case, you'd need four different (odd/even and which of two LNBs) combinations. Wonder what algorithm is used?

    ==================================

    Thanks for the confirmation. I have to go SWM8 anyway, since I have three DVRs. So, I would think that it would be better to use more cables than a single, split cable, assuming that the SWM8 starts out using the lower modulated band and then starts using the higher frequency bands per allocation. In this example, I mean if you have two HR20s, it would seem to be better to use both FTM outputs separately than one FTM (other than splitter losses). That assumes that it doesn't just simply use whatever the SWM8 hasn't already allocated to another band; if that's the case then the only thing you buy by using a single wire for both receivers is that you can use splitters and daisy-chain the coax cable from one room to another (but who else can do this except the CATV and MoCA/UWB/HPNA guys).
     
  7. Nov 2, 2007 #207 of 323
    veryoldschool

    veryoldschool Lifetime Achiever Staff Member Super Moderator DBSTalk Club

    42,684
    349
    Dec 9, 2006
    JustRob
    So whether you use a SWM 5 or 8 you need to diplex out the OTA at the HR20 to the input.
    Connecting a second SWM coax to tuner #2 will do nothing.
    SWM compatible receivers detect the SWM on boot up [and shut down #2 input].
    The 5 LNB dish works the same as the 3 LNB [ 2 control voltages, + tone or not]. The Ku SATs use 950-1450 MHz & Ka uses both 250-750 MHz [requiring the BBC conversion] and 1650-2150 MHz.
    So where there used to be one 500 MHz "block", there are now three.
    Power passing splitters won't have a "back feed" problem since both legs are going to the same input of a multi-switch [with the same voltage].
    This is why you: need a multi-switch, and can't split a SAT feed to multiple receivers [that would have varying voltages]
    The SWM does have an amp [gain] in it. How the output frequencies are determined is still unknown.
    It is made for RG59 cabling to make D* more compatible with houses wired for CATV.
    The SWM does a far amount of conversion. Where a multi-switch will sent all of the even/odd transponders [500 MHz], the SWM send only the required transponder.
    So if filters, converts and distributes the require transponder to each receiver on it's two outputs [FTM 1 & 2].
    Since a double E needs the last word [ :lol: ] what's next? :)

    OH yeah: there was a posting about a SWM 11 and the output were from 950 MHz to 2150 MHz.
     
  8. Nov 2, 2007 #208 of 323
    JustRob

    JustRob Cool Member

    23
    0
    Aug 7, 2007
    Thanks for quick response and detailed answers. I saw that in another thread. BMoreRavens picture didn't show the back-side of the HR20, or if a splitter was required (for both tuners) or diplexer. It's clear now I think.

    Makes sense now. Perhaps not enough people have shown pics of their HR20 coax connections and diplexer.

    Gotcha, so that could mean then, that the HR20s are intelligent enough to tune and pick up what they need from this larger band. I assume the out-of-band signal quality for each band is low enough as to not affect the adjacent bands SNR.

    You'd think it'd be safer to just have a 5-2300MHz splitter, and only one leg with power steering diodes, to be safest, than to use a splitter that is 5-2300MHz and both legs diode-steered.


    Heck, give it to me; I have a 3GHz 75-ohm network analyzer at work!!:D If the D* band starts at 950MHz, sounds like it won't interoperate with either MoCA at 1GHz or the new DOCSIS 3.0 (ends at 1GHz). Oh well... I wonder if the 950 is start of guard band or what the adjacent floor noise has to be in order to not affect D* signal in the same cable.

    :goodjob:

    Hmm, don't think I quite understand how this works entirely then. You have so many bands, and each band is shifted to a frequency within the 950-2150MHz band. However, this can't be the end of it as there is a limitation to the number of receivers each output has. If it's simple (easier said than done) frequency shifting bands, and I guess the HR20 must map each channel to whatever band (and channel within that band) the SWM output has chosen. Must take forever to scan all channels then, but hopefully with a UPS, shouldn't need to do it too often.
    :scratch:

    I'm wondering if there is tilt in response of SWMs. I doubt that will help significantly since you'll still be adding a bit of noise with the amplification. I'd also like to know if I'd still need to keep my 20dB OTA RatShack amp; would be nice if the SWM had a higher quality LNA. You gotta imagine that SWM internally must be an RF headache!
     
  9. Nov 2, 2007 #209 of 323
    carl6

    carl6 Moderator Staff Member DBSTalk Club

    12,520
    958
    Nov 15, 2005
    Seattle, WA
    The SWM outputs each provide access to all channels. On the SWM8 you can put all 8 tuners on output 1, all 8 tuners on output 2, or have any mix between them. The only difference between the two outputs is that the power inserter must be connected to output 1.

    So essentially the two SWM outputs are nothing more than a diode isolated splitter.

    Carl
     
  10. Nov 2, 2007 #210 of 323
    veryoldschool

    veryoldschool Lifetime Achiever Staff Member Super Moderator DBSTalk Club

    42,684
    349
    Dec 9, 2006
    It's my understanding the this is more done in the SWM, where the SWM is a "pre-tuner" using a "IF" for each connected receiver [tuner in each]. There is also one other signal [IF] that seems to be the control. A SWM 5 has 6 & a SWM 8 show 9, on the HR20 test screen.

    The only problem with a single leg power passing splitter is if you use the Zinwell WB68 multi-switch [the most common for Ku/Ka]. These are passive and require the receiver to supply the LNB voltage. If the passing leg doesn't have a tuner suppling the voltage, then the non-passing leg tuner won't be able to tune to the transponder.
    If one of the multi-switches is powered [or a SWM] and is connected to the passing leg, there isn't a problem [duh].

    As ^ I don't think they're "bands" but "A" transponder sent over an IF [about 40 MHz bw] to the tuner.

    From reports, the OTA internal diplexer is [very] lossy, so some use their own external diplexer. I don't figure D* spent much on the OTA path [the SWM5 doesn't even have it, only the SWM8].
     
  11. Nov 2, 2007 #211 of 323
    JustRob

    JustRob Cool Member

    23
    0
    Aug 7, 2007
    Okay, now this is beginning to make sense. But I don't think you could have that many channels in 40MHz b/w. At one time, when I had CATV, they could get 4 SD channels into one 6 MHz carrier, and I'm not sure about HD (presumably more b/w). Since satellite has the fading channel problem, I'm guessing it's not quite as nice of a medium as HFC, therefore it needs more b/w due to lower symbol rate and error recovery techniques. Do ou mean perhaps 400MHz?

    Now that makes a lot of sense. I can see that the SWM8 might not do that good of a job, since it may rely on a strong signal or CATV feed and it might have an insertion loss of perhaps 10dB if un-amplified (not counting the splitter).

    Since the termination must be capped when the 2nd output is not used, this makes me think you'd never want to use just one leg for 8 tuners on the SWM8 as you'd be just wasting power across a 75ohm load (so why not put it to good use and get less loss across a 2-way splitter than a 4-way). Furthermore, I would guess that the OTA, if unamplified with LNA, must be at least 3.5dB or more of attenuation (thus why people said it was more lossy than using their own separate feed).

    This is good; thanks for keeping my brain working! :grin:
     
  12. Nov 6, 2007 #212 of 323
    robq391

    robq391 Legend

    119
    0
    Nov 6, 2007
    I have a question that may have been answered but I cannot find it.

    I have my OTA/SAT diplexed into single feeds w/ BBC's @ output of multiswitch before diplexer, exactly as described here. Works beautifully w/o any discernable SL in the SAT levels. Approximately 150' from dish to M/S & 75'-100' from M/S to receivers. Using WB68 and no in line amps.

    I want to get rid of my splitter for my OTA & increase the number of OTA feeds.

    Quetion is can I use the SDC5401 (made by zinwell & perfect vision) off 4 outputs from my WB68 to achieve 4 diplexed feeds?

    If this would work, I could get rid of a splitter feeding 2 diplexers & combine this into one step as well as give me 2 more feeds.

    Thanks in advance
     
  13. Nov 6, 2007 #213 of 323
    veryoldschool

    veryoldschool Lifetime Achiever Staff Member Super Moderator DBSTalk Club

    42,684
    349
    Dec 9, 2006
    :welcome_s to DBSTalk
    Since you would need to mount the BBCs to the input of that multi-switch, I don't think you can use it as I doubt you could control the BBCs anymore. They need to switch in & out and the control signal won't get through the switch to them, is my guess.
     
  14. Nov 6, 2007 #214 of 323
    robq391

    robq391 Legend

    119
    0
    Nov 6, 2007
    Thanks VOS.

    I wasn't sure because Zinwell calls it a multiswitch & PerfectVision calls it a splitter/combiner. Same model # from both companys.

    They aren't that expensive so I might just order 1 and see what happens.
     
  15. Nov 6, 2007 #215 of 323
    veryoldschool

    veryoldschool Lifetime Achiever Staff Member Super Moderator DBSTalk Club

    42,684
    349
    Dec 9, 2006
    If you do get it to work, please come back and post it. :)
     
  16. ITSec_Guy

    ITSec_Guy AllStar

    68
    3
    Sep 23, 2007
    Tampa, FL
    GOD BLESS YOU! This is the first time I have seen this tread but I have been TRYING to tell people that you can DIPLEX the MPEG 4 signal! I just came to DirecTV last month but I had been doing it with Dish as long as MPEG 4 had been OUT!
     
  17. veryoldschool

    veryoldschool Lifetime Achiever Staff Member Super Moderator DBSTalk Club

    42,684
    349
    Dec 9, 2006
    It's not the MPEG-4 but the use of Ka-lo band that is the issue and where the BBC is mounted [since this is what upconverts the 250-750 MHz frequencies that are the problem].
     
  18. RobertE

    RobertE New Member

    8,024
    0
    Jun 9, 2006
    I still have to wonder if this hack will work when D11 is live.
     
  19. veryoldschool

    veryoldschool Lifetime Achiever Staff Member Super Moderator DBSTalk Club

    42,684
    349
    Dec 9, 2006
    I can't think why if it works with D10, that it wouldn't with D11 [same voltages but just no tone, and same frequencies].
     
  20. texasbrit

    texasbrit Well-Known Member

    5,211
    105
    Aug 9, 2006
    The signals from D11 will look just like the signals from D10. Should have no effect at all.
     

Share This Page