SWM Question

Status
Please reply by conversation.

PBX_Guy

SatelliteGuys Family
Original poster
Dec 27, 2008
105
0
Texas
D* Offers a triple SWM ODU to catch just the 99/101/103 birds. OK 3 birds is all you need when your Locals are on Ka (mine are)

So a question...... If you're working with a legacy AU9 dish can you get away with running only 3 coax downleads to your SWM-8?

I've been told that I've got to have 4, but I'd like another opinion and maybe an explanation why.

Thanks!
 
I think youll be perfectly alright with only one.

You obviously did not read my post. I do not have a SWM ODU. I have a conventional AU9 dish (Pre-SWM technology). It has (and has always had) 4 coax runs coming into the building.

Once more the question, worded more explicitly.... Do I need 3 cables or 4 cables to drive the inputs (plural) of a SWM-8 stacker? I do not need the 110/119 birds. Again the question, can I achieve what I need from just the 99/101/103 sats and drive a SWM-8 stacker with only 3 inputs or do I have to have all 4?

Hint: ONE is not the answer.
 
Thanks. That's what another told me. Do you happen to know why? We've got over 5,000 feet of trunk cabling to put in and it'd sure save a lot of wire & coinnectors & taps if we could drop the backbone to 3 trunks stead of 4.
 
You need four lines because each line carries one of four combinations from the dish. These are 13v/0Hz, 18v/)hz, 13v/22kHz, and 18v/22kHz. These combinations of voltage and tone switch the LNB to a certain satellite.

Without a multiswitch or SWM module it's the receiver that sends these to select the satellite. With a multiswitch or SWM module, each line is locked in to one of the combinations. Then the receiver sends the signal and the multiswitch/SWM selects the correct input. There's more going on with the SWM, but this is good enough for a basic explanation

Now one solution could be to connect the SWM module at the dish and run one line in. The problem with this is that you won't be able to use any legacy (non-SWM capable) receivers.
 
Thanks. That's what another told me. Do you happen to know why? We've got over 5,000 feet of trunk cabling to put in and it'd sure save a lot of wire & coinnectors & taps if we could drop the backbone to 3 trunks stead of 4.

Are you putting in a MDU installation (5000 ft is a LOT of cable)? If so, you should look into a MFH2 system, which uses multiple SWM8s.
 
Are you putting in a MDU installation (5000 ft is a LOT of cable)? If so, you should look into a MFH2 system, which uses multiple SWM8s.

Yes. -and doing it ourselves- but don't roll your eyes as we did our own 24 channel Blonder-Tongue head-end and distribution several years ago with great success. Tight system, spectrally clean, BIDA 450-30 trunk amps and all unused ports terminated. This will not be our first rodeo. :)

My building is unique (aren't they all?). Seriously, it's a 9-story "shoebox" 640 ft long and 85' wide. There are 40 (total) wiring closets, 10 in each of 4 vertical risers (including the basement). I'd like to cable the whole thing out (including passives) but only heat up the front end and only as much trunk as necessary to reach the (so far) first and only client via riser #2 down on the 3rd floor.

MFH2 right now is overkill, but I can see far enough ahead to know that's where I'd like to be at the end of the day.

I can go cheap & dirty with just a SWM ODU and a single run of RG11 plenum to the 1 lonely client and forget all about building a backbone, but that's setting a dangerous precedent because as we all know one client soon leads to two and so on and I'd really prefer not to end up with 50 dishes on the roof 5 years from now. Right now there's just under 100 NTSC televisions on "our" cable system and we've just had our first request for HD to a client who is installing three 65" displays in the master control center of a major energy company (my employer).

I've seen diagrams of several MFH2 systems (Sonora based) and that is unarguably the way to do it, (not necessarily w/Sonora) but regardless whose electronics we use that's spending cubic dollars when today there's only a single prospective end user who for the near term at least will have only 3 tuners. If I'm going to do something like an MDU, I've got to be able to eat that elephant one bite at a time from the standpoint of rolling out the electronics and paying for it all. I have the cable (RG11/Quad/Plenum) & connectors to lay-in that much of it now and taps are cheap enough that they won't break me to do that too. I figure on a single SWM-8 in every other wiring closet (staggered floors, 2nd, 4th, 6th and 8th)

I've got to work up a viable design where I can walk-in the electronics to the system strictly on a 'demand' basis. I have all my loss and slope numbers for the cable lengths. The horizontal run is what's eating my lunch because even with RG11 I have one leg of it (between the #2 and #3 risers) that on paper has 10 db of slope in 15 db of attenuation. All the amps I see have only 5 db slope compensation max, suggesting I'm going to have to put equalizers in or run more amps with lower individual gain. I want to keep the trunk as flat as I can without spending a fortune to do it. I didn't see any designs that addressed this hurdle.

We are the only tenant in the building; 1400 employees, 550,000 sq. feet of office space. (12+ acres under roof). We currently have two D* business accounts and 33 GAEBO IRDs (24 in the head-end with B/T modulators & combiners) plus 9 in dedicated applications around the building (so we're reasonably familiar with the amplification requirements of pushing L-band signals thither and yon down long runs of RG11 and winding up with a presentable signal at the other end.. I was simply choking a bit on the concept of quadrupling the L-band backbone :) and curious as to where we could trim unnecessary expenses. Thoughts welcome.
 
Last edited:
You need four lines because each line carries one of four combinations from the dish. These are 13v/0Hz, 18v/)hz, 13v/22kHz, and 18v/22kHz. These combinations of voltage and tone switch the LNB to a certain satellite.

Without a multiswitch or SWM module it's the receiver that sends these to select the satellite. With a multiswitch or SWM module, each line is locked in to one of the combinations. Then the receiver sends the signal and the multiswitch/SWM selects the correct input. There's more going on with the SWM, but this is good enough for a basic explanation

Now one solution could be to connect the SWM module at the dish and run one line in. The problem with this is that you won't be able to use any legacy (non-SWM capable) receivers.
Thanks for explanation. The existing legacy system will remain as-is, undisturbed for now. Too much visibility to interrupt it.

Running only one line in will put out the fire, but this is a 550,000 sq. ft office complex. I cannot design for one lone drop today without planning for the 100 more that are certain to eventually follow. Besides, if we do it on the cheap for client #1 how then do we go back to management and tell them we can't do the same thing for everyone without winding up with an antenna farm on the roof? If we do it right the first time we won't have to re-do it.
 
For further explanation on the loss/slope issue in the horizontal run, there are 4 risers but not equidistant from each other. Risers #1 and #2 are 150' apart. Similarly, Risers #3 and #4 are 150' apart. The problem stems from the fact that Risers #2 and #3 are 330' apart.

One thought would be to build two separate head-ends with an AU9s dish at each end of the building. While doubling the head-end hardware I would eliminate four 300' runs of coax and at least one amplifier & equalizer and solve (eliminate) the slope problem in the long cable run. I'll have to run the numbers. I'm thinking it'd be less expensive to do that, plus it would allow me to cut the initial $ outlay in half as well.
 
Good idea. Possibly one Slimline on each riser (4 head-ends) would simplify things a bit further. Dishes are cheaper than a lot of the peripheral equipment.
You make a good point Bob. Unfortunately, existing MPOE penetrations are nearest risers #2 and #3, so 4 slimlines vs 2 would not improve signals more than 3.5 db (splitter loss) in risers #1 and #4 compared to using only 2 with branch feeders (The amount of cable would be approx. the same. Either 2X as much downlead or 2X as much for the branch trunks). I can compensate for the slope in 150' of RG11.

4 Slimlines would need 4 PI/PL units plus 4 amplifiers
2 Slimlines would need 2 PI/PL units, 2 HFS splitters, plus 2 amps.

Net gain = equip. savings. Cabling is the same. Net signal loss is 11 db with 4.5 db of slope. I can buy 2 HFS splitters & 2 higher output slope-compensated amps for less than I can 2 more Slimlines, PI/PL's, and 2 additional amps. The problem was the slope. 2 head-ends solves it for less than I'd have spent on 1200 more feet of coax & electronics to try & feed it all from 1.

Sounds like you've done this before. Any recommendations on brand of electronics? (or anything to avoid?)

What's a good source for 2150 Mhz-rated directional couplers/taps? (range of 3 to 20 db) One vendor (!) suggested to use all 2X splitters and add pads on my inputs. Clever, but overlooks the high thru-loss (50%) of the splitter. Obviously he had splitters & pads in stock but didn't have taps... He also didn't understand what slope attenuation was. To him loss was loss. I'll refrain from mentioning names (Lum-n-Abner's Supply House, Bait & Boats Rented) :eek:
 
Status
Please reply by conversation.

highest frequency

splitter

Users Who Are Viewing This Thread (Total: 0, Members: 0, Guests: 0)

Who Read This Thread (Total Members: 1)

Latest posts