DBSTalk Forum banner

Question re; swm splitters

2922 Views 20 Replies 6 Participants Last post by  HarleyD
This may be a dumb question and I may just be in denial of what I already know but how big of a signal loss issue is having "too big" of a SWM splitter with multiple terminated outputs.

Here is specifically why I am asking.

I got a great deal on a SWM-16 on eBay (SWM unit, 29V PI and two 8-way splitters for $59 delivered) which I installed around a month ago. It was packaged with the 8-way splitters even though I knew I didn't need that many outputs. It was a good deal and I jumped on it.

Since that was the first purchase against my "master plan" of incrementally expanding my system and replacing all the non-SWM legacy boxes to build out a righteous MRV system, I presently have only 2 outputs on one of the splitters connected and none of the outputs on the other splitter are connected at all. The two legacy outputs are connected to a 3x8 multiswitch feeding my five legacy STBs.

All of the unused outputs on the splitters are properly terminated but ever since I installed the SWM-16 I am seeing intermittent tiling and pixellation on all channels on the two units that are currently part of my SWM cloud (HR20 & R22). Both of these units are at the end of long cable runs as well.

Would this be the result of execessive signal loss across all of these terminated outputs on the two 8-way splitters?

Even once my plan comes to fruition (one HMC, four SWM capable DVRs and two HD receivers) I would still only need two 4-way splitters. I've been putting off replacing the 8-ways with 4-ways until my next order from Solid Signal (probably an R22) but I may just go ahead and make that change now if that is at the root of the signal breakup I am seeing.
1 - 20 of 21 Posts
Personally I always recommend using the smallest possible splitter. As a test I would connect one of the receivers directly to the SWM1 output and the other to the SWM2 output (as you currently don't need more than two receivers anyway) and see if your pixelation issues are resolved. If not you may have a defective SWM.

I'm not sure what you would consider a long cable run but, even though you "should" get good performance over 250' of cable after passing through an 8-way splitter, I've personally seen problems with runs as "short" as 150'.
Thanks. I will try that.

The runs from the SWM to the STBs are right at 150'. They are also copper clad cores, not solid copper.

I have all the hardware in the attic of a 2-story house with the cables running from there to the individual STBs, all but two of which are on the first floor. The cables from the dish to the SWM are about 30'.
copper coated steel (CCS) is ok for satellite because RF transmissions travel on the outside of the cable. Only DC voltage travels through the entire cable, so the cable from your PI should be solid copper.
The PI is connected directly to the power port on the SWM16.

EDIT to add: I have to check that cable. but it is about 4 inches long.

At first I thought the cable was part of the PI, now I'm not so sure. I may have made it up myself. I'll have to look when I get home.

I have everything mounted on a board within inches of one another. I done fab-ree-cated all the short run cables myself from a long piece of RG-6 I had on hand using the snap-seal connectors and stripper/crimper tools I have.
when i read about your short cables, it made me think of a d2 advantage system, and in a d2 advantage system (which is basically a large swm system) directv engineers spec out no 90 degree angles on cables, and short cables should be at least a 12" length and in a loop, to maintain proper resistance. this may or may not be a cause of your problems, but i know for a fact that 4 inch cables are not to directv specs.
Ken Stomski said:
when i read about your short cables, it made me think of a d2 advantage system, and in a d2 advantage system (which is basically a large swm system) directv engineers spec out no 90 degree angles on cables, and short cables should be at least a 12" length and in a loop, to maintain proper resistance. this may or may not be a cause of your problems, but i know for a fact that 4 inch cables are not to directv specs.
90º "angles" are a fact of life, but it all comes down to the radius of the bend, and larger is ALWAYS better.
I don't think not using "short cable" is because of "proper resistance", but has to do with wavelengths, or more precisely staying away from quarter and half wavelengths, as these can cause very strange things to the RF signal.
"Tuning" RF transmission lines IS DONE with quarter wavelengths, but in the field you shouldn't be trying to do this, since you don't have the equipment.
I can say that I have nothing approaching a 90 degree bend in any of the cables.

I'll try to post a pic later.
Well, I am mistaken. There are actually two cables that traverse a 90 degree turn, but they are radiused pretty gently.

I did take some care not to bend any of the cables too acutely.
HarleyD said:
Well, I am mistaken. There are actually two cables that traverse a 90 degree turn, but they are radiused pretty gently.
In terms of RG6, a 3" bend radius is generally considered the safe minimum. Anything more is gravy (I'm intrigued by VOS's sidebar on "bad" large radius bends) and anything less is likely to cause the cable to oval.
harsh said:
(I'm intrigued by VOS's sidebar on "bad" large radius bends)
Are you having problems reading these days?
[the quote]
larger is ALWAYS better.
veryoldschool said:
Are you having problems reading these days?
[the quote]
Apparently.


A little soft on the focus but its' 120 degrees in the attic and I'm holding a lamp in one hand and my cell phone in the other.

This pic is "close enough for government work". The two cables I spoke of are on the left going frm SWM1 and SWM2 to the splitters themselves.

I could just move the splitters to soften that bend. Since I'm going to ultimately replace them anyway ti wouldn't be a big deal. The two legacy outs to the Eagle Aspen 3x8 are probably sharper bends but I am not having issues with any of those units.

It's not necessarily an installation to be proud of...but I'm not ashamed of it either.
See less See more
A 1" radius is anything but gentle.

I'd suggest you consider Stuart's recommendation of offing the splitters entirely and add smaller splitters as necessary.
I think ultimately I'm going to order the 4-way splitters from Solid Signal since there is no way I need the 8-ways.

Then when I put them in I'll re-orient them 90 degrees from how the 8-ways are installed and place them on either side of the PI with the inputs to the splitters pointed horizontally straight at the SWM.

That way the cables from the SWM to the splitters won't need to be bent so severely. Just curved a little around either side of the PI.

In the mean time I'll just connect the runs from the STBs directly to SWM1 and SWM2
I would order one two-way and one four-way, and if you are not comfortable with making your own cables, I hear they make excellent ones. Stay away from the home improvement stores.

VOS' recommendation of a loop is the right way to go; it's possible you've already done damage to those short cables.
I've been making my own cables for a while. I'm pretty comfortable with it. I've got all the stuff I need to do it.

EDIT to add: The funny thing is, I've got many pre-fabricated 3' RG6 cables that came with things like surge protecters and what not aht come pre-packaged with really nasty bends in them fresh out of the box.

It's like thy coil them up in 4" circle, then squash them flat and put a wire tie around the middle. Nasty.
Yeah, you remember when the HR20 first came out and the B-Band converters had nasty bends in them to fit them in the box? I wonder how much trouble they were responsible for down the road.
Hello HarleyD,

After seeing the picture I think that you might locate the new splitters at the corners of the board thereby making the cable runs longer and the bend less severe.
1 - 20 of 21 Posts
This is an older thread, you may not receive a response, and could be reviving an old thread. Please consider creating a new thread.
Top