Terminating unused taps on a coax splitter

Status
Not open for further replies.

andy365

New member
Location
Norway
Questions like this has been asked on this forum before and I have read through some of them but not understood the details.
In my apartement the incoming tv/internet signal first goes through a 2 way splitter. One of the taps of this is currently open. The other one is connected to a 75 ohms cable that goes to a coax outlet that is terminated with a 75 ohms. My set top box is connected this outlet and it both has a hdmi output for tv and act as a wifi router.

The general advice seems to be this: terminate the unused port with a 75 ohms. This will prevent problems for other people in my apartement building.
What I am wondering then is:
1. will not terminating the unused port reduce the power of the signal at my outlet?
2. how bad could the problems caused by not terminating the port be?


Regarding 1: I would think that if 4mW goes into the splitter each port receives almost 2mW. The unterminated ports effect is completely reflected and divided by the other port and the incoming coax. Therefore 1mW of the reflected energy should go into the active port? If I calculate the reflection coefficient of the open port I get: R=(inf - 75)/(inf + 75) = 1. Does this not mean that the reflected signal has the same phase as the incoming signal and therefore will interfere constructively with the incoming signal? Therefore I should get 3mW of signal power with an open port vs only 2mW with a terminated port?

Regarding 2: There are 16 apartements in my building. I imagine the signal comes into the building goes trough an amplifier followed by a 16 way splitter before it gets into my apartement. The 1mW of reflected power from my example above that goes back to the 16 way splitter would then be split into 16. Therefore my neighbour would get a slightly delayed ghost signal 1/64 of the power of the original signal. Does this matter? Can this do any harm?

Thanks in advance for any answers!
 

GoldDigger

Moderator
Staff member
Location
Placerville, CA, USA
Occupation
Retired PV System Designer
The general rule is that any impedance mismatch causes reflections, and depending on the length of any particular cable with reflections you can get destructive interference at specific carrier frequencies. That in turn can both drop signal strength and deliver a delayed echo at other points in the network.
The typical cable has signals over a wide range of frequencies going in both directions on the cable.
One unterminated port on one splitter may not cause problems, but as you add multiple splitters separated by random cable lengths the chances of severe problems goes way up.
With digital rather than analog signals, instead of ghosts you will see a sudden transition from a perfect picture to total picture loss when the interference pushes the error rate past the ability for redundant information to allow error correction.

mobile
 

hbiss

EC, Westchester, New York NEC: 2014
Location
Hawthorne, New York NEC: 2014
Occupation
EC
I mean this with all due respect, but how is any of this your business? You are not the cable company. They own it, they maintain it- you don't. If you have a problem you call them.

To answer your question (from someone who worked for cable companies) yes, unused ports should be terminated as well as unused drops and wall plates. The reason is to minimize radiation of the cable signal as well as to prevent possible ingress of strong signals back into the cable plant, not anything to do with your musings.

But in real life unfortunately you will rarely find anything terminated unless there was a radiation or ingress problem. The only terminators you are likely to find are the locking ones out on the tap to prevent theft of service.

-Hal
 

gadfly56

Senior Member
Location
New Jersey
Occupation
Professional Engineer, Fire & Life Safety
I mean this with all due respect, but how is any of this your business? You are not the cable company. They own it, they maintain it- you don't. If you have a problem you call them.

To answer your question (from someone who worked for cable companies) yes, unused ports should be terminated as well as unused drops and wall plates. The reason is to minimize radiation of the cable signal as well as to prevent possible ingress of strong signals back into the cable plant, not anything to do with your musings.

But in real life unfortunately you will rarely find anything terminated unless there was a radiation or ingress problem. The only terminators you are likely to find are the locking ones out on the tap to prevent theft of service.

-Hal

If this is at some kind of building demarc I would agree with you, but if it's inside his apartment it is very much his business.
 

hbiss

EC, Westchester, New York NEC: 2014
Location
Hawthorne, New York NEC: 2014
Occupation
EC
If this is at some kind of building demarc I would agree with you, but if it's inside his apartment it is very much his business.

I would disagree with that too. People think it's their business because they make it their business and it's just not worth the trouble for the cable companies to argue with them. So if people cause problems that effect other subscribers they get disconnected. If their handi-work causes their own reception to suffer they either complain about it on the internet or pay for a service call when they finally get fed up.

Just for the record- it was decided long ago that there was no need to terminate unused wall plates unless the location caused problems due to strong RF or proximity to sensitive radio services like aircraft navigation. There is no practical way to keep terminators on them anyway since customers will remove them and never replace them. My own house has probably 10 unterminated wall plates with zero problems. Back when I had my own company doing sub contract work for the cable companies we probably installed thousands of wall plates and the distribution to go with them. It was my policy to terminate unused distribution ports on splitters and taps but we didn't have to. There were no terminated unused wall plates or splitter ports in the customer premises. There was never a problem.


-Hal
 

GoldDigger

Moderator
Staff member
Location
Placerville, CA, USA
Occupation
Retired PV System Designer
I would disagree with that too. People think it's their business because they make it their business and it's just not worth the trouble for the cable companies to argue with them. So if people cause problems that effect other subscribers they get disconnected. If their handi-work causes their own reception to suffer they either complain about it on the internet or pay for a service call when they finally get fed up.

Just for the record- it was decided long ago that there was no need to terminate unused wall plates unless the location caused problems due to strong RF or proximity to sensitive radio services like aircraft navigation. There is no practical way to keep terminators on them anyway since customers will remove them and never replace them. My own house has probably 10 unterminated wall plates with zero problems. Back when I had my own company doing sub contract work for the cable companies we probably installed thousands of wall plates and the distribution to go with them. It was my policy to terminate unused distribution ports on splitters and taps but we didn't have to. There were no terminated unused wall plates or splitter ports in the customer premises. There was never a problem.


-Hal
Unused wall plates are much less likely to be a problem since they will be terminated at their other end by a close match to the line characteristic impedance at the splitter where they originate.
Unterminated ports on a splitter are potentially more of a problem. Note that the right length cable (odd integral multiple of 1/4 wavelength at frequency of interest) will look like a dead short at the splitter end. So instead of more signal on the other taps you will get less.

And in current generation cable installations, with multiple DVRs and broadband cable modems, which have bidirectional digital data on them it is much more critical.
FWIW, the Comcast installer who last worked on my two DVR, one non-DVR smart set top box and broadband cable not only made sure that all splitter ports were terminated, he also put new ends on some of the 10 year old inside and outside cable. Then he retested the signal levels in all directions and pronounced the problems solved. And except for the box software crashing every day or two, they were. :happysad:

It was very clear that Comcast considered all cables up to the wall plate (and to some extent from wall plate to cable box) to be their responsibility.
The TV interconnections are not pretty much HDMI, so there is no way the TV can affect the cable network.
 

JFletcher

Senior Member
Location
Williamsburg, VA
Questions like this has been asked on this forum before and I have read through some of them but not understood the details.
In my apartement the incoming tv/internet signal first goes through a 2 way splitter. One of the taps of this is currently open. The other one is connected to a 75 ohms cable that goes to a coax outlet that is terminated with a 75 ohms. My set top box is connected this outlet and it both has a hdmi output for tv and act as a wifi router.

The general advice seems to be this: terminate the unused port with a 75 ohms. This will prevent problems for other people in my apartement building.
What I am wondering then is:
1. will not terminating the unused port reduce the power of the signal at my outlet?
2. how bad could the problems caused by not terminating the port be?


Regarding 1: I would think that if 4mW goes into the splitter each port receives almost 2mW. The unterminated ports effect is completely reflected and divided by the other port and the incoming coax. Therefore 1mW of the reflected energy should go into the active port? If I calculate the reflection coefficient of the open port I get: R=(inf - 75)/(inf + 75) = 1. Does this not mean that the reflected signal has the same phase as the incoming signal and therefore will interfere constructively with the incoming signal? Therefore I should get 3mW of signal power with an open port vs only 2mW with a terminated port?

Regarding 2: There are 16 apartements in my building. I imagine the signal comes into the building goes trough an amplifier followed by a 16 way splitter before it gets into my apartement. The 1mW of reflected power from my example above that goes back to the 16 way splitter would then be split into 16. Therefore my neighbour would get a slightly delayed ghost signal 1/64 of the power of the original signal. Does this matter? Can this do any harm?

Thanks in advance for any answers!

Welcome. To say any of us are familiar with Norway's telecom and cable methods would be a shot in the dark. 16 apartments here would likely have taps and not splitters, but in any event, capping unused ports is never a bad idea, and the caps and time it takes to install is virtually nil, so why not.

The better option is using the smallest practical splitter(s) so that the signal loss is as low as possible, and forget capping anything other than wallplates.

We have Cox here, I've talked to their techs. A solid 80% of coax problems here are inside customer premises. It is not their responsibility to fix/maintain those here (in my state/area), but they CAN shut off your service should an internal problem cause them problems. YMMVIN (your mileage may vary in Norway).
 

FionaZuppa

Senior Member
Location
AZ
Occupation
Part Time Electrician (semi retired, old) - EE retired.
a 2-way has 50% of the signal on each output regardless of whats connected.

http://support.channelmaster.com/hc...ch-Signal-Do-I-Lose-Going-Through-A-Splitter-

the 75ohm terminators help kill off signal leakage (aka noise) out from the ends of connectors. this is one of the major reason coax F type moved away from the old crude crimp style and into the better compression types. but to note, look at the suppliers of splitters, rarely do they supply 75ohm terminators, or even offer such as an add-on to the splitter.

and i'll note, if the splitter is located in a metal cabinet, then terminators really wont help much. but if you want the finished install to be "Pro" rated, use terminators.
 

GoldDigger

Moderator
Staff member
Location
Placerville, CA, USA
Occupation
Retired PV System Designer
a 2-way has 50% of the signal on each output regardless of whats connected.

http://support.channelmaster.com/hc...ch-Signal-Do-I-Lose-Going-Through-A-Splitter-

the 75ohm terminators help kill off signal leakage (aka noise) out from the ends of connectors. this is one of the major reason coax F type moved away from the old crude crimp style and into the better compression types. but to note, look at the suppliers of splitters, rarely do they supply 75ohm terminators, or even offer such as an add-on to the splitter.

and i'll note, if the splitter is located in a metal cabinet, then terminators really wont help much. but if you want the finished install to be "Pro" rated, use terminators.

The page you cite does not say anything about what happens when you leave a tap unterminated.
With an ideal (no resistive components) passive tap you get sqrt(2) times the voltage at each tap if all are terminated. And then each output gets 50% minus losses of the input power.
Obviously no power will be going out an open tap, so the voltage on the other tap has to be different as a result. And the input impedance will not match the nominal impedance, so there will be a reflection on the supply cable too.
If there are no losses (good approximation for our consideration) there is no magic way to make the effects of the open tap go away.
If you use a resistive divider instead, then you get 1/2 the voltage at each port when terminated, so 1/4 of the input power at each output tap.
But the resistive divider is not as severely impacted by leaving one tap open. For a two way, the input impedance will be twice the characteristic impedance, causing reflections, and the output voltage will still be half of the input voltage after allowing for reflections.
 

hbiss

EC, Westchester, New York NEC: 2014
Location
Hawthorne, New York NEC: 2014
Occupation
EC
The page you cite does not say anything about what happens when you leave a tap unterminated.
With an ideal (no resistive components) passive tap you get sqrt(2) times the voltage at each tap if all are terminated...

I've stayed out of this argument because I have no knowledge of splitter design other than to say that it's my understanding that these issues are no longer relevant with current day splitters used in CATV. The ports are extremely well matched and the isolation between output ports is greater than 40db. So any reflections from an unterminated port essentially isn't going anywhere.

I know that you can put a dead short on a port and not affect the signal on the others. This is a problem that happens all the time.

-Hal
 

GoldDigger

Moderator
Staff member
Location
Placerville, CA, USA
Occupation
Retired PV System Designer
I've stayed out of this argument because I have no knowledge of splitter design other than to say that it's my understanding that these issues are no longer relevant with current day splitters used in CATV. The ports are extremely well matched and the isolation between output ports is greater than 40db. So any reflections from an unterminated port essentially isn't going anywhere.

I know that you can put a dead short on a port and not affect the signal on the others. This is a problem that happens all the time.

-Hal
You can put a dead short (which could actually be an open at the other end of the cable) on a port and not short out the signal on the others. I would be hard pressed to believe that it would have no effect at all.
If the splitter is based on what are known as "directional couplers", then it possible that you could get that degree of isolation.
Unlike simple splitters, a directional coupler will have a minimum working frequency as well as a maximum frequency, and so would be more suitable for a cable system that does not have to process low-VHF (~50MHz) signals.
 

JFletcher

Senior Member
Location
Williamsburg, VA
I will switch a splitter to a smaller one vs capping unused ports every single time. Less signal loss. I honesty do not know the ramifications of uncapped ports at the splitter or end use (wall plate), except to say CATV is extremely forgiving of wiring errors, right up until it totally crashes, or the cable company shows up wanting access to your house.

Anything other than compression fittings are garbage. Cox here will cut them off and install their own fittings, period.
 

hbiss

EC, Westchester, New York NEC: 2014
Location
Hawthorne, New York NEC: 2014
Occupation
EC
Unlike simple splitters, a directional coupler will have a minimum working frequency as well as a maximum frequency, and so would be more suitable for a cable system that does not have to process low-VHF (~50MHz) signals.

See, that's the thing. They aren't "simple" splitters anymore. And directional couplers certainly do go down to 5Mhz. I've used the then available ones back in the 80's to send reverse signal back to the head-end for local origination.

The taps out on the poles are directional couplers by the way.

-Hal
 

FionaZuppa

Senior Member
Location
AZ
Occupation
Part Time Electrician (semi retired, old) - EE retired.
I will switch a splitter to a smaller one vs capping unused ports every single time. Less signal loss. I honesty do not know the ramifications of uncapped ports at the splitter or end use (wall plate), except to say CATV is extremely forgiving of wiring errors, right up until it totally crashes, or the cable company shows up wanting access to your house.

Anything other than compression fittings are garbage. Cox here will cut them off and install their own fittings, period.
reflective waves are the issue. in homes, kinda a waste of $$ to install terminators. it really only becomes an issue under special circumstances (usually for the service provider).
 

egnlsn

Senior Member
Location
Herriman, UT
Occupation
A/V/Security Technician
...the isolation between output ports is greater than 40db. -Hal

That 40dB port-to-port isolation typically is only in the return path, and then not even the low end of it. Outside of the 15-42MHz range (typically), the isolation is mid-20s to mid-30s. That's for 2-way splitters. In 4 and 8-ways, the 40dB isolation rating does go up into the mid-60MHz range.
 

hbiss

EC, Westchester, New York NEC: 2014
Location
Hawthorne, New York NEC: 2014
Occupation
EC
So, after two pages of differing opinions I wonder if the OP has learned anything? Really, if it bothers him that much go out and spend 75 cents for a terminator or better yet just remove the splitter. This is the kind of question that reminds me of the nonsense you get from audiophiles who debate whether $2000 high purity copper speaker cables make a difference over zip cord.

And as far as cable TV is concerned, all this coax and RF stuff running around your house is soon going to be a moot point anyway when they move to IP and your TVs connect via WiFi. Then you'll only have one cable or fiber running in to your house to connect to the router/WiFi access point.

TV sales are way down already because the Millennials watch everything on their iPhones.

-Hal
 

JFletcher

Senior Member
Location
Williamsburg, VA
So, after two pages of differing opinions I wonder if the OP has learned anything? Really, if it bothers him that much go out and spend 75 cents for a terminator or better yet just remove the splitter. This is the kind of question that reminds me of the nonsense you get from audiophiles who debate whether $2000 high purity copper speaker cables make a difference over zip cord.

And as far as cable TV is concerned, all this coax and RF stuff running around your house is soon going to be a moot point anyway when they move to IP and your TVs connect via WiFi. Then you'll only have one cable or fiber running in to your house to connect to the router/WiFi access point.

TV sales are way down already because the Millennials watch everything on their iPhones.

-Hal

Agreed on the bolded part. As for the part I underlined, it's not even close; I'd hazard an EG (educated or experienced guess) that 99% of the installs here are existing, fairly old (at least as old as CATV itself), and definitely subject to 'conventional CATV problems'.

Yeah, people are hooking up TVs via WiFi, and I've seen my fair share of problems with that; mainly, an entry level/cheap router at one end of the house trying to throw a signal ~150' thru multiple layers of drywall, subfloor, doors, etc.

TV sales are way down because Cox here wants to charge almost 200$ for phone/cable/internet, yet their bronze package, which is better than what I pay for now, is 90$. and there is almost nothing worth watching on TV anyway, HDTV or not.
 

kwired

Electron manager
Location
NE Nebraska
TV sales are way down already because the Millennials watch everything on their iPhones.

-Hal
They can have it. Not that I don't look at anything on my phone, but if I want to watch TV, I am not looking a tiny little screen, in fact I recently put in a 100 inch screen and projector in my man cave. Want to watch football, that screen is much more like being at the stadium then that tiny little image on the phone is.

TV sales are way down because Cox here wants to charge almost 200$ for phone/cable/internet, yet their bronze package, which is better than what I pay for now, is 90$. and there is almost nothing worth watching on TV anyway, HDTV or not.

Also correct that even if you get 300+ channels, there isn't much worth watching on most of them, what is worth watching is old reruns of shows that were good when they were new. And if you want what little "good stuff" there is, you must subscribe to higher level and higher cost programming packages:(
 
Status
Not open for further replies.
Top