Control line bridging problem

[justanotherguy59301]

Maybe someone can help me out with an issue. The radio system I maintain used a combination of U-wave, UHF and Leased lines for control of a Base. Three area offices use the base. In order to make a long story short, some of the equipment has been prone to failure and is no longer serviceable. In order to simplify the system I contacted a local phone company to provide all leased lines and bridging equipment. The solution they came back with, including a 6 month minimum, worked in theory so we went with it.

Here is where the problem comes in. Our offices use 2-wire for the control lines, easier to integrate speakers/ paging. The 4 circuits(three offices and radio site) are transported from the Telco CO via 4 wire and terminated on site via a 2w-4w hybrid, not uncommon for leased lines. The Telco CO is using bridging equipment to combine the audio in an everyone-hears-everyone manner. The circuit works, but only when all 4 legs are terminated, and terminated at a perfect 600 ohm.

That isn't what we have though. With various needs of the offices for tone remote numbers, and other equipment on the lines, we have different impedances at each site. This in conjuction with the 2w to 4w hybrid circuits and the bridging equipment, causes feedback on the lines rendering them useless. This doesn't happen with the test equiment since it has a perfect(as close as can be) 600 ohm termination.

Does anyone have or better yet (had) this issue and/or ideas on a cure to make these lines useable.


One thought I had was to come up with a way to isolate my office equipment from the phone circuit equipment in order to have flexability on my equipment while presenting the phone circuit with the required 600 ohm load. It would be possible to get the equipment i have to 600 ohms using various methods but.....1 tone remote fails or is unplugged from the wall, and.... down goes the system.


Wilcom makes a "SB21-K2 Line Powered Amplifier - NID/Wall mount" that may give the isolation i need but it may have to be an externally powered and hopefully less costly version.

 

[davidgcet]

simple, goto rat shack and pick up some 1:1 600 ohm transformers. wire the primary across the telco circuit, then your in house stuff across the secondary. works like a charm, i keep about 5-6 of these in my truck because we still have 2W control lines at a few sites and these tend to pop when lightning comes in on the phone line, which is actually a secondary benefit though by no means should be considered a proper telco lightning protector.

 

[mastr]

Yur post implies that you are double, triple or "who-knows-how-many-times" terminating your 2 wire circuit. That is a basic and still remarkably common mistake.

Only one of your remotes at a given location should have the 600 ohm termination resistor in place, and the others have said resistor removed, and thus are a high impedance across the line. If the terminated remote is to be removed, one of the others should be set up to provide the termination, Other high impedance remotes can be removed with no issue. If others commonly move or tamper with remotes, you could set up all remotes as high impedance and put a fixed 600 ohm termination resistor across the circuit in any handy location.

The transformer idea will reduce hybrid unbalance issues, but with multiple terminated remotes your levels will still jump up considerably when someone unplugs a remote.

 

[justanotherguy59301]

Thanks for the replys.

The termination of the remotes, to the best of my knowledge is:

Office A: 6 tone remotes at 6k, 1 at 600
Office B: 9 tone remotes at 6k
Office C: 9-10 tone remotes at 6k, one telco systems channel bank card at 600
Radio site: GE Master III at 600


Is my thought process flawed

Office A should be somewhere just under 600 ohm
Office B should be just over 600 ohm
Office C is over loaded and should be somewhere around 300

Would I need to know more accurately the impedance of each site and maybe go with an offset transformer, or would the 1:1 600 transformer make the hybrid circuit always see 600 reguardless of what i have on the line so I would only have to adjust the output levels to compensate.

I've been told that changing the input impedance on a transformer does affect the output impedance, if this is true, would it be enough to throw the balance off and start the circuits singing again? Maybe trying it would be the best option.

 

[davidgcet]

i've never had a problem with using the 600 ohm transformers, even a dead short on the customer side does not affect the telco side of the line at all. of course a dead short will kill audio on the customer side, but it will not send hum/feedback out to the other end of the circuit.

also many telco 2w/4w cards have a jumper setting for bridge or termination. this accomplishes the same thing as putting a transformer in place. not all of them have it, it really depends on the card they used, but i have had a few that did and i did not have to use the tranformer.

i used to have a site that had satellite feed to the local paging TX, then 3 2wire telco circuits to other nearby sites. we simply used a transformer series/parallel arrangment on the feed to each of the telco lines so that if any one of them had an issue it did not affect the other lines. nor did unhooking the local transmitter change our output levels. it was either do it with a few transformers for maybe 20.00 or spend a few hundred to a grand on an analog mux to do the same thing.

EDIT: i meant to add that all this does assume you have the remotes configured properly as pointed out above, only 1 serves as a termination and the rest are set for bridge.

 

[SkyPager]

Look here. These guys may be able to assist.

Dataprobe Inc. Uptime Solutions. Remote Site Management.

Give them a call and explain the situation.

 

[ramal121]

First off, don't worry about the phone company transport. If they want to split to 4W for the long haul (which they normally do) that's their business. If you ordered a 2W circuit, it should behave like a 2W circuit as long as you terminate correctly. Also phone company demarcs are rarely terminated, leaving that to the end user.

Now let me say, you have a butt-load of stuff hung off of that line. If I remember from the Motorola manuals, they recommend no more than 10 or 12 remotes on a line and only the farthest one out terminated. Even at high impedance, that many are gonna load down the line and cause problems. Davidcet suggested a 1:1 transformer, but I just can't grasp how that would help unless there was a ground to the balanced line somehow. The remotes and line cards already have a transformer in them. A 1:1 transformer should reflect the same impedance input to output. You could try a transformer with a different ratio, but I suspect the levels will change and the if you cannot compensate by tweeking the remotes, you're at a dead end.

Suggestions? You could install your own distribution hybrid. This would bring each leg to the different offices to acceptable limits. Kinda expensive.

Also you may consider swapping out tone remotes for local remotes. Example, one office have a single tone remote (with an acceptable termination) feeding several local remotes. That will require a dedicated comm cable to each position, but I have done this in the past and works fine as far as the user is concerned.

Needless to say, you have a lot of stuff off that phone line, no wonder it's barking!

 

[ramal121]

Alright open mouth, insert foot. I read through too fast. You have THREE separate legs from the telco.
Each should be considered individually (although they are cross connected at the CO).

A and B are good, however C is a little low. Surprised the phone line at C buckles at three hundred ohms, but hey you are outta spec. Maybe the are running you through some old 150 coils, who is to say. Can you switch out the termination to the line card and see if that helps?

Although 600 ohm is spec, the phone company would really prefer you terminate at about 1000 to 1200 ohms...

 

[davidgcet]

the transformers i am talking about are 600ohm across the primary, they will always present 600 ohm to the telco side no matter what is on the secondary side of them. i've used them many times for just this type of situation where multiple devices and or control lines needed to be tied together.

 

[gmt0000]

Another reason to NEVER deal with the telco......RF it

 

[ramal121]

the transformers i am talking about are 600ohm across the primary, they will always present 600 ohm to the telco side no matter what is on the secondary side of them. i've used them many times for just this type of situation where multiple devices and or control lines needed to be tied together.

If you measure those little transformers with an ohmmeter, they will show about 75 ohms resistive. That is the only constant. 600 ohms is reactive and is the nominal impedeance only when source, transmission and load are equal. When I cut a chunk of rg-58, I measure 0 ohms end to end. It is 50 ohm end to end only in an AC circuit and the load and source are matched at 50 ohm.

An example, a filament transformer 10:1. I plug the primary into 120 VAC. The secondary I'm getting 12 VAC. I put a 1.2 ohm load across the secondary, now pulling 10 amps. The primary humming along at 1 amp (10;1).
Now I put the crowbar short across the secondary. What does the primary do? Just keep up with its 1 amp as if nothing is wrong? No! its current spikes also, the windings smoke, and the primary fuse blows. Why? Because the impedance of the secondary has now been imposed upon the primary (with the turns ratio taken into account).

The little 600:600 transformers really do neat tricks, and the coil resistance provides some load stability (at the cost of some thermal loss), but they will not correct an impedance imbalance that gets too far out of whack.

 

[ramal121]

Another reason to NEVER deal with the telco......RF it

"Nuf said...

 

[davidgcet]

we won't get into the theory arguements, i am speaking from 1st hand experience at dozens upon dozens of sites. this is audio, not power so it does nto have the current available to burn up and a dead short on one side will not affect the load the other side enough to matter if at all.

but i too throw my vote in for RF links, they work much better and you don't have to pay Ma Bell every month.

 

[justanotherguy59301]

Thanks all for the input. For the cost and hopefull effectiveness, i've ordered the transformers. The tiny town radio shack here doesn't carry them.

The Telco circuits will total roughly 2700 a year. In order for us to do the RF links, I would need to get the audio out of the valley's the offices are in(U-wave or UHF link) and from the local hill top towers to the main radio site. Minimally it would be 4 additional UHF radios and the narrowband upgrades for the 2 we already have.


Cost is a toss-up(if we can get them working) depending on the lifespan of the radio equipment.
As for reliability,depends on who you talk to, I've had way more RF failures in the last 5 years than i have had leased line issues. For some people it's the other way around.

10 year estimates
phone lines = 27k
4 UHF Link Radios = 30 to 40k
2 UHF Link Radios 2 U-wave Links(cheap un-licensed) = 45k+

 

[davidgcet]

As for reliability,depends on who you talk to, I've had way more RF failures in the last 5 years than i have had leased line issues. For some people it's the other way around.

that is true, it depends on quality of the equipment, knowledge of those maintaining it, and just pure dumb luck. i've fought with a couple microwave hops that are going to have a fialure of some sorts at least 1x every 90 days. maybe a radio this time, a card in the MUX next time, etc.. but it fails regular enough to set you clock by it. i can say this, DON'T buy GE/MDS LEDR series if you want reliability!

 

[justanotherguy59301]

that is true, it depends on quality of the equipment, knowledge of those maintaining it, and just pure dumb luck. i've fought with a couple microwave hops that are going to have a fialure of some sorts at least 1x every 90 days. maybe a radio this time, a card in the MUX next time, etc.. but it fails regular enough to set you clock by it. i can say this, DON'T buy GE/MDS LEDR series if you want reliability!

MDS 960 with a kilomux - was the "regular" point of failure that caused the initial attempt at a switch. It worked well for years but it saw it's last failure and was removed and used for spare parts on other failing mds equipment.