Noise on balanced line input

Where does this transformer physically live? Close to the siren driver, close to the music amp, somewhere else? If you think you're picking up junk from the wires, it might help to move the transformer as close to the music amp as you can get it. I think (and I'm sure I'll be corrected if I'm wrong) that the mic inputs typically work with a lot lower signal levels than, say, a line or aux input (for a CD player or similar), so the mic input will be more sensitive to this kind of pickup. If the smaller signal only travels a couple of inches, it will be less likely to pick up junk.

If you have a line or aux input on the amp, you may be able to fix it by moving the siren input to that. The suspected induced noise will still be there, but the lower sensitivity of the line/aux input will make it harder to hear.

Are you sure? You might not be the first person to re-use existing wiring in that installation, especially if it runs between two buildings

- grabbing a pair from the old telco wire is way easier than bringing a shovel or a ladder. :) I have seen old Bell 25-pair 1A2 PBX cables repurposed for *Ethernet*; it actually works sort of OK at 10 megabit,

100 megabit is iffy. My point is, there may be something besides the alarm wiring that is radiating junk into the pair that you used for the siren tone.

You might go to where the transformer is, disconnect it from the siren driver, short the two primary wires together, and see if you still hear the noise. If it goes away, it may be coming from the siren driver itself; changing inputs on the music amp might be the fastest way out of that situation. If it stays, then it's getting picked up on the wiring somewhere.

If it's getting picked up on the wiring, an idea: if the alarm "brain" has enough 12 V drive to run two siren drivers, run the 12 volt siren driver supply to where the music amp is. Put another siren driver there, and put an 8 to 16 ohm power resistor (look at the siren driver spec) across its output. Then put the ends of a pot (1K maybe?) across that, and run the pot wiper and one end into the amp. That gives you a local and hopefully "clean" source of siren noise for the music amp. If the 12 V from the alarm is really too noisy, still run it to the music amp, but use it to switch a little 12 V relay. Use a local "clean" 12 V DC wall wart to power the local siren driver at the music amp.

If you have any ferrite beads (those lumps at the end of some computer cords), looping the cable through one a few times right before the mic input at the music amp might help a little. I think those tend to help with higher frequencies than what you probably have, though.

Matt Roberds

,

I appreciate your ideas Matt. The transformer is mounted right on the PA amp with a 6 inch secondary wire to the XLR connector. I'm certain that the noise is being picked up by the cables. none are shielded and the best there is is the CAT3. You also may be right about the chokes. I've used them in the past for RF suppression, but this is at an audible rate.

I made a service call once to a house very near an AM broadcast transmitter. The station was just about as loud on the telephone as a normal conversation would have been. I fed the CO line into two small chokes, (value unknown), and then hung a cap, (again I don't remember what I used. I wish I had written it down), across that. When I connected the premises phone line up to my filter the line was clean.

I did try the other high level AUX input but couldn't get it to work with my 45 ohm transformer. I suspect the step down effect was working against me at that point. Eliminating the transformer and connecting the AUX input directly to the siren driver output didn't work at all either. I'm not really sure why though. I might need a transformer with a less aggressive turns ratio for that input to work. Maybe I could try to capacitivly couple to the aux input but that could be noisy too. Lenny

"klem kedidelhopper"

** If both the lines involved REALLY are *equal level & anti-phase balanced * AND twisted - there should be no such problem.

So I bet that is not the real case.

... Phil

Methinks not. The voltage on a 70v constant voltage speaker system is (insert drum roll...) 70.7 volts rms. No way is it "very low" unless something is broken or you're running at drastically reduced volume. The big advantage to 70v and 100v systems is that they can use cheap, junk, small gauge wiring, such a CAT3 telephone wire for relatively high power speakers. This can be a problem when someone actually does use cheap, junk small gauge wire that sometimes simulates a fuse.

How it works:

--
Jeff Liebermann     jeffl@cruzio.com 
150 Felker St #D    http://www.LearnByDestroying.com 
Santa Cruz CA 95060 http://802.11junk.com 
Skype: JeffLiebermann     AE6KS    831-336-2558

"Jeff Liebermann" klem kedidelhopper

** Absolute nonsense.

70V is the maximum voltage output from the amplifier and there is nothing constant about it.

** Volume pots allow the level to be set at the operators convenience.

It is very common to have a low, background music level and a much higher speech level for announcements.

** More nonsense.

The real advantages include the use of long cable runs with low losses and having many distributed speakers - which do not need to be all of the same rating type or level.

.... Phil

s d e

To further explain this, the amplifier's, (which is located in one building), output is taken off the 70 V tap, however as Phil alluded to the voltage on the line could be and in fact in this case is very low, ,just a few volts. This is all I need because, again this is VERY low background music and the level seems to be satisfactory throughout the building. As far as the lines being balanced, I'm assuming that they still are as I have done nothing anywhere that I can think of to cause an unbalance. As I mentioned before this balanced line output leaves the amp on two pairs of a CAT3 cable. I used two pairs because they are the main feed for all the speakers. This "main" cable runs from this building underground through PVC pipe to the 66 block in the other building. At that location all the other CAT3 wires also appear on that block as well. On each cable that was reused the brown pair is connected to a speaker. in a different room. These brown pairs are all placed in parallel and then connected to the 70V feed. So this is all well. It is not however the problem. The problem is occurring when I try to sample the siren driver output from the alarm and apply it to the low level input of the music amplifier. That's when I hear the data noise which I'm certain is being picked up by the "antenna" which is the sample wire and fed into the mic input of the amplifier. Lenny

Old cable can have a conductor shorted to ground somewhere. I've found wire with badly abraded jacket laying on steel beams, or nicked on the corner of an electrical box.

Is that "Siren driver output" the actual speaker connection, or DC to power the sirens? Is the connection isolated by a transformer?

to

When the alarm is activated, 12VDC is applied to the "siren driver". The driver outputs a "whoop whoop" sound to an 8 ohm speaker. The speaker is located in a different building than the alarm system. The CAT3 cable to the speaker is punched down onto a 66 block near the alarm system control box. This cable serves two functions. It sends the siren driver output to the speaker, and also by cross connecting another of it's pairs on the 66 block also sends the siren driver output back to the PA amp where I connected it to the 70V transformer primary and then connected the 45 ohm secondary to the mic input. That's when I heard the "ticking" or "fluttering" noise. I really do think that the cables are physically in good shape. Lenny

If the level is right, I would add a 12 volt relay to connect the audio to the amp when the siren is on. It sounds like the level to that mic input is way too hot. You say, 70 volt, but not the wattage so I can't calculate the turns ratio. You also don't state the output level from that siren driver. From working with alarms back 40 years ago, the drivers were a minimum of 5 watts, and some were close to 45 watts. You could be feeding several volts to that mic input when the siren goes, and it may destroy the transistor or IC for that input. I always used a pot after the transformer to set the maximum level. I used to make promo tapes for Cinemax, HBO and other CATV channels and mail them out to radio stations each month. They were recorded of C-band and the only place to get the audio was from the external speaker jack on the monitor. I put a 10 ohm resistor across the input, and used a 10K pot to set level. No one ever complained about the audio quality.

You need to use a shielded audio transformer. You can probably get a

1:1 mike isolation transformer, or possibly a mike single-ended to balanced transformer to fix this. My guess is the speaker line transformer is not at all balanced or shielded, and so some unbalanced signals can get into the long bundled cables and then go through the transformer and into the balanced input. A shielded transformer has a shield between primary and secondary, and should prevent unbalanced signals from getting across, only the balanced part could get through. The shield of the transformer needs to be grounded to the PA amp.

Jon

Please tell that to the Wikipedia authors and system vendors that insist on calling it a "Constant voltage speaker system". However, you're correct, the voltage is not constant, but that's what it's called, and what I've been calling it for many years (without really thinking much about the implications).

From the original question: "Each speaker is connected to the line with a small 1 watt 70V to 45 ohm transformer" So, that means that at maximum volume setting, and at maximum amplifier power output, this speaker will be delivering a fabulous 1 watt rms. That's about right for background music, or in-room intercom, but not enough to shake the hallways for an announcement.

Agreed. I didn't want to itemize the benefits and limitations in detail. That's why I included the links with the details.

From my experience, the principal advantage to the service company was the ability to sneak the wiring past the inspectors disguised as telco wire, alarm wire, doorbell wire, "thermocouple" wire, barbed wire, or whatever else could be found or scrounged for the occasion. It kinda balances the added cost of the xformers.

--
Jeff Liebermann     jeffl@cruzio.com 
150 Felker St #D    http://www.LearnByDestroying.com 
Santa Cruz CA 95060 http://802.11junk.com 
Skype: JeffLiebermann     AE6KS    831-336-2558