short circuit safe line amp

[milesdf]

Hi everyone,

 

The radio station where I work had a huge issue today when the tunnels where our studio-transmitter link (~1000' balanced pair, 2 channels L&R) is housed. The link was completely shorted out. This seems to happen about once a year for some reason, and while the link is restored a couple days later our distro amp (which takes one input and feeds multiple streaming servers and the transmitter) has channels fried. This summer we upgraded to a distro amp which claimed to be resistant to this, but alas when it happened for the first time this season instead of just killing one channel it completely disabled the distro amp (which i unplugged)! I didn't have time to disconect the shorted channels and test it again but I will tommorrow, hopefully the whole thing didn't go up in smoke. This is totally undesireable because when this happens our streaming services are especially important, So it looks like a good solution would be to get another line amp to feed the transmitter after the distro amp. Does anyone know of one that is economical and can survive being shorted out for long periods of time? even if it has to be reset after it happens, as long as I don't have to buy a new one its fine.

 

Also is there any other solution that I'm missing? Maybe a transformer would do this well actually? Maybe all it needs is some isolation?

 

Thanks, Miles

 

(I realize this isnt really live sound, but I know theres alot of knowledge here, and I think I've helped enough noobs in 2 years to be lent a hand every know and then )

[agedhorse]

Absolutely, place a 1:1 line level isolation transformer at the driving end, and you should use one with a primary and secondary electrostatic shield that the driver and driven grounds should terminate to. This will provide a floating and really isolated signal. I stock the transformers. I would also put a pair of back to back 18V zener diodes across the differential driving output, which will clamp any large transient originating at the load.

[Ear Abuser]

Is the short across the balanced line or does one of lines short to ground?

If it shorts line to line, I don't think the transformer will help, but if it is one of the lines to ground it could.If the audio line is transformer balanced on both ends, you might get away with a line to ground connection if it only happens on one line.

 

 

It's kind of like using an ungrounded delta connected power system. One reason it is done is so that you can still operate even if one phase gets grounded.

Can you inspect the cable? It sounds like the tunnel might have a moisture problem that is getting into a splice or the cable was damaged when it was pulled in.

[agedhorse]

The reason a transformer will protect against a shorted line is that under normal circumstances, impedances are (pretty much) reflected, but as soon as you short the secondary, the transfer efficiency decreases and you will not be able to transfer a "short" across the magnetic field efficiently. Measure the impedance of a shorted audio transformer and youwill find that it's up around 10-20% of it's nominal impedance.

 

There are also line op-amps that have current limiting built in, and I do hope that there are build-out resistors on the output, maybe it's an oscillation problem as line capacitance increases due to a moisture problem? You might look at the 5532 op-amps, in a balanced output mode circuit, capacitor coupled w/ 22-47 ohm build-out resistors. About as bullet-proof as you can get w/out a transformer. Socket the 5532's and you are good to go.

 

I have also designed a low impedance drive line amp, that has current foldback protection. Rarely is this necessary though.

[milesdf]

Well the link doesnt have a ground, its a telecom dry pair. I don't have access to the actual line, only the ends. Possibly the distro amp is going crazy because voltage is being thrown into it? anything is possible, more diodes would protect against that correct?

 

how would I go about ordering those transformers from you andy? email me at [myfirstname].frangenberg at villanova.edu.

[agedhorse]

Just e-mail me at

 

agedhorse AT aol.com

 

and I'll get you the info. I am on the road right now.

[Ear Abuser]

 

its a telecom dry pair

 

 

I was wondering if that might be the case. A teleco dry pair normally goes to the central office master distribution frame where its directly connected to another pair out to the transmitter. In the old days Ma Bell made you go through a big transformer on each end. Perhaps as Agedhorse has suggested, thats what you need if there isn't already one there.

 

There are plenty of ways that a teleco pair can have problems. There are lots of splices that can get exposed to the weather and start leaking to ground.

I used to work with a telemetry system that used leased lines and DC current loops . Everytime there was a big rainstorm the system started going to hell. The solution was to go to a tone based system.

Also telco people could inadvertantly put 48volts DC or worse, 90 VAC ringing voltage on your line while testing. Those clamping diodes sound like a real good idea.

[milesdf]

 

I was wondering if that might be the case. A teleco dry pair normally goes to the central office master distribution frame where its directly connected to another pair out to the transmitter. In the old days Ma Bell made you go through a big transformer on each end. Perhaps as Agedhorse has suggested, thats what you need if there isn't already one there. There are plenty of ways that a teleco pair can have problems. There are lots of splices that can get exposed to the weather and start leaking to ground. I used to work with a telemetry system that used leased lines and DC current loops . Everytime there was a big rainstorm the system started going to hell. The solution was to go to a tone based system. Also telco people could inadvertantly put 48volts DC or worse, 90 VAC ringing voltage on your line while testing. Those clamping diodes sound like a real good idea.

 

 

well its not a traditional dry pair, its from the university telecom department here which has multipair cables run point to point between all the major buildings. I believe that our 2 pairs run from our studio to the telecom room in the building to the telecom room in a building 500' away and then to the telecom room in our transmitter building 500' away, up a comms riser and then directly into the closet where our transmitter is. one time they accidently messed up one of the lines and to get it back they put a tone generator on one end and listened at all thier blocks in the telecom rooms to find where the error was. the tone came through our transmitter crystal clear and at a suitable line level, but I would like to safeguard against them using something more dangerous.