Audio Amplifier Problems

** Groper Alert !

** Have any idea how to calculate the load presented by all the speakers together ?

Was your amperage test done with a sine wave signal or speech / music ?

The minimum load impedance for a 120 watt 70 volt amp is 41 ohms.

If there are similar 100 speakers, each must be have 4100 ohms load impedance.

Also - beware the existence of low audio frequencies, a big NO-NO with

70volt systems.

....... Phil

Sounds like an overload is causing the amps to shut down. Are all of the speakers working before it shuts down? Any chance there is a short on the secondary side of one of the circuits?

70.7 volt systems only produce that voltage at maximum output power. How are calculating power? Are you measuring voltage and current? Also, what is the test signal (waveform and frequency). The meter might not be adequate.

All you have to do is add up the watts from each transformer. it should be close to the rated output of the amplifier. The transformers have multiple taps to select the watts at each speaker. This requires a visual inspection of every speaker and transformer. Write down the numbers and add them up. If the system doesn't work, test the amp into a good dummy load, or check the wiring for a short to the building's ground system. The whole idea of constant voltage sound systems is to make it dirt simple to install and maintain. If the OP has 100 speakers, the taps should be set to 1 watt per speaker. He mentions "Amplifiers", implying more than one amp. He doesn't say if it is broken into zones, or the amplifiers are in parallel. If they are in parallel, you have to match the gain of each amp, and make sure that all of them are in phase.

--
Service to my country? Been there, Done that, and I\'ve got my DD214 to
prove it.
Member of DAV #85.

Michael A. Terrell
Central Florida

yeah, i does sound like an overload, except that it is inconsistent. too inconsistent to be able to say for sure. all speakers seem to be working fine, though with the quantities involved, i can't be sure. by short on the secondary side, you mean on the secondary of the transformers for the speakers? each speaker has its own transformer and each is tapped at 1/2 watt.

yes, i'm measuring voltage and current. the test signal is a speech & tone. the tone seems to be around 1Khz, though i don't know the waveform. the meter has a min-max function which i've been using to capture peak levels. the actual voltage on the circuit is more like 50. what kind of meter do i need? or will a meter at all even be adequate... do i need an O-scope?

good dummy load, or check the wiring for a short to the building's ground system. The whole idea of constant voltage sound systems is to make it dirt simple to install and maintain. If the OP has 100 speakers, the taps should be set to 1 watt per speaker. He mentions "Amplifiers", implying more than one amp...

what is a good dummy load? there are actually more than 100 speakers, i'm just trying to get an idea for how to fix this and was therfore giving an idea for 1 amp and its approximate load, i'm also learning a lot as i go. the reality of it is more like >1000 speakers divided across 10 amplifiers that are all separate from each other except for a shared pre-amp input. though not all amps are causing a problem, currently only 2 are, so the rest of it i should be able to ignore. i think the main problem circuit is around 175 1/2 watt speakers. that means by my calcs, i should have about 87.5 watts of speakers and should then be well within the amps capacity. any more help is greatly appreciated.

Are the amplifiers in the same room? Can you swap a set of 70 volt speaker leads between a working area and one with problems? If the problem moves, the amp needs repaired. If not, it is in the speakers, wiring or transformers for that area. I have found wiring cable tied to the steel trusses that had worn spots, and was arching to ground when it was above a certain level. In others, the line was shorted to ground, and overloading the amp. The volume is low, so someone cranks up the level, till the amp starts to clip.

How old are the amplifiers? You may have some failing filter capacitors, allowing the amp top oscillate at an ultrasonic frequency.

As far as test equipment, a low distortion audio sine wave generator and a distortion analyzer will show problems before you can hear them.

Yes.

That's good.

Speech is not the correct test signal for measuring power but is OK as a basic check. A 400 Hz sine wave is typical for testing audio systems.

Yes, as I posted , 70.7 volts is realized only at maximum output ... so at

50 vols, the system is operating below its capability (assuming all is normal).

Is there an 8 ohm or 4 ohm output? If so, you can use a resistor of that resistance, rated at 100 W (or somewhat less for a brief test). I use an oscilloscope connected across the dummy load and a triangle wave test signal at 400 Hz.

Thanks for all the input. all the amps are in 2 racks right next to each other. i have tried moving leads between amps to check for amp vs. circuit problems, but its too inconsistent. I switched one and let it run for maybe an hour or more and the problem seemed to be fixed (different amp doesn't trip = original is weak) but then a day or two later it started again on the new amp. about a week before that a different amp had been tripping and after checking speakers i found two that were not functioning. i replaced them and the problem stopped... for a few days, then started again. it's also tricky because these particular model of amps supposedly had a recall around the time that they were manufactured (2004?) but they have since discontinued production. so i need to verify for myself with absolute certainty that the amps are good or bad. also they've been sitting in a construction site with lots of dust, humidity, and uncontrolled temperatures for presumably the last two years.

Thanks for your help. well, i dont know if there are any shorts on the transformer secondaries, and unless there is a way to check from the main circuit, the labor involved to pull down each one to check would be ridiculous. as far as test signals i think i have a 1Khz Sine wave signal on my laptop that i can try to load. is there someplace i might find the test tone of which you speak? also, where would i find a 100 watt resistor? i think the biggest iv'e ever seen is 5 watts. and i may be showing some more ignorance here, but if i load/play the test tone to the dummy resistor, then the scope shows a waveform & freq. that match the test tone, does that mean the amps are ok? if so, then i just need a huge resistor, test tone and scope :) also, do you know if there is a way to find out the min/max frequencies that an amp could push to such a circuit. i think i read that impedance is dependant on frequency, and that if the impedance were to rise or fall too much, either too much current or voltage would be needed and could cause problems. is that to far-fetched?

I know that, and that is why I asked if there was normal output at all locations before the amp shut down. A dead speaker would be a big clue. I installed a 70.7 volt system in a church and it later turned into a nightmare. Carpet installers shorted secondary circuits and ... you do not need to know the rest. S**T happens!

as far as test signals i think i have a 1Khz Sine wave

You can indeed use a laptop because there are free-ware signal generators out there ... Google is your friend.

also, where would i find a 100

Mouser, Digi-Key, Allied and others ... Google is your friend.

i think the biggest iv'e ever seen is 5 watts. and i may

Impedance is certainly not constant in these systems. But, that is not the issue here. I suggested a triangular test waveform because, with an oscilloscope across the output, it is easy to spot clipping and crossover distortion.

I don't want to insult you, but you might want to consider some outside help with this. Not via the Internet. An experienced audio technician to go to the site with you.

Standard practice is one transformer per speaker, and mounted to the speaker frame or cabinet. This prevents the secondary circuit from being exposed.

I started servicing these types of systems in the mid '60s. We didn't have anything fancy to troubleshoot with. Most problems were found to be where someone had tampered with the wiring so try to raise or lower the volume. I have seen messes that were so bad that they had to be ripped out, and start over.

Personally, I would have split a system like this into zones of 10 speakers with a home run back to the equipment room for each zone. A "66" series telephone stake block will handle 25 zones. You remove the bridging clips to isolate a zone. Then if you have problems, you can easily isolate it. A "Test tac" made for telephone work and a speaker with a matching transformer (with a volume control) can be plugged into any circuit to listen to the amp.

There are impedance bridges available to read the impedance @ 1 KHz, but I used the test speaker and a small FM receiver and 5 watt amp with both 25 and 70 volt outputs to feed unknown wiring. It wasn't uncommon for some idiot to find a wiring cabinet and rip all the wiring apart, and rip all the tags off, as well.

for a 120W load on a 70.7 volt system:

Ideally a 150W 41 ohm resistor, you don't need to buy one, you may be able to find one, or something close enough, you want something that doesn't vary in resistance as it heat up nichrome wire is good - it's wat heating elements are made from.

about 345W worth of 120V heating element, or 1382W worth of 240V heating element.

can you swap them round? see if the problem moves with the amp or if it's tied to the position.

Is it possible that a splice in the speaker cable is getting wet?

Bye. Jasen

=========================================== The Don Davis book says a good way to find a rogue speaker that has bypassed the transformer is hook a filament transformer to the 70 volt line... the speaker that is receiving 6vac at 60hz will speak loudly for a while then smoke. I suppose a variac would also work.

Not if there are several ohms resistance in the cable between the amp and the speaker. He said that there are about 100 speakers per amp. What are the chances that it is on the end closest to the amp? If there is no transformer, you can hear the difference, anyway. BTDT for over

30 years.

--
Service to my country? Been there, Done that, and I\'ve got my DD214 to
prove it.
Member of DAV #85.

Michael A. Terrell
Central Florida

Yeah, i can swap them around. i have and right now it's too inconsistent to be sure of any results. why do you suggest a 41ohm resistor instead of a 4 or 8?

its very unlikely that anything is getting wet. all outdoor devices are weatherproof and in weatherproof backboxes and they're somewhat protected from rainfall by the floors above them. everything else is indoors, mostly in electrical rooms with no water.

Thanks charles for all of your help, no insult taken. however, i pretty much am the outside help. this audio system is just a small portion of the rest of the system, of which i am quite familiar. so i'm the guy that gets to figure it out, and by the time its all said and done, i will have learned a lot, so its actually a good challenge. So just to confirm... an 8 ohm resistor rated for 100 watts (does this actually matter as long as its big enough? eg. a 150 watt or 300 watt would be equally effective?) gets tied to the 8 ohm output and if waveform & frequency stay true, then the amp is good??? after that i can go after the circuits. also i have been told an impedance meter is invaluable, but couldn't i just take amp/volt measurments (Z=V/A)

Thanks for the help michael. they're actually split into zones of around 21 speakers each. each zone can be manually turned on, so i don't even need to re-connect anything to isolate a given zone. the problem doesn't happen on any given zone though, its more like a variable combination of zones :( I'm really hoping i don't have to tear out too much. but thanks to all of you guys, i at least have a few directions to go in. is an impedance bridge the same thing as an impedance meter?

100W is 70V across about 50 ohms. All the other transformers are higher impedance and bridge off this one. 1/2W is 2V across 8 ohms. 21 speakers 2W each in a zone need 10.5W, and you could run about 10 zones with that 100W amp, 210 speakers per amp. The voltage gain from the 70V bus to the speaker is 70V/2V or 35, so the impedance seen on the primary is the secondary impedance scaled by the turns ratio squared. The turns ratio squared is 35^2 or about 1200 on the primary x 8 on the secondary, so you each primary is about 9600 ohms, 210 of them is about a 45ohm load on the amp and the 70V bus. Cheap 70V transformers have poor low freq resp... system might work ok from 200hz up, and overload at low freqs. Stick a resistor in series and watch the volts across it as you sweep the freq below 200. Tell me if I identified the prob. The fix is fatter transformers or a hi pass filter in front of the amp.

has several models listed, with some details on the individual pages. Most impedance bridges are at higher frequencies, but if it will do 1 KHz in the right impedance range it may work. The meters allow you to set the line voltage which will help if it is a problem of insulation breaking down, Since you can zone the speakers easily, you may want to check each group to see if there is any leakage to ground. An ohm meter will show some problems, but a "Megger" at a hundred volts or so will will find arc over problems.

--
Service to my country? Been there, Done that, and I\'ve got my DD214 to
prove it.
Member of DAV #85.

Michael A. Terrell
Central Florida

You want to load the 70 volt output. 50 ohms is close enough for testing.

--
Service to my country? Been there, Done that, and I\'ve got my DD214 to
prove it.
Member of DAV #85.

Michael A. Terrell
Central Florida