How do I Find this Intermittent NAC Short?

Greetings Douglas,

I got dispatched to a site today that has a intermittent short on the horn/strobe (A/V) circuit (NAC). The short comes in for 4 to 6 seconds and clears. This site has multiple different electrical switch gear rooms, and going to and from each building are surge suppressors.

I checked that all the surge suppressors are clear. The only thing I noticed is that 2 of the surge suppressors are rated for 1 amp, which I think is low.

Anyways, after I checked out the circuit, which is class A, I decided to take off the field devices to see if the short would still come in.

I waited about 45 minutes - nothing happened - so I decided to just put all new A/V devices. There are 4 A/V devices, but the run has to be about 400 ft to 500 ft with a mix of 14 AWG THHN and 16 AWG FPL cable.

When I finished about 15 minutes later the short came back, so I decided to move the circuit to another NAC output on the Fire Alarm Control Panel.

The short came back and cleared after 4 to 6 seconds.

So I regrouped myself after getting somewhat frustrated. I decided to check the class A board and panel circuit. I took of the panel side of the circuit and waited about 40 minutes (nothing). I took the circuit from the class A module on the control panel, and landed the circuit back on the main board. Then, I waited 30 minutes give or take (Nothing).

So, I landed all circuits back and short came back after 20 minutes. I'm stumped and don't know what to do.

When I was wrapping up, I decided to check the other NAC circuits and I was getting 200-300 kilo ohm.

Tell me what do you think, any input would be greatly appreciated. Thanks in advance!!!

Thank You, AC

Look up: Is Water a Non-Linear Resistor?

Read that discussion, it helps with what are otherwise confusing problems. Non-linear resistances will be referred to in the following discussion.

When you have a problem that seems difficult to fix, remember that anything can cause the panel to show a short. OK, you've tested the panel using the parts-replacement technique, so at least the part of the panel that you tested seems to work.

You're also fortunate that it's a series wired circuit, so chasing down the problem using a divide-and-conquer technique can be used.

When approaching this type of problem, you need to use all three types of tools.

Three Types of Test Tools

--- Electronic Instrumentation ---

You have a DC voltmeter. "Calibrate" the voltmeter by checking the voltage on the panel's screw terminals for a working NAC circuit. This is your standard working NAC voltage.

Measure the voltage on the screw terminals for the shorted NAC circuit when the shorted NAC circuit is wired normally. Look for differences in voltage.

The DC voltages you measure are all compared to the standard working voltage.

• Higher voltage shows trouble on the circuit: a wire is loose or broken wire between the panel and the end of line resistor. In a Class A circuit, the location of the end of line resistor is inside the panel, in the Class A input.
  
  
• Lower voltage shows a shorted circuit: there is a wire to wire short somewhere in the 400 to 500 feet of cable, between the Class B NAC terminals and the Class A terminals of the panel. The short could also be on any component, including the A / Vs and the surge protectors.

The threshold to turn on the "shorted-NAC-light" for the lower voltage isn't a standard level. At least with one company's panel I know of, the voltage can be pulled down to a third of the normal voltage (by a reverse-wired strobe) and still show "normal".

Learn to make this kind of measurement. Your skills will increase.

You have an AC voltmeter. "Calibrate" the voltmeter like you did with the DC voltmeter. There should never be any AC voltage on the NAC circuit, but with the way things are wired, anything could happen.

If the wires are inside the same conduit as an air-handler or other AC power device, for instance, there could be cross-talk between the wires, and AC will show up on the meter. This kind of problem could be marginal, so when something else changed, the threshold inside the fire alarm control panel could be crossed.

Then again, AC voltage on the NAC circuit could be normal. That's why you have to "calibrate" the AC voltmeter on a working NAC circuit - to see what is normal before jumping to conclusions.

When following the circuit, pay attention to any other wires inside the conduit besides the NAC wires. This has been a problem I've had to deal with.

You have an ohmmeter. The ohmmeter shows you the resistance of a standard resistor or wire (those are linear resistances), but the ohmmeter cannot show the resistance of a horn, strobe, or surge suppressor. The horn, the strobe, and the surge suppressor each have many non-linear resistors in them. Both diodes and transistors are non-linear resistors. Ohmmeters don't work with non-linear devices.

--- Senses ---

You have many senses. Included in your senses is a sense of time.

Perhaps, in one of the out-buildings, some piece of non-fire alarm equipment is turning on and off. That could be a problem with AC being introduced. I've experienced it, and others have mentioned it to me.

--- Does-It-Work ---

If the system works while something is removed, but doesn't work when the system is normal, that is a test showing that there is a problem. Usually, the problem is with what was removed, but don't count on it. Sometimes the problem is somewhere else and only shows up with a complete system. The big thing with troubleshooting this way is to not become frustrated. Play with the system instead. Keep trying different things, disconnect different things, connect different things. There're always other things to try.

Divide and Conquer

When the problems aren't obvious, your job is to divide and conquer. From what you say, this is a Class A circuit, so the wires and devices should all be wired in series. Both ends of the circuit should also be inside the panel.

So that the Class A wiring doesn't mess you up while you divide the circuit, disconnect the wires from the Class A terminals of the panel. That way the panel won't send out voltage from both ends of the circuit. Disconnecting the Class A wires will put the circuit into trouble, so during the testing, trouble voltages will be what you will consider to be normal.

Hopefully while calibrating your meters, you've seen something that doesn't make sense. Keep your calibrated meter readings in mind as you progress through the circuit.

Parts to the Circuit

I'm guessing here, but this would be for 4 out-buildings, each with one horn/strobe.

1. Fire Alarm Control Panel with its Class B NAC Circuit Output
2. Wire
3. Grounded Surge Protector
4. Underground Wire
5. Grounded Surge Protector
6. Wire
7. Audible/Visible
8. Wire
9. Grounded Surge Protector
10. Underground Wire
11. Grounded Surge Protector
12. Wire
13. Audible/Visible
14. Wire
15. Grounded Surge Protector
16. Underground Wire
17. Grounded Surge Protector
18. Wire
19. Audible/Visible
20. Wire
21. Grounded Surge Protector
22. Underground Wire
23. Grounded Surge Protector
24. Wire
25. Audible/Visible
26. Wire
27. Grounded Surge Protector
28. Underground Wire
29. Grounded Surge Protector
30. Wire
31. Fire Alarm Control Panel with its Class A NAC Circuit Input

The panel is showing short, but the panel isn't very good at troubleshooting. It calls almost any AC or DC wire-to-wire leakage a short. It may even see a soft ground fault as a wire-to-wire short. Instead of reacting the word "short" on the panel, use your test instruments instead.

Device Polarity

Don't assume that no one was working on the circuit before you originally arrived. If they took out a horn/strobe and replaced it again, there is a good chance it is wired backwards. This happens especially with in-house maintenance people. I've even seen where professional fire alarm technicians and installers have done this.

A NAC circuit reverses polarity when it's sounding the alarm. It's counter-intuitive, but that means that when the fire alarm system isn't sounding the alarm, the wire that measures "positive" on your voltmeter goes to the "-" terminal of the A/V, and the wire that measures "negative" on your voltmeter goes to the "+" terminal of the AV.

Just check that out.

Surge Suppressor

Unless it's an industrial surge suppressor, designed to work on 100-amp or 1,000-amp circuits, surge suppressors are really one-shot devices.

Think of it this way. Surge suppressors are designed to "short-to-ground" any voltage above, say, 25 volts. Without the surge suppressor, a nearby lightning bolt can easily produce a 50-volt pulse (or greater), at 100 amps (or greater). It's a very short duration pulse.

The surge suppressor is designed to short this out. During that short-duration pulse, though, using Watt's Law, there's nearly 5,000 watts of power being shorted to ground by that little tiny surge suppressor. Not many power-limited surge suppressors are designed to repeatedly handle that.

The only method of checking these surge suppressors out in the field is to replace them.



Douglas Krantz