Thank you for giving me such a detailed question to answer.
Good for you. You are the first one I've ever heard of that is trying to confirm that an already installed Notification Appliance Circuit (NAC) works properly. With the method you're trying, you can make a better judgement on whether a NAC circuit is overloaded, or can be added to. Almost everyone else I've worked with, or received questions from, is thinking in terms of - "Can I add just one more strobe?"
However, there is some confusion in your readings, and I hope to explain the causes. When you understand what's going on, you'll find it easier to calculate, and less confusing.
Estimated Voltage Loss
When designing the fire alarm system originally, you're making estimates for what you are going to have for a voltage loss in a building's Notification Appliance Circuit (NAC). Any real-world resistance on the wires and real-world current draw on the horns and strobes isn't available because, when originally designing, nothing is installed in the building yet. Also, the bottom-line, actual reason to do the calculations is not to see how much voltage is lost into the wires, the bottom-line, actual reason to do the calculations to make sure there is enough voltage for every horn, strobe, chime, etc. on the circuit.
At this time, though, the only method of figuring out the voltage loss on the NAC circuit is to estimate the resistance of the wire using estimated length and wire gauge, and to estimate the current use by using the manufacturer's published data sheets for worst-case current draw. Before installation, if you can know the approximate resistance of the wire and the worst-case scenario current use of the horns and strobes, all you need is Ohm's Law to figure out the estimated voltage loss. It's just Ohm's Law.
Real-World Voltage Loss
In the real world, installers don't measure the length of the installed wire to make sure it will match the engineered plans. Sometimes the wire has to take longer routes than the installation plans, and any additions made after the original installation are usually not shown on the as-builts at all. (Assuming that the as-builts are available.)
Even though the installation plans and proposed voltage loss calculations are usually good-enough for the original installations, the real-world voltage loss on the NAC circuit never matches the proposed voltage loss.
The only way to determine real-world voltage loss is to measure the installed circuit. When there are changes or additions made to the NAC circuit, real-world measurements are the only way to have confidence the NAC circuit will still work under worst-case scenario conditions.
To measure the real-world voltage loss, measure the real-world resistance of the wire, and using worst-case scenario current draw methods, measure the real-world current. Then use Ohm's Law to figure out the voltage loss. That sounds simple, and when everything is pre-measured correctly, the calculated results are fairly accurate.
To do the measurements accurately, though, one has to understand what they're up against.
Ohm's Law
You're seeing a difference between what you calculate and what you measure. Ohm's Law, though, isn't the problem. Series and parallel resistors still use Ohm's Law. There are other factors that have to be taken into account.
- Ohmmeter Low Resistance Measurement Inaccuracies
- Forward/Backward Resistance Reading Inaccuracies
- Varying Working Voltage
- Weird NAC Current
All of these gang up to make the measurements harder.
It isn't shown in the rules or codes, I've never seen it in any books on fire alarm systems, but the method you are describing is the only true method of measuring resistance on a wire; the method is to use an ohmmeter to measure resistance.
Using the formulas and the tables in the code books, the calculations are there for the Authorities Having Jurisdiction (AHJ) because they focus on protecting from fire, they don't have time to make a career out of studying electronics.
Your confusion, in the measured end of line voltages, is caused the by non-linear voltage/current ratios in the NAC circuit.
Example of a Linear Voltage/Current Ratio on a single circuit:
- Voltage is 10 volts - Current is 1 amp - - Ratio is 10:1
- Voltage is 24 volts - Current is 2.4 amps - - Ratio is 10:1
- Voltage is 32 volts - Current is 3.2 Amps - - Ratio is 10:1
Here, the linear voltage/current ratios always remain the same, no matter what the voltage is going to be. Ohm's law is based on linear voltage/current ratios.
Check it out - using Ohm's Law - R = E / I, do the math. The resistance remains the same, no matter what the voltage.
Example of a Non-Linear Voltage/Current Ratio on a single circuit:
- Voltage is 10 volts - Current is 8 amps - - Ratio is 10:8
- Voltage is 20 volts - Current is 4 amps - - Ratio is 10:2
- Voltage is 40 volts - Current is 2 amps - - Ratio is 20:1
Here, the non-linear voltage/current ratios vary according to either the voltage or the current. The ratio is inverted, and it appears to be all over the place. See the section labeled Weird NAC Current later in the email.
Check it out - using Ohm's Law - R = E / I, do the math. The resistance is different for each voltage.
But then, use Watt's Law instead - P = I x E. You'll find that the power is the same for each voltage, even though the voltage/current ratio is all over the place.
Now we'll deal with some of the issues causing your confusion.
Ohmmeter Low Resistance Measurement Inaccuracies
One problem I've had to work with, in the past, was that some ohmmeters aren't accurate at very low resistance reading, like readings of less than 10 ohms. Some of them are very inaccurate at less than 2 ohms.
To gain confidence that the ohmmeter being used is accurate at the low readings, purchase some very low value resistors (cheap ones will work just fine), say 1 ohm, about 3 ohms, about 5 ohms, and 10 ohms. Use the ohmmeter to read the resistance on these, preferably at the start of each day you're measuring wire resistance. That's referred to in the testing industry as checking calibration, and that's how you know that your readings will be accurate.
Forward/Backward Resistance Reading Inaccuracies
Inaccuracies can sneak in, confusing the readings. A big problem is the readings can easily be affected by the blocking diode inside the horns and strobes.
All horns and strobes have a blocking diode inside them to prevent backward supervision current from flowing through them. In other words, in one voltage direction, the horns and strobes have essentially infinite resistance, in the other voltage direction, some horns and strobes can conduct some of the ohmmeter current.
Your ohmmeter is polarized. It pushes a DC current, and the DC current is either blocked by the blocking diode, or allowed into the horn or strobe by the blocking diode. The resistance measured by the ohmmeter will be different depending on the polarity of the probes on the ohmmeter.
Some models of ohmmeter can be affected by extra conduction in a horn or strobe; some readings will be confused and some won't. Some manufacturer's horns or strobes on the circuit can be more of a conduction problem; some will confuse an ohmmeter, and some of them won't.
To avoid confusion from the extra conduction issues, take a resistance measurement, reverse the probes from the ohmmeter, and take another resistance measurement. That way, the blocking diode will be blocking the current for at least one of the two readings. With the two measurements, the good reading is the higher resistance measurement. The lower measurement is conducting through the horns or strobes. Use the higher measurement. It's easy to take the two readings, and it doesn't take much more time, either.
Using the higher of the two readings shows the real-world resistance of the wire. Knowing the resistance of the wire is the first step in the calculations. The next step is to decide what voltage to use when measuring the current.
Varying Working Voltage
A 24-Volt battery is the model of the battery; it's 24 volts - Nominal. Nominal means "In Name Only". Take your voltmeter and measure the value of a 24-volt battery. It's not 24 volts. When the battery is fully charged, it's closer to 27.5 volts; when it's used up, it's closer to 20 volts. Current measurements are going to be vastly affected by the charge/discharge cycle.
Being that this is a life-safety fire alarm system, the current measurements have to be performed under worst-case scenario conditions. The worst-case condition for a Notification Appliance Circuit (NAC) is considered to be at the end of a 24-hour power blackout, where the fire alarm system is just standing-by, plus another 5 minutes or 15 minutes of full alarm.
Keep in mind that the main power supply for most fire alarm systems, and most NAC booster power supplies, will be powered by the battery, not utility power during a power outage. The voltage provided by most power supplies will be the voltage on the battery.
The assumption (yes, for life-safety there have to be assumptions) is that at the end of that time (24 hours plus 5 minutes or else 24 hours plus 15 minutes) the backup batteries are at the end of their possible lives. The manufacturers are all a little different, but to provide enough power, the minimum voltage on the 24-volt (nominal) batteries will be somewhere between 20 volts and 22 volts. Look in the installation manual for this minimum voltage value, or call the manufacturer's technical support team and ask them. They will tell you.
Using the working voltage for the measurements is important. This is because the measured NAC current can be higher, or it can be lower, when the batteries are about used-up.
Weird NAC Current
Inside the old mechanical fire horns, because there are no active components like transistors, the current will decrease as the voltage is decreased; the voltage/current ratio is linear. The voltages, the current, and the resistance follow Ohm's Law because they are constant-resistance devices.
We are no longer using the mechanical fire horns.
Strobes and low-frequency hors are the worst. Both the strobes and low-frequency horns are constant power devices. They are driven by 16 volts to 33 volts, they are required to flash at a one-second rate, or sound off at a certain sound level, and still have the same candela, or the same sound level, no matter what voltage is being supplied. Basically, no matter the voltage that they're given (within their useful range), they use about the same power.
There are active electronics inside the them, so the total power won't quite follow Watt's Law for determining power. However, generally, as the provided voltage goes down, the current used by the strobe goes up; the voltage/current ratio is non-linear.
To get an idea of what I mean, look at EST's Genesis Strobe ratings at:
https://myeddie.edwardsfiresafety.com/Media/Catalog%20Sheets/E85001-0557%20--%20Genesis%20Ceiling%20Strobes.pdf
Page 3 shows the current for each candela setting that the Genesis Strobe uses at 16 Vdc, 20 Vdc, 24 Vdc, and 33 Vdc. (Ignore the Vfwr ratings, batteries will only provide DC voltages; batteries will not provide Full Wave Rectified voltages.)
Genesis Strobe Example: A standard output model, at 15 candelas, will use 63 milliamps of current when provided 24 volts. When the voltage is
decreased to 20 volts, the current is
increased to 74 milliamps. When the voltage is
decreased further to 16 volts, the current is
increased further to 94 milliamps.
That is a lot of current change from 24 volts to 16 volts, and the current has increased, even though the voltage is decreased.
Unlike the old mechanical fire horns, this is not a linear voltage/current ratio for the strobes. All strobes and most low-frequency have this same non-linear issue, and many other fire horns also have this or a similar non-linear issue.
Getting the Kinks Out of the Measurements
When making the current measurements, use what the panel's NAC circuit uses during the worst-case scenario condition. However, rather than trying to discharge the 24-volt nominal batteries down to 20 volts, or whatever the fire alarm manufacturer has for a minimum voltage, purchase a 0-30 volt, 3 amp power supply. Then, when you perform your current-while-in-alarm tests, you can set the power supply voltage to test the NAC at the fire alarm manufacturer's minimum voltage requirements.
Google "dc variable power supply". The web has several 0-30 volt, 3 amp power supplies available at well under $100.00. Some of them will even show in how much current is being used so you don't have to measure it yourself.
For the sake of safety in the field, I recommend that the supply be capable of providing no more than 3 amps. The NAC circuit goes all over the building, and it goes behind walls and ceilings. The NAC circuit is rated for 3 amps, so the power supply should provide no more than 3 amps.
Douglas Krantz
