It's taken me a while to re-find all this information, but these are some things to consider when trying to understand what's gone on through the years. Mostly, it's a realization that fire alarm systems weren't always the way they are now, fire alarm systems evolved over time. Technology had a lot to do with that evolution.
Extra Resistor in Waterflow Zone
You mention an ADT fire alarm system that could use a second resistor in series with the tampers so the tampers could be normally open, and still send a supervisory on the waterflow alarm zone. The whole purpose was to send a separate supervisory signal from the waterflow signal. The system was "Listed" so this method was allowed.
Simplex also used that system. I ran into that while troubleshooting and retro-documenting a fire alarm system. That idea apparently did not catch on to become wide-spread, and it pretty much disappeared. I can only assume that was acceptable because of the need to retrofit newer panels to older fire alarm system building wiring.
Backwards Engineering
I'm an electronic technician. Yes, I've had nearly 20 years servicing fire alarm systems, but that is on top of 24 months of intense electronics training, 15 years as a broadcast engineer, and 10 years as a power limited contractor.
I've performed component level troubleshooting in microphone pre-amplifiers, 35,000 watt television transmitters, and microwave transmission/reception systems. I've found and fixed several engineering-design flaws in equipment as diverse as television studio cameras and low power FM transmitters.
One fire alarm manufacture changed the design of their
addressable modules because I found an interference problem that both of us were having. I had provided the necessary research for them to fix their problem.
In all of the service work that I've done, I had to look into the history of what could be done, and what couldn't be done. Some of the conclusions have to be based on conjecture because detailed information isn't always available. However, the details can often be "backwards engineered" from the final designs.
Example
The history of the
trouble relay on fire alarm panels was different with old systems. Nowadays, when power is removed from the fire alarm system, the trouble relay inside the fire alarm panel relaxes and sends a trouble. It wasn't always required to be that way.
In 1998, when I first started to work with new fire alarm systems and install the panels, one system's trouble relays would relax and send a trouble signal when was power was lost to the panel. With a catastrophic failure, the relay would relax and send a trouble.
Some other brands of panel built around that time would only send a trouble signal when the relay activated. In other words, some fire alarm panels would not send a trouble signal when there was a catastrophic failure in the fire alarm panel. If the panel completely died, the relaxed relay said the panel was "Normal".
A few years later, all panels set up their trouble relays just like the first one; when the power is lost completely, the panel sent a trouble. To me, it looked like the rules changed over time.
Dual View - Technical and Legal
These were both "
listed" panels. Technically, the relay could be wired one way or the other, depending on the panel; legally, at one time both methods were allowed because both methods were "listed" and in use.
Later, everyone changed what they were providing. That had to be because the law now said to change the way it is done. You can't have that much conformity without having a law that requires everyone to conform.
Technical History of the Fire Alarm Panel
The original fire alarm panels were relay-based. The transistor wasn't even invented until 1947, so any fire alarm panel made before then could not have been solid-state. Later, as solid-state technically became more stable and proven, solid-state electronics would allow for fire alarm panels to be built to a reasonably small size, and allow for a more complicated system.
I've serviced dozens of relay-based fire alarm systems. (Yes, they are very obsolete. However, lightning doesn't destroy them, and because some municipalities allow them to remain in service due to Grandfather Laws, they stay.)
I never worked with large relay-based fire alarm systems that have staged evacuation, but I did work on a lot of them in medium to small apartment buildings, schools, churches, etc. Being an electronic technician, I paid close attention to how they worked.
What was common to all of them was that they were three-condition systems. They were either green-light normal,
amber-light trouble, or they were in a
red-light fire alarm condition. They did not have the relays or the circuitry to be capable of having a dual-alarm system: in addition to the red-light general alarm, they did not have an amber-light supervisory alarm.
Just as a reminder, electronically, the only differences between a red-light alarm and an amber-light supervisory is the color of the light, and that a red-light fire alarm makes a lot of noise around the building while the amber-light supervisory alarm only sounds a local-to-the-panel buzzer. (OK, usually the supervisory alarm won't call the fire department, but I know of one municipality that even required that the fire department be called for a supervisory, so even that difference doesn't count.)
Beyond those two differences, the alarm conditions are exactly the same.
The bottom line in all this is that in order for a fire alarm control panel to even have the capability to detect a supervisory alarm, the fire alarm panel requires two completely separate alarm systems built into the control panel: General Alarm and Supervisory Alarm. All relay-based panels I've worked with only have one alarm system built in - general fire alarm.
Many buildings were wired using those relay-based fire alarm panels. And yes, even though those systems did not have a supervisory alarm capability, those panels and systems of zone wiring were listed by the testing laboratories as proper fire alarm systems.
I have personally serviced a huge number of solid-state panels. Most of the really older solid-state panels did not have the capability to separate out an initiating switch that closed-on-alarm (waterflow) from an initiating switch that closed-on-supervisory (tamper). Most of these really older solid-state fire alarm panels did not have supervisory capability at all, they only had one alarm capability - general fire alarm.
Research
In 1998, when getting into fire alarm work, I did a lot of reading. I read from the "
National Fire Alarm Code Handbook", published by the National Fire Protection Association, a non-profit publishing house. The Handbook has all the NFPA 72 Code, and it arranges Appendix A so it is included with the Code rather than at the back of the book. It also has the commentary you mention, and pictures and diagrams to help explain the code.
Yes, I saw that formal interpretation you referred to. I concluded, though, that if they felt that they had to say that interpretation officially, sometime in the past, that was not the requirement. To me, it looked like the requirement had been changed.
Another book that I read was "Fire Alarm Signaling Systems", written by Richard W. Bukowski, P. E. and Robert J. O'Laughlin, P. E. The book was published in a joint effort of the National Fire Protection Association (NFPA) and Society of Fire Protection Engineers, Copyright date 1994. The last time I looked, the book, at least in a later version, is available on the
NFPA website.
It might not be there in current versions of the book, but on page 87 in the first version, they have a picture and a cross section diagram of a fire alarm tamper switch used with an Outside Screw and Yoke (OS&Y) Gatevalve. People actually used the switches shown in the book. I remember seeing that particular switch used in at least two systems.
As part of the caption under the pictures, the authors say: "The switch is adjusted so that, in the open valve position, the contact ring closes the electric circuit between the contact blades. When the valve is closed not over two turns, the plunger tip rides out of the depression in the valve stem, opening the electric circuit."
This tamper switch also used a conductor cord instead of three screw connectors, and it was only a two-conductor cord. There was no option; it opened up the circuit when the gatevalve itself was closed.
OK, the engineers should have used a different switch in their 1994 publication. However, the switch they did use showed me that at some earlier time, the switch opening up the circuit to send a trouble signal was an acceptable practice.
Grandfather Clause
I've worked with fire alarm systems that were originally installed in new buildings constructed in the 1950s, 1960s, 1970s, 1980s, 1990s, 2000s, and 2010s. Over time, many of the buildings' original control panels had been changed. Some of them were originally relay-based, then changed to discrete component solid-state based, then changed to Central Processor Unit (CPU) based. Many of them have been changed from conventional systems to addressable systems.
Much of the time, when the panel was relaced, the building was not rewired from scratch; only the devices and panels were changed. Even when upgrading addressable systems, some of the buildings were not rewired. The old building wiring was grandfathered in - and the grandfathering was acceptable to many fire marshals as long as the fire alarm system worked as advertised.
That explains the series wiring for tamper switches. Rather than tearing apart finished buildings to run separate wiring, if the system of connecting the waterflow and tamper switches had been acceptable to the NFPA at one time, the system of connecting the waterflow and tamper switches was re-used.
Capabilities
No, I have not actually seen where at one time in the past the NFPA 72 specifically accepted series wiring for tamper switches - I don't have a library of old
NFPA Code books. However, based on the technical capabilities of older fire alarm panels, it can be seen that a "Supervisory Alarm" could not have been in existence much earlier than the 1970s. The circuitry inside the panels made at that time (the panels having been listed as having been tested and found to work according to the rules of the NFPA Code at the time) just did not have the technical capability of a providing a separate supervisory alarm.
As a fire alarm technician, the biggest problem that I have is that I have to deal with what other people have done. In this case, I have to deal with fire alarm systems that have the tamper switches wired in series (so they send a trouble signal on a waterflow zone). They still exist. There are quite a few systems still around that are series wired.
When I work on a fire alarm system, I can't change the tamper switch wiring without re-certifying the systems. If that is the way that the system was certified and accepted by the fire marshal, and it produces the proper results (in the case of a valve tamper switch that requires the valve to be opened again to get rid of the trouble indication), and it isn't dangerous, I can't change the wiring without re-certifying the system.
The best I can do is work with the fire alarm system the way it is; I can't bring the system up to current code. I wrote the article because others also have to deal with the archaic system for the tamper/waterflow combination, and it would help them to know how it actually works.
Douglas Krantz 
P. S. - - Paraphrased from the U S Constitution: If it was legal at the time something was done, it is still legal at the time it was done. There is no "reach-back" for illegality. Of course, the government can say it has to be upgraded, but they can't say it is wrong.
