Do they mean they are prohibiting electronic, as in computers, networks, logic and all the scope for error that gives, or also pure electro-mechanical where a switch contact, or ideally two of them both operated together, really is in series with the actuator/lock  (E-stops that are for use in safety of life cases have two force -guided contacts that are mechanically arranged that the failure of one would be reported, but not prevent the power being cut.)

Personally I'd be surprised if anyone who understood the problem would not accept a variant of the second as being a lot safer than the first. I'm wary of things where the push button by the door calls some central controller, and that in turn issues some network commands to release the doors.

Mike.

lyledunn said:

Can any one conceive of a situation where the interruption of power to the electro-magnets might not occur and how that might be overcome?

Short in a component on, or between tracks of, the local Access Controller PCB if only one switched "hold" is used. (Issue here is that the power is supposed to be removed to release, the short or faulty transistor prevents this happening until the PSU ... or PSU and battery ... fails.)

"Wired AND" control (two relays or transistors per mag-lock "hold") is all I can think of to remedy that problem ... alternatively, regular (weekly or daily) testing of the release operation of each lock to help check there is no fault.

I’m not sure it answers your question but on a site I work on they have Paxton access control to most doors. Next to each door is a green push button and a break glass unit. I asked what the difference was and apparently the push button issues a command to the entry system to unlock whereas the break glass is wired directly into the supply to the electro magnet.

yes - that is how we do it too - the break glass always works even if the network is lost, while the exit button or keyfob readers rely on the connection to the computer and each entry or exit request is logged before opening.  It is what Paxton suggest for that situation.
Mike.

The supply to the locks is connected through the N/C contact of the green break glass and the N/C contact of the MICCO and the N/C of the access control panel. Thus, when the fire alarm is activated, the supply to the locks is interrupted. The fire alarm has a backup battery, so it would still function during a power outage. Are you implying that the FA battery was also removed?

Yes, the battery was also removed from the FA.  The requirement is that the doors should be released to open on failure of either or both the supplies to the FA panel.

I don’t know about that requirement, so use the emergency green break. But what if the button doesn’t work when you press it? Maybe you should leave the backup battery out of the access control panel so then the lock will fail open. The Access CP would have to be on the same circuit as the FA.