Digital circuit breaker

I've read this whole thread, and the article, and can't imagine what this product can be used for. In residential, if my wife turns on the hair dryer while the space heater is on, I guess this new tech saves me a trip to the basement. But for maintenance work, you still need an air gap. Also, @MNicolai makes a great point. How you gonna start switching amp ratings on any circuit you want? Any consumer device/appliance that is rated lower than a standard circuit is going to have some sort of fuse/breaker built into it. Even extension cords have fuses. So downgrading your circuit at the breaker to protect a device would be redundant. And even if you chose to do it, what application would require doing it so often that you'd need the convenience of doing it from your phone? Also no one is considering INCREASING the rating of the circuit, right?

someone help me out. I don't know much about power outside of residential, what are the possibilities of this?
 
Last edited:
Someone mentioned air gap and there is an air gap. You might look at the data sheet: https://docs.wixstatic.com/ugd/1baea3_0d836aba4fb143219cbc0f452352cb92.pdf

And its UL 489 - which is tougher than a 1077 - a basic breaker stabs into a panel. https://www.c3controls.com/ul489-ul1077-guide/

And only 50 and 100 amp in 2 and 3 poles.

And I'll bet pricey - so don't worry about what in your basement - I doubt if any of us can afford it in our homes. From their web site: "The Atom Switch is intended for the commercial and industrial market only, please do not submit requests for residential applications."

I think the motor control applications might be interesting, perhaps a better current sensing transformer because its faster to modify current limits.

I expect if CB had been here 57 years ago and said - look at the LED thing - there be little support for it ever being the source for a theatre light. (I do think it will dim on a resistance dimmer though.)
 
One very important capability in the residential market would be combining OCD’s, Arc Fault Interrupter, and Ground Fault Interrupters into one smart device that does all 3. If the digital circuit could sense arcing and ground faults, than you eliminate those devices from needing to be installed where code requires, as well as being retrofitted. Possibly a labor savings upgrading a panel, rather than a panel as well as needing to install AFCI and GFCI’s all over a house on an upgrade to an electrical service.

As well, possibly the controlling devices would be better at powering LED’s, then what an existing Triac can do. Making a home “smart” can go a long way towards saving electricity, which I think is the goal.
 
Last edited:
It seems like they're targeting a world where you would need to flip back and forth often from multiple sources. Solar, grid, battery, generator, etc, while maybe reducing the amount of outboard switchgear required.

They could be thinking about load shedding during peak hours or while on battery/generator. I can't see a utility wanting to leverage a building owner's system in this way, but my utility already has remote on/off of my HVAC if they want to stagger loading. That's a device that they own and provide an incentive for that they place between my HVAC and the circuit breaker. But let's say you've got a couple EV chargers in your garage and you only want 1 running at a time, and don't want them to run at the same time your AC is ramped up.

For an industrial application, maybe they're talking about switching mechanical loads on/off during peak hours. For example, my company does a lot of consulting on central energy plants where the owner wants to use their cooling towers and chillers to make ice for a chilled water loop. They have an agreement with the utility where when the utility goes off peak-demand hours and the electricity rates go down overnight, the chiller system fires up and runs overnight on cheap power. Then they make ice all night and store it in tanks. Then they switch off again before peak demand the next morning. During peak demand they only run the pumps to cycle water through the chilled water system. Saves on energy costs while being a little friendlier to the power grid. Not sure hard-switching of mains power through a third party's control system is how they would ever want to do that in practice though.

One of the big issues I think they'll have is that there aren't electrical rooms anywhere in the world with enough wall space to handle how low-density these panels are. You aren't going to slay the huns with 14 switches per panel.

And only 50 and 100 amp in 2 and 3 poles.

I thought that was weird and just a matter of them prioritzing their product offering during initial testing. Taking another look at the cut sheet, I think it's because the idea is that you buy a panel full of 50A modules and through software you can can configure them to be whatever you want. Of course there's nothing to prevent someone from forgetting to set this or from setting it incorrectly, and suddenly your OCPD doesn't match your load wiring's capacity. The primary benefit appears to be that they can achieve economies of scale by making a few modules that cover every voltage/phase/current/frequency/trip scenario. Same module for 20A, 60Hz, 120V can do 30A/277/50Hz.

upload_2019-5-24_23-11-16.png


upload_2019-5-24_23-15-11.png
 
Unless I'm misreading the NEC definition of a CB at the top of this article, it doesn't qualify anyway:

https://www.mikeholt.com/mojonewsarchive/NEC-HTML/HTML/ElectricalCircuitBreakers~20020419.htm

"Non-automatic Means".

Well, definitions in codes and standards are not requirements. And I'm not sure your point since this device does have a mechanical switch and air gap. And I'm not where I can look but I believe the NEC references the UL standards for these devices like it does for so many devices Determining if a specific product is acceptable for code is what UL standards do.
 
Well, ok, Bill, but if the formal definition in part 70 says "non-automatic", that's kind of controlling, isn't it? You wouldn't say it if you didn't mean it...

And, presumably, it *means* "cannot be automatically operated except to/when tripping", no?

I'll have to go look at the actual section.

We've confirmed this device is *not* an SSR, as people thought, though?
 
Upon further investigation, this looks like it has good possibilities. One of its most powerful endorsements is that Siemens, ABB, and Eaton are all investors in the company. They are major circuit breaker manufacturers that have been working on their own solid state breakers for some time. In my mind, that makes Atom a potentially serious player going forward, and gives them access to a large body of "institutional knowledge".

The Atom breaker does indeed have the correct UL Listing to UL489--molded case circuit breakers. For some reason, I can only find the 60A frame size in the UL online certifications directory. BTW, if you want to know what standard applies to every type of device in the NEC, see Informative Annex A in the back of the Code.

The most compelling feature of these devices appears to be their intelligent and fast trip sensing, which will reduce incident energy and significantly reduce or eliminate arc-flash hazard on fault currents up to 100kA. This will have the most application in larger frame sizes.

The downside is low density, 2x-5x cost over traditional breakers, and power dissipation of the solid state silicon carbide switching devices. I seriously doubt that this device will displace single-pole branch breakers in the 15-50A range, if only based on cost.

ST
 
Reading further into the specs, it appears that the device is indeed a combination of Mechanical AND Solid State relay. The airgapping is provided by the mechanical relay, which to me anyways, makes this product very interesting. There are also separate states for the SSR tripping and the airgap being tripped.

Yes, you can LOTO the airgap.

The breaker has seperate states for being:
- completely airgapped
- a mode they call "standby", which closes the mechanical relay but leaves the SSR off
- normal operation

As well as several tripped states of the SSR and Mechanical Relay depending on the fault type.

With most of the work being done by that SSR, that mechanical relay should almost never experience arcing / pitting or otherwise damage caused by switching under load.

The "standby" mode as i understand it essentially trades speed for that small leakage current. the SSR can change states incredibly quickly compared to a mechanical switch.
My understanding is that you can configure it to airgap such that your LEDs don't start flickering downstream.
 
Well, ok, Bill, but if the formal definition in part 70 says "non-automatic", that's kind of controlling, isn't it? You wouldn't say it if you didn't mean it...

And, presumably, it *means* "cannot be automatically operated except to/when tripping", no?

I'll have to go look at the actual section.

We've confirmed this device is *not* an SSR, as people thought, though?

There are motorized circuit breakers. How is his any more or less non-automatic? digital - SSR or some other solid state device - and air-gap. The virtues of both. Probably with the cost of all and more.
 
@Jay Ashworth, in my reading of the NEC's definition, they're just stipulating that in addition to having an automatic means for opening the circuit as a result of overcurrent, a circuit breaker also needs an additionl means of being operating which someone can readily open or close the circuit. They use the term "non-automatic". That's not to say you cannot have an external controller like you would with motorized circuit breakers. It means if you open the panel up there's a way to switch a circuit on/off -- that the only way to operate the circuit breaker is not by tripping and resetting it.

upload_2019-5-25_19-36-15.png
 
So, you're saying the standard *requires* a non-automatic actuator, but does not *forbid* some other automatic one, in addition to the actual overcurrent-trip mechanism.

Yeah, I guess that's a sensible reading of that. :)
 
It looks to me the air gap can only be reset manually at the device, which seems to be the point of non-automatic. You can turn this off and it only can be reset at the device by manually.

It would not be listed if it didn't meet the requirements for a breaker. Interesting the backing it has.
 
It is Listed. Granted, they don't say what standard. And I did say possibilities (def:. a thing that may happen or be the case) not certainties.

Late to the party as always, but these things are technically Molded Case circuit breakers, listed to UL489 (Molded-Case Circuit Breakers, Molded-Case Switches, and Circuit-Breaker Enclosures) and UL1557 (Standard for Electrically Isolated Semiconductor Devices). As for NFPA, it does not "approve" anything, it issues requirements which must be fulfilled.

The questions and considerations raised here are valid to some degree, but the key here is the speed with which these breakers can interrupt and fault and the complete lack of an arc inside the breaker because there are no parting contacts that interrupt the current. The speed is critical because the goal is to reduce the "let through" current which feeds the fault and which can result in an arcing fault downstream of the breaker. Arcing faults are bad and high-speed interruption could potentially be considered "other approved means" under NFPA 70 240.87 (Arc Energy Reduction). If they pass the requirements of UL489 (and there are a lot of tests that are required), they work.

They are constructed from solid state switches but not SCRs or TRIACs. They have a non-zero ON resistance and dissipate power when current flows through them. ALL circuit breakers have a non-zero ON resistance. The reason why we only see smaller (100A) breakers is that it's difficult to make very low resistance solid-state current paths. I have no doubt we will see more of this sort of thing in the future. My business is circuit breakers and we would LOVE to replace the huge contacts, arc chutes and spring-driven mechanisms with a big a$$ solid state switch that can interrupt 100kA at 690V.

The fact that they can be networked and remotely controlled is neither good nor bad, it just is. There are lots and lots of power system controls that are connected over a TCP/IP network and, when poorly managed, accessible from the greater Internet. We have firewalls and access controls for a reason. But just because you can network them, does not mean that you must.
 
They are constructed from solid state switches but not SCRs or TRIACs. They have a non-zero ON resistance and dissipate power when current flows through them. ALL circuit breakers have a non-zero ON resistance. The reason why we only see smaller (100A) breakers is that it's difficult to make very low resistance solid-state current paths. I have no doubt we will see more of this sort of thing in the future. My business is circuit breakers and we would LOVE to replace the huge contacts, arc chutes and spring-driven mechanisms with a big a$$ solid state switch that can interrupt 100kA at 690V.
not SCR's, so that leaves only IGBTs (most likely) or MOSFETs.
 
Or something a little newer... SiC devices. Try to to keep up, people. :angryoldman:

The SiC device in question is a MOSFET with a different semiconductor substrate than the typical (pure) silicon. You could also make a germanium MOSFET, I suppose, though it wouldn't likely be of much practical use; and probably one with various other substrates as well.

Besides MOSFETs, SCRs, TRIACs, and IGBTs, there are a couple of other basic semiconductor devices that conceivably can be used to switch power, such as plain old BJTs (or Darlington pairs of them) and JFETs. BJTs are generally inferior to MOSFETs and IGBTs for high-power applications these days, but some new power JFETs might be suitable for this application and have quite low on resistances and hence parasitic losses. They are not particularly inexpensive, though, and likely have some other drawbacks.

SCRs and TRIACs would seem to me unlikely to be used for a solid-state circuit breaker because they can't be triggered to turn off, but only to turn on; they turn off when the current through them is zero (or very nearly so) for a sufficiently long time (tens to hundreds of microseconds typically). This means that the device could not react faster than a half-cycle of the AC waveform; and worse, I suspect it may be possible for some fault conditions to prevent commutation entirely for somewhat longer periods of time.
 

Users who are viewing this thread

Back