GFCI for lower than 120V

Status
Not open for further replies.

charlie b

Moderator
Staff member
Location
Lockport, IL
Occupation
Semi-Retired Electrical Engineer
Without going into details (I don't know them anyway), is there a product out there capable of acting as a GFCI for a circuit that operates with a voltage that varies within the range of 0 to 70 volts? I would not use a standard GFCI receptacle outlet, even if it would work, for fear that someone would unwittingly plug a 120 volt component into that outlet. Any ideas?

By the way, this is not a DIY project, but rather a question raised by one of our clients.
 
Only thing I can think that would work is a GFCI relay with a separate CT. The 0-70 volt wires pass through the doughnut while a separate 120 volt circuit is brought to the GFCI relay.


Also, curious minds are curios... :p but what equipment is looking for 70 volts?
 
Without going into details (I don't know them anyway), is there a product out there capable of acting as a GFCI for a circuit that operates with a voltage that varies within the range of 0 to 70 volts? I would not use a standard GFCI receptacle outlet, even if it would work, for fear that someone would unwittingly plug a 120 volt component into that outlet. Any ideas?

By the way, this is not a DIY project, but rather a question raised by one of our clients.

The differential current transformer would not care in the least about the utilization voltage of the equipment, and the power supply to the electronics might well work over a range of 70 to 120V nominal.
But there is no way that it (either receptacle or breaker) could operate all the way down to less than 5V.

The product capable of doing that, probably not packaged and definitely not cheap, would be a current sensor wired to a sensitive relay that then operates either a contactor requiring manual reset or a shunt trip breaker.
I could do it for you easily with a PLC or smart relay and a current sensor wired to a differential CT. :happyyes:

The key requirement is that the power to operate the unit must not be variable.

Thinking again, there is also the slight chance that if you control the voltage in a way that does not create an SDS you might be able to put a conventional GFCI on the input to the voltage converter.
Depending on the location of the fault you might get a variable threshold current that increased the lower your voltage setting.
 
161121-1940 EST

Can it be done probably. Does one exist? The probability is almost zero unless your customer's application is one that someone else already has had solved.

You did not specify frequency which is one possible problem area. DC could not likely be done with a current transformer. This becomes a frequency response problem. Low frequency probably requires a DC shunt method, or a Hall sensor. Zero volts won't work. Some power source is required.

.
 
161121-1940 EST
or a Hall sensor.
.

Many Hall effect sensors have serious problems with zero shift at high sensitivity applications, especially in the Earth's magnetic field. If you want to have a set-and-forget installation you may have to build in zero calibration by shutting off the load circuit periodically.
 
Where is this 0 - 70v coming from ?
Can you not put the GFCI on the main power feed to what ever it is that is producing this voltage ?

In most cases that will not work.

A GFCI supplying a transformer will not see a ground fault in the secondary side as anything more than load current.
 
In most cases that will not work.

A GFCI supplying a transformer will not see a ground fault in the secondary side as anything more than load current.
Except if it is an autotransformer or an isolation transformer with the two low sides wired together and no ground neutral bond on that combined terminal (non-SDS connection.)
In the latter case fault current will travel on the EGC all the way back to the system bond and then back out the primary neutral to get to the secondary neutral. That will cause an unbalanced current on the primary neutral which the GFCI can detect.
 
Something like this should work, you will need a contactor or relay to go with it. There is list of relays that is approved with it. I see this is a 1999 dated document, some of those relays are under different names now, but the device with an external CT on the back was mostly what I was trying to point out. Took a little searching to find one, key word that finally yielded results I was looking for was "high amperage" GFCI.
 
Something like this should work, you will need a contactor or relay to go with it. There is list of relays that is approved with it. I see this is a 1999 dated document, some of those relays are under different names now, but the device with an external CT on the back was mostly what I was trying to point out. Took a little searching to find one, key word that finally yielded results I was looking for was "high amperage" GFCI.

X2. Anything with a separate CT will do the job. Only the electronics care about a steady 120 volts.
 
The 'external ground fault sensor/relay' that others have described above is IMHO the only one that would plausibly work.

The leviton 6895 looks like a nice packaged version of the sensor, however it appears to be discontinued.

If you can find one, then you would supply 120V power to the 6895, and use its 120V output to control whatever is supplying the 0-70V source. The 0-70V leads would be threaded through the CT hole on the back of the 6895; in the event of a ground on the 0-70V side the CT would trigger the unit and cut off the 120V power, and this would have to somehow cut off the 0-70V output.

The CR Magnetics CR7310 http://www.crmagnetics.com/relays/cr7310 is a plausible sensor to use, but I don't think it is a 'class A GFCI'. Minimum sense current is 10mA. I have used this relay in applications where I was looking for about 3A ground faults, not in the same class at all.

-Jon
 
The leviton 6895 looks like a nice packaged version of the sensor, however it appears to be discontinued.

May be that it was replaced with something else with similar changes/listing requirements as has occurred to receptacle type devices and also resulting in new catalog number? That document I linked to did have 1999 copyright on it.
 
Thanks for the replies, everyone. There is very little I can share on this one. The component under discussion is something that my client purchased. It gets a 120 volt supply and has its own transformer and secondary voltage controller. Their concern is that if the item were to be accidentally damaged in a certain way, it could become possible that live parts would be accessible to the touch of a person’s hand.

I would not be concerned if the GFCI device does not function at the low end of the voltage range. There would be an insignificant risk of shock at that voltage.

My inclination is to advise the client to look to the manufacturer for a solution. They should be able to build something into the voltage controller that can detect leakage current downstream.

Here is another question. I understand that the component draws less than 2 amps. So the total wattage is below 140. I do not understand any of the code articles that deal with “intrinsically safe” power systems. Is it possible that this component, and its power supply, fit into that category? If so, is that enough to alleviate my client’s concerns?

 
Thanks for the replies, everyone. There is very little I can share on this one. The component under discussion is something that my client purchased. It gets a 120 volt supply and has its own transformer and secondary voltage controller. Their concern is that if the item were to be accidentally damaged in a certain way, it could become possible that live parts would be accessible to the touch of a person’s hand.

I would not be concerned if the GFCI device does not function at the low end of the voltage range. There would be an insignificant risk of shock at that voltage.

My inclination is to advise the client to look to the manufacturer for a solution. They should be able to build something into the voltage controller that can detect leakage current downstream.

Here is another question. I understand that the component draws less than 2 amps. So the total wattage is below 140. I do not understand any of the code articles that deal with “intrinsically safe” power systems. Is it possible that this component, and its power supply, fit into that category? If so, is that enough to alleviate my client’s concerns?

AFAIK "intrinsically safe" relates only to ignition hazard in classified environments. It does not insure against shock hazard.
And, in any case, 140W is far above intrinsically safe.
 
Something like this should work, you will need a contactor or relay to go with it. There is list of relays that is approved with it. I see this is a 1999 dated document, some of those relays are under different names now, but the device with an external CT on the back was mostly what I was trying to point out. Took a little searching to find one, key word that finally yielded results I was looking for was "high amperage" GFCI.
No, not on the LV side of whatever this is.
Cat. No. 6895 is UL-recognized and requires a 120VAC line-side feed for sensor electronics to operate.

And that's the problem here, a GFCI will have an electronic PC board that requires control power. Because the only thing that REQUIRES a GFCI per the NEC is going to be at least a 120V circuit, there is zero incentive for the mfrs to make a PC board that has it's own wide range power supply on board. The only other valid use of a Class A GFCI would be in an industrial control panel built under UL 508A that requires the use of an unevaluated device, in which case you must feed that device through a Class A GFCI and an isolation transformer, and even then, UL shows that you put the GFCI device on the 120V feed ahead of the transformer. But below 30VAC RMS, you do not need the protection of the GFCI at all, you can just use a Class 2 transformer, so there is no reason to make one.

So even if they did make a GFCI that could accept a voltage lower than 120V, it might go down to 48VAC, but never "zero to..." anything. I say 48VAC just because that is the only standard control voltage level in use that is greater than 30VAC RMS but less than 120V. Does that device exist? I doubt it. I looked at the common brands; Leviton, LeGrand/P&S, Hubbell none of them say anything other than 120VAC.

Can someone MAKE one for you? Maybe. Bender makes a wider range of options for GFCI devices, here is one:
http://www.bender-us.com/solutions/gfci.aspx
Buried in the brochure under the available voltage codes (starting at 120VAC by the way) is this statement:
Other voltages available upon request. Contact Bender for more information.
If they need to have it, might be worth a call. But be prepared, you are looking at a $800 device, minimum (last time I had to buy one).
 
The search term 'GFCI sensing module' appears to be the key.

This leads to several Hubbell products, eg:
http://www.cesco.com/b2c/product/Hubbell-Wiring-GFSM120M-Manual-Set-GFCI-Sensing/93853

Bryant: (the same thing under a different brand???)
http://ecatalog.bryant-electric.com/literature/BSS112.pdf

North Shore Safety:
http://www.nssltd.com/pdfs/Layout for sensing module cutsheet.pdf

These all are 120 or 240V devices with external sense donuts. That means that the sensor itself is powered by, and is switching a 120V (or 240V) circuit, however what it is measuring is the current on any arbitrary wires fed through the sense donut.

It is not clear (and IMHO not likely) that the sensor would detect DC leakage, and it is not clear how the sensor would deal with anything other than 60Hz AC (if the device generating the 0-70V was using PWM for example)

The Bender units that jraef are specified for AC or DC systems, which suggests that they have lots more flexibility, probably for lots more $$$$.

-Jon
 
The differential current transformer would not care in the least about the utilization voltage of the equipment,
That was my thinking too. You would need a seperate power supply to operate it but I wouldn't see that as being a show stopper.
 
But... the term "GFCI" has a very narrowly specific definition here in North America, as a device intended to protect personnel from shock hazards at levels not above 6mA of current and listed as "Class A" under UL 943 (or any other acceptable NRTL test standard, if there were one, but there is not). So "custom" solutions will not technically be a "GFCI" in the strict sense of the term, they would be, at best, custom shock hazard protection devices that would have no official recognition.

I would think that, given the unlikely ability to get someone to create a UL 943 listed GFCI for low voltage for you, it might be better for them to just alter the equipment somehow to limit that voltage to 50VAC or less rather than 70VAC (if possible), then make that entire portion of the equipment into a Class 2 circuit.

725.
Class 2 Circuit. The portion of the wiring system between
the load side of a Class 2 power source and the connected
equipment. Due to its power limitations, a Class 2 circuit
considers safety from a fire initiation standpoint and provides
acceptable protection from electric shock.
 
Last edited:
So "custom" solutions will not technically be a "GFCI" in the strict sense of the term, they would be, at best, custom shock hazard protection devices that would have no official recognition.
I think even that could satisfy my client's concerns. I suggested use of a GFCI, if there was such an animal that could operate with the component's voltage range. But they will take any solution that would reduce the risk of shock, in the event someone bumped into this component and damaged its outer surface.

 
Status
Not open for further replies.
Top