AFCI compatibility with Electromagnetic Interference Filters

Status
Not open for further replies.
al hildenbrand

al hildenbrand

Senior Member
Location
Minnesota
Occupation
Electrical Contractor, Electrical Consultant, Electrical Engineer
Do electromagnetic interference (EMI) filters have an effect on the efficacy of the combination type AFCI?

I looked at the White Book classification, FOKY, for EMI filters, but it has little content. I've looked at a product line by a manufacturer, but have not scared up much technical information yet. I did find a representative data sheet for the small cord connected unit that utilization equipment would be plugged into.

I have seen EMI filters offered as an alternative to layered SPD (whole house & outlet).

I am wondering if the EMI filter will block part or all of the "bad" arc signatures generated by a fault on the load side of the filter. Any ideas?
 
It would seem they would if the arc signatures were a voltage amplitude type modulation, but from my understanding it is the current and frequency modulated signature that the AFCI is looking at (75 amps at that), Good question as there are many loads that could mask either of these. The question is how deep has UL done it's testing with as many types of loads or devices also on the protected circuit?
We know that a current modulated signature can be masked by a simple capacitor in the circuit, whether it is part of a motor starter, or an LC in a TVSS strip?

I posted a couple of years ago to the problem of the circuit impedance with AFCI and posted the replies from UL which went no where. but it still remains that I think much more testing needs to be done!

At least this is my opinion

Edited to add:

The filters in most of the SPD's I have run into filter both up and down stream from the device, as I found out using X10 technology
 
Last edited:
al hildenbrand said:
Do electromagnetic interference (EMI) filters have an effect on the efficacy of the combination type AFCI?

I am wondering if the EMI filter will block part or all of the "bad" arc signatures generated by a fault on the load side of the filter. Any ideas?
Click to expand...

Hi Al,

That is a great question. Most people say that if you push the test button on the AFCI and it trips you are good to go :D

I think this would require some sort of external "indicator" or perhaps even a "tester" to sort out, but there is no such thing as a real AFCI tester:cool:


Just kidding ....

It would be my opinion that EMI filters could have an effect on an AFCI's ability to detect an arc downstream of that EMI filter. It would vary based on the type of EMI filter used. EMI filters with large inductances (designed to
attenuate lower frequencies) would have the most negative effect.

The EMI filters you pointed to I think are "Series mode" and I believe tend to have larger inductors since their claim to fame is no MOVS.

Some typical values for commercial EMI filter inductances (general purpose) might be in the range of 1 -10 mh. The inductive reactance of these coils at 20K -100Khz I think might be on the edge of attenuating the high frequency components of the Arc enough to be an issue.

These coils might also reduce the 60 hz di/dt rate (when arc is established and extinquished) enough to be an issue.


This is one of the reasons I beleive in the external tester idea. What about testing of the downstream circuit impedance!


UL and AFCI manufacturers mention that use of a AFCI indicator may indicate that a AFCI is defective when it is not really. As long as the test switch on the AFCI works then the breaker is good they say.

One of the reasons that the external tester may not trip the AFCI is due to circuit impedance. Wouldn't it be nice to know if there is a circuit impedance issue?

Once again it makes me wonder if these devices are really doing their job - in all installations - if all you are ever required to do is to push the test button.
 
One of the reasons that the external tester may not trip the AFCI is due to circuit impedance. Wouldn't it be nice to know if there is a circuit impedance issue?

Once again it makes me wonder if these devices are really doing their job - in all installations - if all you are ever required to do is to push the test button.
Click to expand...


I wish I had my other computer set up so I could get the e-mail responses from UL off it to post here, but I think Al remembers them, they were funny, when I first brought this up to UL, it sent them in a frenzy, and it was like how could we have missed this, and after several long E-Mails back and forth one of there engineers tried to tell me that the circuit impedance would not affect the detection of the arc fault device, but could never answer, when I ask if the IDEAL ARC tester could not trip them then how could a real ARC @ 75 amps trip them. well thats why I said "it went no where"

What gets me is computer software manufactures have to test the software with many computer configurations, hardware configurations ETC... to make sure it preforms as intended.
But life safety devices only need a test button to be sure its working:-? to them it doesn't matter if devices and loads on the circuit , and even the circuit impedance it self might render these life safety devices useless:mad:

Don't get me wrong, we need UL as it would be very much worse without them, but I think they need to think out of the box a little to research the "what ifs" a little more to help make the home more safe from products that might not preform as intended.:grin:
 
hurk27 said:
ELA
I did find my old post regarding the exchange between UL and myself over the issue of circuit impedance, it's long been closed but it kind of shows as to what happened.

http://forums.mikeholt.com/showthread.php?t=69614
Click to expand...

Interesting read from a while back as you said. As George pointed out AFCIs now also detect series arcs above 5 amps.
As far as I know the ideal only tests parallel arcs up to this point?


Wptski was kind enough to provide the scope trace of the ideal test waveform in this other thread:

http://forums.mikeholt.com/showthread.php?t=111163 (see post #9)


From my investigations and limited testing I know the arc signature of the series arc varies from the parallel arc. My guess is that it is more difficult to simulate in the electronic form than the parallel. In my testing to this point I have only simulated the parallel arc test.

Ideal will probably need to adjust their devices to produce a series arc as this should help overcome the issue of excessive line Resistance. Excessive line Reactance may still be an issue for detecting a low level series arc.
 
Status
Not open for further replies.
Top