AFCI Headache

[iwire]

Because there is no standard on how the AFCIs detect the arc...there are only standards on what arcs to detect and when to trip the circuit.

Well I agree with you there as far as the AFCI manufacturers should be able and willing to provide info to the appliance manufacturers as to the trip characteristics of AFCIs.

I don't see this as even remotely close to the GFCI issue.

We disagree there, IMO you are just choosing what conveniently fits in with how you feel about AFCIs.

I have never been happy about AFCIs either but the code is in place, all manufactures will have to deal with it.

But lets forget about GFCIs, lets talk about standard breakers.

Say I start 'Bob's Appliance Company' and I design an air compressor to run on a 20 GP amp circuit, its RLA are well within the 20 amps circuits capacity. However being cheap I do not include an unloader so the start up current exceeds the trip curve of certain manufacturers breakers.

Is that the breakers fault or my fault for not designing my equipment to work with all breakers?

to work with another device that is really based only on voodoo science?

I do not agree that AFCIs are 'Voodoo science' the fact that the current AFCIs have not fully matured does not make the concept junk. IMO it only means they where required to early.

 

[George Stolz]

I do not agree that AFCIs are 'Voodoo science'...

Bob,

...I have no way of diagnosing the problem.

We have a black box that we can't open and we can't determine with any meter what ticks it off. We can't test a circuit in the field with a device to confirm the breaker's conclusions. Isn't that kind of voodoo?

I'm not saying it's snake oil, but I'd go along with voodoo.

 

[iwire]

Bob,

...I have no way of diagnosing the problem.

The fact you cannot diagnose an issue does not mean that issue is based on 'voodoo science'.

I could not diagnose an electron micro scope that does not mean it's operation is not based on real science.

We have a black box that we can't open and we can't determine with any meter what ticks it off. We can't test a circuit in the field with a device to confirm the breaker's conclusions. Isn't that kind of voodoo?

I'm not saying it's snake oil, but I'd go along with voodoo.

IMO you do not understand the term 'voodoo science'

Voodoo science, is a neologism referring to research that falls short of adhering to the scientific method. The term was popularized in the book Voodoo Science: The Road from Foolishness to Fraud, by professor and scientific skeptic Robert L. Park.

Park uses the term voodoo science as a catch-all concept covering four categories sometimes difficult to distinguish from each other:

* pathological science, wherein genuine scientists deceive themselves
* junk science, speculative theorizing which bamboozles rather than enlightens
* pseudoscience proper, work falsely claiming to have a scientific basis, which may be dependent on supernatural explanations
* fraudulent science, exploiting bad science for the purposes of fraud

http://en.wikipedia.org/wiki/Voodoo_science

IMO AFCIs do not fall into any of those categories.

I will go with 'secretive technology' making life difficult for many.

 

[George Stolz]

IMO you do not understand the term 'voodoo science'

Yeah, I'll go along with that.

:Slips back into the inky shadows:

 

[don_resqcapt19]

...
I do not agree that AFCIs are 'Voodoo science' the fact that the current AFCIs have not fully matured does not make the concept junk. IMO it only means they where required to early.

In my opinion it is voodoo science both on the need and the actual operation of the device.

 

[don_resqcapt19]

...

Say I start 'Bob's Appliance Company' and I design an air compressor to run on a 20 GP amp circuit, its RLA are well within the 20 amps circuits capacity. However being cheap I do not include an unloader so the start up current exceeds the trip curve of certain manufacturers breakers.

Is that the breakers fault or my fault for not designing my equipment to work with all breakers?
...

Here I also see a major difference. The operation and trip curves for standard breakers are readily available to Bob's appliance company and his engineers can design to fit the curves. The algorithms for the AFCIs are not available and Bob's engineers cannot design to the them even if they wanted to.

 

[ramsy]

I have just heard that CH has introduced an RFI filtering AFCI breaker. Any word or spec on it?

Welcome dstryr. The trolls are at play because no data exists on AFCI tests using RFI filters.

In 1997 Square D published a testing brief of AFCI's that used RFI filters to mask the signal.
See last page. Test 15.1
http://www.cpsc.gov/LIBRARY/FOIA/meetings/mtg98/afci2.pdf

RFI filters may have bean used to disable this AFCI entirely.

 

[Ohmy]

Here I also see a major difference. The operation and trip curves for standard breakers are readily available to Bob's appliance company and his engineers can design to fit the curves. The algorithms for the AFCIs are not available and Bob's engineers cannot design to the them even if they wanted to.

That is frustrating because its our biggest issue with Arc Faults. We had been telling people there is a problem with their appliance (hair dryer, vacuum, etc.) but you look like an idiot when they plug in a brand new whatever and it trips the breaker.

 

[dstryr]

I just had a conversation in my office with a breaker manufacturer's product engineer who is right now at a home to test an installation where a newer LCD flatscreen TV is tripping the AFCI breaker. He has testing equipment to capture the data on the circuit to determine what the TV is doing. They will make a second stop later at another home to do similar testing.

Using their diagnostic tool a few weeks ago, the local sales engineer tested both the customer's TV and his own flatscreen, same basic model but different size. The customer's TV produced an arc signature while the sales engineer's TV did not.

The product engineer says that UL and FCC have standards for noise and arc signatures for consumer products but with manufacturing process variations some new devices will fall outside of the parameters set by UL for the AFCI breaker manufacturers' trip curves. Additionally, the accumulative noise on a circuit from multiple loads all within tolerances can cause a nuisance trip.

Defect rates are in the neighborhood of .001%, one in one thousand, based on testing of returned units, meaning nearly all of the time the returned breaker is in perfect condition. They continually collect data from loads to build a better breaker but UL's standards must still be met by the AFCI device.

 

[ELA]

I just had a conversation in my office with a breaker manufacturer's product engineer who is right now at a home to test an installation where a newer LCD flatscreen TV is tripping the AFCI breaker.

Using their diagnostic tool a few weeks ago, the local sales engineer tested both the customer's TV and his own flatscreen, same basic model but different size. The customer's TV produced an arc signature while the sales engineer's TV did not.

The product engineer says that UL and FCC have standards for noise and arc signatures for consumer products but with manufacturing process variations some new devices will fall outside of the parameters set by UL for the AFCI breaker manufacturers' trip curves.

.

Hello Dstryr,
If you talk with that engineer again please ask him where we can find this standard that contains the "arc signature" requirement for consumer products.
It is understood that both UL and the FCC have EMC (RFI) standards that equipment must meet but I have never heard of the "arc signature" requirement.

If we are ever going to have comsumer equipment and the AFCI devices play nicely together we surely do need such a standard.

It is my opinion that the AFCI manufacturer-designers did the best they could while keeping the cost per unit down. The current designs were forced on the public prematurely. To make the AFCI much more nuisance trip prone would require a much more complex and expensive design.

 

[don_resqcapt19]

...
Defect rates are in the neighborhood of .001%, one in one thousand, based on testing of returned units, meaning nearly all of the time the returned breaker is in perfect condition. They continually collect data from loads to build a better breaker but UL's standards must still be met by the AFCI device.

Just because the breaker itself is in perfect condition and meets all of the standards, does not mean that the device that is causing the trip is not in perfect condition. An AFCI that trips as a result of the normal operation an appliance that is in complete compliance with the standard that covers the appliance is, in my opinion, a defective AFCI even if the AFCI is in "perfect condition".

What they are trying to do is to create an algorithm that can tell the difference between all of normal "arc signatures" of all of the devices on the market and the "arc signature" that is really a problem. That task is almost impossible.

 

[ramsy]

the "arc signature" that is really a problem. That task is almost impossible.

I'll take a shot.

Using a TDR it is possible to; 1) pinpoint a fault to within centimeters, 2) differentiate higher-impedance appliance cords from building wiring, and 3) expect loads to be centered along appliance cords and near outlets mapped during TDR installation.

Most buildings require more than one AFCI circuit breaker, and a central TDR module makes AFCI breakers a low-cost relay. If it wont fit on the breaker, one place to employ a TDR chip is from a Smart Meter, but a wireless controller could address AFCI relays from anywhere.

 

[cschmid]

the afci debate continues..are they not wanting to require them in more places in the next code cycle?

 

[don_resqcapt19]

I'll take a shot.

Using a TDR it is possible to; 1) pinpoint a fault to within centimeters, 2) differentiate higher-impedance appliance cords from building wiring, and 3) expect loads to be centered along appliance cords and near outlets mapped during TDR installation.
...

It is not an issue of pin pointing a fault...it is an issue of seperating "fault arc signatures" from "arc signatures" that are a normal and safe part of the equipment operation.

 

[ramsy]

Separating "arc signatures" is not necessary if one can be excluded.

NEC AFCI CODE
210.12(B) "..15- and 20-ampere branch circuits supplying outlets .."

By definition, appliance cords, and perhaps fixture wires feeding loads, --do not supply outlets--, and can be excluded.

 

[don_resqcapt19]

Separating "arc signatures" is not necessary if one can be excluded.

NEC AFCI CODE
210.12(B) "..15- and 20-ampere branch circuits supplying outlets .."

By definition, appliance cords, and perhaps fixture wires feeding loads, --do not supply outlets--, and can be excluded.

No they can't. The only reason we have the "combination" type AFCIs is to provide protection beyond the fixed wiring of the dwelling unit. The code rule tells us that we have to use the combination type and the UL standard for them specifies protection beyond the outlet.

 

[ptonsparky]

My wife has been dealing with CHs BR style combination AFCIs since I insisted they be installed to replace the older non combination style. The local inspector suggested a temporary, and he meant temporary, installation of different mfg that would fit, then try the offending appliances again.

He also suggested at one of his code classes that it was our duty as electricians to test, record and forward all data to the mfg. :roll: I won't get that done at my own house let alone a customers that only wants his problem solved. I just will not use that mfg again.

 

[ramsy]

the UL standard for them specifies protection beyond the outlet.

Thanks for that clarification. My 2007 UL White Book does show AFCI's protecting "Cord Sets (Extension Cords)."

Then AFCI's using TDR's could ignore "arc signatures" from loads, only if centered along appliance cords, or fixture wire, or confirmed during setup.

 

[don_resqcapt19]

...
He also suggested at one of his code classes that it was our duty as electricians to test, record and forward all data to the mfg. :roll: I won't get that done at my own house let alone a customers that only wants his problem solved. I just will not use that mfg again.

That is exactly what the AFCI manufacturers want to happen. They forced a product that was not ready for the marketplace into the marketplace using the NEC so that the mandatory users become their product testers. Really cuts down on the R&D costs.

 

[ELA]

ramsy,

TDRs do not recognize "Arc Signatures". They estimate the distance to a discontinuity in the characteristic impedance of a transmission line.
They do not work on an energized line that I am aware of.
I do not see how you could use a TDR in place of an AFCI since they have to be monitoring "Real Time". (not all faults are continuous, some are intermittent).

TDRs are an expensive piece of equipment to be dedicated to monitoring each individual circuit in a home.

The point I was making in post 50 is that the current AFCI designs are way too primitive to distinguish between a real and potentially dangerous arc and normal loads with irregular and non linear current signatures.
If the manufacturers employed a much more sophisticated detection scheme than they now use they may be able to make them compatible with all existing loads. To do this would probably make the cost of each breaker -out of reach- to be required in each home on many circuits.