DIY Noise Filter

[johnnyt]

Hi,

 

I was browsing for an inline noise filter at smarthome and found the ZNF10A. The manual (found here http://www.smarthome.com/manuals/22956.pdf) suggests to me I can make my own noise filter just by getting a transformer and capacitor then wiring it according to the diagram in the ZNF10A manual.

 

Is that right? Since the parts in the ZNF10A are only rated for 10A (even though the instruction for the wire-in model refers to a 15A inductor - ??), what do I need to ask my local electronics store for if I want to support a 15A circuit with these parts?

 

Is there something in the 15A AFT120, or for that matter the 10A Filterlinc, that makes either a better filter than using the parts mentioned in the ZNF10A?

 

Any info would be appreciated.

[Brian H]

Yes they are specifically tuned to isolate the X10 120 KHz and Insteon 131,65 KHz power line signals.

The one in your example is for UPB and I don't know what it actually is tuned to.

You would have to select the proper part values for it to work correctly.

 

There are a few filter schematics on this web site. Including the ACT AF120.

http://www.idobartana.com/hakb/index.htm

 

FilterLinc: http://home.comcast.net/~gafield/family ... Filter.pdf

[johnnyt]

The ZNF10A says it blocks everything above 4 kHz, and, according to the manual, "should be compatible with all power line technologies".

 

I'm no expert but the schematic for the Leviton 5A 6287, the AFT120 and the Filterlinc seem to show 2 transformers and 2 or 3 capacitors. Does that make those better than this one transformer / one capacitor solution and perhaps justify paying more for them?

 

On a related note, if the filterlinc is a better noise filter, can I simply replace the 10A fuse in it for a 15A fuse? I thought I read that you could do that with the older 5A version to turn it into a 10A version but not sure if the jump to 15A was / is too much with either the older (5A) one or a current (10A) one. I ask because I do have a 5A version with a blown fuse but haven't gone out to get a new fuse yet.

[Brian H]

The AF120 and FilterLinc use the added coils and capacitors to tune them to a frequency band. Anything above and below the frequency band will pass through.

The ZNF10A blocks all frequencies above 4KHz and as shown. Should stop any noise and signal absorbers from effecting power line signals.

 

One important thing the schematic of the ZFN10A does not provide is the value of the inductor. Without knowing its value. Finding the correct inductive value would have to be calculated.

Also the 15 uf capacitor. Has to be one that is rated to be used across an AC Power Line safely. This type of capacitor is not cheap. Most capacitors are not made for connection directly across the power lines and may fail in a flame out.

 

I have looked at the 5 amp and 10 amp Filterlincs. The parts and PCB look the same and changing the 5 amp fuse to a 10 amp one should be OK. Going to 15 amps may stress the components and PCB runs.

The fuse is a 5mm X 20mm fuse with wire leads, ceramic body. I believe they are normal acting and not time delayed. From what I was able to see on the fuse. F5H250 in the 5 amp. F10H250 in the 10 amp. Soldered to the PCB.

[IndyMike]

Hello JohnnyT,

 

The ZNF10A is designed to filter/isolate low frequency signals in the UPB range (4k - 40KHz and above). For Insteon signals (130KHz) It's rather like hitting a finishing nail with a sledgehammer (10lb). It will absolutely Isolate/filter problem devices from the powerline. Assuming that this device uses a 500 uH inductor, it will also draw measurable current - on the order of 81 Volt-Amps or 0.04W of real (utility chargeable) in phase power.

 

Other than the low cost (on SH), I would not consider the above for an Insteon installation.

 

Beyond that, it would be best to understand what type of device you are trying to isolate/filter. The filters that Brian referenced have different goals:

 

1) The Leviton 6287/Filterlinc are "notch" filters. As Brian described, the units filter/isolate (extremely well) a narrow frequency range around 130 KHz. This is the primary frequency of concern for Insteon.

2) The Leviton 6288 filter is a "Low pass" filter. It will attenuate noise sources will below the Insteon range ( 60 KHz if memory serves) but will not isolate/filter at 130KHz as well as #1 above.

3) A "side benefit" of the 6287/Filterlinc is that they use large capacitors that provide "powerline loading" below the Insteon frequency range. At 60 KHz a would provide a 5 Ohm load whereas a 6288 would provide a 51 Ohm load. In short, the 6287 would load noise from "nearby" devices than the 6288.

 

In short, we need to understand if you are trying to "filter" or "isolate" and what type(s) of devices you are dealing with...

 

IM

 

Edit: Corrected choke value (500uH) VA and wattage. At 60 hz the load appears almost completely capacitive and the power is therefore not chargeable.

[ELA]

Hello IndyMike,

I agree that the ZNF10A is probably not the best choice filter for Insteon.

Your assumptions about the inductor size caught my eye. Is that Milli or Micro? Surely the device would never dissipate 40 watts.

 

I found this post for reference:

http://cocoontech.com/forums/topic/21991-upb-filter-parts-lc-values/

[IndyMike]

Hello IndyMike,

I agree that the ZNF10A is probably not the best choice filter for Insteon.

Your assumptions about the inductor size caught my eye. Is that Milli or Micro? Surely the device would never dissipate 40 watts.

 

I found this post for reference:

http://cocoontech.com/forums/topic/21991-upb-filter-parts-lc-values/

 

ELA,

 

The inductor size was intended to be uH (typo). The V-A numbers and dissipation were totally hosed. Too many mistakes and typo's to shake a stick at... They're corrected now.

 

Thank you for the link to the cocoontech post. The 220uH inductor quoted won't provide much isolation at 4Khz. I guess it's a good thing that UPB starts with a 40V level.

[johnnyt]

Thanks for the info, guys.

 

I want to isolate an entire lighting circuit with about 10 CFL bulbs and 1 fluorescent fixture (basement rec room). For this circuit I have an insteon relay right at the panel to turn it on/off. I use the wall switches (which don't have neutrals anyway) to tweak the lighting manually when needed (rarely).

 

As I chip away when time permits at improving intermittent comm problems in some areas of my house, I thought this whole circuit of CFL bulbs plus a cheap electronic ballast for the fluorescent fixture would be a good thing to isolate. I wanted something I could put in the same junction box as my insteon relay (just after it, of course.)

 

Is my best bet to take apart a 15A AFT120 and wire the board inside into my junction box? I've never looked inside an AFT120 so I'm not sure that's possible/practical but it would support the full 15A of my circuit. Unless I change to a 10A circuit breaker and use the filterlinc I have (with the blown 5A fuse) and "short" the fuse ends together - counting on the circuit breaker to protect it...?

 

By the way, can someone confirm an assumption I've been making for a couple of years now that CFL bulbs and fluorescent ballast are NOT a problem when they are OFF; they are only a problem when they are on, and not all of them are a problem even then. Is that right? My comm problems cannot be linked solely/directly to the basement rec room lights being ON so I'm not thinking this is "the" fix for my problems, just maybe a contributing factor, and only when they're ON.

 

Thanks again.

[Brian H]

When they are Off. They probably will not be a problem. As they are not connected to the power through their On Off Switch.

Not all CFLs and electronic ballasts are problems, but high on the suspected offenders list.

 

Not sure if this avilable any more. As X10/X10Pro has closed their factory in China and are having some core devices made by outside factories.

The X10Pro XPF is a 20 amp wire in filter. Though it is quite large in physical size.

 

http://www.x10pro.com/pro/pdf/xpf.pdf

[ELA]

The XPF filter is a nice heavy duty choice. They do appear to be available out there on the net.

 

I had bought one for my refrigerator but elected to use a dual band device to "get around" the dead spot behind the refrigerator's signal sucker tendencies instead.

 

One caution. I put an ohm meter on my XPF prior to looking at installing it. And it measured an intermittent open circuit. Took it apart and the solder joints were cold. Would not have wanted to install it behind the refer and learned that the hard way.

 

Also note that this filter is very large and would require a large box to install into.

 

All of the CFL's I have tested were signal suckers to some extent. Some worse than others. There may be some better brands that incorporate power factor correction that are not signal suckers.