Print

Please login or register.
1 Hour 1 Day 1 Week 1 Month Forever
Login with username, password and session length

[Geewiz]

I have a Signaliinc Coupler/Repeater plugged in to my electric dryer plug up on the second floor of the house. 

My computer, located in the basement, is plugged into an X10 filter, which is plugged into an APC 500 UPS, which then plugs into the wall plug. 

When my computer is on (which is basically 24/7) the LED receive light on the Signalinc repeater is glowing perpetually.  Turn the computer off, and the receive light goes off. 

Being a neophyte in the X10 game, I'm thinking it's safe to assume my computer is pumping out a lot of line noise.  I've experimented with placing the line filter between the wall outlet and the UPC supply, and between the computer and UPC supply with the same effect. 

Short of turning the computer off, which is not an option, how can I eliminate this noise in light of the fact that I already have a filter?

Thanks for any ideas!

[Brian H]

Wow. The APC also has a filter in its AC Input; so with the filter on its AC cord you have two filters in-line. I have my APC BX1000s AC cord on a 10 amp X10 filter myself.
Are there any RF receiving Transceivers near it? I doubt it would be a problem; but if you have transceivers; maybe unplugging them as a test may show something.
In another thread I believe you said you have a CM15A try unplugging it.
I have only seen one computer power supply so dirty when on it killed all X10 signals and wiped out local radio stations also.

[Geewiz]

Are there any RF receiving Transceivers near it?
In another thread I believe you said you have a CM15A try unplugging it.

I do have a DS7000, but I powered that down and it had no effect. 

I just finished unplugging the CM15A and, again, no effect on the noise signals reflected on the Signalinc booster/coupler. 

Seems it's just a dirty power supply?  I am surprised that the filter(s) don't seem to help.  Anyone have any ideas as to how to approach this, short of trying to replace the computer power supply?

Thanks again!

[JeffVolp]

When using an UPS with a computer, the correct placement for the filter is between the UPS and the wall socket.

I have several APC UPS's here.  They are excellent units, and contain filters to prevent conducted radiation from reaching the powerline.  Unfortunately, that filter also attenuates any X10 signals being sent over the powerline.  Installing a X10 filter between the powerline and the UPS will isolate the unit.

After re-reading the original post, this may be a more complex issue than it first appears.  To meet FCC specs on conducted radiation, a UPS is normally a "signal sucker", not a noise source.  The same is true of most computer power supplies.  Every one of my computer power supplies has a capacitor directly across the power input.

X10 filters are tuned to eliminate noise in the X10 frequency band.  It is possible that switching transients from the UPS are still leaking onto the powerline, and causing the SignaLinc LED to glow.  The X10 filter should prevent that noise from corrupting X10 communication.  However, it is possible that the SignaLinc operation can be influenced by noise not inside the X10 frequency band.

The obvious question is when the UPS is isolated from the powerline with a filter, does your X10 system work differently when the computer is on or off. 

Jeff

[Geewiz]

The obvious question is when the UPS is isolated from the powerline with a filter, does your X10 system work differently when the computer is on or off. 

Jeff

The answer here is: Yes.  X10 works differently when the computer is on from when it is off.

I placed the filter back between the powerline and the UPS leaving the chain plugged-in but with the computer off.  Checking the Signalinc LED's, I find them all dark.  No sporadic, unexplained lighting.  Checking 4 lamp modules throughout the house, all on the same unit code, with 2 each on each phase of the house circuit, I find good, solid on/off reception to the 4 modules.

Now, turning the computer on, I find the Signalinc LED's glowing red with a little flickering.  Checking back the 4 lamp modules, the two that are on the same phase as computer will not turn on/off. 

So, as you proposed, is what I have a signal sucking problem?  If that's the case, where to now...

Thanks!

[JeffVolp]

So, as you proposed, is what I have a signal sucking problem? If that's the case, where to now...

With the filter isolating the powerline from the UPS, it should not be a signal sucking problem.  The X10 filters do a good job of isolating devices in the X10 frequency band.  Also, since the line filters in both the UPS and computer are in the circuit whether the computer is on or off, their "suckability" would not change when the computer is switched on.

This may be more an issue for the SignaLinc booster/coupler being confused by out of band noise.  Assuming you are using a X10 filter now, I'm not sure what to do other than adding a broadband filter.

It would be interesting to see if removing the UPS from the circuit (run the computer directly from the powerline through the X10 filter) eliminates the problem.  When the computer is on, is there any chance that it might be sending X10 traffic over the powerline - perhaps via a CM11A or CM15A?  Also, which X10 filter are you using?

Jeff

[Geewiz]

It would be interesting to see if removing the UPS from the circuit (run the computer directly from the powerline through the X10 filter) eliminates the problem.  When the computer is on, is there any chance that it might be sending X10 traffic over the powerline - perhaps via a CM11A or CM15A?  Also, which X10 filter are you using?

Jeff

Well, the saga continues...
I removed the UPS from the chain, and plugged the computer directly into the powerline, using the basic X10 5amp wallwart filter.  Before doing that, I took a snapshot of the X10 Code Usage chart under the "Find other computers" menu tool.  Then took similar snapshots showing various results.

I'm not fully up to speed on uploading/posting pics,so I will do that separately in subsequent postings...

Happy New Year!

[Geewiz]

X10 Code Usage

#1: Computer plugged in to APC UPS-500, with X10 filter between UPS and powerline...

[Geewiz]

X10 Code Usage

#2 After turning off Computer...everything left plugged in...

[Geewiz]

X10 Code Usage

#3 Computer plugged back in...this time, directly to powerline with X10 filter.  This snapshot taken about 15 minutes after turning the computer back on...

[Geewiz]

X10 Code usage

#4: Same as previous, but taken at about 30 minutes after turning the computer back on...

[Geewiz]

My conclusion:  Houston...I have a problem

I really don't want to keep my computer turned off - that's really not an option!  Is there a super-duper, heavy-duty filter that I can get to isloate the computer from the powerline?

[JeffVolp]

My conclusion: Houston...I have a problem

I really don't want to keep my computer turned off - that's really not an option! Is there a super-duper, heavy-duty filter that I can get to isolate the computer from the powerline?

I have to agree with your assessment.  So, it is clear that something coming from your computer is causing the problem.  A single X10 filter between the computer and the powerline should block any X10 communication.  There are industrial wide-band filters that block virtually all conducted noise.  Electronic supply houses such as Mouser sell them, but they are not convenient "plug-in" units like the X10 filters.

Before we go there, the one issue that has not been investigated is your SignaLinc.  That may be picking up weak noise radiated by your computer, and re-transmitting it.  It would be very interesting to repeat that test with the SignaLinc out of the circuit.  In the mean time I'll do some research on other filters.

Jeff

[JeffVolp]

Is there a super-duper, heavy-duty filter that I can get to isolate the computer from the powerline?

After picking through filter catalogs and running some simulations, one that stands out above the rest is the Schurter FSS2 2-stage filter part number 5500.2056 available from Mouser for about $50.  This is not a plug-in unit, and it would have to be mounted inside some type of enclosure.

The advantage of this particular 10A filter is that its "Load" side has no capacitor across the input pins.  When connected backwards (Load to Powerline, and Line to UPS/Computer), it would not attenuate X10 signals on the powerline.  It has a very low cutoff on its "Line" side.  When connected backwards, that should prevent virtually anything from reaching the powerline.  It also has high-frequency bypassing to ground at each filter stage.  When mounted inside a grounded metal chassis, this should be a very effective "super-duper, heavy-duty filter".

But, before taking this step, I think some further investigation into the cause of your problem is warrented.  One thing that pops to mind is to insure that your computer case is effectively connected to "earth ground" through the power cord's third prong.  And run that test with the SignaLinc removed to see if that might be a factor.

Jeff

[Geewiz]

In desparation, I've been trying multiple variations, and I seem to have created some improvement.  The downside is, I've tried so many variations, I'm not sure what single modification cause the improvement, or if it was a combination of things.  Here's what I did...

1. Change the power cord from the computer box to the UPS-500
2. Change the UPS-500 unit to another same make/model that I have here.
3. Plug the unit, via a short 3' heavy duty shielded extension cord, into a different electric outlet that is on a different circuit.
4. Take spare side panels from two old computer boxes I've been holding on to (I knew there would be a time I would need them!) and place one on the side of my computer and another in front of the electric outlet nearest the computer.  My rationale here may be totally off base but I'll try and explain...you tell me if I'm wacko   When I built my current computer, I went for a large box that could handle 4 hard drives.  Only one I could find that didn't cost a fortune was from Tiger Direct.  It's mostly made of aluminum and has a clear plastic side panel - not metal.  This, I'm thinking, might not contain RF frequencies?  So, I'm thinking, placing a real metal panel on the side might offer some type of shielding?  I'm really guessing here.  Any validity to this?

I've attached a screen shot of the X10 Code Usage tool after about 1 hour of running time.  Quite a difference!  I'm running my modules and CM15A on Housecode I, and running my DS7000 on Housecode M.  Based on this Usage log, it's looking like Housecodes C through G are running particularly "clean."  I'll keep this current config running for a good 24 hours before investing any more time in changing codes, though. 

Will be interested in hearing your observations, and thanks so much for the time you've invested in helping out with this.  You got me thinking outside the box!