Miscellaneous notes on the
Western Digital "Livewire" ethernet modem


Additional notes:

It should be pointed out that devices such as these are covered under FCC Part 15.  If one examines the literature accompanying devices such as these that have the potential to cause interference to other radio/wireless devices - or even your TV, DVR or computer - you'll find a notice included that is worded something like this:
This device complies with part 15 of the FCC Rules. Operation is subject to the following two conditions:
(1) This device may not cause harmful interference, and
(2) this device must accept any interference received, including interference that may cause undesired operation.

(See 47CFR - FCC Rules, part 15)

What this means is that anyone who operates such a device, by using it, has agreed to be responsible when it comes to any interference that the device under question causes - that is, if someone complains that said device causes interference to them, it is incumbent on the owner of that device to mitigate that interference - even if it means that you have to stop using it!  The second part means that if radio operation causes the device to not work properly - such as when a Ham operator transmits - that the user of that device has no legal recourse when it comes to stopping such operation as this agreement also stipulates that they are using it at their own risk.

The simple fact is that many devices are just susceptible to being interfered with by RF signals:  If you have a GSM telephone (like an IPhone (tm), Blackberry (tm), etc.) you may have noticed that if you are using it you'll sometimes hear your stereo, TV or computer speakers buzz when you receive or send a call, text, data, or browse the web - and this buzzing is a result of circuits in those devices unintentionally picking up radio frequency (wireless) signals.

While it's true that devices could be designed to resist such interference (stronger rules in some countries required a degree of resistance to this effect - but not necessarily so in the U.S.) doing so would add a few cents or dollars to the device, so it isn't done.

Therefore, someone using a device such as this is obliged to not interfere with other users of the spectrum - even if the device is operating as designed!

Purpose of this page:

Recently, one of these modems - Western Digital "Livewire" - crossed my path:  An acquaintance of mine bought one of them and a co-worker and I were curious about their mode of operation, so, during a bit of "off" time at work, we placed it on the bench, coupled it to a spectrum analyzer and looked at its RF utilization.

A little about "BPL" - or "Broadband over powerline" communications:

One of the reasons why we were curious has to do with the recent kerfuffle about "BPL" - that is "Broadband Over Powerline" communications.   The proposal was that, using existing wires that ubiquitously connect our lives - namely powerlines - RF signals could be conducted across those same lines that would allow broadband data connectivity with a relatively minimum of outlay and added infrastructure.  To this end, a number of pilot projects and some full-time services were implemented over the years to demonstrate and bring to fruition (possible) commercial applications of the technology.

The proponents of the system had stipulated that the conductors of power were adequately efficient for this task and that minimal radiation of the RF (Radio Frequency) signals would occur from these powerlines.  This last point is an important one as that property is necessary for such a system to work effectively and the conducting wires not radiate signals into the air (as an antenna might) and cause interference to other services - or itself by cross-coupling of one "branch" of BPL service/customers into another.

For the most part the proponents are right:  Very little of the signal is actually radiated from the lines along any given segment - but how much is too much, and if a little bit of power is radiated from each segment of powerline, how much is radiated over the entire run when you add it all up?

Since these same powerline literally connect us together, they go everywhere and there is the real possibility that such signals conducting the data will interfere with existing services:  After all, there's no real way to "route" the RF energy - except by virtue of the where the powerlines conducting this RF energy just happen to run unless, of course, some sort of filtering was installed on the powerlines themselves.  This latter point is an unlikely prospect as suitable filters, to be effective, would likely be quite large and expensive were they to be placed on feeder lines and their installation would require an interruption in power to customers being served on those lines:  Their use by power companies or their internet partners seemed unlikely!

In addition to these RF signals being conducted, there was the possibility that the overhead power wires on which these RF data signals would be imposed would, in fact, act like antennas.  Particularly concerned the possibility of this happening were Amateur Radio Operators ("Hams") and shortwave listeners who use portions of the "HF" radio spectrum from 1.8-54 MHz (among other places) - the same general frequency range expected to be used by BPL operations.  Also concerned were broadcasters - AM, FM and TV - who worried that customers' signals may be degraded by the presence of these "other" signals.
Figure 1:  The modem and its testbed.
Top:  The modem itself
Center:  The modem, signal coupler, and spectrum analyzer used.
Bottom:  The makeshift "Current Transformer" used to couple the RF from the power cord.
Click on an image for a larger version.
Western Digital "Livewire" modem
Setup used to check the spectrum of the modem
"Current Transformer" used to couple signals from the powerline

This concern sparked a debate among the amateur radio groups as well as some concern among government and commercial users of the HF radio spectrum that frequency ranges over which BPL operated would essentially become unusable.  On the side of the proponents, statements made by the FCC indicated great optimism that the technology was feasible and would not result in significant disruption to users of the "shared" spectrum.

Amateur radio operators, on the other hand, took it upon themselves to review the literature and make their own in-field measurements, dragging radio gear and test equipment into communities in which BPL systems had been installed.  To the dismay of the amateur radio operators, their worst fears were justified when it was observed that moderate to extreme interference was observed and that some of the recommendations made by the technical arm of the FCC were seemingly dismissed in favor of moving forward with deployment.

Here are a number of videos related to the "Hams-versus-BPL" debate:
As can be seen from the above videos, there is some understandable concern that systems that use standard wiring - either power distribution or in-home wiring - could cause interference to shortwave (and other) communications.

Again, being an Amateur Radio operator, I tend to be concerned about the use of such systems that have the strong potential toward causing interference to existing services - even when used in the manner in which they were intended to be operated!

The difference between "BPL" and the type of device discussed here:

At this point it is worth mentioning that there is a difference between BPL and the devices intended for providing in-home connectivity:
What these devices share in common are the fact that they both use mains to conduct the radio signals used to carry the data and their potential to cause interference to users of the HF spectrum.

The Western Digital "Livewire" device.

Refer to the pictures on the right, in Figure 1.

This is a consumer device that is intended to be used within the home to provide connectivity by using the in-house wiring to conduct RF signals amongst computers.  Based on what was already known about the potential for such devices to cause interference, we decided to put it on a spectrum analyzer to see what it's spectrum looked like.

The test setup:

The test setup was very simple.  We had, on hand, a 3-way power tap (the orange thing) that had been modified to allow the use of a clamp-on ammeter to measure current consumed by a device plugged into it.  These modifications were very simple:  Simply cut away the outer jacket of the power tap and expose the three individual wires (each being independently insulated) so that current in each of the three leads - Hot, Neutral and Ground (Black, White and Green, respectively) could be measured.  Due to the very nature of current flow and magnetics, simply clamping a meter on the entire cord and surrounding all three conductors would not yield usable results as the magnetic fields would cancel out - hence the requirement that each wire be individually accessible.

To couple the HF (High Frequency) energy transmitted by the modem into the power line several turns of wire were wound around the black (Hot) conductor to act as a very simple current transformer - with the wire passing through it acting as the primary.  The two ends of this wrapped wire were terminated in a 15 ohm resistor (this low value being used to minimize the effects of stray inductances on the pickup) and then this was conducted, via coaxial cable, to the spectrum analyzer.  In this way the RF signals could be picked up while detecting a minimum of other signals - except, of course, those that might also be on the powerline itself.

Figure 2 (below) shows the initial plots of the occupied spectrum of the modem.

Figure 2:  Wide-bandwidth plots of the spectrum occupied by this modem.
Top Left:  0-30 MHz plot.  Top Right:  0-10 MHz plot  Bottom Left:  10-20 MHz plot  Bottom Right:  20-30 MHz plot.
Click on an image for a larger version.
0 to 30 MHz plot of the modem's output
0-10 MHz plot of the modem
10-20 MHz plot of the modem
20-30 MHz plot of the modem

Initial analysis:

We were both surprised and pleased to see that there were many "holes" in the spectrum occupied by this devices.  Closer analysis indicated that these holes aligned nicely with the HF amateur bands and the AM Broadcast band.  Except for these holes, this modem produced spectral energy over the frequency range from just above 2 MHz to just below 28 MHz.


Closer analysis:

Based on these results we decided to "zoom in" using the analyzer and see, exactly, how wide and deep these "notches" were - an indication as to how effective they may be in preventing interference to amateur, HF communications.  To do this, individual plots were taken of each of the HF amateur bands.

It should be pointed out that these "notches" aren't due to filtering within the device, but just that the device has been programmed NOT to use certain frequency ranges.

Figure 3:  Plots centered on the 160 and 80 meter amateur bands
160 meter centered plot
80 meter centered plot

160 and 80 meters:

Figure 3 shows a spectrum +/-500 kHz (a total of 1 MHz span) about the center of the 160 and 80 meter amateur bands.

As can be seen from the plots, the energy is about 28dB down at the edges of each amateur band - a significant reduction.  Into the AM broadcast band, the signals drop off dramatically:  The slight "peaks" are due to leakage of nearby AM stations into the power cord of the test setup.

Figure 4:  Plots of the 60 and 40 meter amateur bands
60 meter plot
40 meter centered plot

60 and 40 meters:

Figure 4 shows 1 MHz wide spectrum plots of the "60" meter band and centered on the 40 meter band.

We were surprised to see a "notch" in the general area of the "60 meter" amateur band.  At the time of writing, this isn't an "amateur band" per-se, but a number of individual channels on which restricted amateur radio operations are permitted in the 5.2-5.4 MHz area:  The exact frequencies of operate vary depending on the country and are subject to change.

The 40 meter band is also "notched" out to the tune of 30-33dB at the band edges.

Figure 5:  Plots centered on the 30 and 20 meter amateur bands
30 meter centered plot
20 meter centered plot

30 and 20 meters:

Figure 5 shows 1 MHz wide spectrum plots of the 30 and 20 meter bands.

The measured "notch depth" on 30 meters - a rather narrow (50 kHz wide) band just above 10 MHz was not as deep, but still over 25dB.  At the edges of the 20 meter band notch depth was about 27dB, but well over 30dB in the center.

Figure 6:  Plots centered on the 17 and 15 meter amateur bands
17 meter centered plot
15 meter centered plot

17 and 15 meters:

Figure 6 shows 1 MHz wide spectrum plots of the 17 and 15 meter bands.

17 meters - like 30 meters - is a rather narrow band and the depth of the "notches" at the edges weren't as deep - but still over 25dB.  As with 20 meters, the notches at the edges of 15 meters were about 27dB deep and well over 30 in the center of the bands.

Figure 7:  Plots showing the 12 and 10 meter amateur bands
12 meter plot
10 meter plot

12 and 10 meters:

Figure 7 shows a 1 MHz wide spectrum plots of the 12 and another plot spanning from 27 to 30 MHz.

The results on 12 meters were very similar to those on the other two WARC bands (e.g. 30 and 17 meters) with slightly lower "notch" depth owing to the relative narrowness of the band - but still over 27dB within the band itself.

The final plot shows a 3 MHz wide swath from 27 to 30 MHz, encompassing much of the CB radio frequencies and all of the 10 meter amateur band.  At the very bottom edge of the 10 meter amateur band the depth of the "notch" is only about 25dB, but it approaches 40dB by the time one reaches the upper edge.

Above 30 MHz there is very little energy:  It would appear that the highest frequency being used would be just below 28 MHz and by the time one gets above 50 MHz or so, residual energy from this device was unmeasurable with our simple setup.


Additional notes:

Any comments or questions?  Send an email - but please remember that I really can't answer any questions about if these will cause interference as I don't actually own these devices, nor have I used them.

Return to the main KA7OEI page.

This page maintained by Clint Turner, KA7OEI and was last updated on 20110124.  (Copyright 2010-2011 by Clint Turner)

Since 12/2010: