Sending Control Signals over Power Lines

[jimgineer]

A little while back I was trading emails with a PE who claimed to have come up with a way to effectively transmit data over 60Hz (or whatever frequency for that matter) power lines. I have kind of lost touch since then but wanted to get some feedback on thoughts as to whether or not this would be feasible and what kind of regulations or restrictions might be placed on doing so.

One of the big things I'm wondering about is power quality - how much 'divergence' (THD?) from a pure sine wave is one allotted and what standards and terms should I become familiar with to quantify this? Additionally, what is the relationship between THD on for example, the low side of a utility transformer (where control signals might be useful), and the high side of that transformer? Would the reatance of the transformer act to filter out fast rise times on the secondary? I'm picturing putting a square wave directly into the low side of a transformer (no 60Hz signal), would the high side of the transformer essentially be the signal having passed through a bandpass filter?

Additionally, I have run into applications where engineers decide that they need to run control signal wires separate from power wires, and it sounds like the major driving factor behind this is due to electromagnetic interference that is introduced in the signal cables when in close proximity to the power cables. Are there any additional restrictions or codes that govern this or is it simply good engineering practice, depending on the case..?

 

[alehman]

Power line communication has been around for a long time. Typically the signals do not make it through transformers so repeaters are used to bypass them.

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

 

[opmgr1]

Transformers were a problem back then. I was reading an article that indicated that Quadlogic has overcome this hurdle without repeaters (patented). These systems are in the states and a couple of south American countries and Mexico. You could check out their sites for additional info.

 

[itsmoked]

There were several proposals to run data of high voltage power transmission lines. It got pretty far along but radio astronermers went up in arms over it. I haven't heard the ultimate result.

My company did it for Sea Land Systems. We transmitted refrigerated container data over all the 460VAC power lines on ship and shore.

There are no THD issues with respect to data over power lines as you are talking about a small HF modulated signal on top of a very large noisy LF power signal.

Keith Cress
kcress -

http://www.flaminsystems.com

 

[jimgineer]

Wikipedia as usual gives some good general information. I found the bit about the FCC Chairman making BPL (Broadband over Power Lines) a high priority. After some more digging, it looks like the FCC got their hand slapped by a committee in congress, so it's not as high of a priority as the FCC originally stated.

I am wondering what might be causing the issues with interference with some of the bands dictated by the FCC. It seems as though a number of groups are up in arms over this and this is the big stumbling block right now for this technology. It also looks like it is suprisingly mature, in the sense that there are already a number of people on this board who have mentioned they've used something like it in their design or have come across it in work.

There's two design considerations I would like to learn more about concerning these technologies:

1. Cause for the creation of interference of other bands that groups are complaining about. There shouldn't be any 'new' EM waves created on a large scale that I would think - it sounds like the carrier is 20-200 kHz but it's modulated on a power cable, which (especially for HV/MV systems, I know they are'nt all this way especially as you get into the distribution level) there should be a bit of insulation that will mitigate these effects, I would think. And if it's just a matter of getting the right insulation in place (my Emag is a bit sketchy), then I don't see this as being a limiting factor at all. Of course from the sounds of the complaints I think I'm off base on this.

2. Data rate/ Throughput of these systems. I can't even really begin to comfortably talk about communication systems and use the right terminology, but I'm hoping to build some intuition for what kind of constraints there are on transmitting data (speed of transfer, bandwith, etc) when using a carrier wave on top of 60Hz wave, vs just transmitting the data directly over the line (using the line JUST for the data).

 

[Skogsgurra]

The EU has declared that all electric power meters shall be read remotely and that it shall be done latest june 31st 2009. There are several techniques used to do this and PLC (Power Line Communication) is one of them.

A popular technique is to use dual frequency signalling with alternate frequency pairs in the band below 150 kHz. Another one is the "Turtle" technology, which is in the 90-95 and 105-110 Hz bands. Turtle is very slow, a typical meter reading takes around 27 hours.

There are problems with both techniques. The <150 kHz is being disturbed by frequency inverters and, to get the signal through, many systems now operate above their allowed power levels as "temporary" solutions. That makes the total radiation from these systems a problem for some tele and radio activities.

The Turtle system used to be very reliable. But the use of frequency inverters to couple wind turbines to the grid has introduced interharminics that sweep the 90-95 Hz and 105-110 Hz bands continuously. My understanding is that Turtle - for obvious reasons - does not check and retransmit every character sent. Instead, the whole telegram is checked when transmission is complete (after 27 hours). Since it has been subject to interference many times during the 27 h period, it is mostly (always) so corrupt that is is scrapped.

I have been involved with several cases here in Sweden where about six thousand meters (seems to be a practical limit for one Turtle system) cannot be read because of this. Measurements always show either frequency components sweeping in and out the communication frequencies or, more common with the larger turbines, an elevated noise floor with fairly constant energy/Hz across the signalling frequency bands.

Gunnar Englund

http://www.gke.org

--------------------------------------
100 % recycled posting: Electrons, ideas, finger-tips have been used over and over again...

 

[itsmoked]

That's great Skogs..

Guess they'll turn off all your power on June 31. Stock up on candles.


jimgineer; The bandwidth of power line communications is absurdly low for broadband to a bunch of customers. One fiber will carry vastly more data with much less trouble OTShelf.

The right of way is always the issue. If a power company wants to get into the data business they should either lease right-of-way to someone who knows what they're doing, or run copperless fiber tower to tower, or buried and not involve their HV systems.

This is just what the railroads did, ran fiber next to the rails on their right-of-ways. They actually used locomotives to try to plow the trenches.

Furthermore, I believe there is a glut of bandwidth because the lots of fiber was run to cover projected needs, then we came up with wave division multiplexing which upped the bandwidth capacity of any fiber more than 10x.

So what's the purpose of your data?


Keith Cress
kcress -

http://www.flaminsystems.com

 

[jimgineer]

At this point it's mostly ideas bouncing around in my head - sometimes I get really excited about an idea, and most of the time, I learn later why either it won't work or that someone has already done it. So I'm keeping it tempered and not going spout it out. I'm hoping the one time it is a great idea, this will come in handy (because I know a lot of you out there are a lot better equipped to execute a great idea than I at this point).. but I'm not thinking of transmitting data on transmission lines. This doesn't have anything to do with smart metering or the like, but some background on when utilities find it useful to use some of these techniques is useful.

Can you give any more information on how exactly this 'turtle' system transmits data using 90-95 and 110-115Hz bands? Also, does a transformer effectively shield against transmission of control signal (essentially isolate), but still (obviously) transmit power to the other side? Or does it depend on the signalling method to be able to answer this question.

From a utility perspective, if they wanted to transmit data on a power line they would prefer not to have to buy repeaters to get the signal through the coils, to the other side. I guess I'm thinking about the reverse of this-I don't want to transmist any data to the other side of the transformer ---

 

[jraef]

I used this company on a project at Boeing where we did control of bridge cranes via the power bus bars, all of which had wiping contacts feeding VFDs on the carriers. Harmonics were terrible, but this system worked flawlessly.

http://www.data-linc.com/powerlcm/lcmapp.htm

"If I had eight hours to chop down a tree, I'd spend six sharpening my axe." -- Abraham Lincoln
For the best use of Eng-Tips, please click here -> faq731-376

 

[waross]

Transformers have inductance. This leads to inductive reactance which is proportional to the frequency. Although a transformer is designed for a low reactive impedance at 50Hz or 60 Hz, at high frequencies, the inductive reactance is so high that so little RF power passes through the transformer as to be unusable.

Bill
--------------------
"Why not the best?"
Jimmy Carter

 

[davidbeach]

Transformers have inductance. This leads to inductive reactance which is proportional to the frequency. Although a transformer is designed for a low reactive impedance at 50Hz or 60 Hz, at high frequencies, the inductive reactance is so high that so little RF power passes through the transformer as to be unusable.

Very true, when you think about the transformer in terms of the power system frequency. There is also an appreciable amount of capacitance between the windings that can usually be ignored at the power system frequency but can have a very low impedance at high frequencies. At the frequencies associated with close in lightning strikes the transformer can capacitivly couple most of the transient between windings without regard to the inductance.

 

[waross]

Thanks David.
Is there also energy transfer from the effect of the transformer acting as an air core transformer?

Bill
--------------------
"Why not the best?"
Jimmy Carter

 

[davidbeach]

I would suspect there would be, to the extent that there is any energy in the high frequency signal.

 

[jimgineer]

Help me understand here: When we're talking about inductance, (or capacitance for that matter), in terms of the frequency response we are effectively talking about a filter. It then makes sense that the higher frequency carrier which has the actual data on it won't make it through. But if there's capacitance as well as inductance, the capacitance would effectively cancel out the inductance (if in parallel - well actually, either way right?), unless there is much more inductance than capacitance.

I've always understood pf correction you to place the capacitance in parallel with the line. I know this is not fully related to what we were talking about before, but still.

 

[alehman]

At high frequencies simple lumped impedance models for transformers are less accurate. The windings must be modeled with distributed inductance and capacitance elements. Transformers typically have numerous resonances. In general, inductance elements are greater magnitude than capacitance. Also iron cores are very lossy at high frequencies and may dissipate much of the high frequency power.