Mystery

[gar]

211222-2111 EST

With my scope on the bench I am getting an interfering signal.

The carrier frequency of the damped oscillations is about 20 MHz ( 50 nanoseconds, and some + or - a small amount ). It seems that part of the signal frequency modulation is factor in the signal. Then there is also modulation of the amplitude of the carrier. Further there is modulation of the rep rate of these pulses.

Light travels about 982 ft/microsec.

I doubt this signal is propagated on the power line. The power line might be an antenna.

When I tune my shortwave receiver to around 20 MHz I get several spurious signals ( possibly generated in the receiver ) but nothing that might correlate with what I see on the scope.

A mystery to bug me for a while.

.

 

[Fred B]

Get Scully and Mulder on the case. Guys looking for ET would hop on a spurious signal.

More seriously, if the powerline is acting as an antenna, could you be getting some readings caused by recent solar flares?

 

[gar]

211223-0927 EST

Fred B;

Not ET

The damped oscillation has a carrier about 20 MHz ( 50 nano sec period ), damps fairly quickly. The pulse rate is about 49.75 kH. Then this pulsing lasts for about 10.6 mS ( 0.0106 S ), and for about 1 mS no output. Thus. the period of the pulses plus off time is about 11.6 millisec, or a frequency of about 86 Hz. Somewhat near freq of submarine communication. But 80 Hz or so is their carrier frequency. This is not likely related to that.

So what is the purpose of this signal, and why the structure?

.

 

[Hv&Lv]

Could you post a picture of the Oscope display?

 

[gar]

211223-1214 EST

Hv&Lv:

No, I ran out of space several years ago.

.

 

[synchro]

The damped oscillation has a carrier about 20 MHz ( 50 nano sec period ), damps fairly quickly. The pulse rate is about 49.75 kHz.

Perhaps this 20 MHz damped oscillation is the ringing of a resonant circuit (e.g., a length of wire, transmission line, etc.) that's being driven by the energy from a switching supply operating near 50 kHz. For example, the driver of an LED light could be switching at a ~50 kHz frequency.
About how many cycles are there during the 20 MHz oscillation interval before it is essentially gone? If it's a lot (like more than 30 or so) than that would require a resonant element with a relatively high "Q" (quality factor), which is less likely to be present.

... The pulse rate is about 49.75 kHz. Then this pulsing lasts for about 10.6 mS ( 0.0106 S ), and for about 1 mS no output. Thus. the period of the pulses plus off time is about 11.6 millisec, or a frequency of about 86 Hz.

That 1 msec drop-out interval would be harder to explain. However, the 86 Hz frequency at which the drop-out occurs might be tolerable for dimming purposes. If so, further dimming could be applied using more frequent dropouts with the same 1 msec drop-out interval (up to a point). But I don't know if such pulse-density modulation is used for dimming light fixtures. Of course pulse-width modulation is commonly used.

 

[gar]

211223-1535 EST

Synchro:

The about 20 MHz oscillation seems to have three discrete frequencies. The modulation shape of the oscillation changes. Does not always look damped. I see this signal when virtually anything in my home is off that could be a source. My home and my next door neighbor are the only ones on a 50 kVA transformer that is fed from a delta source. Thus, only common with other neighbors is via water line.

Some other periodicities 4.5 mS, 1. 6 mS, 21 microS, and about 42 microS. The in-between pulse at 21 microS is lower in amplitude than the 42 microS pulses.

There are other signals at various times that could be echoes. Whether what i am seeing is power line related, or out in the sky stuff.

Direct short at scope input eliminates the signal.

Is there anyone else out there with a capable scope that sees anything like what i see?

.

 

[gar]

211223-1635 EST

You do not get a signal like what I am describing from random noise. This is a designed signal, but for what purpose I have no idea. Randomly you just can not have such a signal created.

My scope has a large amount of memory. Thus, I can take a rather long sample and go back and see a lot of fine detail.

.

 

[Hv&Lv]

211223-1635 EST

You do not get a signal like what I am describing from random noise. This is a designed signal, but for what purpose I have no idea. Randomly you just can not have such a signal created.

My scope has a large amount of memory. Thus, I can take a rather long sample and go back and see a lot of fine detail.

.

Does your POCO use PLC communications for their smart metering?

 

[hbiss]

Does your POCO use PLC communications for their smart metering?

That would be my question also. Even if not, power line communication has been widely used for a very long time. Consider home control systems.

-Hal

 

[Hv&Lv]

Here is some light reading for tonight…

https://users.ece.utexas.edu/~bevans/papers/2012/powerline/powerlineCommNoiseIEEESPMDraft.pdf

 

[synchro]

... The damped oscillation has a carrier about 20 MHz ( 50 nano sec period ), damps fairly quickly. The pulse rate is about 49.75 kH. Then this pulsing lasts for about 10.6 mS ( 0.0106 S ), and for about 1 mS no output. Thus. the period of the pulses plus off time is about 11.6 millisec, or a frequency of about 86 Hz.

So what is the purpose of this signal, and why the structure?

The about 20 MHz oscillation seems to have three discrete frequencies. The modulation shape of the oscillation changes. Does not always look damped.
...
Some other periodicities 4.5 mS, 1. 6 mS, 21 microS, and about 42 microS. The in-between pulse at 21 microS is lower in amplitude than the 42 microS pulses..

It sounds like the 10.6 mS interval could be a "frame" of power line carrier data. Does the very beginning of this 10.6 mS interval always look the same? Often a fixed "preamble" is used to identify the beginning of a data frame and to properly synchronize the data detection process.

 

[Hv&Lv]

See if it looks similar to this…

In the comments section below:

The TWACS system is a command-slave system. Commands (called Outbound) are sent from the substation to all meters downstream from the sub. The module in the meter (called a transponder) receives those commands and, if it determines that a response is needed, sends a reply called inbound. Outbound messages are sent by perturbing the voltage waveform as seen at about 10-13 seconds in the video. When the transponder responds, it sends encodes a response to the substation by drawing large amounts of current on various half cycles. The higher frequency "noise" seen at about 37 seconds on the waveform are very characteristic of the effect on the voltage waveform that the inbound current pulses can produce. Inbound is generated by turning on a triac that sits across the power line at the meter. For each signaled half cycle, this triac turns on at approximately 40 degrees before local zero crossing of the voltage waveform. There is a .8 ohm (approx) current limiting resistor in series with the triac. This results in an instantaneous current draw of around 100 amps on each signaled half cycle. This is what causes the "noise" pulses seen at about 37 seconds in the video. In extreme cases, It can collapse the voltage waveform from the point where it turns on until just after zero crossing. In prior TWACS generations, there can be up to 256 transponders responding to a single command. In the latest generation, there can be several thousand responses. These responses are phase and time domain multiplexed. Since the substation is looking for pulses of current and since distributed power sources downstream of the substation may be supplying the current needed by the transponders during signaling, the following (over simplified) scenario is very common: 1] The substation sends an outbound message to the transponders. 2] The transponders reply to the outbound message. 3] Since the substation could not detect a response, it re-sends the original command. 4] Again, the transponder resends its' response. Due to built in thermal safeguards, the transponder may stop signaling for a period of time causing additional failed outbound responses. 5] Trouble tickets are issued for "non-responding" transponders. 6] A service trip is made to the "Out of service meter". Local test equipment indicates the "Bad" meters are functional. But to be safe, the meter is replaced with another one. 7] The above cycles repeat unless the service tech notices a distributed power source nearby. As far as the effect on lighting is concerned, this is a known problem with the TWACS system. It generally results from undersized service at the premises. The waveform seen at 2:29 and the effect on lighting would be expected in most undersized installations or distribution systems. The general solution is to increase the distribution/service size. This means upgrading the wiring and distribution transformer. This is also the cause of the Enphase invertor problem. Since you have to be downstream of the meter, there is no real solution. This is another known problem with the TWACS signaling method. It is so severe in the Southwest that many utilities and are no longer selecting TWACS for their AMI/AMR solutions.​

 

[gar]

211224-1030 EST

Hv&Lv:

I do not perceptibly see a signal on my 60 Hz waveform like what your video shows.

I see a much more continuous stream of signals. But these are signals from clip leads in free space. Looking at the high frequency portion of my signal I have seen a clear modulation of the carrier frequency in a pulse. One pulse has a 200 nanoS duration of 12.5 MHz carrier or possibly 16 MHz, followed by a 180 degree phase shift of the carrier for about a comparable duration. Not all pulses have this 180 shift.

I am not seeing a relatively constant amplitude of the signal. What appears to be the sync portion of the signal is almost always the strongest part. Sync being the first part. Clearly some signals have the phase reverasal at 200 nanoS, and others do not. Clearly a man made signal with information contained in the signal.

.

 

[ELA]

Gar,
You have chastised others for making incomprehensible posts.

Are you simply looking at signals on your scope with floating leads?

You cannot post a screen shot why. Because " I ran out of space several years ago???"

You asked:
"Is there anyone else out there with a capable scope that sees anything like what i see?"

Test Setup details are important.

I fail to see exactly what you are troubleshooting ( the section you posted in) here.

 

[gar]

211224-1512 EST

ELA:

Why my post in Troubleshooting? Because what I am seeing is noise to measurements I am trying to make. Noise because it is interfering with what I am looking for. But it is not random Gaussian noise.

My scope is a Rigol DS2072A, and quite capable. It is powered from my 120 V 60 Hz power company power. The connection is via the scope 120 V 3 wire cable, and EGC is not isolated from power company. This scope has very little DC resistance from computer internal common and the AC cord EGC. This is in contrast to my many years old Tektronix scopes where there is some resistance that is noticeable.

My Rigol screen is 8 major divisions vertical, and 14 horizontal. Data storage is quite long in time, and only 8 bits in vertical. For my present measurements Y is 20 millivolt per major division, and X is 1 microsecond per major division. However, for recording the trace time base was set to 1 millisecond per major division for the single shot recording, then switched to the finer resolution for viewing.

Background noise between noise pulses is around 2 millivolts. These are noise pulses to me, but are a valid signal to somebody else. These signals can not be random, there is too much structure to them. They are clearly not white Gaussian noise.

The Y axis has a BNC to binding post adapter installed. The input voltage source is a 5 ft long twisted pair shorted at the input end with about a 6" diameter 1 turn loop. Over a small range loop diameter did not much change the signal level.

There seems to be about a 17 microsecond spacing between what appear to be repetitive similar pulses. This is a 58.8 kHz rep rate. A pulse appears to have a carrier frequency somewhere near 15 MHz, a pulse duration of 200 nanoseconds, and sometimes followed by at least one 200 nanosecond pulse that is in phase with the first pulse carrier, or 180 degrees out of phase

Hv&Lv has presented an interesting signal. So far I don't see such a signal on my power line.

Had my Christmas last night with kids and grandkids. They will be with big families of inlaws on Saturday.

.

 

[ELA]

Gar,
What you have is an antenna picking up all kinds of electronic RF emissions.
If what you are trying to measure is so low amplitude that it is buried in the background noise floor you will need to find another approach.

Try some experiments changing the length of your antenna and also isolating the scope from EG and you will see it is not that big of a mystery.

 

[gar]

211225-1919 EST

ELA:

Fundamentally I have a tuned LC resonant circuit connected to the input of my scope. This resonant circuit does not define the frequency of the signal I see. That frequency is generated at the source of the signal. I am not seeing an impulse triggering my resonant circuit. On battery power I see the same kind of signal.

The signal is a structured signal generated by some humans. It is not just random noise. Is the origin of the signal from my power lines, or is it an RF radiated signal?

The mystery is that I do not presently know the origin of the signal, or its purpose.

Why is not troubleshooting a good place to discuss this signal?

This is not lots of signals out in space. This is some specific signal that pretty much stands out by itself. Possibly my next step is to go out in the country and see if this disappears.

Apparently you see no similar signal. You would know it if you saw it. To see this signal you have to get rid of a lot of low frequency clutter.

.

 

[gar]

211225-2124 EST

ELA:

The timing between similar pulse bursts generally varies from 14 to 19 micro seconds. A pulse rate frequency range of 52 kHz to 71 kHz. The pulse amplitude changes, and the amplitude modulation shape changes from one pulse to another. The present inductor associated with the scope input internal capacitance is tuned somewhere near the pulse carrier frequency.

.

 

[ramsy]

I was going to ask if a new cell-phone tower went up somewhere close by, but If background noise sources can interfere with each other, as ELA suggested it would explain the variability as interference patterns change.