Unusual topic - Wireless frequency and interference

[#1]

Here is a technical question for someone who understands wireless technology.  Health warning : This has nothing to do with broadband or Plusnet !  I hope this may be of interest to someone.

In my house in France I have an electric robot which cuts my grass constantly.  It roams around the garden, nibbling the grass in a random fashion, but has a digital memory so that it knows where it has NOT cut.  Its freedom to roam is limited by a wire buried in the ground all around the garden.  This wire transmits a wireless signal to stop the robot going out of bounds, and also to guide it back automatically to its charging station when the battery needs re-charging.  The robot is quite sophisticated and its memory has many features, including basic programming, a weather / rain sensor, GPS so that it can be tracked by mobile phone from the UK, GSM  so that it can be remotely instructed  (i.e. "Go Home") by mobile phone, Bluetooth so that it can be "manually driven" and programmed by the mobile phone, and basic settings like how close to the perimeter cable it should cut, zone identification and path back to the charging station etc.

This is the second season during which it has cut an acre of grass up hill and down dale in three zones without major issue.  On a couple of occasions, it has tried to commit a sexual offence with a mole-hill or two or got its knickers in a twist with a bramble, which has meant that someone has had to intervene !

However, recently there has been a major issue.  Suddenly, for no apparent reason, it has re-set itself to its default settings for some reason.  That has meant that it has also lost contact with the base station and the perimeter wire as the frequency among other things has to be set up.  There is no obvious explanation but two possibilities come to mind.

In recent years, 400 KV overhead power cables have been installed by EDF in the vicinity with huge controversy, and the nearest pylon is 350 meters away - the cables are not visible as we are in a valley fortunately and there is no line-of-sight.  As the electricity is only being exported to Spain from the new nuclear plant at Flamanville in the North of France (Normandy), it is scheduled to be upgraded to 800 KV, and this may have already happened for all I know.  This is a classic example of political hypocrisy, as Spain has outlawed nuclear energy but is willing to import electricity generated in a nuclear plant in France !  France is self sufficient in nuclear energy already, and has no need for the new plant in Normandy; however, they have built it to obtain technical development expertise of the latest technology in order to be able to sell their expertise to Eastern Europe as consultants.  But they need to find a use for the unwanted electricity, so the Spanish and UK markets are a convenient excuse.

So I am wondering whether there could have been some sort of spike from these cables, either from general use or maybe the switch-over to 800 KV which may have knocked my robot out.  So my question is ; Could this be a possibie explanation ?

The other explanation may be a little more mundane.  We have a lot of land and a local farmer keeps cows in our field.  He uses an electric fence to keep the cows in the field.  The fence is powered from the mains supply in my house by a transformer which generates a 10,000 volt electric pulse - dramatic but harmless to man or beast.  However, the system does cause interference on my Long Wave radio, which means I cannot listen to the BBC Radio 4 broadcast except by Internet Radio.  Therefore, is it possible that this interference may be affecting the robot ?  The electric fence is effectively one long antenna !

The fly in the ointment for that explanation is that the robot has worked without this fault for nearly two years, despite having the electric fence operative.  However, until three weeks ago, only one of the two fields was being grazed by the cows; this field is about 5 acres.  Recently, due to the drought, the cows were moved into our paddock which is only about half an acre, and it was while they were there that the settings on the robot suddenly changed to default.  As this paddock is so much smaller than the field, any signal from it would be more concentrated as the electric fence is substantially shorter.  Maybe this concentration of signal caused the robot to re-boot ?

In a nutshell, can anyone tell me if either of these explanations are likely to be the cause or would wireless spikes not have this sort of effect ?

[#2]

Both are theoretically possible although I would very much favour the electric fence option, it's unlikely that the HV transmission lines emit any frequencies in the band that would upset a processor.. It all depends on how the software has been written and what the internal processor does if it's memory gets screwed up. If it loses it's configuration data then resetting it to the factory default seems a sensible thing for it to do.

The question has to be how the RFI is getting in, there are two possible options:

1. It is conducted interference, that is whenever the fence 'fires' a pulse is being conducted through your electricity cabling from the fence controller through the mower power supply and is addling it's brain. In case it's caused by conducted mains spikes you could use a mains surge protector as comes built into extension leads. You could also try running the mower controller through an RFI filter such as this   http://www.reo.co.uk/product/157 , which you would need to install in an insulated box. Or of course both, they don't do the same job.

2. It is radiated interference from the fence to the mower controller. This would be very hard to combat short of the farmer moving the fence further away

One further thought, when the farmer moved his fence did he move the controller nearer to your property and/or supply it from a shorter cable? If so get him to move it as far away as possible.

Of course the fault might be elsewhere, it could have been a lightning strike or the mower just having a temporary nervous breakdown etc.

[#3]

Baldrick1

Thanks for a very helpful reply.  Firstly, the position of the controller has not changed, only the electric fence itself.

However, it has made me aware that both the fence controller and the mower are on the same electric circuit - that is to say, the electric sockets for both of them are in the loft.  In France, they do not have ring mains but sockets are fed straight from the electric panel through individual circuit breakers, usually about three or four sockets per MCB.  In this case, there are just two sockets, one for the mower and one for the fence.  So your suggestion of the pulse being conducted to the mower from the fence controller is quite plausible.  I could try connecting the fence controller to a separate MCB, like the garage.

Although it has been as dry as in the UK these last few weeks, there has been no thunder or lightening.  So I doubt if that has been the cause.

The more likely is the radiated interference from the fence because the only thing that has changed is the fence itself which is far shorter than when the cows are in the large field.  I am assuming that being shorter, there would be more concentration of radiated signal (??).  The paddock would involve only about 85 yards of fence while the field would involve 700 yards.  This problem has only arisen since the electric fence was moved from the large field to the small paddock.  However, the mower is actually further away from the small paddock than the large field.  In the case of the large field, the mower's buried antenna runs directly underneath the electric fence for about 30 yards in one place.

In the meantime, I will try the garage socket and a surge protector and see if it happens again.

Many thanks for those ideas.

[#4]

@Baldrick1 has summed it up well..... and I agree with his surmise, that it is probably the electric fence that has caused the problem.... particularly as your last paragraph has mentioned that the Robot antenna is buried beneath the fence itself...  this would probably pick up any voltage/current spikes, in the same way as the original "spark transmitters" did when Marconi, and others, were experimenting  with radio transmission and reception.

Would it be possible to re-locate the robot antenna to be further away ( opposite side of the mowing area to the cows fields  )  

[#5]

Shutter

Thanks for your reply.  However, I probably did not explain myself clearly.  When I said that the mower's buried antenna runs directly underneath the electric fence for about 30 yards in one place, I should have made it clear that is only in the larger field where there has not (yet) been a problem.  In the case of the paddock where I have experienced the problem, the closest the robot gets to the electric fence is a good twenty yards.

So that is why I suggested it may be the concentration of signal in the shorter fence, because where the fence is ten times longer, there is no problem, even when the robot runs directly underneath the electric fence !  So is my idea of more concentrated signal in the short fence wishful thinking or could the short fence indeed make a difference ?

Whichever is the case, it is clear that the consensus of opinion is that the electric fence is the root cause, one way or another.  My difficulty is that I am only here for a week and when I get it going again, the problem may not occur again before I leave.  So it is difficult to experiment.  Murphy's law is that it will be fine until I leave and then.....

[#6]

Moving the mower onto a different circuit is an excellent idea.

When you say the field is more concentrated, my suspicion is that on the long run the voltage spike is losing its high frequency components due to the additional length increasing the capacitance to earth both from ground clearance and the dielectric properties of the insulating supporting posts. It's these higher frequencies that tend to do the damage.Add this to the closer positioning and away you go

[#7]

Electric fences generally use an electro-mechanical gadget with a coil in it to give a brief massive voltage from a battery, every few seconds.

Have you thought of employing the services of a sheep to cut your grass?

Very cheap, no running costs & you can eat it when you no longer need its services.

[#8]

---

@VileReynard wrote:

Electric fences generally use an electro-mechanical gadget with a coil in it to give a brief massive voltage from a battery, every few seconds.

Gosh that sends me back a few years, they had a huge battery and timing was done with a spring return flywheel.. Modern mains driven devices use new fangled electronics the same way as cars no longer use contact points in the distributor.

---

[#9]

This is Normandy, the World's home of dairy.  Normandy doesn't do sheep !

[#10 · Accepted solution]

Baldrick1

That makes a lot of sense because on the longer run, there are places where grass has grown up and is touching the electric fence, thereby presumably again reducing the strength of the pulses due to shorting.  On the short run (the paddock) the electric fence has not been in use before this year, and therefore the farmer checked for vegetation before switching to it.  If the paddock has a stronger pulse than the large field, it is supported by the fact that since the cows were moved back to the large field two days ago and the fence in the paddock has been turned off therefore, there have been no more issues.

The paddock is only used for a few days a year, because the cows soon graze it bare.  However, I have asked the farmer next time he uses the paddock, to leave the electric fence in both the paddock and the large field connected in order to dissipate the current.  I will also try different power circuits.

So thanks very much for the advice.  Back to broadband issues ......

[#11]

Baldrick1

That makes a lot of sense because on the longer run, there are places where grass has grown up and is touching the electric fence, thereby presumably again reducing the strength of the pulses due to shorting.  On the short run (the paddock) the electric fence has not been in use before this year, and therefore the farmer checked for vegetation before switching to it.  If the paddock has a stronger pulse than the large field, it is supported by the fact that since the cows were moved back to the large field two days ago and the fence in the paddock has been turned off therefore, there have been no more issues.

The paddock is only used for a few days a year, because the cows soon graze it bare.  However, I have asked the farmer next time he uses the paddock, to leave the electric fence in both the paddock and the large field connected in order to dissipate the current.  I will also try different power circuits.

So thanks very much for the advice.  Back to broadband issues ......