Signal fading in the cable

[JPMalef]

Hi,

When you mentioned 600 meters of wire, you mean boundarie wire only, or boundarie and guide wire?

One way to isolate the problem is to connect the guide wire #1 in place of the right boundarie wire and check the station's Led.
Then do the same with guide wire #2.
Then same test with the left boundarie wire.

If you have a multimeter, measure the resistance in each test case.

This will help you to isolate the segment of boundarie wire which cause you this trouble.

[LOUISXIV]

Hi kolaf,

With 10 ore more Scotchlocks in the loop I am not surprised that you have problems with the loop. Personally I dont trust their long time stability, as contacts are only made at points, which may oxidize with time and increase contact resistance. Therefore I have soldered and enveloped all my ground connections.

You say that the mower complains about signal loss. Then the control lamp at the charging station should be blinking, is it? if it blinks this means that the total resistance of the loop is to high and the reason for this lies in increased transfer resistances at the contacts within the numerous contacts. Hard to find them all!

I would first measure the resistance with a multimeter. It should be less than 20 Ohm. You should disconnect the loop cable at the station and make a safe contact to the ohmmeter. I personally push small strips of metal into the connectors to reach to the contacts quite far inside. Can you do that and inform us about the value? Otherwise let us at least know if the test lamp blinks and the blinking sequence.

Do you also have guide wires installed? In order to localise parts of the boundary loop one can also make resistance tests between a guide wire and one branch of the boundary wire.

May I kindly ask you to bring your signature (mower type, your region, other information) up to date.

Do you also understand French, or only English?

LOUISXIV

[kolaf]

Hi,

I mean the boundary wire. I have a guide wire of around 70 m in addition. Problems I'm seeing are only with the boundary wire.

I'm not sure I understood your suggestions. Should I attach the boundary wire to the guide wire plug on the base station instead? And how would measuring the resistance help me? Wouldn't I measure the resistance of the entire boundary wire, which wouldn't help me determine where in the wire the break was.

[kolaf]

Hi LOUISXIV,

Sadly I only understand English. I will try to fix my signature once I figure out how (with a a lot of help from Google translate) :-).

The lamp on the station is blinking slowly which is indicative of a wire break if I recall correctly. I will be sure to do a resistance measurement when I have a time. Soldiering the brakes instead of using scotchlocks sounds like a good idea, maybe I will go around and try to a number of the joints in order to improve the signal.

I now understood the point with a resistant test with the guide wire, I will give it a try.

Thank you for your input. I'll be back soon as I have done the tests.

[JPMalef]

At the base station, you disconnect the right boundarie wire and your guide wire and you switch them.
Same for the other side.

[kolaf]

I have made the measurements and found the following:

80 kOhm for the boundary cable end-to-end
110 kOhm between the guide and one end of the boundary cable
43 kOhm between the guide and the other end of the boundary cable

This is more than 1000 orders of magnitude (kOhm vs Ohm) larger than what you indicated. Something is apparently seriously wrong with both segments of my boundary wire. Perhaps the best is to simply replace the whole thing? There could be a weakness in the cable connector. After all, it has been under the snow for two winters. I will try to make a new connection to see if that improves the conductivity.

[JPMalef]

A small detail with big impact BD.

When you measure the resistance, both cable ends SHOULDN'T be attached to the base station.

[kolaf]

Of course, I disconnected all cables before measuring :-)

[LOUISXIV]

kolaf! Found 40-100 kOhm? Did you not take the wrong scale? And did you check the meter if you shorten the 2 inputs that it reads 0 ohm? Further: the classical analog meters have a potentiometer to adjust 0 ohm. But I have the impression you understand quite well what you do...

The resistances that you found are far to high and correspond practically to open circuit with the boudary loop connected. At these resistances, the control lamp should blink at 2 seconds interval. Can you confirm?

I would try to detect the break using an portable AM radio (LW preferrable to MW). If you hold it near the boundary wire, you will here a knattering noise. Now step along the wire until it disappears: here is the break. You can possibly also remove one connector at the station and localize better. I confess, I never had to do this job. An electric sheep fence is also recommended by certain people: if you connect its HV-outlet to the loop, you will get quite a spectacular electric arc/explosion at the place of highest resistance. This second procedure is better suited for the case where the cable is not really interrupted but only shows locally a high resistance. But not many people have access to such a tool.

I just realize that you have about 600 m of boundary wire, which is slightly more than the admitted 500 m by Husqwarna. But it worked for 2 years. Question: do you have rodents in the ground? They might thin the cable.

LOUISXIV

[JPMalef]

Remember the radio method, one of the wire extremity shouldn't be connected.

[kolaf]

I have tracked the wire with an AM radio (disconnecting one of the connectors), and as I initially described there is no clear point where the signal dropped drastically. It was more like a general fade. If it faded too low, I could typically go back a few metres to pick it up again, and continue tracking it much further. The lights on the charging station is blinking slowly. I didn't time the interval, but two seconds sounds about right.

I do not have a sheep fence, but I have an electric fence for horses. I could try to connect that and see if it makes any flashes or sounds I can follow.

As for the measurements, I'm pretty sure they are correct. I have a digital multimeter, and I had to set the dial on 2000k in order to get anything at all. Of course, when shorting the probes I get zero resistance. I suspect still that maybe the connectors have gone a bit bad, so I will try tomorrow to cut and strip the wire to get to fresh metal. I will have to do this to try the electric fence anyway.

I'm not certain about my 600 meter length, I don't remember how many cable rolls are used when laying it. I am aware that it should be on the edge of what is supported, but as you said, it is worked for two years, even with more than 10 breaks. I guess we have some rodents (I live in the countryside), but not an exceptional amount, I believe.

[JPMalef]

If (for sure you will) you have to replace your cable, try to use the simple electric wire of 1.5 mm2 (available in any "do it yourself" shop).

We use them a lot, they are robust (against rodents), and not as expensive as "the original".

[Zazou]

It remains for me one question: 4000 square meters are 200 x 200 meters.
Then i guess the perimetric length of cable is rather 800 meters than 600.
What do you mean about ?

[Zazou]

Sorry , the calcul is idiot !

[JPMalef]

If the surface's shape is a cercle , it will have a length of only 225 meters. :o)

[LOUISXIV]

Kolaf, answers to your post of 09/04 22:40:

a) AM-radio tracking: Did you also follow JPMalefs advice to connect only one side of the loop? Keep in mind, that the HF signal is probabably just on the "high" output of the station (the other beeing ground or low), therefore try both outputs. Nevertheless, due to the high resistance of the loop, the HF-signal is probably going all around the loop. The electric fence method might in this case be the method of choice and fortunately you have one. At least, using it, the high resistance part of the boundary wire will burn through and you might switch back to the radio method.

b) Resistance measurements: from your explanations I conclude also that they are correct. However, before making further tedious steps please verify that the end connectors are really making good and stable contact: a little waggling during the resistance measurement should reveal this.

Clarification of the recommendation of JPMalef about using "simple electric wire of 1.5 mm2":

He means the solid electric wire for electrical house installations. I.e. not braid wire like the original wire supplied by Husqwarna. According to my information/calculations the standard tinned braid wire supplied with has only 0,75 mm2. So if you replace by solid 1,5 mm2 dont worry anymore about the length. But please solder the connections!

LOUIS XIV

[kolaf]

Thanks for the updates, LOUISXIV.

Yes, I collected only a single side of the loop. I connected my electric fence generator, but I was not able to directly determine where any break was. I have yet to do and AM radio tracking after this, though.

To connect the electric fence I had to strip new ends on the wire, and I did a second resistance measurement using these fresh ends. I got the same ridiculously high resistance values. I guess, though that resistance is not the whole picture since we are talking about a HF signal maybe impedance is more correct. In this case, though, I guess it doesn't matter :-)

Thanks for the clarification about using the solid electric wire. I will definitely order a long roll of this. In my local gadget store they also have a manually operated "wheel" that can be used to cut the wire a bit down into the earth. This should also help minimise the potential for future breaks.

The cable I have now has been driven over regularly (it even goes underneath my driveway, cut by a snowblower several times, cut by the robot mower several times, an excavator broke the cable in at least one place, and my father-in-law trying to trim the edges also cut the cable a few times. I'm guessing it will be best roll if I just replace it :-). It is just such a horrible job.

As to the shape of my lawn, it is in three patches. I have one quite irregular patch divided from everything else by a road on all sides. The robot reaches this using a guide wire. I have a quite regular patch which is an island in my courtyard with a road all the way around, and the third patch is highly irregular with several buildings on it. I am now isolated all these patches separately using the boundary wire. When I do this again I'm seriously considering just making one big circumference of everything and let the robot go wherever it wishes, crossing the road as many times as it wants to, and bump into various buildings. I previously have the boundary wire blocking off the buildings as well, but maybe I'll just skip this as well to save myself a lot of work. I'm not certain, though. If you have any advice/ideas around this it is very welcome.

[kolaf]

I just went out and did a new test with the AM radio. This time a clear break in the signal was detectable. This was just where I had two scotchlocks quite close to each other. I cut these two out of the cable, made a temporary join just twisting the cable together (after pulling up much of it to get them close to each other again) and redid the test. This time the signal went much farther. I think I could do this multiple times, connecting the electric fence generator to break the cable and then find the break using the AM radio, but I suspect this will just lead me to every join I already have on the cable. Still, maybe it is easier than putting down a new one :-).

This leads me to a very important question. If I were to solder the new joins, how do you waterproof them? Just wrapping them on tape is insufficient. Would hot glue work? Or maybe some kind of vulcanising tape?

[LOUISXIV]

Fine - so you get along with localizing the the critical points. What is the difference in testing now compared to the first time where you did not get any sharp break with the radio test?

Waterproofing of the soldered contacts: I use a piece of THIN plastic tube into which I inject silicone rubber. I take the sanitary silicone that I have usually at home, this material splits off acetic acid that has to diffuse through the tube (therefore thin). Others recommend shrink tubing, others shrink tubing equipped with hot glue. The newest information by one member in this forum is that he even abandoned the putting of plastic tube filled with silicone rubber with no negative consequences, i.e. no protection. This fits well into the image that due to the very low resistance of the loop (compared to the conductivity of the soil around) the outer insulation is really of minor importance. One might also consider using the silicone method and replacing the plastic tube by a cardbox tube...

Anyhow, in your case with so many points of repair I would put some protection around the solder.

Concerning your plans of installing a new boundary around the mowing area, I would prefer that JPMalef gives you an answer. The question is if this problem should not be better discussed in the section "Installation et configuration du robot". But may be there arises a problem because of the English language...

You should also consider the fact that the 230 at standard speed is quite fast on the way. There were reported broken bodies due to this high speed in the past. Although the body has been reinforced in the meantime I would not rely on bumping to often.

LOUISXIV

[kolaf]

Hi guys,

I just wanted to give you a short update. Thanks to your help and suggestions I have made some progress on debugging my cable. So far I have covered around quarter of the length and mended six breaks. Most of them were old scotchlock junctions, but there were also one of two new ones.

Thanks to this success I have disbanded the idea of replacing the entire cable. Instead I will mend portions of the cable that have multiple breaks. Currently I'm just cutting the poor junctions, stripping the wires, and twining them together as a quick debug solution. Once I have full connectivity all the way around I will determine which pieces I should replace and then use the soldering iron to make the junctions permanent.

I have passed the point where my guide cable is attached, and when I include this in the loop I have only 5 ohms of resistance on that segment :-)

Thank you very much for your help and guidance!