Take the 2-minute tour ×
Electrical Engineering Stack Exchange is a question and answer site for electronics and electrical engineering professionals, students, and enthusiasts. It's 100% free, no registration required.

This story must be quite familiar:

From the start, I've had a problem with my phone/ADSL line: Cracks, hisses and other noise on the phone - enough to make it occasionally unusable, along with an ADSL connection that is slow and as far from reliable as it can get without failing completely. Naturally, the ISP and phone company insist that the problem is on my end.

My house was renovated about 3 years ago and all indoors wiring was replaced, before we moved in permanently. When my phone went dead three months ago, I finally replaced the last pre-renovation length of cable, the one that went under the garden to connect to the telecom entry point.

Admittedly, that cable was a mess. It was a heavy-duty Cat-3 cable, but it had been in that underground channel for almost 20 years - the insulation had been obviously damaged in some spots, exposing the wires. I think that something may have actually eaten through the cable here and there.

Unfortunately, replacing that cable did get the dial tone back, but it did not solve any of the other issues. I have disconnected and re-connected all accessible joints in the line, and done all the usual tests like disconnecting the house wiring (including the alarm system), to no avail.

Rather than have another bout with the phone company or start going after the newer 3-year old phone cables, I was wondering if there was a way to test the phone line from the telecom entry point to the distribution point inside the house.

Of course, I am talking about something more useful than a simple continuity/resistance test, which does not reflect how the line will behave under an AC signal. I know that there are specialized cable testers, but they are very expensive and very few technicians have them. Ideally I'd use an oscillograph along with a frequency sweep generator to test the frequency response of the line, but I do not have access to any of these instruments at this time.

Is there anything else that I could try, preferably something that does not require a lot of cash, or tearing cables out of the walls?

PS: I don't mind building a circuit or two, as long as I have the schematics - I have not designed a new circuit for quite some time, and I'd rather not do it if I can avoid it.

share|improve this question
    
Nothing to do with electronic design. Question will be closed. –  Leon Heller Nov 30 '11 at 19:32
11  
I disagree. The fact that it's a phone line specifically is irrelevant... at core it's a question about how to test a circuit to see if it's introducing noise. That generalizes to many other things... and the constraint of not having access to specialized test equipment makes it an even more interesting question since one may want to build a quick and dirty, one-off bit of test kit for this scenario. I say this is absolutely on topic. –  mindcrime Nov 30 '11 at 19:38
12  
I'm sure Leon is a robot... Or at least a perl script. –  Majenko - not Google Nov 30 '11 at 19:41
1  
Perhaps he's a Javascript? (@Leon: If you're not using that tool, you ought to give it a look!) –  Kevin Vermeer Dec 1 '11 at 0:04

4 Answers 4

Testing the wiring is an interesting project in and of itself, but there may be a simpler and more effective way to address the overall system, at least before resorting to that: Can you take the DSL modem and a laptop to the service entry point and try it right there, thus eliminating the wiring issue? If it works there, then you work on testing your wiring; if it doesn't the problem would seem to be either with the carrier/drop or with the modem itself.

share|improve this answer
1  
+1 - nice and simple way to check things. Using a modem known to work (e.g. try it on a working line first) would make even more sure of things. –  Oli Glaser Nov 30 '11 at 20:33
    
I'll +1 this one, as this was one of the first things I tried. Unfortunately, I seem to have a Heisen-fault in my hands - both times I tried to do this kind of test, the problem had (mostly) gone away regardless of the location of the modem. I'll see if third time's the charm, as soon as the weather and my time allows... –  thkala Nov 30 '11 at 21:31

A Time Domain Reflectometer is the tool you need. However they are big bucks. If you can borrow a decent o'scope, that is half the battle. The only other thing you need is a pulse generator. By sending a pulse onto the line and looking at the pulse with the O'scope, you can tell the condition of the line. Shorts, opens, length of cable, etc all can be calculated. There happens to be an article in Elektor December 2011 magazine on how to do this, also includes plans for a cheap pulse generator.

share|improve this answer
    
+1 I'll try to get my hands on an oscilloscope - using a pulse generator to perform a basic TDR test sounds interesting... –  thkala Nov 30 '11 at 23:27

Some good suggestions here. Just to hazard a guess at the cause, it's sounds like it could possibly be what is known as a "high open". This is when there is a high resistance partial open that lowers in resistance when current flows through it.
In cases like these where the DSL is down, you may find that taking the phone off the hook (causing current to flow in the lines) brings it back up again.
If this happens for you, then this almost certainly the case. We had this happen here, and it took a lot of perseverance to get the phone company to admit the problem lay with them. I lost count of the number of times we were warned about how much it would cost us if it turned out to be on our property. We checked things thoroughly (luckily have the equipment here to do so) and made sure it wasn't on our end. We then got Ofcom involved (as the repeated phone calls were not doing the trick), at which point they resolved things immediately ;-)
I think as well as the other tests suggested, it could be a simple case of contact oxidisation so it's worth a basic resistance test also, as it may flag something.

share|improve this answer
    
+1 An HR fault is exactly what I am suspecting, due to the occasional existence of audible noise in the line when the ADSL modem is operational, along with the usual cracks and hisses... –  thkala Nov 30 '11 at 21:42

Well... You could disconnect the circuit from the telco owned line at the demarc box, and then inject a signal into the line that you can simply listen to on a phone in the house, and see if you get any crackling and popping. If you deliver around 6-12 volts to the line - and you have a generic hard-line desktop telephone - you can just take said phone off-hook manually, and you'll be able to hear the signal you deliver. You could generate a constant tone any number of ways, (555 timer circuit, SN76477 IC, or use an Arduino and the "tone" library) or you could just play music from an MP3 player or something. You could also take another old telephone, wire it up in series with the circuit to the house, give the whole thing 6-12 volts (DC) and you can talk across the circuit (just take both phones off hook and start talking... there's no ringing or anything involved if you do this).

If you want a slightly more technical approach, and you don't have an oscilloscope handy, you might be able to use xoscope1 on a PC. It's a software oscilloscope that can use your sound card for input. It's nowhere near as capable as a real oscilloscope, but for the kind of simple test you're talking about, I think it would work.

share|improve this answer
    
+1 For some reason it had not occurred to me to try an audio signal. It might not catch an HR fault, since I'd have an upper limit of at most 15 KHz, but it would be better than a DC continuity test. –  thkala Nov 30 '11 at 21:52
1  
There's also the very good Soundcard Scope for Windows which is free for personal use. –  AndrejaKo Dec 2 '11 at 14:45

Your Answer

 
discard

By posting your answer, you agree to the privacy policy and terms of service.

Not the answer you're looking for? Browse other questions tagged or ask your own question.