Interesting Experiment With Adsl Splitter - 2

[thinhnghiem]

I have been considering what happens if I make the feedback from out put to input of a ADSL splitter, as well as change input to output and vice versa

Below are my steps:

- Take the first splitter, plug the phone line to the Net output (NOT the input) of it

- Plug the input of the first splitter to the input of the second splitter

- Plug the Phone output of the second splitter to the phone output of the first one

- Plug the Net output of the second splitter to Modem, and connect to laptop

 

See diagram in my attached screenshot

 

Here I made some abnormal things:

- Telephone signal transmits into the output, not the input of the splitter

The output of second splitter is connected back to the first one

 

Surprisingly, my laptop can access internet normally. Nothing wrong happens

Why my extraordinary actions do not affect the operation of the splitters?

[JMJones0424]

I do not know enough about electronics to state one way or the other that the filter in the DSL splitter is bidirectional, though it seems it could be from your experiment.

 

There are two possibilities (that I know of). The filter is either bidirectional or it isn't. If it is bidirectional, than your experiment as given in the above image preserves the higher frequencies for DSL communication. If it isn't bidirectional, then this experiment would fail to show any meaningful data, as the first filter did nothing to the signal.

 

I would be interested if you carried this experiment one step further and plugged in a phone (or fax or dial-up modem or some other device that operates in the voice freq range) into the "phone" output of the second splitter. If you receive a busy signal, than this is evidence that the filter is bidirectional. Plugging the "phone" output of the second splitter into the "phone" jack of the first splitter doesn't seem to me to accomplish anything.

 

EDIT: To distinguish the difference between a dial-up modem and a DSL modem.

Edited November 20, 2013 by JMJones0424

[DFINITLYDISTRUBD]

Ehrm, the filters have to be by-directional as data travels both to and from the modem, flipping them around will have at best a negligible affect, more likely no effect on the signal.

 

But looping the phone as in the pic will do nothing as you cannot actually loop the input to the output short of shorting the circuit or placing a sufficient load between the outbound line and the inbound. To create a feedback loop you would have to cut into the cables and splice the red to the green and other green to the other red. (or black and yellow depending on the line you are working on)

 

 

Upon giving it more thought even that is not likely to work either....I suspect that a transformer would have to be added in line with the red wired to one coil and the green to the other this would cause the input to "hear" the output and the output to "hear" the in.

Edited November 21, 2013 by DFINITLYDISTRUBD

[JMJones0424]

Ehrm, the filters have to be by-directional as data travels both to and from the modem, flipping them around will have at best a negligible affect, more likely no effect on the signal.

 

Again, I claim ignorance, but it is at least conceivable that the filtering could be done on the sending end and that this second experiment alone is not enough to show that the filter is bidirectional.

 

Additionally, the transmit and receive frequencies are different, and both are covered in the range of the DSL filter, so I don't particularly understand your assertion. I'm not disagreeing; I believe the filter is bidirectional. I just don't know that for a fact, and I am working in the constraints of the experiment as it has been presented.

 

PEDANTIC EDIT: If I was a better grammarian than I am, then I wouldn't have previously used "than" when I meant "then".

Edited November 21, 2013 by JMJones0424

[thinhnghiem]

Great, I will try again with your suggestion. Thank you

[thinhnghiem]

I do not know enough about electronics to state one way or the other that the filter in the DSL splitter is bidirectional, though it seems it could be from your experiment.

 

There are two possibilities (that I know of). The filter is either bidirectional or it isn't. If it is bidirectional, than your experiment as given in the above image preserves the higher frequencies for DSL communication. If it isn't bidirectional, then this experiment would fail to show any meaningful data, as the first filter did nothing to the signal.

 

I would be interested if you carried this experiment one step further and plugged in a phone (or fax or dial-up modem or some other device that operates in the voice freq range) into the "phone" output of the second splitter. If you receive a busy signal, than this is evidence that the filter is bidirectional. Plugging the "phone" output of the second splitter into the "phone" jack of the first splitter doesn't seem to me to accomplish anything.

 

EDIT: To distinguish the difference between a dial-up modem and a DSL modem.

 

I am using DSL modem

 

I follow your suggestion, I receive no busy signal, and I can make the call normally

[thinhnghiem]

 

To create a feedback loop you would have to cut into the cables and splice the red to the green and other green to the other red. (or black and yellow depending on the line you are working on)

 

 

Is there any danger when doing this, for example, the blow up

[C1ay]

Only the phone tap is filtered to keep you from having to listen to the data. Internally a typical filter is just some inductors and a capacitor or two on the phone side to attenuate bands used on the network.

 

[JMJones0424]

Only the phone tap is filtered to keep you from having to listen to the data.

 

Thanks. This is an explanation that I hadn't considered, but it explains the results of this experiment.

 

It also matches my trials here at work. We have DSL with one splitter. While I didn't have access to multiple splitters necessary to replicate Thinh's experiment, I tried reversing a single DSL splitter in numerous configurations.

1) Input of splitter to phone, DSL output of splitter to POTS line = phone worked.

2) Input of splitter to DSL modem, DSL output of splitter to POTS line = DSL modem worked.

3) Input of splitter to phone, phone output of splitter to POTS line = phone worked.

4) Input of splitter to DSL modem, phone output of splitter to POTS line = DSL modem failed.

 

This agrees with the conclusion that the DSL splitter is both bidirectional and only filters out the higher frequencies used by ADSL. But this doesn't seem to me to agree with the findings in Thinh's first experiment: Switching DSL and phone line outputs gave him a busy signal when trying to dial out on a fax. Do you know if there are different standards internationally for fax machines, such that some faxes use frequencies above voice? Or are some DSL splitters capable of filtering both the voice and the data channels?

Edited November 26, 2013 by JMJones0424

[JMJones0424]

OK, now I've confused myself even more. While Thinh's second experiment and my tests here at work agree with Clay's comments, Thinh's first experiment appears to me to be contradictory. Here, Thinh simply swapped the output of the DSL splitter between the fax machine and the data line. When he dialed out on the fax, he received a busy signal. Thinh did not give any information regarding the DSL modem's performance in that experiment when it was hooked up to the voice side of the splitter. Regardless, if the spitter only filtered the voice side, the fax machine, hooked into the data side, shouldn't have been affected (if my understanding were correct).

 

We do not have a fax machine here at work, and I neither have a fax machine nor DSL at home. When I replicate Thinh's first experiment with a phone, the phone plugged into the data side of the filter has a dial tone and can complete a call without any problems. However, when I run a speed test at http://www.dslreports.com/speedtest?flash=1 on the DSL modem when it is plugged into the voice side of the modem, I only show 104 Kbs/s down and 263 Kb/s up while when the DSL modem is routed correctly through the data side of the splitter, I get 853 Kb/s down and 353 Kb/s up.

 

Clearly my understanding of the situation is lacking, and my response to Thinh's first experiment may be entirely false.

 

I did a few more tests under Thinh's first experiment's constraints, with the voice and data lines swapped on the DSL splitter. While calling my home phone through the data side gave no issues whatsoever, I was unable to even start the DSL speed test at dslreports.com while the phone was ringing. Once the phone company's automatic answering service had picked up, I was able to start the DSL speed test and got results that matched my above report, about 100Kb/s down and 250 Kb/s up. EDIT: My hearing is notoriously poor, after years of abuse from exposure to extremely loud environments, but when conducting this test, I heard no bleed-over on the phone from the DSL modem.

 

It's at least possible that standards are different internationally for fax machines, and it may be possible that standards for DSL splitters are different for countries outside of the US. I do not understand, given my tests and Clay's explanation, why Thinh was unable to complete a connection with a fax machine while the voice and data outputs of the DSL splitter were switched. I also do not understand why my ADSL modem is able to operate at all, much less at decreased performance, when hooked up to the low-pass filter side of the DSL splitter/filter. Is this simply a symptom of a filter that needs to be replaced? What other tests can I perform to better understand what's going on? What am I missing?

Edited November 26, 2013 by JMJones0424

[thinhnghiem]

Hi all,

 

There is recommendation to me that it must be careful with these kinds of experiments, since they may be illegal to do like this.

 

I do not take this recommendation seriously, until I posted the my experiment to a canadian forum. Surprisingly, my account was blocked immediately by administrator.

 

By the idea

"To create a feedback loop you would have to cut into the cables and splice the red to the green and other green to the other red. (or black and yellow depending on the line you are working on)"

 

It is impossible for me to do this in physical, since it is very difficult to accomplish. It demand for very skillful person

 

Anyone can tell me how to do in different way to create the same result?

[thinhnghiem]

Hi DFINITLYDISTRUBD

 

"To create a feedback loop you would have to cut into the cables and splice the red to the green and other green to the other red. (or black and yellow depending on the line you are working on)"

 

Finally I can do this with the help of a technician

 

See my attached screenshot

 

In a), I loop the net output to input, and connect phone output to a telephone machine. No signal at all

 

In B), I loop the phone output to input, and connect net output to a modem, this time I can access internet normally

 

Is it working correctly?

[DFINITLYDISTRUBD]

Sorry for the long wait. Too many pots in the fire lately.

 

If you're really dead set on creating a feedback loop (Just out of curiosity, why?) the easiest solution would be to get your hands on a cheap simple old phone (the kind that has a cord to the handset) remove the cord from the handset then connect the mike leads to the speaker leads.

[thinhnghiem]

Sorry for the long wait. Too many pots in the fire lately.

 

If you're really dead set on creating a feedback loop (Just out of curiosity, why?) the easiest solution would be to get your hands on a cheap simple old phone (the kind that has a cord to the handset) remove the cord from the handset then connect the mike leads to the speaker leads.

 

Hi DFINITLYDISTRUBD,

 

Yes, I implemented this experiment just for curiosity. I create the looping back like this, just to know what happens. Who knows, if it can result in any effect that we ever know, and many application can go further after that

[thinhnghiem]

Hi DFINITLYDISTRUBD,

 

Yes, I implemented this experiment just for curiosity. I create the looping back like this, just to know what happens. Who knows, if it can result in any effect that we ever know, and many application can go further after that

 

Hi all,

 

 

I have been going ahead with more experiment for the ADSL splitter

 

Following is all of the experiments I have done, including those I have posted before, and the new ones

 

Their schemas can be found in the attached screenshot

 

Experiment a)

Plug the telecommunication line into the input. Then in the output lines, I make the swapping. I plug the phone line to the internet output, and the internet line to the phone output

Then I dial the phone to call to a fax number

Surprisingly, I heard the busy tone, instead of fax tone.

Note: When I connect in normal way, plug the phone line to the phone output, and internet line to the internet output, this effect does not happen. That means I heard the fax tone as usual

Experiment B)

I make the looping back from the net output to the input. To do this, I cut into the cable and heal the wire inside, with the same color correspondingly (green to green, red to red)

Then, I connect phone output to a telephone machine.

Result: No tone heard when I picked up the phone machine

 

Experiment c)

I make the connection between phone output and net output, by cutting into the cable and healing the wire inside, with the same color correspondingly (green to green, red to red), similarity as I did in experiment B)

Then, I connect phone output to a telephone machine.

Result: No tone heard when I picked up the phone machine

 

Experiment d)

Wall line is plugged into the net output (Not the input). Then, I make the looping back from the phone output to the input. To do this, I cut into the cable and heal the wire inside, with the reverse color correspondingly (green to red, red to green)

Then, I connect input line to a telephone machine.

Result: No tone heard when I picked up the phone machine

Note: If I repeat the same experiment by healing the wire with the same color correspondingly (green to green, red to red), no issue happens. The tone can be heard from the phone machine

Experiment e)

Wall line is plugged into the phone output (Not the input). Then, I make the looping back from the net output to the input. To do this, I cut into the cable and heal the wire inside, with the same color correspondingly (green to green, red to red)

Then, I connect input line to a modem to emit Internet signal

Result: No internet signal is created