Radio Frequency Interference from NF4V NetComm VDSL modem

Forums › LAN (ethernet/Wifi/routers/Bluetooth) › Radio Frequency Interference from NF4V NetComm VDSL modem

61339

17 posts

Geek

#171565 22-Apr-2015 21:34

Amateur radio operators be warned. If you are thinking of getting an Orcon Genius VDSL account, even one that delivers relatively slow speeds of say 20 Mb/s down provided through an Orcon-supplied NetComm NF4V modem, likely both 160m and 80 m bands will be wiped out completely by RFI. At higher download speeds expect the 60, 40 and 30 m to be progressively unusable as well because of RFI up to 12 MHz - directly radiated by the modem. With only the two vital cables, PSU and DSL, wound around ferrite chokes and everything else unplugged, no reduction in the radiated interference could be detected. The NF4V is the only modem Orcon provides for VDSL, so nothing else was offered (or tried). Have had to revert to ADSL: all interference vanished while still using the NF4V. Would be most interested to hear of any other experiences in this vein since NetComm have given assurance that their NF4V has passed compliance testing to both AU and NZ standards. Maybe the bar is simply set too low, or there are too few active hams still operating the low HF bands to worry about properly silencing these VDSL monsters? If the problem is encountered, prepare for a battle of the 'ageists', few younger folk understand the nuances of analogue equipment anymore, and few oldies (like me) want to be swamped out by digital hash.

hyperman

418 posts

Ultimate Geek

#1289799 22-Apr-2015 21:54

sounds to me that the RFI you are experiencing is from the line

if you check out http://www.geekzone.co.nz/forums.asp?forumid=90&topicid=105744

you will see that the line operates upto around 7mHz on a 8b profile and this could explain why you don't see the problem on adsl as it only reaches 2.2 mHz

The views expressed by me are not necessarily those of my employer

61339

17 posts

Geek

#1290421 23-Apr-2015 21:42

Yes, you are correct about the spectrum requirements for VDSL. Let's look at the theory and testing done to see why it should not be a problem and the design of the modem is a problem.

The balanced phone line configuration, as in 'twisted pair', is sufficient to cancel most all radiation from so-called 'push-pull', or differential signals, which is the mode in this, indeed all, high-impedance telephone-line methodologies. My antenna feedlines are balanced for exactly the same reason. Interference as discussed here is therefore not from the line itself, a contention with which both the manufacturer of the modem and Orcon techs agreed. Any significant radiation from the line equates to unacceptable losses, these in addition to the small capacitive losses causing attentuation intrinsic to the twisted pair.

Common-mode radiation from the line, pick up from inside the modem, though not a contributor to signal loss could indeed be the source of the radiation.   But it was not, because had it existed, it would have been noticeably, if not entirely, reduced by winding the phone (DSL) line through a ferrite toroid, but there was no reduction at all when this was done. The hash interference began only after DSL sync was achieved and was preceded by clearly heard sync tones through the HF receivers. Digital handshaking data prior to sync was not radiated.

Interestingly, radiated interference from the modem was reported during compliance testing, but the modem passed anyway. Two possible conclusions to draw from that: 1. the testing bar is set far too low, but then why report the interference at all if it is deemed tolerable, or 2. the modem was passed for commercial, or possibly political expediency, more likely the former.   There could be a far more useful outcome for our knowledge-base if several VDSL-capable modems were offered for testing in the same RF environment, but that's not going to happen based on only one complainant.

hyperman

418 posts

Ultimate Geek

#1290439 23-Apr-2015 22:21

61339: Yes, you are correct about the spectrum requirements for VDSL. Let's look at the theory and testing done to see why it should not be a problem and the design of the modem is a problem.

The balanced phone line configuration, as in 'twisted pair', is sufficient to cancel most all radiation from so-called 'push-pull', or differential signals, which is the mode in this, indeed all, high-impedance telephone-line methodologies. My antenna feedlines are balanced for exactly the same reason. Interference as discussed here is therefore not from the line itself, a contention with which both the manufacturer of the modem and Orcon techs agreed. Any significant radiation from the line equates to unacceptable losses, these in addition to the small capacitive losses causing attentuation intrinsic to the twisted pair.

Common-mode radiation from the line, pick up from inside the modem, though not a contributor to signal loss could indeed be the source of the radiation.   But it was not, because had it existed, it would have been noticeably, if not entirely, reduced by winding the phone (DSL) line through a ferrite toroid, but there was no reduction at all when this was done. The hash interference began only after DSL sync was achieved and was preceded by clearly heard sync tones through the HF receivers. Digital handshaking data prior to sync was not radiated.

Interestingly, radiated interference from the modem was reported during compliance testing, but the modem passed anyway. Two possible conclusions to draw from that: 1. the testing bar is set far too low, but then why report the interference at all if it is deemed tolerable, or 2. the modem was passed for commercial, or possibly political expediency, more likely the former.   There could be a far more useful outcome for our knowledge-base if several VDSL-capable modems were offered for testing in the same RF environment, but that's not going to happen based on only one complainant.

I defiantly agree with you that the Netcomm units are not of any great quality (I have one sitting in it's box as it dose not meet my requirements)
however I am running Netgear instead I will see if I can do some RFI testing for comparison

The views expressed by me are not necessarily those of my employer

raytaylor

4014 posts

Uber Geek

Trusted

#1290508 24-Apr-2015 00:44

We have a similar problem with our customers.
The switchmode power supplies and ethernet will interfere with highly sensitive FM radios. We get around that by moving the power supply and router away from the stereo.

But HAM guys still complain about the other bands it interferes with.

To them I say you can either have broadband, or you can have your amateur radio - not both at the same time.
And I am in no way going to troll the depths of the jaycar bargain bins for transformer power supplies, but if anyone wants to supply us one of the right specification, we will use it instead of our standard switchmode at the time of the install.

There are often just no way to work around conflicting technologies.

We dont expect the wifi to work when the microwave oven is going, and so we therefore need to accept in this case that VDSL will interfere with bands that the VDSL operates over.

I believe chorus have emmission standards that they need to keep within so the interference should drop off as your radio equipment gets further from a copper pair carrying VDSL.

Ray Taylor

There is no place like localhost

Spreadsheet for Comparing Electricity Plans Here

ubergeeknz

3344 posts

Uber Geek

Trusted

Vocus

#1290515 24-Apr-2015 01:09

Not speaking specifically in defense of the NF4V here, but I doubt another VDSL modem would produce significantly less RFI.

Given the frequencies that VDSL is operating at (0.14-12Mhz), and the design of telephone twisted pair cable, which was designed to carry voiceband only, there is going to be RFI leakage no matter how well the modem is designed.

Putting a toroidal filter/choke on the VDSL line doesn't make sense, because you are filtering the actual signal that VDSL is using to transmit and receive data. Shielding the chip inside the modem doesn't make sense either, because you are then attaching it to what amounts to a bloody great antenna.

I imagine the powers-that-be did indeed decide that the possibility of someone wanting to opearate a HAM Radio in close proximity to a VDSL line/modem was fairly low and not a reason to disallow the technology.

61339

17 posts

Geek

#1290669 24-Apr-2015 10:44

It is gratifying indeed to see this forum is widely read and helpful advice abounds, however some contributions are just plain unhelpful. Many thanks are due to ubergeeknz for, at least, vindicating in his third and fourth sentences, complete with swear word, my contention that there exists today among the self-titled ubergeeks, little or no understanding of long established electronic theory. The difference between differential and common-mode line transmissions was explained to partially overcome just such a problem in the well-intended readership. Perhaps in support of his fourth sentence, ubergeeknz can explain how the raison d'etre, and the well established functionality of not just any old shield, but Faraday shields, has suddenly disappeared in the 'digital' 21st century. In contrast, hyperman has recognized an opportunity to further our understanding.

Beccara

1469 posts

Uber Geek

ID Verified

#1290709 24-Apr-2015 11:21

Things generate EMI sometimes, sometimes right out of the box and part of poor design but it only impacts a small subset of RF users and RSM having limited tools go after the things that impact more people like Emergency services com's (Like a switch mode supply from Motorola that went wonky and knocked out rural com's for police a while back).

Any half decent ham will check known beacons or clear channels for any signs of EMI, Moreso when they bring in new kit because every ham and his dog knows cheap switchmode PSU's are everywhere

- A ZL1xxx HAM

Most problems are the result of previous solutions...

All comment's I make are my own personal opinion and do not in any way, shape or form reflect the views of current or former employers unless specifically stated

AliExpress

Shop now on AliExpress (affiliate link).

61339

17 posts

Geek

#1290743 24-Apr-2015 12:05

Beccara, of course you're right in principle and thanks for mentioning it, however this problem has NOTHING to do with switched mode power supplies. Testing for radiation from the SMPS (wall wart) was the first, easiest and most obvious point of call ... and it proved to be clean. Unexpectedly all seven SMPS units running concurrently in the house are not contributing any detectable raising of the noise floor on any ham band. In fact hearing lively QRN crashes on 80 m is quite a bonus after the artificial AGC-induced limiting caused by 10 dB over 9 hash from the NF4V at a distance of 30 m from the antenna, as far away as physically possible.

ubergeeknz

3344 posts

Uber Geek

Trusted

Vocus

#1290825 24-Apr-2015 13:45

61339: Perhaps in support of his fourth sentence, ubergeeknz can explain how the raison d'etre, and the well established functionality of not just any old shield, but Faraday shields, has suddenly disappeared in the 'digital' 21st century. In contrast, hyperman has recognized an opportunity to further our understanding.

I do understand enough to know that in response to higher and higher transmission rates and thereby signal frequncy, the design of twisted pair (tpi, insulation, conductor diameter and so on) has changed, with the express goal of reducing crosstalk between pairs in the same bundle - that is to say, reducing RFI and the impact thereof. So your assertion that any old twisted pair is going to not leak at all provided the pair is driven symmetrically regardless of the frequencies being used cannot be the case.

At any rate, I'll be interested to see how other VDSL modems on the market perform, and whether my hypothesis that they won't be any better prove true or false.