PLT, Homeplug and mains Networking etc.
27 October 2009 - 11:44am
Some of these claim 200Mbbps or even 1Gbps. BT's own engineers have found they can interfere with ADSL, but the BT marketing guys selling Homevision don't want to talk about that.
RSGB has taken Ofcom to court alleging lack of enforcement of standards. Some produce more interference than others. It's hard to beat Shielded Cat5e along the edge of the skirting board.
Sweden has now taken action: http://www.southgatearc.org/news/october2009/sweden_withdraws_pollluting_plt_equipment.htm
If installing plugtop mains networking, do check your DSL Modem stats before and after and UK LW & MW Radio. SW too if you use it.
Some Energy saver bulbs completely obliterate R4 LW and others have no effect. Not all Electronic gear is created equal, but PLT/Homeplug etc is inherently a DSL like up toi 30MHz transmitter. DSL is on twisted pair wiring designed to minimise interference. Mains wiring is not, and the lighting and switch circuits particularly can act as aerials as "live" and "neutral" can be in different places in the ceiling or wall rather than side by side. The fuse box does not block the signals and the Electricity meter has only a small effect to block the signal.
See also http://www.rsgb.org/emc/plt.php
__________________
tap, tap.. Is this thing turned on?
29 October 2009 - 11:36am
#2
I doubt there is much difference between Ring and Radial Twin&Earth for PLT/Homeplug. It's the length of wiring that affects both speed and interference. The lighting circuits are also in parallel as the fuses/thermal trips have minimal or no blocking effect and the the ELCB feeding sockets has very little filter effect, especially below 10MHz.
31 October 2009 - 11:08pm
#3
Zilog Jones;2226 wrote:
At least 802.11b/g/n only interferes with one already congested band
Are ring mains better or worse than radial circuits for causing interference? Most of the original wiring in my home (built in late '60s) is all rings.
They are all bad since the RF energy is present on all parts of the wiring in your home (& quite possibly next door too..). your lighting circuits for example are usually loops that make effective antennas. neutral goes around in a ring to all the bulbs and the Live takes quite a different route via the swtiches
Some PLT devices are worse than others because they leave all their carriers on all the time (DS2 chipset based devices such as comtrend)
ALL of units that were tested recently by the Norwegian regulator failed to meet the EN55022 standard by which the vendors claim CE compliance.
Some Units submitted for testing in the UK to an accredited test house were found to exceed EN55022 by over 30dB
These devices might be convenient, saving the installation of a few M of CAT5 and be available currently on the market in the EU but they cause horrendous radio interference and greatly exceed EMC standards.
Quite apart from the interference issue (which is a Genuine and severe problem) there is the issue of allowing devices to remain on the marketplace that do not meet EMC standards whilst operating as intended.
EMC (ElectroMagnetic Compatibility) is the art of compromise such that devices do not cause each other undue interference.
If these devices are allowed to continue to remain on the market whilst greatly exceeding the EMC limits that they are supposed to adhere to;
Why then should PC power supply manufacturers continue to fit expensive mains filtering ?
Why should CFL bulbs have filtering, why should any device be filtered in any way?
In a scenario where absence of these power supply filters was common then the mains would then be far too noisy for the PLT systems to operate at all (quite apart from all the radio interference caused of course ) They only operate successfully BECAUSE other devices adhere to the standards dictated by EMC laws and these laws have been enforced to date.
I have evaluated samples of various power line adapters here, at best in a near ideal set-up I was able to get a bit under 60Mbs out of a pair of '200Mbs' Units. once I went real world with these I managed between 6M and 20M
Not even as good as cat5 running at 100M in any scenario I tried. in all cases, even on a specially EMC filtered mains spur with around 60dB of isolation from the household ring main and lighting circuits there was enough radiation of RF noise from the comtrend devices to severely compromise Short-wave broadcast reception.
Some links from am EMC perspective as to why these things are a supremely bad idea are here courtesy of The EMC Journal
http://www.nutwooduk.co.uk/default.aspx?id=17
regards
Brendan
Member, IEEE EMC society
2 November 2009 - 10:03am
#4
This one is interesting:
http://www.compliance-club.com/PLT/Tim%20Williams%20EMCJ%20Issue%2080.pdf
Quote:
A HomePlug network was established. One terminal was
a laptop PC using a USB-to-mains-PLT HomePlug
device. The latter was plugged into a mains extension
lead and thence into the mains wall socket. A set of
Christmas-tree lights was also plugged into the same
mains extension lead. The PLT network functioned as
expected, communicating with a second terminal that
was plugged in elsewhere. When the mains extension
lead was then unplugged from the wall, so that the
laptop PC’s HomePlug device was no longer physically
connected to the mains, the HomePlug network
nevertheless continued to function. It was now
functioning in effect as a Wireless LAN, using HF
frequency spectrum. The lights acted as an antenna for
the first terminal. This is possible since the particular
USB-to-mains-PLT device draws its power supply from
the USB connection and not from the mains and thus
can still inject PLT signals. The mains wiring acted as
the antenna for the second terminal. It could also be
made to work (at lower capacity) with less obvious
‘antennas’ than the lights, e.g. by simply holding an
exposed pin of the plug of the ‘unplugged’ HomePlug
device.
2 November 2009 - 12:05pm
#5
Review here of new Belkin http://www.techtir.ie/forums/forumdisplay.php?f=54
It's just a "200Mbps" adaptor with a Gigabit Ethernet port.
No proper test of performance with CFL lights, Plug top chargers, distance, effect on DSL line if modem phone cable is near, or interference produced.
You can buy a crimp tool, a few drums of STP Cat5e cable and connectors for less than the price of these and have real gigabit ethernet.
2 November 2009 - 8:44pm
#6
I have a Netgear 200mbit Homeplug setup. Maximum throughput from PC upstairs to the NAS drive downstairs is about 2.6MB/s. Take from that what you will. What I do know is that it takes a long time to transfer media files to the NAS drive!
2 November 2009 - 11:09pm
#7
Surely the only reason they are getting away with it is because LW/MW/SW radio is now only a niche .
Any device causing RFI on VHF /UHF or satellite IF is likely to be dealt with more severely.
Complaints have to be made before action is taken!
3 November 2009 - 3:45pm
#8
Signal;2470 wrote:
Surely the only reason they are getting away with it is because LW/MW/SW radio is now only a niche.
On a global scale they are hardly a niche. There are at least 3 national MW stations in the UK, IIRC there's still areas of the UK with poor BBC4 FM coverage, MW commercial radio is still very common in the US, the LW band is far from empty and SW is an invaluable resource for those in remote areas.
I've also noticed LCD displays are terrible for causing AM radio interference (not like CRT monitors are bad enough).
3 November 2009 - 4:52pm
#9
Quote:
I've also noticed LCD displays are terrible for causing AM radio interference (not like CRT monitors are bad enough).
What you are hearing here is the Power supply driving the Illumination (usually a Cold cathode Tube or an Electroluminescent panel) or possibly the digital drive circuitry for the display. LCD technology it's self is very benign from an RF standpoint.
The EL illumination on the Sony ICF-SW07 Shortwave receiver interferes with it's own reception which certainly rates a crappy EMC design in my book!
.brendan
3 November 2009 - 7:02pm
#10
My laptop screen interferes with handheld on 145MHz on TRANSMIT, not just Receive! If it's within about 0.25 metre.
Some models of LCD TV have a metallic screen on face.
My Sony ICF-2001D Radio is fine, it has an unusable "press button to light up with feeble green LEDs" backlight. I've actually heard USA 80m Amateur stations on its whip in my dad's basement on it. In Limerick in bedroom on internal ferrite rod it gets 5 or 6 LW stations. I obviously live outside the City.
I think though the Multiplexing on video type LCDs can create interference, but small cold cathode / CFL type backlights in older portable LCDs TVs and some older laptops are terrible as they use step-up transformers very like miniature 1970s Hong Kong Radio audio transformers and no shielding. If you are brave you can convert some CFL/Cold cathode lit old laptops to use White LEDs. Small portable devices could be easier or harder. Soon is the season to get cheap LEDs.
3 November 2009 - 8:14pm
#11
Watty;2543 wrote:
My laptop screen interferes with handheld on 145MHz on TRANSMIT, not just Receive! If it's within about 0.25 metre.
Sounds like the handheld has it's own set of EMC compromises too then..
What do you think the ingress path might be ?
.brendan
3 November 2009 - 10:49pm
#12
Could be RF pickup on Audio/FM in, or at IF, if it uses IF conversion on TX. I have schematic someplace.
Alinco DJ-V5
Hmm... no TX IF.
It has discrete VHF & UHF VCOs with audio modulating the PLL feedback of Synth.
Buffer all, 2SC5065 or 2SC5066
uPC277?? pre Driver
Driver (can't read part, looks like VRF9745, poor Resolution PDF of service manual)
PA is 2SK2975
Then PIN diode switches (two sets, one VHF, one set UHF),
LPF + Duplexer in VHF path
and 3rd PIN switch to same Duplexer in UHF path.
If the RF interference was very strong the PIN diodes can act as modulators..
Nothing is IMMUNE, there are only degrees of immunity. Also if frequency is pulled off at rate only of 40KHz or so via pickup in microphone amp/PLL/VCO inputs, only at a low level, then you will get massive distorion on the very narrow Limerick Repeater IF filters.
The FT817ND is only affected on receive, if too close using Rubber Duck aerial. It's all metal, with some plastic on front panel. Quite a serious little radio. It out performs the Racal Syncal30 for intermod on receive and sensitivity. The Syncal30 was $12,000 each new. A lot used by Iraqi in 1st Gulf war, but mine is Grinel S.A. built and origin.
Reading up on the PLT there seems to be a 60dB gap between what the BBC thinks is acceptable and what the PLT home mains networking folk want.
The fact is that if a laptop screen creates the Max permitted interference level you can mitigate by moving away. I have an old DELL CRT that makes no interference at all on HF in my shack. The Laptop or a modern LCD screen or my philips TV PAL/RGB Monitor in the shack does noticeably raise the noise floor. If people have PLT / Homeplug/ comtrend you can't get away from it anywhere. "Normal" interference sources are usually quite localised.
A poorly filtered (most? or Many? of them in the past?) Dimmer switch is similar. It "transmits" all over the house. In fact the guy that looked after the Radio Museum in Cork (old Jail?) had a huge interference issue that was display lighting.
9 November 2009 - 2:59pm
#13
On The Register comments about the Belkin gigabit model (which links to this thread...) I found a reference to this -
http://www.leacom.fr/niv3uk-pays59-id66-Powerline_NetSocket_NetSocket200+_Euro.html
a Homeplug AV compliant device which claims to "embeds over current protection and a filter to protect the Powerline Communications (PLC) network from electrical noise generated by Consumer Electronic equipment."
Spec sheet here - http://www.leacom.fr/userfiles/file/66_uk.pdf
Anyone have any experience of this, any tests??
Bye, Barry
9 November 2009 - 3:57pm
#14
Snake oil
If you have a noisy CFL (electronic ballast bady filtered on lamp), Halogen lamp PSU, PC PSU or plug top PSU the only place filtering helps is at the noisy device, filtering at the PLT network node will not help at all. Think of car radio and interference from Alternator/ignition. Only removing the interfence at source works.
In real world PLT (Homeplug or its competitor) on home mains is very suspectible to interference and generates a huge amount itself. Real world speeds are 20Mbps to 80Mbps HALF duplex for ANY mains product, even if it has a Gigibit port.
WiFi speeds for one user can be 1/3rd of stated max speed, typically.
Multiple users of WiFi or Homeplug share the headline speed.
Cat5e cable supports 100Mbps FULL duplex per person or even 1Gbps. You use a Switch. So you can have 4 pairs of users doing 100MBps full duplex, that's 1200Mbps bandwidth at 75% utilisation compared with 25Mbps half duplex on a 200Mbps Homeplug!
Obviously any particular user is often limited by sustained throughput of the slowest hard drive at either end or the Broadband connection.
But you can't beat Cat5e (I use shielded for exposed cable or cable near mains) for a cheap reliable solution. We use WiFi too, with one 108Mbps Turbo G (supports a,b&g) and one basic 54Mbps 802.11g WiFi points. 802.11n frequently drops to 802.11g (54) or 802.11b(11) speeds if the Airpoint and client are not in same room.
All our fixed location PCs are on Cat5e. The Netbooks & Laptops are Cat5e or WiFi depending on location or application. The Wifi is handy for the two smart phones, PSP and Archos605 and the netbooks.
A fridge/freezer motor, vacuum cleaner, mains power tool etc will all affect Mains Networking. By definition the amount of interference any electrical device generates can be equal to the maximum signal of Mains Networking, as their signal counts as interference. The PLT / mains networking / comtrend / homeplug devices are in reality transmitters and receivers using DSL like signals up to 30MHz. They have not got a Transmitter licence or CE mark. Their approval is on basis of not causing more tan permitted interference.
But the approval/measurement process may be flawed as it was to test unintentional interference generation, not deliberate wideband transmission. The measurement method may also mean that many produce more RF outside the wiring than is intended by the people that drafted the legislation. The BBC seemed to think the interference is 30dB too high. See earlier posts.
22 February 2010 - 9:14am
#15
http://www.rsgb.org/news/statements.php?id=0036
Quote:
The Society has now completed a review of its options in conjunction with its lawyers and Counsel.From this review it is clear that there are a number of avenues to be further explored and until this work is complete the Society will defer for the time being the question of legal action.
So no legal action right now.
28 June 2010 - 3:05pm
#16
Current Power Networking AKA "homeplug" or "power line Networking", Network adaptors work up to 30MHz and may create and illegal amount of interference but be approved via a loop hole in laws.
But newer high speed plug units use more of the Radio band and can interfere not just with LW, MW, Shortwave, CB and Amateurs but FM radio, Mobile and Aircraft
6 July 2010 - 2:43pm
#17
PLT interference range competition!
See here at
emcia (opens new tab/window)
23 July 2010 - 11:43am
#18
Recent Ofcom report on PLT produced by "PA Consulting" seems largely to be a paper exercise and very flawed. See August 2010 Radcom.
Also
http://www.amateurradio.org.za/plcs...
Quote:
A PLT Project Team (PT) was formed in 2005 to
produce an amendment to CISPR 22 to cover special requirements
for PLT equipment and its first Committee Draft (CD) was
issued in February 2008 (doc. CISPR/I/257/CD). However, the
comments of 23 IEC members National Committees (NC) and the
European Broadcasting Union (EBU) showed insufficient support
for the selected approach (doc. CISPR/I/266A/CC) as only 6 NCs
supported the draft: Belgium, France, Israel, Italy, Spain and
Switzerland. Interestingly, the major European PLT technology
providers, developers and manufacturers reside in 5 of these
6 countries.
8 NCs strongly opposed the draft - Australia,
Austria, Cyprus, Denmark, Finland, South Africa, Sweden and the
United States of America - and some well-founded comments
revealed its true purpose: to camouflage an intended 18 dB
relaxation of the present PLT disturbance limits by introducing
a revised method of measurement with an estimated Longitudinal
Conversion Loss (LCL) of 24 dB in contrast to 6 dB in CISPR
22:2005.
At least 802.11b/g/n only interferes with one already congested band
Are ring mains better or worse than radial circuits for causing interference? Most of the original wiring in my home (built in late '60s) is all rings.