ARRL

Forum Home - Rules - Help - Login - Forgot Password
Members can access, post and reply to the forums below. Before you do, please first read the RULES.

Linksys WRT54GS 20 Meter Interference

Jul 16th 2011, 14:17

N5LB

Joined: Apr 4th 1998, 00:00
Total Topics: 0
Total Posts: 0
Title summarizes the issue. Here are the details.

This router is about 20 feet from my TS850. The router is connected to the Comcast cable modem. The router provides WiFi in the house as well as two wired Ethernet connections, on for the ham station and one for my office computer.

This router basically blankets the 20m band with interference with broad peaks (3-4 kHz) every 14-16 kHz. With all of the Ethernet cables disconnected including the connection to the cable modem the interference remains. Powering down the router stops the interference.

The interference is basically the across the HF spectrum. Unfortunately one of the peaks is 14.070 making digital in that area impossible.

Has anyone tamed this Linksys router? Or has the experience been such as to require a new router and if so what has worked best?

73

N5LB, Lionel
Jul 16th 2011, 15:25

N5LB

Joined: Apr 4th 1998, 00:00
Total Topics: 0
Total Posts: 0
Here is an update:
Putting the TS850 in AM and observing the fldigi waterfall showed clear lines 120 hz spacing. Surprise, surprise. I added 3 clamp on RFI cores on the power cable (really 2 x #20 or so wires) by making multiple loops though the cores (snap ons).

Result: One core dropped the noise 10dB three dropped it about 20 dB. Clearly the wall wart is generating a huge broadband signal. More work can be done to further cut the noise, including getting a better power supply for the Linksys.

Lesson learned: keep a few snap on chokes around the shack and try the simple solutions first.

Cheers and back on 14070. I didn't realize how many BPSK signals were really there!!
Jul 18th 2011, 12:12

w1rfi

Super Moderator

Joined: Apr 4th 1998, 00:00
Total Topics: 0
Total Posts: 0
What you have made with those ferrite cores is a common-mode choke.

Although you did get some supression of noise from the clamp-on ferrites, most of the time, on HF, multiple turns of wire are needed to provide enough supression. The best all-around material for making ferrites is #31 material. For the output wire of the wall wart, I'd recommend and FT-140 size core, onto which you should be able to get about 10 turns of wire. You could accomplish the same general effect with about 10 of the clamp-on ferrites. A core made from #43 material would also work pretty well. Of course, once you have supressed the noise below the level of other noises you are probably also hearing at your station, more supression isn't needed.

If the ones you tried are the square-shaped snap-together cores once sold by Radio Shack, from the reports I've received, they really don't work as well as "real" ferrite toroidal cores.

If you have some "unknown" ferrite material, it's okay to try it, although if it doesn't work, don't assume that a common-mode choke won't work. You may be able to get some additional suppression by ordering and using an FT-140-31 ferrite core. Amidon, Dan's Small Parts and Palomar are 3 vendors that come to mind.

The power supply may also be putting noise onto the AC mains, and an additional filter may be needed there. In this case, I would recommend first trying a differential-mode filter. Many "surge supressors" also indicate that they are EMI filters, and one of those may help.

And of course, as you have noted, a different power supply may work, although watch those voltages and polarities!

At this point, the ARRL Lab has not (yet) tested any of the surge supressors/EMI filters for effectiveness, so if anyone else has, or can provide some information about their experiences, please post here.

73,
Ed Hare, W1RFI
ARRL Lab
Jul 26th 2011, 17:43

NX9A

Joined: Jan 12th 2010, 11:05
Total Topics: 0
Total Posts: 0
This sounds very similar to the issue I've been having with an older WRT54G. A very clear audio tone at multiple frequencies across the
20 meter band. Probably elsewhere as well that I have not tested yet.
I'll see what I've got for ferrites and see if it helps much. Thanks!
Jul 26th 2011, 22:46

W1RFIAdmin

Joined: Jul 25th 2011, 14:25
Total Topics: 0
Total Posts: 0
I must admit that i am very surprised at the attention this thread is getting. There have been over 250 views, roughly four times the number on any other post. I've been watching the counter and it is steadily increasing, so it's not an anomaly.

One of the things that the ARRL staff plans to do is to take the discussions here, winnow out the essence of what hams may need to know from good end results and use that to improve (or write) FAQs for its other pages.

Keep the good information flowing. It will help us build a world-class set of techical information. Although the HQ staff does some great work, I've always noted that there are roughly 700,000 licensed hams and 4 engineers in the ARRL Lab. I can guarantee that those 700,000 hams can collectively accomplish more than the 4 guys in the ARRL Lab! Our job will be to collect the knowledge that results and make sure it's available to Amateur Radio.

73,
Ed Hare, W1RFI
ARRL Lab
Technology forums moderator
Jul 28th 2011, 22:40

AB7ZU

Joined: Apr 4th 1998, 00:00
Total Topics: 0
Total Posts: 0
Ed, I think one of the reasons for the numerous visits to this thread is that many-a-router are probably doing the same thing.... I know mine did and it is a completely different manufacturer (netgear). Again, the wall wart was the culprit. The router itself doesn't seem to be producing anything I can identify. I could be wrong, though.

Anyway, what ham doesn't own at least one router these days? he he he. Very applicable stuff here. This is a great forum idea.

Thanks
Mike AB7ZU

Jul 30th 2011, 11:10

KB0HAE

Joined: Apr 4th 1998, 00:00
Total Topics: 0
Total Posts: 0
Hi Guys. I have found that many routers and hubs use cheaply made switching power supplies. These are usually the culprit when a router or hub produces excessive interference. My solution here is to use a non-switching supply to replace the cheap switching wall wart that came with my router.

I have two hubs (its only a 4 port router) that each have non-switching wall wart power supplies. I do have 5 snap-on ferrites on the cable from the cable modem to the router, all next to the cable modem end of the cable.

I am lucky in that I have been able to eliminate most interference generated by anything in my home, and the neighbors homes are at least twice as far as is normal here. I still get some noise on isolated frequencies, but nothing too bad.

I did find that using a non-IBM supply on the Thinkpad T42 that is on the shack desk for logging and digital modes caused a lot of interference. Replacing it with the original IBM supply cured that interference.

The chargers for my Kenwood TH-G71A, Kenwood TH-K2AT, and Wouxun KG-UVD1P are all switching types. They are usually not plugged in unless they are being used, usually at night when we are all asleep.

I am sure that many others could tell stories of horrendous interference caused by cheaply made switching power supplies, especially wall-warts.

I think it is long past time for the FCC to set much stricter limits on the amount of RFI that home electronics (especially those using cheaply made switching power supplies) are allowed to produce! I'm not holding my breath though...
Sep 17th 2011, 19:43

K8WHB

Joined: Jan 11th 2011, 19:02
Total Topics: 0
Total Posts: 0
Gentlemen,

I've had similar issues with the Linksys router I used to have (a power surge caused that router to dump the EEPROM & go "brain dead") and with the new Netgear WGR614 I got to replace the Linksys.

As KB0HAE mentioned I was also able to reduce the EMI by about 75% by substituting a linear regulator "wall wart" I had in the junk box (from an old Motorola cable modem) for the switching regulator one Netgear had provided. It appears from talking to a number of my computer system engineer friends that most consumer routers have switched to switching regulator power supplies both due to cost and the manufacturers chasing the EPA's "Energy Star" rating.

The other switchers I have in the shack for my Toshiba laptop, the Kenwood KPS-15 that powers the TM-D710A and the charger for my Wouxun KG-UV2D all are quiet, at least on 6 meters & above (I'm not presently running any HF). I'll bet I'll have to revisit this issue next year when I get the HF station up and running.
Dec 4th 2011, 16:17

AC0XU

Joined: Oct 30th 2011, 02:35
Total Topics: 0
Total Posts: 0
My neighbor's COMCAST cable seems to self-broadcast RFI from 3 MHz up to about 20MHz, causing severe HAM band interference on 75M to 17M. It pegs my Icom RX-7 S-meter from a block away. The neighbor let me in his house, and even with all his cable equipment powered off and disconnected, just the bare cable coming out of the wall is transmitting the signal!

In this case, the peaks are spaced about every 60 KHz from 3 MHz to 20 MHz. On 20M, they are about 20-30dB above the ambient noise floor, totally drowning out comms signals.

The only saving grace is that the RFI is bursty - very short pulses with a repetition interval of about 1.3 seconds. The result is corrupted, but usually copyable PSK31 and the like.

My best guess is that there is some bad active component in the cable system that is going nonlinear and generating this wideband noise. Why it would have this periodic/bursty character, I don't know.

I am trying to work with Comcast, but their attitude is that they won't even talk with me. They only want to deal with the homeowner where the RFI is being generated.


Back to Top