Ethernet Switch RF Problem

[Mexicanuck]

Shaw Cable was by today saying they had identified a problem in our area and it appeared to come from our house. Upon investigation it appears to be RF interference from a D-Link 8 port gigabit switch, so now I am looking for a new switch.

Can anyone recommend an 8 port switch that is likely to not have an RF interference problem?

Thanks in advance.

[recneps77]

I'm using a trendnet switch I got a couple years back. It's a solid unit with all metal case and good performance.
They're regularly on sale these days for $25-30 IIRC.

My switch is literally next to all my OTA gear and I haven't noticed any interference issues.

[hugh]

Had no problem with Netgear gigabit switches. I have the 8 port home version and an 8 port Prosafe POE version.

BTW, have you contacted D-Link? They might send you a replacement

[Alex Wright]

Well if the source of interference is the wireless switch that you are using, then the best option would be to buy a wireless switch which is tuned to any other set of frequency then of 8Mhz which your current wireless switch is tuned too.

[JamesK]

^^^^
WiFi uses frequencies well above those used on cable, so it's unlikely that's the cause. However, defective equipment can cause interference over a wide range of frequencies. Did the OP even mention wireless?

[Mexicanuck]

The problem wasn't with wireless interference. Our wifi is fine.

Apparently the switch is putting of radiation that is causing problems with cable signals, affecting others in the neighbourhood. The technician had a hand-held graphic meter. He showed me the display with the switch unplugged. It showed a fairly straight line near the bottom of the display.

He then plugged in the switch. The line quickly became more ragged and jumped about three-quarters of the way up the display. He said that was the interference. We swapped power supplies (see below). The interference was still there. I disconnected another switch that is in a different room and the interference was still there. We unplugged the switch again and the line again evened out and dropped to near the bottom of the display.

I think it's the switch.

I may have caused my own problem. When we changed the power supply, I realized I had bee using the wrong PS. The PS for the switch was 5V 2A output. The PS I had been using was 7.5V 1000 mA. I put in the correct PS, but the problem remained.

[ScaryBob]

The PS for the switch was 5V 2A output. The PS I had been using was 7.5V 1000 mA.

The switch should be robust enough to handle that. Output from wall warts is notoriously variable and most devices have voltage regulation circuitry built in. If the required P/S is a regulated switching unit, that could cause a problem but not necessarily so.

[majortom]

I didn't see the model # mentioned. Which is it?
I have heard a certain Model Dlink wireless router (maybe it was a switch and not a router???) is notorious for causing out of band interference due to a defective component.
Pretty sure they know all about the issue and have been replacing them for customers.
I can dig up the specific info I saw sometime next week if you'd like.

[Mexicanuck]

This is a DLink Green 1008D.

Here is a link.

http://www.dlink.ca/products/?pid=230

It is past "end of life". I'm guessing that is corporate-speak for "don't bother us".

[majortom]

thanks, I'll dig up the info next week and post back here. That may be the one (ie how many can there possibly be with the same issue).
It mentioned they had traced the issue back to a defective capacitor and clearly said they are taking care of existing customers.

[scotta]

Originally Posted by ScaryBob

Output from wall warts is notoriously variable and most devices have voltage regulation circuitry built in.

That's true for older equipment but not for most current stuff.

Almost all recent wall warts contain switching regulators, which makes them small, light, efficient and well regulated. The device being powered may rely on this regulation and not contain regulation circuitry of it's own (unless it requires a different or multiple voltages, which it then has to convert from the wall wart's source voltage).