Mini displayport stops 802.11g from working? how? why?
February 9, 2010 9:53 AM Subscribe
Mini displayport to DVI adapter stops 802.11g from working? What?! We have a Macbook5,1 which is plugged into a 21" Dell LCD using the Monoprice mini displayport to DVI adapter. When the displayport is unplugged wireless works perfectly. When it's plugged in, the wireless NIC remains associated to the WAP but there is instant 100% packet loss.
The WAP is a Linksys WRT54GL which is located about one meter from the laptop. We have three other laptops in the house and various iPhones and other 802.11g devices, none of which have connectivity problems even as far out as 30 meters.
The laptop is a Macbook5,1 running the latest updated version of Snow Leopard. When the mini displayport to DVI adapter is plugged in and the 21" LCD is turned on, the laptop can see various WAPs in the area but cannot associate to them. If I unplug the display, connect the laptop to a WAP and run packet loss tests to our NAT gateway, there is an average 0.5 to 1.1ms latency and 0% packet loss. The instant I plug back in the displayport adapter there is 100% packet loss and the laptop falls off the net.
If I leave the displayport adapter plugged into the laptop but power off the 21" LCD, there is about 60% packet loss to our WRT54GL's NAT gateway IP (192.168.101.1). That is particularly weird.
This seems like some kind of weird voodoo - any chance we have a defective macbook motherboard that glitches when the video output circuitry is used at the same time as the wireless NIC? Any other ideas?
The WAP is a Linksys WRT54GL which is located about one meter from the laptop. We have three other laptops in the house and various iPhones and other 802.11g devices, none of which have connectivity problems even as far out as 30 meters.
The laptop is a Macbook5,1 running the latest updated version of Snow Leopard. When the mini displayport to DVI adapter is plugged in and the 21" LCD is turned on, the laptop can see various WAPs in the area but cannot associate to them. If I unplug the display, connect the laptop to a WAP and run packet loss tests to our NAT gateway, there is an average 0.5 to 1.1ms latency and 0% packet loss. The instant I plug back in the displayport adapter there is 100% packet loss and the laptop falls off the net.
If I leave the displayport adapter plugged into the laptop but power off the 21" LCD, there is about 60% packet loss to our WRT54GL's NAT gateway IP (192.168.101.1). That is particularly weird.
This seems like some kind of weird voodoo - any chance we have a defective macbook motherboard that glitches when the video output circuitry is used at the same time as the wireless NIC? Any other ideas?
following up on the grounding idea.. what if you just have the adapter plugged in to the computer but not to the monitor? is the monitor connected to a properly grounded outlet? (you might want to get a tester for that - it physically having a ground pin or a grounding adapter doesn't necessarily mean it's actually grounded.)
posted by mrg at 10:44 AM on February 9, 2010
posted by mrg at 10:44 AM on February 9, 2010
Response by poster: That was truly odd. We changed the WAP from channel 3 to channel 11 of the 2.4GHz spectrum and now everything works fine. There -should not- be any significant interference in the spectrum in our area, since our flat has concrete walls and floor... There are four other WAPs visible from a laptop but they all show up with 1 bar of signal strength, even the open one without any security is difficult to associate to unless we stand with the laptop at a window and rotate it at a certain angle. Maybe someone bought themselves a 500mW 2.4GHz amplifier and aimed it at our flat...
My theory is that the power brick for the Dell 2001FP monitor is putting out some kind of interference that interferes with channel 3, located as it is about 20 cm behind the laptop by the base of the monitor. Grounding should not be a problem as the AC side of the Dell's power brick (110V-240V to 19V DC) has a 2-pin power plug, there's no way for it to be grounded as far as I know... If the Dell display has a ground fault and is shorting internally to ground through the macbook it's certainly not shocking us. The Dell 21", Macbook, two other laptops, a cheap inkjet printer, a WAP, a DSL modem, two external hard drives etc are plugged into a pair of 6-receptacle grounded UK type surge protectors which in turn are plugged into a properly grounded UK dual socket wall receptacle.
posted by thewalrus at 10:52 AM on February 9, 2010
My theory is that the power brick for the Dell 2001FP monitor is putting out some kind of interference that interferes with channel 3, located as it is about 20 cm behind the laptop by the base of the monitor. Grounding should not be a problem as the AC side of the Dell's power brick (110V-240V to 19V DC) has a 2-pin power plug, there's no way for it to be grounded as far as I know... If the Dell display has a ground fault and is shorting internally to ground through the macbook it's certainly not shocking us. The Dell 21", Macbook, two other laptops, a cheap inkjet printer, a WAP, a DSL modem, two external hard drives etc are plugged into a pair of 6-receptacle grounded UK type surge protectors which in turn are plugged into a properly grounded UK dual socket wall receptacle.
posted by thewalrus at 10:52 AM on February 9, 2010
There are four other WAPs visible from a laptop but they all show up with 1 bar of signal strength
Your wireless-g application isnt going to show you wireless-n interference or anything from cordless phones, microwaves, wireless video, etc. Not to mention wireless-g networks that dont send SSIDs. There's a lot out there that's pretty much invisible to you but can cause you a lot of problems.
My theory is that the power brick for the Dell 2001FP monitor is putting out some kind of interference that interferes with channel 3
That could be true too. Wireless-b/g/n exists on the ISM spectrum, which is pretty much a safe dumping ground for noise from electronics. Putting wireless communication there is something of a hack that usually works, until people run into situations like yours.
posted by damn dirty ape at 11:45 AM on February 9, 2010
Your wireless-g application isnt going to show you wireless-n interference or anything from cordless phones, microwaves, wireless video, etc. Not to mention wireless-g networks that dont send SSIDs. There's a lot out there that's pretty much invisible to you but can cause you a lot of problems.
My theory is that the power brick for the Dell 2001FP monitor is putting out some kind of interference that interferes with channel 3
That could be true too. Wireless-b/g/n exists on the ISM spectrum, which is pretty much a safe dumping ground for noise from electronics. Putting wireless communication there is something of a hack that usually works, until people run into situations like yours.
posted by damn dirty ape at 11:45 AM on February 9, 2010
i really hate wifi for any serious application. i made my way through powerline networking (never lived up to the speed promises) and am finally using MoCA. works great - almost 100mbit over coax, reliably. no problem with microwaves, telephones, phases of the moon or barking dogs. just works.
posted by joeblough at 11:50 AM on February 9, 2010
posted by joeblough at 11:50 AM on February 9, 2010
This thread is closed to new comments.
posted by joeblough at 10:39 AM on February 9, 2010