Posted: Thu Jan 13, 2011 10:52 Post subject: WNDR3700 - What to do ?
Ok bear with me, a complete novice here..
I have a Netgear WNDR3700 V1 router (atheros based I believe) but my PC's and laptops are quite away away and only get 1 bar signal when connected.
The router is used primarily for its wireless capabilities from all over the house but my office is at the furthest point away.
The WNDR3700 cannot be moved from where it is.
I need to get better coverage/throughput so someone suggested loading DD-WRT firmware on the router as you can "tweak" the power on the radio's to give a better signal.
Is this feasable or would it be better to add a range extender such as this ?
At this time I can't recommend installing dd-wrt on your wndr3700, dd-wrt just isn't ready for it yet. Either go with the range extender or get a Broadcom based wireless router.
Posted: Fri Jan 14, 2011 16:40 Post subject: How to achieve above average 5Ghz performance.
I've been troubleshooting my install of DD-WRT on my WNDR3700 for about the past week.
I'm running the WNDR3700 with DD-WRT. I have 2 Gigabit LAN clients, 1 wireless-G client, 1 wireless-N Dual band client (usually on 5Ghz.)
I monitor connection strength client side using the Xirrus wi-fi inspector. http://www.xirrus.com/library/wifitools.php. I monitor connection strength router side using DD-WRT Status tab.
I have a great connection on the Wireless G client (2.4Ghz) and the Wireless N client (2.4Ghz or 5Ghz). The 2.4Ghz is a bit more powerful on G or N in all tests, but only marginally.
According to Xirrus (on the clients), the signal strength on each radio 5m away is between -50dbm and -40dmb . I always have full green bars reported on both clients on both radios. Xirrus doesn't monitor noise.
According to DD-WRT, signal on the 2.4Ghz radio is between -50 dbm and -40dbm. Noise is between -105dbm and -120dbm. SNR is between 50 and 60.
According to DD-WRT, signal on the 5Ghz band is between -60dbm and -50dbm. Noise is solid at -96dbm. The SNR is usually between 40 and 50. Marginally less than 2.4Ghz.
I don't monitor the "signal quality" bar since it often stays around 50% even with a SNR of 50.
*In order to achieve this signal strength on the 5Ghz radio, I have noticed 2 "bugs":
1.) Auto channel selection doesn't appear to be working correctly. "Wireless Channel" should be manually selected. Auto will result in about 50% less signal strength.
2.) Whenever you make a change in DD-WRT and "Apply Settings" it degrades the 5Ghz radio strength about 50%. Go back to Wireless settings for ath1, switch the channel > apply > switch back > apply.*
Please forgive the double post in the 194 pg. WNDR3700 thread.
Joined: 06 Feb 2010 Posts: 7401 Location: Little Rock
Posted: Sat Jan 15, 2011 0:47 Post subject:
68rustang wrote:
buddee wrote:
I don't understand statements like that though, why get a broadcom based unit...
Now i dont have the 3700 to say how it is...
Because dd-wrt sucks on the WNDR3700. Wireless speeds are generally less than half of what they are with stock Netgear firmware.
I think im gonna purchase one tonight and see how it is, i read to many of these styled posts about the buffalo wzr-hp-g300nh as well, i bought one of those, and its probably one of the best units i have so i am going to have to see this for myself. _________________ Wireless N Config | Linking Routers | DD-WRT Wiki | DD-WRT Builds | Peacock - Broadcom FAQ
Joined: 05 Dec 2009 Posts: 377 Location: Toronto, Canada
Posted: Sat Jan 15, 2011 0:54 Post subject:
I ran DD-WRT for just a few hours on my WNDR3700 and flashed back to stock. Wireless performance was just awful and not the performance I expected at all from my ~$200 device.
I think the performance is awesome with DD-WRT as long as I manually specify the 5Ghz broadcast channel and restart the radio after any other settings change is made in DD. Even if the router is rebooted, the 5Ghz needs to be turned off and back on or channel switched back and forth and then re-applied..
There is a version of OpenWRT already pre-compiled on the net. Both radios are enabled by default and the lights are set to factory (green for gigabit, solid when connected but not transmitting).
I find OpenWRT's interface isn't as intuitive and it lacks a few features that DD-WRT has. I can post the link later.
Previously I had the Cisco E3000 and if you know anything about electronics, heat = death. I doubt those units will last more than a year. Yes, I did put rubber feet on the bottom. They ingeniously chose a chipset with a short life span, I guess it's just part of the "designed for the dump" theory.
Joined: 06 Feb 2010 Posts: 7401 Location: Little Rock
Posted: Sat Jan 15, 2011 3:32 Post subject:
SonicFrequency wrote:
Previously I had the Cisco E3000 and if you know anything about electronics, heat = death. I doubt those units will last more than a year. Yes, I did put rubber feet on the bottom. They ingeniously chose a chipset with a short life span, I guess it's just part of the "designed for the dump" theory.
I have to disagree with this here on one note, its not the chipset they choose, its the case design that makes it where the unit can't breathe.
I have an Asus RT-N16 with the same chipset and because its case is designed somewhat better, it stays cooler than my E3000. The hottest unit i have is a wrt320n, now those units get HOT, but mine has lasted more than a year now and still works fine. _________________ Wireless N Config | Linking Routers | DD-WRT Wiki | DD-WRT Builds | Peacock - Broadcom FAQ
The bottom of the E3000 is perforated, it should allow for decent airflow. With 1/2 inch feet on it (adhesive backed rubber feet from Lowes), it still gets very hot.
The bottom of the WNDR3700 is not perforated. With 1/8 inch feet (factory attached), it barely gets warm.
I'm just trying to be factual since I have tested both routers.
Joined: 06 Feb 2010 Posts: 7401 Location: Little Rock
Posted: Sat Jan 15, 2011 5:21 Post subject:
SonicFrequency wrote:
The bottom of the E3000 is perforated, it should allow for decent airflow. With 1/2 inch feet on it (adhesive backed rubber feet from Lowes), it still gets very hot.
The bottom of the WNDR3700 is not perforated. With 1/8 inch feet (factory attached), it barely gets warm.
I'm just trying to be factual since I have tested both routers.
Ok lemme ask it like this then.
Do you believe that the thin design case on the newer linksys (yes the one that you put the feet on to help for cooling) gives better spacing airflow then the box design of the 3700?
Honestly I don't think either case design is any better than the other. They both look like they allow for moderate airflow. I'm no engineer though. I couldn't tell if it's chipset or case design that causes the excessive heat. It might not be the chipset. The Linksys doesn't seem to have much space between the mobo and the top of the case. Maybe the case getting hot is actually a good thing...meaning the case is dissipating the heat...
Joined: 06 Feb 2010 Posts: 7401 Location: Little Rock
Posted: Sat Jan 15, 2011 6:08 Post subject:
It takes no engineer to know, the chip causes the heat, but also it doesn't take an engineer to know that air cools off the further away from the heat source that it is, so if there is no space for the airflow then the hot air is trapped, and creates somewhat of a convection oven.
My RT-N16 and this WNDR3700 and a few other units i have use the box design, so they have actual spacing away from the board to allow for airflow so that the air inside cools off a little, which helps, the RT-N16 has the perforated design as well, on both sides instead of just the bottom where it pretty much does very little to help cooling (heat rises) and like i said, it stays MUCH cooler than my E3000 and they both use the same chipset.
I never disagreed or even had a speck of doubt of what causes the heat, my debate was on the design of the newer linksys cases don't help it in the factoring of helping cool the unit off.
Ten-four. I was just debating the possibility that Cisco was using the case as a giant heatsink. I mean, there is nearly no room for air at all inside, I wonder if the top of the case contacts the chipset to conduct heat away from the components? I thought plastic was an insulator. Im sure if this was their intent the top would at least be aluminum.
