Wireless Driver Issues & Settings for N speeds

Discussion in 'Tomato Firmware' started by Toastman, Sep 24, 2011.

  1. pharma

    pharma Network Guru Member

    While I don't have a E3000, I do have an Atheros radio in one of the laptops and I have no problem connecting to the E4200 -- you probably won't have any problems either. One disadvantage with Atheros is the availability of their drivers. One link you might find useful is below where basically all routers are tested over a variety of tests with some very interesting results. The nice thing I like about the Intel Centrino E6200+ is the wireless adapter driver has a setting for "preferred radio", or something like that. If you have a preference for connecting at 5ghz you can select that choice and it seems to try connecting at that first and will fall back to 2.4 ghz if the signal is too weak. The other choices are "No preference" and "2.4 ghz".

  2. bucher

    bucher Networkin' Nut Member

    Is it recommended to use the same SSID for both 2.4 and 5?
  3. jsmiddleton4

    jsmiddleton4 Network Guru Member

    "same SSID for both 2.4 and 5"

    Depends. When I have my setup to only use the 5 ghz channel for router to router I have the 5 ghz channel on a different ssid than 2.4. Clients connect only on the 2.4 ghz based on having that SSID be the only ones the clients pick. Keeps router to router traffic on its own channel.

    Right now however I have it set so both channels are on same SSID.
  4. jsmiddleton4

    jsmiddleton4 Network Guru Member

    Wow the E4200 V2 is no where near the performer the V1 is....
  5. Toastman

    Toastman Super Moderator Staff Member Member

    Just to remind everyone to try to use this thread for wireless issues, it's all reverting back to the various modders' release threads again :eek:


    Forum Moderator
  6. jsmiddleton4

    jsmiddleton4 Network Guru Member

    And didn't someone post a newer version of the RT-N driver already?
  7. Toastman

    Toastman Super Moderator Staff Member Member

    I thought someone posted that the new RT-N16 will use the same RT-N driver that we have now :-(

    Also I see there IS a new driver in one of the newer routers, looks like a progression of the "N" driver - not too useful at this moment though.
  8. Hogan773

    Hogan773 Networkin' Nut Member

    Thanks - I had seen these charts before and SmallNetBldr is a great site. The only thing I never can find is any tests or reviews of the WLAN CARDS (not the routers). Perhaps the reality is that the performance of the WLAN cards within a given tier (1x2, 2x2, etc) are similar. Or maybe its because the vast majority of people focus mostly or solely on the ROUTER and just take whatever they get when it comes to the other end for the laptop.

    Unless someone can convince me that Atheros is significantly BETTER, I'll probably stick with Intel just because I feel the drivers will be easier to find and I know the laptop is already working with a cheapie Intel in it.
  9. Toastman

    Toastman Super Moderator Staff Member Member

    Mito - The RT-N16 has a SOC chip which has support for two radios, but be aware that the rest of the support hardware for the second (unused) radio is not included in the router circuitry. You can't use this router on 5GHz.

    Hogan773 - if I were to have a choice, I'd ban Intel wireless clients from coming within half a kilometer of any of my sites :) Looking at the thousands of complaints on Google, I think an awful lot of people agree with me. If even the cheapest of the cheap Chinese wireless manufacturers can get it right, I can't imagine why Intel seem unable to make anything that works properly with other people's hardware.

    Think about this. I've seen certainly 5 to 10,000 different residents over the last 3 or 4 years move in and out of several apartment blocks. I haven't been counting - but that's a hell of a lot of machines. And when I get a call from the admin staff because someone has a problem guess whose wireless card they are invariably using?

    There's a famous IT store which is the place to go for anything PC related, repairs, knockoffs, whatever. Go look at the "Mr. Fixit" stalls and you see cardboards boxes on the floor full of discarded Intel wireless cards. Go figure ...

    The simple fact is - Atheros, Ralink, the cheapest stuff you plug in and it just works. Period. Intel stuff - you plug in and it gives you grief.

    Up to you which scenario you like best :D
  10. jsmiddleton4

    jsmiddleton4 Network Guru Member

    "Unless someone can convince me that Atheros is significantly BETTER"

    Not sure how one would do that. What I can tell you is the laptop we have with the Atheros based chipset connected at 300mhz by default and required no fooling around. Did and does so without WMM being enabled in the router. The Intel based ones require fiddling and fooling around and making sure stuff like WMM is turned on in the router. All kinds of little odds and ends stuff that if you don't know about you can't get it to connect at anything other than 54G. In other words out of the box defaults usually require tweaking. I don't have a Ralink based chipset laptop wlan card to try but have read those are getting rave reviews as well. Then there's the psp power polling stuff that Intel does that you end up having to turn off or you drop connectivity.

    For me IF I have a choice when buying a laptop I'd stay away from Intel's WLAN cards. Often you don't have a choice.

    "The nice thing I like about the Intel Centrino E6200+ is the wireless adapter driver has a setting for "preferred radio", or something like that."

    Broadcom driver does same.
  11. bucher

    bucher Networkin' Nut Member

    Not to beat a dead horse, but I recently got my E3000 in the mail, and was wondering if anyone can help compare the stability/performance of 5ghz with RT vs RT-N firmware? It seems like some people are using RT-N on E3000 successfully. I'm noticing lackluster performance with 5ghz. I live in a smaller two bedroom home that was built in 1929 and has plaster walls. My router is in my living room in the front of the house and in my bedroom, I'm getting like 18mbit in my bedroom, probably 30 feet away. If I use 2.4ghz, I get much better signal and range. I know by nature that 5ghz does not have the penetration that 2.4ghz has, but I wondered if RT-N would be any better.
  12. jsmiddleton4

    jsmiddleton4 Network Guru Member

    "RT-N on E3000 successfully."

    Define "successfully". Does it work? Mostly. Some stuff like router to router tx/rx speeds, etc., do not report correctly. Any difference between wireless client connections with "N" version compared to "non-N" version? None that I could tell. For me for all functions the "non-N" version works best. I did not test vlan or vpn stuff and compare versions.
  13. jsmiddleton4

    jsmiddleton4 Network Guru Member

    "I live in a smaller two bedroom home that was built in 1929 and has plaster walls."

    I live in 1800 sq foot house. Combination of brick and sheetrock. I get N speeds everywhere. Slowest is 150. You might want to look at your roaming, etc., preferences in your wlan driver for your client.
  14. bripab007

    bripab007 Network Guru Member

    With the E3000, you should be using the RT build, not the RT-N build. The RT build will support simultaneous dual-band on the E3000 just fine. The E4200 5GHz (and perhaps others) requires use of somewhat buggy RT-N firmware.

    After recently purchasing an E3000, I, too, was confused by the "for use with newer dual-band routers such as E4200" verbiage while looking at the different firmware versions. But I eventually found that RT is the one to use with the E3000. Only at the far end of my house is there a significant different in signal between the 2.4 and 5.0Ghz bands.
  15. bucher

    bucher Networkin' Nut Member

    5ghz N speeds everywhere? 2.4ghz is crowded but I get much better speeds than with 5ghz. I barely get a signal in my bedroom with 5ghz.
  16. jsmiddleton4

    jsmiddleton4 Network Guru Member

    "5ghz N speeds everywhere?"

    Yep. For those 5ghz devices I have. I have 2 laptops that are 2.4ghz N and 2 that are dual band type.
  17. bucher

    bucher Networkin' Nut Member

    Maybe I have lead based paint on the walls still.
  18. Hogan773

    Hogan773 Networkin' Nut Member

    Where do I read about/find these Atheros or Ralink (others I should consider too?) products? Unlike routers, I don't see SmallNetBuilder or Anandtech posting "New Review Ralink XYZ62772". Its all about the routers. So I search and tunnel through forums to try to get info. Basically thus far have seen a lot of people praising the Intel 6200/6300 as making a noticeable impact when they upgrade (from their existing laptop cards). And benefit that you can easily download the drivers right from Intel. And others like you and Toast saying to stay away from Intel. But I get the sense that this is more of an OEM product category so it is harder for Joe Consumer (me) to go out and choose for my upgrade.

    The whole reason I'm into this "project" is because I got a new E3000 yet I'm annoyed by the lackluster speeds on my laptop "N" card which is an Intel 1000. I may be chasing numbers and ghosts but dammit I want to see that Speedtest needle hit 20 Mbps on my laptop with my shiny new router, and I'm now thinking that I need 5ghz to get there and/or a better WLAN card since my 2.4ghz neighborhood soup is too thick.

    And then there is the whole issue of counterfeit WLAN cards on Ebay......yay! I just got hosed on a counterfeit PS3 controller and I was amazed at how well done it was in general (even though it didn't hold a battery charge). Now I have to worry that my well-researched WLAN card buy will end up being some Chinese fake......

    me and my "projects"........:eek:
  19. jsmiddleton4

    jsmiddleton4 Network Guru Member

    "So I search and tunnel through forums to try to get info."

    About the only way to do it. Go to laptop maker's user forums and read. Go to hardware forums like dsl reports and read. It isn't hard to find issues with Intel chipset wlan cards. They are in abundance.

    And by the way I'm not bashing Intel. It is very likely the case that Intel, because after all they are Intel, is following the standards exactly. Their products and mostly the drivers could be designed spot on with the engineering specifications. The other companies could easily be fudging the deal and their engineering, for both hardware and drivers, is done with some wiggle room. End users see easier to use products because of that wiggle room. When truth be told it is Intel that is actually doing the thing "right".

    As an end user it comes down to stability and ease of use BOTH. That is where Intel loses out. The other companies can put out products that work. They just work. Would we have the same frustrations with all products if they were following standards and protocol guidelines exactly? Maybe and probably. But the reality of products in the public's grasp is that wiggle room tends to win. At least for now.

    And it isn't just in this kind of product. Don't get me started on the frustration home theater stuff creates when things like HDMI is engineered tightly to exacting standards. The litany of hdmi issues is long when one product allows for some wiggle room and another, Panasonic anyone, is designed to exact standards. And again wiggle room wins. Folks complain about the Panasonics but the problem is really created by wiggle room engineering.
  20. buggage

    buggage Network Guru Member

    Any possibility of a RT-N test build with the new driver?
  21. pharma

    pharma Network Guru Member

    I could be wrong but up to now the major contributor towards incorporating "New Drivers" has been Fedor (TeddyBear). I believe there was a personal tragedy last year and he took some time away from development. Hopefully one day he will return and rejoin the development of TomatoUSB.

  22. pharma

    pharma Network Guru Member

    Sure you saw it ... SmallNetBldr did have a review of nine different types of Wireless 802.11n Adapters. It was done back in 2010 and since that review he adopted the best performing to use in his tests (Intel 5300). However, in his forum the review thread is an interesting read regarding the 5300 and it's successor 6300, and with the most recent comments dated a week ago you might gain alittle more insight there.

    G'luck ...
  23. Toastman

    Toastman Super Moderator Staff Member Member

    pharma - The smallnetbuilder review tested selected adapters that luckily worked with the router used. It unfortunately, as I recall, did not address or even mention the fact that the Intel clients have problems working with many routers.

    buggage - Given the issues we have just been through with (a) the reversion to driver and the many varied bad experiences with it, and (b) the RT-N driver being used for the E3000 where it isn't appropriate, I would be very loath to use this newer driver that has been mentioned on the forums. Generally, newer does not mean better, and people should not assume that the newer drivers will improve anything. I'm coming to believe more and more in the old adage "if it ain't broke don't fix it". Don't forget that Teddy Bear used to try dozens of wireless drivers and he rejected most of them for various reasons. Any one of them may be fine if only we knew exactly how to configure it, there are literally HUNDREDS of variables that can be changed and we have no documentation to do so, and we have no source code or documentation for the drivers themselves. So we may well be running a new driver with the wrong configuration, and getting altogether bad performance from it.

    Wifi. Huh. It's a can of worms. A big can. What is really needed, from the point of view of a somewhat disillusioned engineer, is a completely new system that works on a dedicated data band with wide bandwidths, proper antennas, free of all of the absolute garbage that has been stuck onto the present wifi standard to try to make it work in the presence of other band occupants, and the half baked power saving schemes that screw up what barely worked in the first place. And preferably designed by engineers and not politicians :)
  24. bucher

    bucher Networkin' Nut Member

    There should just be a tomato or linksysinfo irc channel. Is there a way to add bitchx to tomato or dd-wrt?
  25. Dr Strangelove

    Dr Strangelove LI Guru Member

    Your prays have been answered. Now there's 802.11ac. So no more can of worms.
    Now we just have a bucket of eels!

    Dell DW 1520 works will on 2.4GHz and 5GHz for stability using E4200v1 with Tomato firmware (80-216Mbps connect speed at 15m with max 80Mbps[2.4GHz] transfer rate). Signal strength and throughput are a bit on the weak side, but still very usable.

    My E4200 about page states Broadcom Wireless Driver updates

    I have also disabled all 802.11g attributes such as Afterburner and etc at both ends and made sure my WiFi on my notebook has full power in power/battery options.

    Also note the notebook angle to the WiFi AP can add/reduce up to +/- 10dB with a built in antenna in the monitor lid.
  26. Toastman

    Toastman Super Moderator Staff Member Member

    Yep, the new standard is built on top of all the old crap and will be backwards compatible with it. Expect thing to get much, much worse before it gets better. And 1.2Gbps won't happen, any more then 450Mbps did :) Almost all the industry "insiders" are expecting around 200Mbps.

    I'm just adding this:

    On the various forums it has now become popular to add so-called "high-gain" antennas to improve range. Here's some issues for you to think about:

    • Most of the antennas people are advertising are actually for 2.4GHz ONLY. If you have 5GHz these antennas DO NOT - AND CANNOT - WORK PROPERLY. You must get a design certified for the band you want to use. Most of the information on the various forums is misinformation. There is even a direct recommendation by one poster to use ASUS antennas for 5GHz - with a picture of the box clearly stating it is for 2.4GHz. THIS IS WRONG!
    • Antennas can increase the signal in one direction at the expense of others. They cannot produce extra power from nowhere! Think about it.
    • The usual add-on vertical antenna sold for routers is known as a vertical collinear array. They concentrate the signal into a very flattened "donut" shape. So, for example, if you have a router UPSTAIRS and you use these antennas on a router DOWNSTAIRS, then you will increase the signal downstairs, but it will obviously DROP in the upstairs rooms. Think about this before you buy them, or use this knowledge to position the antennas appropriately to cover the areas you need.
    • The antenna gains quoted in the advertisements are mostly fictional :cool:
  27. pharma

    pharma Network Guru Member

    Actually he did address the issue of Intel clients(intel 5300) not working with many routers -- in his router review tests the intel 5300 was the only wireless adapter was used. With each reviewed router there is also a detailed review providing alittle bit more depth. I can't say whether his results are correct or skewed but a check with other router reviews would reveal any inconsistencies.

    I admit many people have had problems with intel wireless adapters, but people also have issues with other brands. I've been fortunate not to have had alot of issues, but I love to tinker and usually adjust default settings on any hardware I use -- in this case my intel, broadcom, atheros, and belkin wireless adapters. :)

  28. buggage

    buggage Network Guru Member

    No problem, just wondering.

    Do have another somewhat related question. So from what I've read around various places online, N chanels (especially top/higher channel(s)) allow for more power and was something like up to 1W. Is the E4200 (or Tomato) capable of that? In the Tomato settings the TX power stops at 400mW or something (not in front of it currently), but is the hardware actually capable of going higher?
  29. Toastman

    Toastman Super Moderator Staff Member Member


    Be careful because this whole subject is cursed with misinformation from people who have not conducted any controlled tests or even looked at the specifications. The hardware in the current crop of routers cannot even give 400mW. Maximum power on 2.4GHz is 250mW and on 5GHz about half of that. The power level is set by the SiGe 2528L power amplifier chips in use and cannot provide more than the manufacturer's specification. Almost all the current routers use the same SiGe 2528L power amplifier chips on 2.4GHz as the old WRT54GL, and that is why they have broadly the same power output. There is a high powered router, the E3200, which people sometimes get confused with the E3000, but the wireless hardware is VASTLY different. I've never seen one yet though. Do they really exist? :confused:

    The maximum power setting on the RT-N16 and E3000, and probably all similar routers using the same wireless driver, is achieved with a setting of about 60. Above that does not increase power, as you can test easily with InSSIDer, but it will probably cause more "noise" to be generated as the power amplifier is driven into nonlinearity. You should view the tx power setting as a relative level and not as a true power setting.

    Read through this thread:


    You'll see also that DD-WRT and earlier Tomato drivers used to give a slightly better signal strength. It was only around 1.5 dB - which is pretty negligible, and certainly not the 10-12 dB claimed by many posters. That would be completely impossible.
  30. mito

    mito Network Guru Member

    Hi Toast, many thx for your explanation, would be great N16 with 5Ghz, i'll keep on watching.
  31. Hogan773

    Hogan773 Networkin' Nut Member

    I moved my power from 42 up to 150 and I watched live on InSSIDER and the connection dropped down about 3db stronger. I didn't leave it there for long. I am not trying to go mano-a-mano with an expert like Toastman by the way! Just passing along my observation on my E3000. This was reading InSSIDER from a laptop about 5 feet from the router. Further away might not have seen the same 3db drop.
  32. jsmiddleton4

    jsmiddleton4 Network Guru Member

    "driven into nonlinearity"

    That is the first left past sine curve and then hang to the right when you see diminishing returns.
  33. Toastman

    Toastman Super Moderator Staff Member Member

    Hogan, you may well see weird things when you are so close to the router, as the wireless receiver on which InSSIDer is based overloads. If the reading were genuine, then you would notice exactly the same change in dB level no matter how far you are from the router.
    I would always measure at least 10 metres away and preferably something like 25 metres. Taking measurements on a WRT54GL closer than that gives incorrect readings, as an example.

    You should have seen a tiny increase of around 1dB as the power increased from 42 to the maximum of 60 (above this does nothing) but using InSSIDer it's very, very hard to see such small changes. Also remember that InSSIDer itself works on the information provided to it by the wireless receiver. That varies between manufacturers and the granularity of the RSSI reading is extremely variable. The only way to make meaningful tests with InSSIDer is to take many readings over a long period of time, remove the obviously wrong ones, and then average the remainder. I'm talking here about hundreds of readings, not one or two.

    While on the subject, you should also realize that InSSIDer is not a very good tool for making measurements. Read this:


    BTW - being too close to the router, causing overload of the receiver, is one of the main reasons why some people experience slowdowns when they increase transmitter power.
  34. Hogan773

    Hogan773 Networkin' Nut Member

    Thanks for that - that's interesting.

    Here's a question - in our upstairs bedroom with the laptop I generally see 6-8 other networks - most of them are far away with RSSI about -85 ish. A couple are in the 65-75 range (including a couple that are trying to go wide channel as I see them popping back and forth between 6+2 and 6, etc.

    From there my E3000 is either -60 or sometimes -50 - it goes back and forth pretty regularly. I assumed at that distance I should leave my power high (I've generally left it at 55) just to do whatever I can to be the stronger signal in that soup. I assume at that distance (its 30-35 feet and upstairs) there isnt going to be any overloading of the receiver so might as well leave power up.

    Hopefully this will all be moot once I get a dual band card to drop into the laptop, and then I can ride high above the soup on my 5ghz spectrum.
  35. Toastman

    Toastman Super Moderator Staff Member Member

    The situation in your bedroom is that your router is often only 5dB or so stronger than the 2 other strong routers you hear. That means you are, if on the same channel, probably causing the PC's in that room to hold off transmission until the channel is clear - you are sharing the band nicely. When the others switch to 40 MHz band that is a time when you may experience drop in throughput because the N signal will likely walk all over yours.

    Yes, I'd leave it at the stronger setting, personally.
  36. mito

    mito Network Guru Member

    Hi, before now i found Transmit Power "0" had the best range, later on Toast explained us that 60 was the best because of range and steady signal, now testing a lot around that number my RT-N16 with 7493 and tree 9db antennas on it, a transmit power at 62 and 50 feet from router, i found that this setting for me has the best range and speed up to now and signal rock hard steady.
  37. LanceMoreland

    LanceMoreland Network Guru Member

    I have not finished reading the thread so I am not sure if you have solved this issue yet, but you can only connect at N speeds (144) on 2.4 ghz channel with an Intel card if you enable WMM in the Advanced wireless section.

    Now I am off to finish the thread.

    Edit: After further reading, I see you have found the solution.
    BTW: I see new drivers were released from Intel 1/12/12. I haven't tried them yet but here is hoping for some improvement.
  38. pharma

    pharma Network Guru Member

    Thanks for the heads-up!
  39. Gaius

    Gaius Networkin' Nut Member

    It's just the addition of the My Wifi feature.

    Currently I find that using the Windows Update Intel driver with the default settings ensures that Intel Wireless cards don't mess with my Tomato router. I suggest everybody do the same if they have an Intel card in their laptop, or just buy a reliable dongle.
  40. LanceMoreland

    LanceMoreland Network Guru Member

    Will the windows update driver allow the intel card to connect at N speeds without having WMM enabled on the router?
  41. Hogan773

    Hogan773 Networkin' Nut Member

    So you are saying NOT to go find the actual Intel drivers? That the Windows update works better?

    I am having some strange issues with my E3000 on 2.4ghz. Today after updating the drivers from Intel it seems the speeds are even slower than before.....only 3-4 Mbps down even when close to router.

    If I have already installed the REAL INTEL driver then how do I go get the "Windows" driver?
  42. TomatoE2000

    TomatoE2000 Networkin' Nut Member

    in latest 4494(vlan version) i see that wireless cannot connect greater than 144Mbps. it use to be 300Mbps in the 4493 version. has anything changed on wireless side? i use an e2000.

    EDIT: Never Mind. i changed to 40Mhz Channel and it can now connect at 300Mbps
  43. Cynamiter

    Cynamiter Networkin' Nut Member

    It seems, that there is a bug in eth2 (5Ghz) for E4200v1:

    I just switched from the lower control sideband (channel was 36) to the upper control sideband (channel set to auto), but the channel didn't change and still is at channel 36. Would you please take a look into this?
  44. pharma

    pharma Network Guru Member

    Mine changes with no problems .... are you saving your changes ? (page bottom )

    However, if you have and the problem persists flash back to the previous release. I flashed but chose to retained my previous settings and could be the reason why I don't have your issue.

    Edit: I reflashed with 2494 to test your problem and was able to change the upper & lower sideband, and channel with no problems using default firmware settings. Afterwards I reloaded my saved config file (from retained 2493 firmware settings but saved with 2494) and did not get the "Invalid hardware type" error as I did earlier when attempting to load saved 2493 config settings.

    It also seems one of the developers changed how the file name is written when saving config files (tomato_v128_2494) ... this probably is causing the error.

    I guess Jon had a reason for doing things the way he did. What's that famous saying ... "If it ain't broke .. "
  45. Cynamiter

    Cynamiter Networkin' Nut Member

    So, does your channel changes automatically, when you switch from lower to upper control sideband (remember to have channel set to "Auto")? I am able to change channel by hand, though.
  46. Oleg Repin

    Oleg Repin Networkin' Nut Member

    Hello, Toastman,

    I am new at this forum. My router is E3000, computer i3-550 4GB Windows 7 64-bit, 1Gbit wired to the router. I-Net connection Ethernet 100 Mbit, static IP, effective 90 MbitPS download, 60MbitPS upload.

    Up-to recent time I used stock Cisco firmware I had some complains:
    1. Slow USB access, 60 MbitPS upload, 36 MbitPS download for both file sharing and FTP.
    2. No VPN.
    3. Media server does not work.
    4. No firmware upgrade for more than an year.
    5. Effectively zero customer support. It is kind of funny, I live a couple blocks away from Cisco headquarters, but I talk with guys in India :)

    So, a week ago I have upgraded firmware to your release tomato-E3000USB-NVRAM60K-1.28.7493MIPSR2-Toastman-RT-VPN-NOCAT.bin.
    The most of settings are default. I set wireless to off, added USB, file sharing, FTP, and media server.

    The first results:
    1. It eats a half of my I-Net speed: 45 MbitPS down, 30 up (stock firmware eats about 5%)
    2. File sharing speed 16 MbitPS up/down to local computer. FTP 12 MbitPS. Yes, I increased buffer size to 64kBytes. If I increase it to 128 kBytes or more, router hangs and cured only with reboot.
    3. QoS eats additional 30%.
    4. Media server works and shows incredible speed! I do not know how to measure speed, but I can play 1920x1280 files. I could not do it with file sharing using Cisco stock firmware.

    A couple days ago I upgraded to the newest release tomato-E3000USB-NVRAM60K-1.28.7494MIPSR2-Toastman-RT-VPN.bin.

    The results:
    1. I-net speed now matches Cisco stock firmware 85 MbitPS down, 55 MbitPS up.
    2. USB speed is the same as for 7493 release.
    3. If I turn QoS on, my down speed is 15 MbitPS. I set up limits for all classes to 100% to effectively disable QoS, it does not help. I tried several times: 15 MbitPS with QoS on, 85 MbitPS with no QoS.
    4. Media server works the same as in 7493 release.

    I have some simple questions: Did somebody sometime compare DD-WRT/Tomato/TomatoUSB/Toastman firmware speed against Cisco's?
    I spent several hours surfing the Web trying to answer this question. No luck. Or just nobody cares? If it is not so, may be I did something wrong, or, more exactly, did not do something right?

    Thank you in advance,
    Best regards,
    San Jose, CA
  47. Hogan773

    Hogan773 Networkin' Nut Member

    I am interested in understanding the implications of what you're saying. I have been trying to get my laptop to connect to my E3000 with better speeds. Using 2.4ghz seemed the best I could get was about 15mbps even when my wired connection is about 20-23. Now with 5.2ghz it gets closer - up to 17-19.

    Was there some sort of a major change in version 7494 that would increase Internet speeds? Something with cut-through forwarding? Or is this just another one of those wifi anecdotal vagaries which makes things tricky?
  48. pharma

    pharma Network Guru Member

    With sideband set to either lower or upper and channel set to "Auto", the channel does change automatically. I'm not sure about the algorithm it uses, but the router will determine which combination to use based on your circumstances. I believe when you use "Auto" for channel it determines the "optimal" settings to use for lower/upper control sideband and will override any manual setting you make. Over the course of a day (timespan) it seems to automatically change channel, channel width and sideband settings to accomodate your wireless devices and environment. If you want to specify a particular control sideband then it should be done manually (not using "Auto" for your channel selection).

    I also believe whatever settings you have for 2.4ghz, it will influence your 5ghz settings. So if you play with both you will see they both seem to impact each other and what settings they seem to chose when channel is set to "Auto".

  49. mito

    mito Network Guru Member

    Hi, on my lappy Dell D520 i witched from 3945 abg card to Intel 4965agn and instaled the corresponding Intel drivers on it, with my home RT-N16 always get 144 Mbps but at my office i have a WRT600N with DD-WRT firm working with two different wifi nets one at 2.4 Ghz and one at 5 Ghz , at 2.4 Ghz the Intel 4965 card conects easyly at 144 Mbps and to 5Gghz it conects at 270 Ghz, so Intel card is capable of getting 270/300 Mbps at 5Gghz router, but never at 2.4Ghz
  50. jsmiddleton4

    jsmiddleton4 Network Guru Member

    "I suggest everybody do the same"

    I'm a little confused too. The drivers, Intel's and Windows update, should be the same.
  51. Hogan773

    Hogan773 Networkin' Nut Member

    I can now get better speeds on my E3000 using 5ghz. Depending on how far away I am, it connects anywhere with a stated rate of 40mbps to 300mbps. Actual throughput can still max out my Comcast connection so I get 20-23mbps on Speedtest.

    I still get cruddy and variable connects on my 2.4ghz. It will sit there and say "144mbps" and then "6.0mbps" or "9.0mbps" and then flip back up to "144mbps". And throughput stinks - generally 4-6mbps.

    I feel like that is some sort of setting on the E3000 and/or the Intel 6205 that is conflicting or otherwise not liking each other. Any suggestions? I have the "typical" Tomato settings ie WMM Enabled, Throughput and Afterburner disabled, 20mhz channels vs 40, etc.

    Someone else was complaining that the Intel drivers were being "too much by-the-book" in terms of standards, etc for enterprisey type things, etc. Is there such a thing as a "non-Intel" driver that would still drive the 6205 card?
  52. Toastman

    Toastman Super Moderator Staff Member Member

    When comparing speeds, it is important to note that these routers are resource-limited low-power devices. You wouldn't buy a PC running at 480MHz, would you? They do pretty well considering. If you have highspeed internet connections around 100Mbps, (especially if you have 100Mbps or similar in BOTH directions!!! = 200Mbps in effect) - and you want to get the best internet speeds, turn OFF everything that uses resources. e.g. QOS, Encryption, display of Graphs, Statistics, any pages that you have onscreen that are refreshing in your browser. All of the USB stuff. And so on. Each thing that you turn off will make resources available for routing. This doesn't just apply to Tomato, it applies to ANY router. For speeds of around 100Mbps I wouldn't personally dream of using such a device. Pretty soon the manufacturers will be forced to wake up to this fact and make some better hardware. RT-N66U's and the like are interesting evolutions, but they aren't a quantum leap in performance and they are grossly overpriced for what is inside. 1GHz plus WOULD have been a bit more interesting :)

    Example - one poster I had PM discussion with was measuring throughput with QOS enabled, 60 rules in use, 8 of them L7 filters, every USB function enabled, the most processor intensive encryption, and was monitoring his router with 3 open windows refreshing the monitoring graphs and statistics as fast as possible.

    To turn off QOS please tick the button to actually turn it OFF. Putting high values in the boxes doesn't turn it off, it just wrecks the config and prevents the limits from working properly, hence crippling it. It may even increase the load.

    Monitoring features on routers are for diagnostics purposes only. If you wish to keep them onscreen all day because they look pretty, then that's fine as long as you understand the implications. Similarly, we use QOS to ensure that everybody gets a slice of the internet pie. There comes a time when too much processing will affect the throughput and be worse than the problem you were trying to solve with QOS. 8 L7 filters is going to make a big dent in throughput if you have a highspeed line. If you have lower speeds, then you probably won't notice anything.

    The facilities are there to use if you need them. If you don't need them, turning them off is a good idea. Things that aren't running consume no resources.
  53. pharma

    pharma Network Guru Member

    There is a difference between versions you find on Windows update and those on your hardware vendors site. New Drivers available from vendors may not show up in Windows Update for weeks/months because of Microsoft Windows certification stds. This applies to any hardware, ie drivers for video cards, monitors, etc...

    For most people it's easier to go directly to your hardware vendors site and download the newest drivers, especially if you want to be on the "bleeding edge".

  54. jsmiddleton4

    jsmiddleton4 Network Guru Member

    This Intel driver was just released from Intel and then showed up at Windows Update.
  55. Oleg Repin

    Oleg Repin Networkin' Nut Member

    Hello, Toastman. Thank you for your answer.

    OK, I understand the general idea. If you have industrial-grade I-Net connection, you should use industrial-grade router to handle it! :-( I suspected something like this...

    I understand that symmetric 100 Mbps connection is not very common at this moment. In our home community we spent 4 years trying to establish sizeable I-Net connection (up-to symmetric 1 Gbps) for reasonable monthly payments. It took a lot of effort and a bunch of money. But times change. I am sure, in a couple of years you will hear a number of complains like mine.

    It might also explain some Cisco's vs Tomato firmware comparison staff. As I understand, Tomato was developed in the direction of complexity, adding new features with no real consideration of performance penalty.

    Nevertheless, some mysteries are still unsolved. I will try to repeat some measurements in other conditions and then let you know.
  56. Oleg Repin

    Oleg Repin Networkin' Nut Member

    In my home area I commonly see more than hundred Wi-Fi servers, so speed measurements have no sense. Never saw more than a couple of Mbps. 5 GHz range is less jammed because of less devices on the market yet. So, the common things to do are:
    1. Always switch to 5 GHz if your hardware allows it.
    2. Make sure that 40 MHz channel width is enabled on both sides.
    3. Make sure that automatic WiFi channel selection is enabled. But I am not sure that auto-channel feature works properly, at least my Cisco stock firmware often made stupid decisions. So, you could also try to manually select WiFi channel. I commonly see in my neighborhood that a half of WiFi devices jammed into the single channel while other channels are relatively free. Probably because of default settings in the WiFi-capable routers.
  57. Oleg Repin

    Oleg Repin Networkin' Nut Member

    OK, my next tests.

    The first thing I discovered that I did not change my administrator password from the default. So, how I commonly do it, I saved the new password to clipboard, then pasted it to Adiministration/Admin Access/Password fields, then tried to reenter. No luck. Tried to type the password. No luck again. So, 30/30/30 reset. BTW, I use 32-bit IE 9 during my experiments.

    This time I physically disconnect everything from the router but my computer and USB drive. Disabled WAN and both 2.4 and 5 wireless. Enabled USB and Samba. Discovered that the drive is not visible. Tried the common tricks: unplug/plug USB connector, turn off/on power on disk drive, reload router. No luck. Connected USB stick instead of hard drive. No luck again. Then I enabled USB 1.1. This time it helped. The first with USB stick, then with my hard drive. But speed is disappointed:

    SanDisk Cruzer 4GB
    Connected to PC: 37.7/127.1 up/down Mbps
    E3000 Samba: 6.9/ 7.0
    E3000 FTP: 6.5/7.7

    Seagate FA GoFlex Desk 2TB
    Connected to PC: 120.0/180.0
    E3000 Samba: 6.7/7.4

    So, in short, what I achieve (in USB part) during this week:
    1. I complained about too slow speed of stock Cisco firmware: 60/36 up/down Mbps.
    2. Upgraded to Toastman Release 7493 and got 16 up/down Mbps.
    3. Upgraded to the Release 7494 (no NVRAM reset) and got the same speed as the previous.
    4. Reset NVRAM, disconnected my too fast I-Net, and got 6.7/7.4 Mbps.

    But my USB somehow works (yet), so I still have a room to move forward :-(

    Any ideas?
  58. Jacques

    Jacques LI Guru Member

    Hi Oleg Repin,

    What is CPU Usage during your test ?
  59. Hogan773

    Hogan773 Networkin' Nut Member

    I am now getting my full internet speeds 20-23 mbps using the 5ghz connection up in the upstairs corner of the house (even when signal strength is only 2 out of 5 bars). So I guess my 20 bucks to buy the new laptop wifi card were not wasted. :D

    As an indication of the interference on 2.4ghz in my suburban neighborhood - last night I could only get about 3-4 Mbps on Speedtest. Just now, @ 10am on a Monday, I could get 13-14! I believe that more and more neighbors are getting wifi cable boxes or something (lots of "2WIRE330" type SSIDs which I assume are not someone's personal PC router). And they probably all have them turned on in the evenings. I live in an area with pretty densely packed single family homes - but still 50-100 feet away from other houses. I can't imagine what it's like in a highrise condo building or a densely packed 2 or 3 story apartment community!

    5ghz actually seems like a great option unless you have a really big house, because neighbors can have it too and the signal will attentuate faster so each person can have their own little clear connection. Plus, more channel options.
  60. Gaius

    Gaius Networkin' Nut Member

    Just do a driver rollback until you get the Microsoft driver and let Windows Update manage updating it from now on. Intel's newer drivers cause problems while the Windows Update one works flawlessly as long as you use ONLY the default settings on the card.

    They are not the same. Windows Update is always way behind Intel's offering on the Intel website. on Intel.com on Windows Update as of today according to my update check.
  61. Toastman

    Toastman Super Moderator Staff Member Member

    Gaius has a good point - in the past I've had many weird problems with various video drivers from the OEM motherboards suppliers, yet Windows Update's drivers always worked. But they may not necessarily be (and usually aren't) the latest versions.

    Microsoft has a responsibility towards its users - I would guess that before loading any drivers onto anyone's PC they almost certainly do their best to test and approve such drivers. Otherwise they could very well be sued :)
  62. Gaius

    Gaius Networkin' Nut Member

    Yes, the approval process is pretty useful, especially when it comes to Intel, which regularly ships junk drivers. Microsoft's update makes sure the PC user experience is more bugfree, and some people may not like that because they think they DESERVE the latest, but when it comes to wireless, stability > features.
  63. Hogan773

    Hogan773 Networkin' Nut Member

    I did an uninstall and it went and found the 14.2.0.something itself......

    Separate but related question. Does anyone use the INTEL PROSET WIRELESS software? I guess it is software that takes over the network connection process from Windows. Wondering if that somehow would make things better? (given the view toward Intel drivers, you'll probably tell me to stay the he!! away from it though)
  64. Oleg Repin

    Oleg Repin Networkin' Nut Member

    Hello Jacques, excellent question!

    So, my 4-th update on my Tomato quest. :) The following numbers are Transfer speed (Mbps)/Free Memory (MB)/CPU Usage (%):

    Idle: 0.0/50/1.2
    Speedtest.net download: 75/49.98/86
    Speedtest.net upload: 43/47.98/42
    Samba download: 41.3/49.1/100
    Samba upload: 15.7/49.2/100

    I do not know why my Samba speed then suddenly increased. I did not reload my router, I did not change my settings. If somebody has some crazy idea what happens, please let me know :-(
  65. Jacques

    Jacques LI Guru Member

    Hi Oleg Repin,
    You have to wait until someone will do a better router than you have because you have a fast internet, or do what you Toastman said.
  66. pharma

    pharma Network Guru Member

    Intel Proset is software from Intel to help manage your wireless connection. You have choice to install or not, but is not needed. If you just want to install driver only that is fine.

    I've never had any problems using drivers from Intel's site. Even though drivers you get in Windows update are usually months older, do not expect any magic "cure" to appear for any problems you have by installing newer drivers outside windows update. You should only do it because you want to be using the latest drivers (and do not have to ask questions about INTEL PROSET WIRELESS software).

  67. jsmiddleton4

    jsmiddleton4 Network Guru Member

    "I've never had any problems using drivers from Intel's site."

    Me either. And I'd much rather use Nvidia's video drivers from their site than Windows Update.

    " on Intel.com on Windows"

    I shoulda looked more closely at the latest version numbers. I only have access to one laptop with the Intel chipset WLAN and its with my daughter at college. Will check when she comes home to do laundry this next weekend.
  68. Hogan773

    Hogan773 Networkin' Nut Member

    Yes this is what I did - I just downloaded the "drivers only" from the Intel website and have not used PROSET (hence my question of whether there was a benefit to using PROSET vs Windows networking manager or whatever its called)
  69. AMurderOfCrows

    AMurderOfCrows Network Guru Member

    Hey all

    So, i have run into this problem and need to make a choice.

    I'm on my second Intel Centrino Ultimate-N 6300 card on my notebook, and my first new router in about 6 years, the ASUS RT-N 16.

    i can't keep consistently good connected speeds unless i turn off mixed mode.

    My choices are:

    A) Get another card for the notebook, something internal with a more reliable driver, that will properly handle N on 2.4ghz on my current router

    B) Get a newer router that can take Tomato and offers 5ghz and works more stable with my current Intel card

    Since i JUST got the new card in september, and i JUST got the router in april, i really don't think i want to replace both of them

    what would you all do? Keep in mind, i want an internal PCI-E card for the laptop

  70. Hogan773

    Hogan773 Networkin' Nut Member

    In my case it seems the 5ghz allows for a faster connection because on 2.4ghz I have about 8-10 competing networks in my neighborhood. I still can't tell if my weak 2.4g performance is due to some conflicts between my Intel 6205 and my E3000 router. But it seems they talk to each other just fine on 5ghz so not sure why that would be different on 2.4ghz. I'm just going on anecdotal stuff from these boards where people say that Intel clients sometimes are fussy.

    Obviously it is cheaper to buy a new wifi card but I'm not sure if that is your best move. Might be better to invest in the ability to go 5ghz, which would imply an E3000 or E4200 or other similar. Perhaps you can Ebay the RT-N16 and get a new simultaneous dual router.
  71. AMurderOfCrows

    AMurderOfCrows Network Guru Member

    E3000 was on the frontpage of slickdeals recently. As long as this is an actual fix to the issue, i'm okay spending the money. My motto has always been that wireless was for convenience, wired was for performance. 300+ N really changed that for me.

    i guess i had better look at the difference between those routers and make a choice here pretty quick.


    Okay, i need a fast consensus opinion. I have the intel 6300, and i saw the cnet review saying that only the 6300 and another card were capable of getting full speed out of the e4200.

    Does this also apply to the tomato builds as of late?

    Tonight, i could buy either the e3000 refurbed at $50 on Newegg, the e3000 new at $70 on Newegg, or the e4200 refurb for $95 on ebay from the linksys store.

    which way should i go?
  72. eahm

    eahm LI Guru Member

    AMurderOfCrows, first of all, thanks for killing those annoying creatures.

    E4200v1 if you can, it's far better than my second favorite, the E3000.
  73. AMurderOfCrows

    AMurderOfCrows Network Guru Member

    I get the sense that i'm being trolled here....

    How much better? is it $30-$50 better? I'm running a shibby version of tomato on my RT-16 right now, same as yours i think. I need stability and really really good transfer from my laptop, as not having that transfer speed means installing another hub in each of my bedrooms and running wired, not something i want to do.

    thanks for the reply

    Now this is a VERY interesting statement. I can afford it, i just need to quantify where the value comes from and justify that value vs the cost. Do you lose any tomato features when moving from the RT-16 to the E4200?
  74. jsmiddleton4

    jsmiddleton4 Network Guru Member

    How much better?

    According to the testing the E3000 is better then the E4200 V2. The V1 is quite a unit however. I am of the opinion however the E3000 is the best price/value/performer. I have 2 of them.
  75. eahm

    eahm LI Guru Member

    That's why I said if you can. I think the E4200 is at least $40 better. I sell E3000 for businesses and E4200 for the owners' homes. They are both very good.
  76. AMurderOfCrows

    AMurderOfCrows Network Guru Member

    Yeah, i'm leaning heavy towards a refurb E3000 just because of the price. I started this off back with the RT-16 because i needed N and gigabit, but didn't do all that much actual research. Now, i'm almost kicking myself about it. I *REALLY* don't want to do another router upgrade for a LONG while. I mean, my wrt54gs is STILL running, but it simply can't handle what i need any longer. On paper, with the exception of the lack of 5ghz, the RT-16 just seems like a better unit.

    Now this is a VERY interesting statement. I could do the E4200, i just need to quantify it. do i lose anything tomato wise with this one over the RT-16?
  77. eahm

    eahm LI Guru Member

    Here you can find more test: http://www.smallnetbuilder.com/lanwan/router-charts/bar/58-2_4-ghz-dn . Make sure to switch between the choices from the drop menu.
  78. AMurderOfCrows

    AMurderOfCrows Network Guru Member

  79. Hogan773

    Hogan773 Networkin' Nut Member

    I came home last night and my wife was trying to use the laptop and it was connected to the router but "No Internet Connection". Turns out the PC wired ethernet to the router also had no Internet. We were connected to the router in both places because I could get into Tomato thru 192.168.....

    So I assumed it was the cable modem and cycled that. Nope - still down. So I power cycled the E3000 and when it came back up, after about 10 seconds, then voila - Internet again!

    Just a random thing or are there settings that can cause it to lose its connection to the Internet? It wasn't as if the router itself had overheated or anything - it was still on and connecting things to the WAN - just wasn't talking with the Internet at all....

    I had enabled Jumbo Frames - don't know if that could affect it. Otherwise I don't think I have too much funky going on from a settings standpoint.
  80. pharma

    pharma Network Guru Member

    If you do buy a E4200 make sure you get a V1 and not the V2. The V2 uses a Marvel chipset which Tomato will probably never support. The box looks the same but you can tell by the serial number on the box and bottom of router. It should also have the mac(s) on the same label:

    01C10C = a V1
    01C106 = a V2

    eahm likes this.
  81. jsmiddleton4

    jsmiddleton4 Network Guru Member

    The drop in performance with the V2 is surprising.
  82. AMurderOfCrows

    AMurderOfCrows Network Guru Member

    Well, i ordered the refurb'd v1 last night from linksys' website. The stats were good enough that i *believe* i should be solid for the next few years with it.

    on the down side, i do have a very nice, kinda expensive RT-N16 mounted to my entertainment center that i will be rotating out of service once the E4200 comes in.

    thanks for the assist guys.
  83. hawkmat

    hawkmat LI Guru Member

    Asus upgrades wireless driver in RT-N16 firmware. I keep one extra RT-N16 loaded with Asus latest firmware to use as a comparison. Yesterday Asus released firmware Up until now they have used driver, same as the one used in the RT-N Tomato builds. With the release on they now use driver I wonder what this driver would do for the newer routers?

    Update: I have been testing it for an hour now. The signal strength is similar to previous drivers but it is the most stable driver I've seen for the RT-N16. With isSSID the signal is strong and very smooth and steady. I always test the router in the same location to keep the variables to a minimum.
  84. Gaius

    Gaius Networkin' Nut Member

    Any noticeable improvements over the latest Tomato builds?
  85. hawkmat

    hawkmat LI Guru Member

    Hi Gaius! I have tested the RT-16N with Tomato and the factory firmwares and found the signal strength to be similar with all current wireless drivers. The difference with the drivers is speed and stability. I have found that the Tomato driver is the best with the Asus RT-N16. The Tomato RT-N driver was a little slower and a little less stable. The new Asus firmware and driver is very stable and the fastest driver I have experienced yet. I have done a few Windows 7 download transfers from our local server and the tranfer rates were 11-13 MB per sec. With all other drivers I would usually get 9-10 MB per sec. These test are always conducted with the notebook and router in the same locations. isSSID shows a very smooth and steady signal. Broadcom must have fixed a few bugs or tweaked a few settings since the driver came out. So far I'm very impressed. I'll keep testing it. Thanks!
  86. Gaius

    Gaius Networkin' Nut Member

    Sounds like we need a Tomato build with the driver so we can confirm.
  87. Hogan773

    Hogan773 Networkin' Nut Member

    Would this driver also be the same for an E3000 since it works on a Broadcom chip? (I'm always hoping for better/faster/more stable for my E3000!!!)
  88. Toastman

    Toastman Super Moderator Staff Member Member

    The increase in speed would of course be expected from any increase in processor speed and probably nothing to do with the wireless driver. I doubt it will be any different.

    EDIT - sorry, you are testing the new firmware on the RT-N16? Apologies.

    There is much talk of the new router having "beam forming" technology and other interesting goodies built in. High power amplifier chips amongst other things. However, nobody seems to have posted any details of this or any useful photos yet. If this is true and not marketing hype (and so far I'm inclined to think the latter) then the driver may well be rather more advanced than is needed for the RT-N16 and E3000's. That isn't to say that it won't be to some extent backwards-compatible, we'll maybe have to find out. A transplant is needed.

    Now, if anyone cares to do it, you may have to change a few files. For example:

  89. hawkmat

    hawkmat LI Guru Member

    Asus is using the driver with there latest RT-N16 firmware. They were using the driver up to this point. I would believe that they saw something worth while in this update. Just tested it again and Windows 7 file transfer is still 11 to 13 MB per sec. The router been up 24 hours now and it is solid and stable.
  90. jsmiddleton4

    jsmiddleton4 Network Guru Member

    Let's hope the driver will be available and can be complied into at least a test version.
  91. hawkmat

    hawkmat LI Guru Member

    This is a list of what Asus fixed with their newest firmware with the new driver. Number 3 is the one that caught my eye.
    ASUS RT-N16 Firmware Version
    1. Applied patch for WPS PIN code security issue
    2. Fixed media server related bugs.
    3. Fixed wireless compatibility issues for specific client chipsets.
    4. Javascript error in English and Danish
    5. Fixed Download master UI bugs for HTTP download
    6. Modified download task information for e2dk download.
  92. AMurderOfCrows

    AMurderOfCrows Network Guru Member

    ....ugh....does this mean i might not have had to buy a $100 router? lol
  93. Gaius

    Gaius Networkin' Nut Member

    I think by "specific client chipsets" they mean Intel chipsets. Every client I had with my RT-N16 worked fine except for an Intel laptop.

    To any devs reading this, how easy would it be to compile the new driver into a test release of Tomato?
  94. dailyglen

    dailyglen Networkin' Nut Member

    Hi Hogan773,

    I also had somegthing like this happen too where my wired network connections dropped (not sure about wireless) with my E4200v1 running:


    The log messages at the time were:

    Jan 20 20:15:53 potato user.info kernel: vlan2: dev_set_allmulti(master, 1)
    Jan 20 20:15:55 potato daemon.notice miniupnpd[922]: version 1.6 started
    Jan 20 20:15:55 potato daemon.notice miniupnpd[922]: HTTP listening on port 42997
    Jan 20 20:15:55 potato daemon.warn miniupnpd[922]: no HTTP IPv6 address
    Jan 20 20:15:55 potato daemon.err miniupnpd[922]: setsockopt(udp, IPV6_ADD_MEMBERSHIP): No such device
    Jan 20 20:16:01 potato user.debug kernel: vlan2: add 01:00:5e:00:01:3c mcast address to master interface
    Jan 20 20:16:23 potato cron.err crond[555]: time disparity of 22118475 minutes detected
    I'm not running VPN, I have IPv6 off and I'm running NTP for time. I'm not sure if these log messages mean anything. To get the internet back up I had to refresh my internet connection on the computers and reboot my phone ATA box (I didn't have to reboot the router). Right now I'm trying to figure out if this is hardware or software related. It happend one other time about 2 weeks ago.

  95. Hogan773

    Hogan773 Networkin' Nut Member

    My issue was that the router wasn't getting an Internet connection. Recycling the router worked. I haven't had the issue for several days. Maybe was just a glitch between the Cable Modem and the Router.
  96. mito

    mito Network Guru Member

    Hi, my RT-N16 is an "old" one with v1.28.7493 MIPSR2-Toastman-RT K26 USB Ext with printer and pen drive on it, and my lappy Dell D520 has an Intel 4965 wifi card and never had an issue at all.

  97. dailyglen

    dailyglen Networkin' Nut Member


    My wireless ports on my E4200 went down again with the log messages:

    Dec 31 19:01:30 potato user.info init[1]: Linksys E4200 v1: Tomato 1.28.9007 MIPSR2_RAF K26 USB VPN-NOCAT
    Jan 22 15:14:59 potato user.info kernel: vlan2: dev_set_allmulti(master, 1)
    Jan 22 15:15:01 potato daemon.notice miniupnpd[913]: version 1.6 started
    Jan 22 15:15:01 potato daemon.notice miniupnpd[913]: HTTP listening on port 3722
    Jan 22 15:15:01 potato daemon.warn miniupnpd[913]: no HTTP IPv6 address
    Jan 22 15:15:01 potato daemon.err miniupnpd[913]: setsockopt(udp, IPV6_ADD_MEMBERSHIP): No such device
    Jan 22 15:15:08 potato user.debug kernel: vlan2: add 01:00:5e:00:01:3c mcast address to master interface
    Jan 22 15:15:28 potato daemon.info dnsmasq-dhcp[543]: DHCPREQUEST(br0) 00:16:76:e1:88:d4
    Jan 22 15:15:28 potato daemon.info dnsmasq-dhcp[543]: DHCPACK(br0) 00:16:76:e1:88:d4 mythtv
    Jan 22 15:15:29 potato cron.err crond[556]: time disparity of 22121054 minutes detected
    I suspect it goes down when vlan2 is doing something but I don't have a VPN configured and not sure why this is happening. Does anyone know how to disable vlan2 or make it so my router doesn't disconnect all wired clients. A refresh of the network on client computers will bring the network up but my phone boxes need to be power cycled.

  98. AMurderOfCrows

    AMurderOfCrows Network Guru Member

    The e4200 arrived today and it doesn't have mounting holes. i'm in absolute shock over this...how does a modern router not have a means of being wall mounted????

    gonna try to call cisco/linksys to find out what they recommend, but i wanted to stop in here real quick to see if anyone had solutions or ideas that might work, and also to find out if anyone who has successfully wall mounted this unit knows if there are any issues with horizontal signal when it's mounted vertically.

    thanks guys
  99. bucher

    bucher Networkin' Nut Member

    Any news on this new broadcom version?
  100. 777Braveheart

    777Braveheart Networkin' Nut Member

    I will take the blame for this. I'm sure I am doing something wrong. I liked the RT-N12B so much I bought the RT-N16. I am using the tomato-K26USB-1.28.RT5x-MIPSR2-083V-AIO.trx Shibby build on that one. Before I loaded it I did the 30-30-30. After I loaded it I did the clear nvram+thorough also. I am fortunate. The only other routers near me are a long way off. I am surrounded by trees. Often, I do not see the other routers unless I walk outside with my laptop. All are on channels 1, 6, and 11. The furthest is on 11. Therefore, I am also using 11 with 40MHZ Upper. WPA2+AE5. WMM enabled. APSD disabled. Transmit Power set to 60. (Yes. I read Toastman's recommendations.)

    I currently have 1 client; a desktop with a rosewill usb wireless nic. The chipset in it is the Realtek RT8191SU. I always test with this client as it is stationary and the same distance from the router every time I test. In Windows it always shows 300mbps. However, the router itself says 65Mbps for rate unless I reboot it and it says 300Mbps but then drops quickly to the 65. Even with the client off, the router says the same 65Mbps. Either way... if I pull up a youtube to watch on the client it stutters at 720p. Before this I had a WHR-HP-G54 and a WRT54GL. Both of those ran Victeks 1.28.9006 build. When those were running I had no problem viewing 720p youtube videos on the client. In addition my ISP speed has not changed in years. Hope all this makes sense. In essence, my rate is slower on this RT-N16. Should I throw the latest RTN build for the RT-N12B that Shibby released recently on the RT-N16 and try that? Or try one of Toastman's builds. Or try the older Victek build? I'm trying to avoid going back to the ASUS FW. Or is there something else I should try?
  1. This site uses cookies to help personalise content, tailor your experience and to keep you logged in if you register.
    By continuing to use this site, you are consenting to our use of cookies.
    Dismiss Notice