I heard a rumor that the Linksys consumer product line (routers, cable modems etc.) cannot handle ISP access speeds higher than 10Mb/sec in either direction. Anyone know if there is truth to this?
I upgraded my Charter cable connection from 10Mb to 20Mb (down, 2Mb up) and am using a Linksys WCG200 Cable Gateway. All speed tests show nothing higher than 8.5Mb, about the same as before the upgrade. Cable guy says all signals are nominal and better than many, but tests with two computers (one hard-wired to the Gateway, one Wireless-G) show nothing higher.
If this is true, anyone recommend the cheapest upgrade that can handle 20Mb as a minimum?
If you really have a computer without the linksys box and don’t see higher speeds it is difficult to blame the linksys box.
Linksys makes a lot of things all of them consumer products. I would be very surprised if the newer (within the last 2 years) boxes could not handle up to 50 Mbps. Certainly not anything the was G.
I don’t know what you mean by “have a computer without the linksys box”. The WCG200 is in the data stream from ISP to computer and back, so if it can’t handle the thruput, it will be a bottleneck.
It looks like that’s exactly the situation. I think Linksys’ units, perhaps just their older models, top out at 10Mb/s for the cable link.
The G refers to the wireless link, not the ISP link, and the Ethernet is 10/100Mb, so that’s not where the weak link is. And this unit is about 3 years old, so it’s possible it is not up to today’s 20Mb, but I wish I could get the official word from an informed party (the tech support I received so far from Cisco doesn’t fall in that category).
I misread what you wrote. I thought you tried one box without the linksys box not one box with a wired connection to the linksys box.
I really don’t think that the linksys box is the issue if it supports wireless G. That is a maximum throughput of around 50 MBps I wouldn’t think that the WAN side would be any less. A quick perusal on the internet does not show complaints for slow through put on the WAN side. A fair number of people don’t seem to like the wireless part of that router.
You only real option is to try another router or connect one computer directly to the internet and do some speed tests.
I personally have a netgear WG614 that gets 19Mbps. When I check here.
He’s asking if you bothered to hook your computer directly to the cable modem. It may be your cable modem or connection itself that’s slowing things down (oversold node, etc.)
Have you tried your connection at different times of the day? Even your 20Mbps connection is going to be affected by how many users are on the system.
When i test my 10Mbps connection on a variety of speed test sites, as well as with large files from known fast servers, my download speed can be anywhere from about 6.5 to 10 Mbps, depending on the time of day. In the middle of the night, i get basically every megabit that i pay for; at 8 in the evening, it’s likely to be a bit slower.
The time of day in my neighborhood has never been a factor. There are very few computer users here and most are only doing weekend granny email.
Yes, the WCG200 is a combo cable modem + router, so it is impossible to bypass the router without hacking the insides. Seemed like a good idea at the time – it reduces the number of boxes on the desk from 2 to 1 – but now I wish I had separated them.
I tried a chat line last night at the linksysbycisco.com site, and the tech said none of their products go above 10MB on the ISP side. However, I don’t think English was her first language and I had to ask the same question many times before I got any kind of answer that related even remotely to the question, so I don’t trust that source much. And it seems beowulff’s experience contradicts that.
I have 2 other units of the same model here. One I know has a very weak wireless signal, so I won’t try that one, but the other should be OK. The only problem is I have to call Charter to change the MAC address and that can be a hassle. Worth a try, tho.
While it is true that these cheap routers are pretty slow, 10mbit/s seems quite low. Does it have any sort of “smart” routing turned on like QoS? I have a Linksys WRT54GL and it craps out around 40mbit/s but with QoS turned on it only gets about half of that.
I’m not familiar with QoS and don’t recall seeing it anywhere in a config screen. Where should I look?
I replaced the WCG200 with another, identical, used unit, did a hard reset and set up the minimum configuration (no encryption) and got Charter to change the MAC. So far, the speed test on one computer shows no difference (about 3-5Mb when 20 is the theoretical max, but it’s a slow computer). I will do the same test on the faster computer in a few minutes, but it’s tied up right now.
While changing the MAC, the Charter tech said this model will handle 20Mb but no cable modems exist that will go faster. I doubt both of those statements, but he said it with authority. :rolleyes:
Trying the fastest computer I have, I get about 8.3Mb/sec download, 1.7Mb/sec up. The up value looks like it is a nominal 2Mb minus CPU overhead, and the down value looks like 10Mb, but Charter doesn’t provide a 10/2 pair, only a 10/1 or 20/2. So it looks like the problem here is the WCG200 (or else Charter isn’t giving me what they are billing me for). I guess the only way I can find out is try a different cable modem, but I don’t see anything in advertised specs of modems for speeds to/from ISPs.
But neither spec applies for a WiFi G. And besides, the 10/100 Ethernet spec is for the LAN, not the WAN. The WAN (Wide Area Network) is the cable to ISP connection, the LAN (Local Area Network) is the local network.
I spoke too soon, googling your model number shows this:
1 x 10 Base-T/100 Base-TX • 1 x RJ-45 for ADSL
>one hard-wired to the Gateway,
Are you sure this computer is set to do 100mbps? How are you testing? You may be throttled by your testing site. You may need to tweak your tcpip settings or mtu to get that 20mbps. What OS are you using?
Lastly, they might be advertising their burst speed. You could be getting 10mbps with a 20mbps burst.
The wireless one will almost never get close to your actual speed. Wireless has tons of overhead. A “solid” 54g connection may only give you 8-12 mbps of usable bandwidth. You can experiment with iperfif you like.
>But neither spec applies for a WiFi G. And besides, the 10/100 Ethernet spec is for the LAN, not the WAN. The WAN (Wide Area Network) is the cable to ISP connection, the LAN (Local Area Network) is the local network.
First off, like I wrote above 54g isnt a useable 54mbps. Wireless has tremendous overhead.
Secondly, the WAN port determines the max speed of the WAN connection. I didnt realize that this one was an all in one cable modem. I thought the router was plugging into the cable modem via the WAN port, thus causing a bottleneck, like in this scenario.
100mps isn’t a factor here, 802.11G is the connection to the fastest machine. WiFi may have a lot of overhead, but not that much.
Nope, 20Mb thruput (DL) is guaranteed and what I am being billed for.
I’m not getting above 8.3. I guess the only way to test that would be to connect that computer with a wired connection instead of wireless. I have a very long cable, but I don’t have a spare Ethernet card handy at the moment to try that.
OPsys is Win XP SP2 for the faster machine.
What good would a testing site be that throttled the tester?
Possibly tweaking some params would improve things, but I see no difference in the 10Mb and the 20Mb speeds supplied by Charter, so I don’t think that is the problem.
And maybe you realize that now, but this unit is a GATEWAY, or a cable modem/router in one package and cannot easily be separated.
I’m gonna pick up another cable modem; I saw one that advertised a 38Mb “upstream” connection and it was $36. And I’ll get another Ethernet card, then I can test two theories: the modem is holding us back, and the wireless is holding us back. One or the other will have to give.
But before I can get that stuff I’m going to run a peer-to-peer LAN speed test thru the wireless and see what kind of speeds that will reach. If it’s much greater than 10Mb, then the wireless is not the culprit.