Most cheap D-Link routers I test are radiating in excess of 42mw. It's no wonder I and many of my friends are having problems with DD-WRT. So, 42mw should be the MINIMUM as the DEFAULT!! Also in locations where there are draft-n Mimo router, 28mw causes FREQUENT wireless DROPS!

[Jack]

I mean make so that in routers that have 28mw as default, make so it's 42mw. That's one of the primary reason for Tomato's stability, as many do not know to change from the defaults.

[BrainSlayer]

i will consider it. the reason for 28mw is the linksys history since linksys uses 28 mw

[Jack]

Most cheap D-Link routers I test are radiating in excess of 42mw. It's no wonder I and many of my friends are having problems with DD-WRT. So, 42mw should be the MINIMUM as the DEFAULT!! Also in locations where there are draft-n Mimo router, 28mw causes FREQUENT wireless DROPS!

[Jack]

Sorry, I accidentally re-opened it!

[BrainSlayer]

the new default value is 70mw (buffalo uses this too)

[mixmasta]

Sorry to open this back up, but do you know why the default value was set to 28 mW? Was it because the Linksys transmitter gets excessively warm? If not, was it because there is no point in setting the "default" value for a router with a standard 2 dbi gain antenna to transmit a signal that can't possibly be returned by a standard wireless pc nic with a sub 25 mW transmitter and low gain antenna? Someone is suggesting that by increasing the power that you get a better signal in crowded wifi areas. This may be true, but now you are causing everyone else (and likely new people too) the same problem. If it is truly the case that different 802.11 standards work better or need different signal strengths to function optimally, then this default value should be set based on chipset or transmitter type, or some combination of both. The default value should be the setting that gives the most users the best experience. It should NOT be a setting that increases interference with others with little real world gain. It should not be a setting that tricks your client wireless nic into indicating it can connect to an access point 1/2 mile away without any special equipment. Adjusting the radio output power has its place in specific cases; like when trying to create a long range link or to compliment and fine tune a network with high gain antenna(s). I just don’t think making routers with dd-wrt firmware transmit at a minimum of 2x their current EIRP by default is the right solution for the apparent problem. If 42 mW is so great, why not start there?

[Jack]

42mW is GREAT, but 70mW is EVEN BETTER, especially in areas where there are cheap MIMO routers. Remember, you can get a CHEAP MIMO router for about $50US at Fry's Electronics. 28mW were used when there were few wi-fi gear around, in the 802.11b days!

[anonymous]

251mw would be even BETTER :-)

[BrainSlayer]

251mw is critical. but 70mw is safe for all routers. the transmitter is basicly the same for all models. very old routers can just handle a maximum of 75mw. so 70mw is safe for all existing phy's. there is no chipset available which can't handle this power.

[anonymous]

Why not 84mW?? It seems to be the most commonly used value being quoted all over the internet. 14mW x 6 = 84mW

[soulstace]

Wake up! Read what BrainSlayer? just said..

[J4k3 aka jacobsuter@…]

Replying to BrainSlayer:

251mw is critical. but 70mw is safe for all routers. the transmitter is basicly the same for all models. very old routers can just handle a maximum of 75mw. so 70mw is safe for all existing phy's. there is no chipset available which can't handle this power.

I disagree. My spectrum analyzer (a real one) says most wrt54g's start blooming profusely above 50mW (15 dBm) and some start even lower. Sure some are just fine (for a while) up to 100mW, but thats temporary.

Running around turning up the txpower beyond its original setting is basically turning your average dd-wrt geek into a criminal in the eyes of the FCC and pretty much every other governing body that handles RF management/usage. If a user is too resource-limited to properly inspect the RF output of their device (ie - own and know how to operate a proper lab-grade spectrum analyzer) they really shouldn't be jacking around with the txpower setting.

If this is made policy for future releases I'll have to suggest/deploy something other than dd-wrt. Also, in the radio world, just because the transmitter doesn't catch on fire doesn't mean it can properly support the transmit power you're requesting...

So, please, reconsider this. As for the power-whiners - quit buying junk/practically untested hardware and expecting so much from it. If you want to play, buy a spec-a and prove your gear isn't causing interference at the power levels you set.

[anonymous]

Replying to J4k3 aka jacobsuter@gmail.com:

I disagree. My spectrum analyzer (a real one) says most wrt54g's start blooming profusely above 50mW (15 dBm) and some start even lower. Sure some are just fine (for a while) up to 100mW, but thats temporary.

err, I meant 32mW/15 dBm, not 50mW/17 dBm. Sorry for the confusion...

[mixmasta]

Replying to anonymous:

Replying to J4k3 aka jacobsuter@gmail.com:

I disagree. My spectrum analyzer (a real one) says most wrt54g's start blooming profusely above 50mW (15 dBm) and some start even lower. Sure some are just fine (for a while) up to 100mW, but thats temporary.

err, I meant 32mW/15 dBm, not 50mW/17 dBm. Sorry for the confusion...

Would you be willing to post some charts and analysis of the data you have collected, and maybe some pictures of the equipment you used and items you've tested? I have a feeling no one here will listen otherwise, although I am in complete agreement/belief of what you have said. It would quite frankly be educational way beyound the scope of dd-wrt as well, as I believe there are quite a few people trying to use these lower end home user type units as cheap replacements for enterprise equipment.

[mixmasta]

Here are some spectrograph images showing a wrt54g at different power levels.

​http://explorer.cyberstreet.com/wrt54g/WRT54g-spectraloutput.html

It does bloom, period.

[tgos]

Since the link posted by mixmasta is dead, here is a link into the Internet archive, that has kindly archived the images:

​http://web.archive.org/web/20080403081757/http://explorer.cyberstreet.com/wrt54g/WRT54g-spectraloutput.html

You can clearly see that the WRT54g peak gets more and more noise to the left and right of it, which will cause interferences on frequencies where it shouldn't. Compared to the last image, which broadcasts at 200 mW, but has a very clean signal.