Re: Sveasoft on WRT54GS continued

SoEven after you agreed to leave the list you're still trying to so that 
you know more than EVERYONE else, but doing it offlist. You just made my 
blacklist, anything further from you goes directly where it belongs, in 
the crapper.
I will make few comments though, I ALWAYS said this was for P2MP, I 
never made any representations of P2P.
I never made any reference to out of band emmisions, I only provided the 
math. Use of certified equipment is the responsiblity of the user & I 
made no attempt to discuss a subject that has been discussed to death. 
Also it was beyond the scope of his question.
yes the FCC talks in dbi not dbd and since the antennas used WiFi are 
dipole not isotropic radiaters, I provided the conversion to account for 
the difference.

Jim Thompson wrote:

> Ran across this:
>
> http://www.seattlewireless.net/pipermail/dev/2005-April/021734.html
>
> Comments in-line:
>
> *-----------------*
>
>> Well the math is "pretty" straight forward. The FCC allow you to put 
>> 1W (1000mW = 30db) into a 6db antenna for a max radiated power of 4W 
>> (36db) on either end of a P2MP circuit.
>>  
>>
> Yes, this is part of it, but you can't exceed spectral limits for out 
> of band emissions,
> either.
>
>> You are allowed to subtract cable & connector losses between the 
>> final rf ampand antenna input & compensate for this difference with a 
>> higher gain antenna. For simplicity, I will ignore any losses from 
>> here on & assume that the rf amp is directly connected to the antenna.
>>  
>>
> Note that any RF amp *MUST* be certified with the rest of the system.  
> Its against
> the rules to just drop one in.
>
>> For EVERY 1db of antenna gain ABOVE 6db, you must reduce the final rf 
>> output power by 3db on a P2MP circuit. You may NEVER exceed 4W (36db) 
>> erp on either end of a P2MP circuit per the FCC.
>>  
>>
> For a P2MP link, you have to reduce the tx power 1dBm for every 1dBi 
> of antenna gain
> above 6dBi.   (Not 3:1 as you state.)
>
> You are correct in that the 4W (36dBm) EIRP limit applies.
> Note that the FCC rules specificly cite dBm, EIRP and dBi.
>
> 'dB' is a unitless measurement, its use above is meaningless.
>
>> P2P circuits are different.....Just looked I don't have them on this 
>> computer, I'll get them for you later & post them.
>>  
>>
> The rule (in 2.4GHz) for P2P circuits are that you're allowed up to 1W 
> (30dBm)
> into 6dBi of antenna gain.  Then for every 3dBi above the original 
> 6dBi of antenna gain,
> you must reduce tx power by 1dBm.
>
> Thus, a 9dBi antenna couldn't have more than 29dBm of power into it. 
> (Legally).
>
>> dbm = 10*LOG(mW) so 64 mW = 18.1 dbm, 48 mW = 16.8, 84 mW = 19.2 dbm.
>> So the MAX antenna gain you can use is 36-18.1=17.9db, 
>> 36-16.8=19.2db, 36-19.2=16.8db.
>>  
>>
> This is for p2mp links.
>
> it also does't account for how the antenna gain will tend to raise the 
> sidelobes, and this
> can cause out of band emissions, especially when the radio is operated 
> toward the band edges.
>
>> Now one last thing, antennas are typically listed in dbi, you should 
>> calculate using dbd to get ERP otherwise you will get EIRP. Since dbd 
>> is 2.2dbm lower than dbi you can add 2.2 to the above antenna gain to 
>> get the correct ERP, so you final antenna gains may be 20.1dbd, 
>> 21.4dbd, and 19dbd respectively.
>>  
>>
> dDd and dBi are mostly interchangable, (with the conversion you 
> supply) but the FCC speaks in reference to dBi.
>
>> It is not hard to figure setup spreadsheet to do the math for you. 
>> Calculating Fade Margins (otherwise known as System Operating Margin) 
>> is more difficult.
>>  
>>
> Not really. 
>
>> You next question, since we are talking about higher gain antennas, 
>> should be(if you don't already know) "How does the HalfPower 
>> BeamWidth affect the circuit?"......
>>  
>>
> this is potentially more complex than you know, due to mac-layer effects.
>
>> Bryan
>>
> *-----------------
> *
>
>> / On a slightly related topic, when Sveasoft lets one specifiy one 
>
> />/ antenna for transmit and the other for receive, which is the right 
> />/ antenna when facing the front of the box?
>
> Sveasoft are idiots.
>
> When you do this, you induce a T/R mismatch, and the receiver has no 
> way to compensate.
>
> The Sveasoft power 'hack' is also worse than useless.  The EVM of the 
> signal goes completely non-linear
> when its applied.  You'll end up with lots of 'S', along with a lot 
> more "N", so SNR will go down.  This probably isn't what you wanted.
>
> Feel free to forward this to the dev list if you want.  I dropped off 
> months ago.
>
> jim
>
> /
>
>
> *
> *

_______________________________________________
Dev mailing list
[email protected]
http://seattlewireless.net/mailman/listinfo/dev