Why are so many people opposed to using amps? Is there a drawback that I am not aware of?
Well, I was using amps.
In the beggining I thought: āThis is cool, this is greatā 
BUT!!!
You will have big problem with noise, user disconnections, system hangings, wireless card stop working, low ping, bad throughput etcā¦You name itā¦
All of this was happening with different amps, different cards, MT versions, antennas, different locationsā¦
I even blamed MT for thatā¦
The bad thing is that somebody could spend time trying to figure out what a hell is happening (hm maybe not just time, nerves too )
Is this enough drawbacks? 
Cheersā¦
OK. I can see this if the AP is what is called an Aligator in the HAM world. (all mouth and no ears) But if the AP is amped to get the reach and the client devices are amped also to get back to the AP, will these same problems still appear?
Thatās even worse situation.
I told ya, noise is the worst drawbackā¦
If you have time and money, you can try to play with that.
I played for a long time and almost lost my customersā¦
Now everything is working superb, online gaming, voipā¦
amps not only raise signal levels they raise your noise floor as well. so now instead of having to get above -80dBm (non-amp) you have to get above -70dbm (ampād) to maintain a decent connectionā¦
You also lose bandwidth and performanceā¦
Check out the 400mW ubiquiti mini-pci cards - http://www.ubiquiti-networks.com/ , those almost eliminate the need for amplification.
Here is the rest of the story on RF Power Amps. If they are bi-directional they will not only increase the signal, but also increase the noise floor. The only time you should amp is if 1. The min level of signal for receive is below what is needed, and that the signal to noise levels are good to start. Then you may be able to raise the siganl level high enough for the radio card to receive. (This is a combination of both the sent signal strength and the receive amplification).
Second, Most amplifiers on the market are built using MMCI chips. They are cheap, but have some problems. The biggest of these problems is the noise that these chip make themselves. If you have a noise level of -93db and then you amp it by 14db, your noise is now at -79db, and most of these MMCI chips have an internal noise level of 5 to 6 db, making your noise level -74db to -73db.
If you signal was at -79db and you added 14db you get -65db receive signal.
Now let us look at the difference. Before amplification you had a signal level of -79db and a noise floor of -93db. This gives you a signal to noise difference of 14db. After amplification you have a signal of -65db and a noise floor of -73db. This give you a signal to noise difference of 8db.
14db signal to noise is much better then 8db signal to noise. Moreover, if you have a good radio card which can receive at -79db you are better off not amplifiing.
There are two ways to correct this trouble. Increasing the signal level out on both ends and not amplifing the receive is the best of all worlds. You can do this by using a transmit only amp with no receive amp; none have been built for sale that I have ever seen on the market, or increase the gain of your antenna system; This is the best way to do it.
For an omni type system using a Slotted Wave Guide antenna that has a gain of 14 to 16db is better then using a Coliner Omni with a gain of 8 to 10db.
For directional antennas, Solid Dishes have the greatest gain. Next is a Grid Dish; Then a Yagi, Etc.
Thatās maybe why the cards are usually having two connectors? One connector can be used for transmitting, and one - for receiving.
I will say that I have had used may types of amplifiers. There is definately a difference between good and bad amps. I totally recommend RFlinx. They have performed very well for us.
Also, I didnāt realize there was a 400mw Atheros based card. Iām guessing it is MT supported.
Eric
Joe Janins and group are some top RF eng. If you are going to AMP your system, RFLINX (headed up by Joe) builds the best ever. Even their Ham and Military Amplifiers are top quality.
The CISPA WIRELESS LAB is still waiting on their new āGā and āAā series with AGP!
I hope to report back sometime soon that they are as GREAT as the āBā series with AGP!.
Joeās amplifiers have some of the lowest noise fig. I have seen around. You will not find a compairison unless you spend Thousands of Dollars!!
If I can reach 20km with 200mw radios and 24db good antenas without any amplifiers , why do we need amplifiers?
Someone thinking to have clients further than 10 km (even that is too mutch) with a 802.11 PTMP wireless infrastrukture, heās making big mistake.
For PTP links, sometims is reasonable to use amplifiers, if you canāt get the link work otherwise.
FOR PTMP, lets say you have interference problems and you amplify your AP. I dont think you can achieve good results with aplifiying only the AP. If clients can recieve good signal from AP, that doesnt mean that they can send good signal to AP, yess you can use bidirectional amplifiers but then you raise the noise floor.
If you want to get good results you need do amplify both sides (all the clietns). Thats expensitve, why not spending that money in better antennas?
The wors thing is if you are in a ānosiyā area you cen benefit a bit from amplifiers only for a limited time until the competitors find out that you ākill the airā with your amps and they start using amps too, and make a big mess in the air.
I use amps 4Wt for 2Ghz - NO ANY NAMED PROBLEMS. Exellent signal, exellent throughput.
vazir, tell us something about your links, antennas used and throughputs.
One thing you need to be very carefull about when amping is that the input of the amp has certain limits.
For example my MT unit has a prisim 200mw card and my 1w amp has an input limitation of a max of 100mw.
Thus I had to drop my prisim output to 100mw or even as low as 30mw. I am running mine at 30mw into the 1w amp. And I get a very clean signal and I can connect to the client sites easly.
I am using the YDI amp located at
http://www.fab-corp.com/
You a right, input must be limited. There is a problem with mikrotik AP features. So far iāve been usink MT for backbone only and just in few cases for AP for 2.4GHz (short distance links) -
As someone mentioned in some of the psts here on longer distances there is latency/throughput problem with MT on 2.4Ghz when distance exeeds 3km, if you connect NON MT CLIENTS.
I did use Lucent bridges and HOST AP Orinoco APās. Everywhere with amps. So I always try to keep 30db SNR on the links - then throughput can reach maximum possible for 802.11b
you know, 802.11g requires DIFFERENT types of amps, due to modulation specifics. And i can say nothing about amps for 802.11g
I use MANUS amps.