Amplifing wifi

Thread Starter

Overclocked2300

Joined Apr 24, 2005
124
Alright, not illegal, but speak in terms on db. Ive seen many sites that use dB as a measure for signal strength. Recently Ive come across a passive wifi repeater. I though what if I could use a small amp to amplify the signal to compensate for losses?

If the orignal signal was 15dB and the signal coming in to the repeater was 10dB and I had a 5 db amp, the signal would be restored to it orignal db.

Now, we all know wifi is either 2.4Ghz or 5Ghz. What is Fc when speaking in terms of a transistor? I see Fco of 500Mhz and 300Mhz. What is it? Is it the max freq the transistor can amplify before gain is less than 1? Whats Ft? (not feet) It has a number in Ghz next to it.

So If I were to design this amp, I would need a transistor of a Fc greater than 5Ghz Correct? I was going to use (err bare with me here..it uses a voltage divider circuit on the base, and the output is taken off the collector..) Common emiter? amp.

Would I also need a Bandpass Filter(class C amp) for 2.4Ghz or 5Ghz so no new noise gets introduced? When they say 2.4Ghz do they mean the carrier signal?
 

Brandon

Joined Dec 14, 2004
306
Actually, it is illegal to boost the wifi signal above a certain level. All public broadcast signals must be under a set power rating including 802.11.

With a repeater, it doesn't really matter if it enters it at 5db or 15db, it takes the signal, normalizes back to a full peak to peak waveform and retransmits it. Its like owning multiple access points all sharing the same MAC. (Are you sure its passive? To transmit you need to power a signal and have it propagate off of the antenna)

There are some software hacks you can find online which will let you alter the firmware in your AP and boost the signal from the inside. Additionally, you can pick up a few db by just getting a better set of antennas for the AP.

Rather than just a single transistor hoping to get the frequency you want, you are probably going to want to use a multistage amp with huge gain so you can sacrifice enough of it to push the bandwidth out into the gigahertz and not attentuate the signal... but unless you have an oscilloscope or frequency counter which can detect signals at that rate, you kinda SOL. You would probably be stuck simulating the circuit and trying to build it to the sim specs.

Yep, carrier is 2.4Ghz.

<Hope this is a correct memory from fields n waves>
Lastly, this is not something you can just throw together. Frequency has a huge role here. at 2.4Ghz = 2.4E9 cycles/s. c=3E8 m/s. So we get that c/f=meters per cycles and we see that 1 cycles of 2.4Ghz is only 12.5 cm long, so that in only 3cm we have a 1/4 wavelength which is a signal 90 degrees out of phase than the original. You would need to surface mount your circuit with very short traces to counter this. Ever wonder why the memory clock in the computer is typically only 400 Mhz where as the CPUs is in the gigahertz?? (the 800 in Intels is just a double wide bus) This gives 75 cm for 1 wavelenth(400Mhz), 1/4 at 17 cm about. The traces between the RAM and your CPU much shorter than 17cm to keep the signal accurate. Now in the CPU, everything is tightly packed so distances are small, hense you can go much faster.
 

Thread Starter

Overclocked2300

Joined Apr 24, 2005
124
but also I just remembered, wifi is bi directional, meaning it goes 2 ways (or full duplex). So it would be kind of impossible to build since the signal going out of 1 amp will be picked up by the other, thus creating a loop (impossible meaning at my level of electronics it is).

thanks though! I was dealing with theoritcal work, and wasnt really going to build it. But it does come in handy If I want to build a transmitter on the Higher Freq spectrum.

Yes that info is correct btw. I remember reading about it.
 
Top