How to improve 2.4 Ghz sensitivity and output power (EIRP) on TP-Link TL-WDR4900v1 (N900)
This router is quite old but it is very powerful:
PowerPC CPU @ 800 Mhz, 16 Mbytes of flash, 128 Mbytes of ram, 2x2.0 USB ports, 1 Gigabit/s ethernet, OperWRT fully compatible.
This router has 6 MIMO antennas, three external (5 Ghz 5 dB) and three internal (2.4 Ghz about 2 dB)
Since I don't use 5 Ghz at the moment, but only 2.4 cause I use this router to connect to my internet provider with a Python script.
The actual signal is just -76 dBm (50%) and so the internet speed obtained is only 5 Mbit/s down/up, any dB gained will mean some more megabit/s gained
My plan is to exchange internal connections by desoldering the wires, exchange and solder them again to have the 3 external antenna connectors connected to the 2.4 Ghz module, then I want to use three 5 dB (or more) external antennas designed for 2.4 Ghz.
The right section is the 2.4 Ghz radio module (gray wires), the left one is the 5 Ghz module (black wires).
Since the wire for the lower-right internal copper antenna is too short to be connected to the 5 Ghz module (on the left side of the mainboard), I'm planning to exchange this antenna with the center-left copper antenna that has a very long wire.
The upper-right antenna has also a short wire, I don't want to change it, so I will place this antenna somewhere on the other side.
It seems the chassis has place for six copper antennas, three on the left and three on the right, I will check it and let you know.
Maybe later on I will add three RP-SMA connectors for external 5 Ghz antennas to have something similar:
Another option is plug a high power 2.4 radio into the USB port, and use internal wifi only for local connections if at all.
I would just move the black cables over to the 2.4 side and leave the 5 radio shut down and without antennas for now. The internal antennas with gray cables are clearly tuned for 2.4.
This is exactly what I did
I moved the black cables over to the 2.4 Ghz module, paying attention to avoid short circuits (it was a very difficult soldering) and leaved the 5 Ghz radio shut down and without antennas (no 'gain', no pain ).
The first results are quite impressive, I think I moved from 20 dBm to at least 22 dBm EIRP (50% more of irradiated power) considering about 1 dBm loss for the SMA connector, also I greatly increased the sensitivity of the receiver.
I don't know if moving the three antennas between them will change anything, actually I didn't pay attention to the original antennas order (left, center, right).
Thanks
I agree, but by using a high power USB radio you can increase output power but not always the sensitivity.
Most of the times you can get much better results by increasing the antenna(s) gain than increasing the electrical output power.
Too much signal will overwhelm the remote router and cause signal distortion, receiver saturation and poor performances.
I think you meant ERP, not EIRP (which is hypothetical).
I don't think you will get a better antenna than the one installed by the OEM; but that doesn't matter, as I see you're re-configuring them. Since you're asking about increasing power, I assume beyond the hardware's current limits - both of which are illegal anyway (at least, in the US).
True, antenna gain is always the first thing to consider in a point to point application.
A remote radio often allows mounting the antenna (with radio) in a better location to reach the other point. It's very hard to overload the other radio unless you're within a meter or two of it.
The primary difference between them is that for ERP, the antenna gain is expressed relative to an ideal half-wave dipole antenna whereas with EIRP, the antenna gain is expressed relative to an ideal (theoretical) isotropic antenna.
The concept of an isotropic antenna is often used as a reference antenna for the antenna gain.
Here in Europe the limit is 20 dBm, but measured outside from your private property, in my case I'm using the router inside my home, outside of the walls the ERP/EIRP power is less than 20 dBm
Anyway, noboby will never mind or come to measure if I'm trasmitting 2 dBm over the limit.
I'm wondering if the problem you had with low tx power was the same that I'm seeing? For my regulatory domain, it's resulting in reductions of txpower for both the 2.4GHz and 5GHz bands. It might depend on the regulatory domain you've set. Details here, I'd be interested to see what output the commands detailed in that bug report are for you, and whether they are consistent with your measured rx powers? https://github.com/openwrt/openwrt/issues/10830
n.b. for WDR4900 phy0 is 5GHz, and phy1 is 2.4GHz.
).