You are not logged in.
Hi, I am working on WARP V3 currently with 802.11 Ref_Design v1.5.3. I want to measure what's the minimum tx power of AP. I configured one WARP board as AP and changed the min tx power band in C program to -30dBm. But when I set the tx power below -10dBm in wlan_exp and replaced tx antenna with coaxial cable connected to spectrum analyzer, I found the tx power is larger than 8dBm. Is there any other band that I need to change to get a lower tx power? What's the minimum tx power can WARP provide?
Any reply and advice would be very helpful, thanks a lot!
Offline
The minimum Tx power is -9dBm (TX_POWER_MIN_DBM). Any value you specify lower than TX_POWER_MIN_DBM this will be interpreted as a value of -9dBm. In our testing, we achieved around -10 dBm at the lowest point. Did you mean to write -8dBm or did you see +8dBm?
You can't lower the Tx power any further using the the MAX2829 Tx gain index (which is what we map the power value on to). So, the options for even lower power are:
1. Add an attenuator in between the RF connector and your antenna or coax cable. This will also attenuate your Rx power.
2. Add an RF circulator and have a separate Tx and Rx antenna. Then you could attenuate the Tx side without affecting the Rx.
3. You could disable the PA on the Tx side, which is adding 32 dB of gain. We don't directly support disabling the PA from software, but you can effectively disable it by raising the second argument of radio_controller_setTxDelays() to its maximum value of 255.
4. You could reduce the digital scaling on the output of the PHY transmitter by changing wlan_phy_tx_set_scaling(). Note: there is a tradeoff here since you are creating quantization noise as you start to lower this value. Obviously, at the extreme edge, you'd be using 1-bit of output. That would be very low-power, but there is nothing in that signal that could be decoded at the receiver. So I'd recommend only changing this parameter as a last resort.
Offline
So can you tell me what is the maximum Tx power of Warp v3? Thank you
Offline
See Tx Characterization in the 802.11 ref design user guide.
Offline
Sorry. The data is quite confusing. It says the highest Tx power is 17dBm, but then the graph show it can reach 21dBm. In my application, we want to generate a cosine wave at 2 antenna with max gain (63), so what is the expected Tx power?
Thank you
Offline
We limit the maximum Tx power in the 802.11 design because of the steep increase in Tx EVM at the top of the power range of the PA. I'm not sure what the maximum Tx power of a sinusoid would be. You'd need to use a spectrum analyzer to measure it.
Also, keep in mind that transmitting a sinusoid on 2 antennas simultaneously will beamform. Depending on the relative phase of the two sinusoids at a receive antenna, the signals will constructively or destructively interfere.
Offline
Thank you. Actually we are doing things with beamforming. Currently I can control the beam but the initial phase is random. So I have 2 questions:
1. The phase of the 2 transceivers on WARP v3 are randomly different, am I correct?
2. Is there any way to cancel out the phase difference of the 2 transceivers? or at least to compute the phase difference?
I tried the nodeSync example and I want to use the similar way to compute the phase difference between the 2 transceivers. However, the phase difference is different every time I run the code (with InitNodes commented out)
Offline
vutran wrote:
1. The phase of the 2 transceivers on WARP v3 are randomly different, am I correct?
The relative phase between the transceivers is arbitrary but fixed as long as the board is not powered off or the center frequency of either interface is unchanged. Either of those conditions will re-lock the PLLs to a different relative phase. If you are using WARPLab, you'll want to make sure you only call "wl_initNodes()" once and then continue to use the initialized node objects. Calling that function will reset the PLLs.
vutran wrote:
2. Is there any way to cancel out the phase difference of the 2 transceivers? or at least to compute the phase difference?
I think you'd need a third transceiver on a second kit to act as a reference. You could rig up an experiment through coax cables and a power splitter to adjust the relative phase of your two sinusoids until Rx power at the other node is maximized. You would then know they are phase aligned. If you do this make sure to use enough attenuation between Tx and Rx to not damage the receiver. I recommend 50dB.
Offline