because I need to reduce tx power until my device fails a radio throughput test. On some radios and devices I'm testing, 1dB is a good tx power for differentiating known good and known bad units. On others, the signal at 1dBm is too strong, throughput is not adequately affected.
The above command seems to reject values below 1 (dBm). Is it possible to set lower tx powers using these uci commands?
(There is a hardware solution to this problem, attenuation, but this is less flexible and more expensive.)
I could be wrong - but I just don't see this happening as WNIC's typically expect a value between 1-30 to operate. Consumer grade WLAN hardware is not designed and tested to operate outside of that range that I have seen. Therefore it doesn't matter what UCI commands you send. Maybe someone knows of some test drivers? GNU radio? Aircrack?
For a test I tried entering a negative number into /etc/config/wireless/txpower --> AR750S / DUT ignored it. Also I tried greater than 30 with the same result. It actually broke LuCI-->network-->wireless - the wireless option disappeared. Uninstalling/reinstalling just LuCI didn't fix it. Factory reset worked.
I won't argue too much with you here as test config / method is your call. But it doesn't cost anything to put the DUT in another room or cupboard (or some sort of makeshift faraday cage) to reduce the RSSI to the desired range. I have used aluminum foil to reduce EIRP by wrapping antennas with it. I think it is around 2-3dBm per layer. NIST/ISO9001 compliant? Probably not.
Yes, I think you are right. Our engineering team has told me that the txpower of our system becomes non-linear at lower powers, so I will use hardware attenuation (in my case SMA inline attenuators) to achieve the lower powers I require. Thanks!