Hi all,
I recently bought an AirGrid M2 with a 20 dBi antenna and loaded the 3.16.1.1 version AREDN software. The configuration here is Channel -2, 10 MHz Channel Width. For some reason, I cannot increase the Tx Power to 28 dBm. If I change it to 28 dBm, select "Apply" followed by "Save Changes" and "Reboot," it goes back to 19 dBm once I open the Setup page. I read that the output power is limited on some units to prevent non-linearities, but a 9 dB reduction in output power seems excessive. Am I doing something wrong? Do the Bullet M2 units have this same problem?
Thanks,
Dave K3GX
Livermore, CA
I recently bought an AirGrid M2 with a 20 dBi antenna and loaded the 3.16.1.1 version AREDN software. The configuration here is Channel -2, 10 MHz Channel Width. For some reason, I cannot increase the Tx Power to 28 dBm. If I change it to 28 dBm, select "Apply" followed by "Save Changes" and "Reboot," it goes back to 19 dBm once I open the Setup page. I read that the output power is limited on some units to prevent non-linearities, but a 9 dB reduction in output power seems excessive. Am I doing something wrong? Do the Bullet M2 units have this same problem?
Thanks,
Dave K3GX
Livermore, CA
I suspect you have an unknown hardware revision for the version you installed (you will see a warning banner on your nose at the top) but a support info will give us a lot of answers.
I bought this AirGrid unit used on eBay and it looked fairly weathered (yellow). I guess there's a possibility it is quite old.
Thanks,
Dave
OK, here's the file. I was trying to do it from a Mac rather than the Raspberry Pi. The Mac downloads .tar archives while the RPi downloads .tgz.
Dave
(I scanned though some IEEE papers on this linearity issue. Here's what I found.) To keep costs down and increase manufacturing reliability, the PAs in these devices end up with a limited power range. the Peak-Average Ratio (PAR) of OFDM modulation increases with higher rates and bandwidth. The higher the PAR, the bigger the PA's power range needs to be. If it's not big enough, then the PA is no longer linear, then spurs and other bad out of band energy start to appear. The device would fail the FCC part 15 tests which limit how much off channel power can be emitted. Consequently, as shown in the vendor specs for these devices, the power is stepped down to keep the PAR down as the modulation rates step up. Long story short, nonlinearity isn't so much an issue between devices, rather an issue of a given device as data rates increase.
Joe AE6XE
Dave, can you telnet in to the device and pass back the output from the commands below?
cat /sys/devices/pci0000:00/0000:00:00.0/subsystem_device
dd if=/dev/mtd7 bs=1 skip=12 count=2 2>/dev/null | hexdump -v -n 4 -e '1/1 "%02x"
Joe
Thanks for your help. After telneting into the device and using the first command, the response was: 0xe212
The second command just came back with a > prompt but no hex dump. I did check and there is a /dev/mtd7 directory.
Dave
Dave
http://bloodhound.aredn.org/products/AREDN/ticket/71
This is an interesting hardware, originally we got a board.info file for it that said there was no external amplifier and it was only 18dbm output rated. Someone had reported it as a AirGrid HP but it didn't actually appear to be so based on board info files given. The fact your limited at 19dbm seems to imply the lower driver is rejecting the value which we have seen when we exceed the power limit settings in the past and is a new datapoint backing up the max level setting of 18 be applied. Its still not known if these have an undocumented amplifier, or if they are an older hardware that just didn't have the same output power, but the ticket system would be the place to have that discussion.
Dave, this device 0xe212 says from the AirOS firmware, that it is max 18dBm xmitter power -- I don't think we have any basis to not believe the vendor's released data config. We can also see from your support data, that the driver is limiting to 19dBm. No one's tracked down to confirm where this limit is defined, but believed to be somewhere on the flash in a vendor config partition -- not aware of any other plausible locations.
It wouldn't be wise for us to try and exceed these power limits as it would be unknown what out-of-channel emissions we'd see. Rather, the question is, what to make the limit, 18dBm or 19dBM, to remove the "STILL BEING TESTED" message.
As I recall it was chosen to close it as invalid at the time as it was already beloved not to be in production and to be already be no longer available hardware in 2014.
That's a bummer... It sounds like I should reload the Ubiquiti firmware (assuming that can be done) so I can put it back on eBay and sell it to someone wanting to use it in the public bands. I'm not sure how to do that, but presumably I can download a file from UBNT and load it from the AREDN software. This makes buying a used unit a little suspect for me... I should probably go for a 28 dBm bullet and 15 dBi antenna as that would be better than a 19 dBm transmitter with a 20 dBi antenna, except for the receive path.
Is the 19 dBm limit considered maximum power for the public wifi bands with a 20 dBi antenna, so the limited power of this unit would not be a problem for them?
Thanks,
Dave
http://site.microcom.us/ag2_datasheet.pdf shows an AirGrid with a 20dbm output at one time, it’s conceivable (based on how Ubiquiti has silently changed devices) that at one time it was actually only 18dbm.
This is a lot like the NanoStation M series (NSM). Vs NanoStatation Loco M (I forget the shortcode for this)
Each hardware was intended to fill a different price point need, why pay for a 28dbm device if you only need 18dbm. Cheaper to make cheaper to sell.
Personally I recommend you avoid the bullet, go with a dual chain device like a NanoStation or a Rocket.
Dave
Check out this link, nicely consolidates the part 15 power limits on page 2, which are different depending on the channel and if PTMP or PTP. For example, a NanoStation (sector assumed multipoint link), between 5.725-5.85Ghz, the power limit with a 20dBi antenna is 16dBm.
https://www.scc-ares-races.org/mesh/doc/WiFi_Part_15_Power_Limits_v15042...
Joe AE6XE