You are here

TP-Link CPE220 v3.0 support in nightly build

6 posts / 0 new
Last post
AE6XE
AE6XE's picture
TP-Link CPE220 v3.0 support in nightly build
Andrew KK4ZUZ strikes again!  The CPE220 v3.0 will be in the Nightly build as of Dec 6th, 2018.  If you purchase a CPE220, these are the latest versions manufactured.   Thus, depending on distributor inventory, you may get this version.   Again, AREDN support in advance of OpenWrt.  Andrew is on a rampage knocking these devices down like clock work.   

The cpe220 device commands:
1) 2GHz
2) 30dBm xmit power specs
3 12 dBi antenna gain
4 MIMO dual polarization
5 $40 on Amazon today

In comparison the  Ubiquit NanoStation M2:
1) 2GHz
2) 28dBm
3) 11.2dBi
4) MIMO dual polarization
5) $80 on Amazon today

Joe AE6XE
AA7AU
AA7AU's picture
25dbm Max on slider with CPE220v3

Running a CPE220v3 and the TX slider setting maxes out at 25dbm: nightly build = current version: 586-6728fba hardware type: cpe (cpe220-v3) - it was the same as when running build #566. Is this normal, it's less than the CPE210s  I've setup.

TIA,
- Don - AA7AU

AE6XE
AE6XE's picture
Yes, some discrepancy between
Yes, some discrepancy between what the tp-link specification and what we are seeing on the device.    The vendor spec says max 30dBm.    We are seeing the device only doing 28dBm max.    But more to the story...  TP-Link for some reason is limiting the power on the outside channels for most of their devices.  Thus, ch 1 is only 25dBm while ch 2 is 28dBm.   Thus, we've kept ch -1 and ch -2 the same as ch 1.    This is why you are seeing 25dBm max.    

We do see the driver limiting the max xmit power by channel.   The current thought is there has to be some data in the vendor flash partition that defines these channel power limits.  No one has dug though the source code to confirm the specifics yet.     Because ch -1 and -2 aren't out-of-box, we could take the power up higher than 25dBm, but then it is unknown what negative effects there might be -- we could speculate why is the vendor is limiting ch 1 lower than max power of other channels?    If someone has ability and interest to measure the signal quality, there is opportunity to investigate increasing the power above 25dBm on ch -1 and -2.

Joe AE6XE
AA7AU
AA7AU's picture
Speculation

I wonder if the ch 1 limit is due to it being on the lower edge of the overall band for them. Do you see the same kind of limit in the CPE210 or is that a different device inside? Seems like I was getting more than 25 with CPE210v2 but none of them is immediately available to check right now. Kind of disappointing to buy a higher power device and actually get lower power.

Thanks for all your work on this,. All the expansion into TP--Link and MikroTik is fantastic. I love my little hAPaclite (except for the actual model#- grin).

Happy Boxing day to all!

- Don - AA7AU

AE6XE
AE6XE's picture
The cpe devices drop power on

The cpe devices drop power on the channel edge as follows -- currently in the AREDN definitions.  The driver will not allow a setting above these values.  We have set ch -1 and -2 power limits the same as ch 1.   I'd speculate the drop of power at band edge is to be conservative to ensure the device passes the part 15 certification tests -- the lower power keeps the PA in limits and minimizes the out-of-band power emissions.   

If further signal quality checks are done, we might be able to increase the ch -2 and -1 power xmit values.    Ubiquiti and Mikrotik are flat xmit power across all the channels, and the exact same Qualcomm-Atheros chipsets are in use.  However, the Power Amp may be different.

device             ch 1               ch 2            ch 9           ch 14
cpe 210 v1.0   22 dBm        22 dBm       22 dBm     23 dBm       
cpe 210 v1.1   22 dBm        22 dBm       22 dBm     23 dBm
cpe 210 v2.0   27 dBm        28 dBm       29 dBm     27 dBm
cpe 210 v3.0   21 dBm        25 dBm       25 dBm     18 dBm      
cpe 220 v2.0   25 dBm        28 dBm       28 dBm      27dBm
cpe 220 v3.0   25 dBm        28 dBm       28 dBm      27dBm       

It's not necessarily a handicap to use 25dBm vs 28dBm.  we are seeing in the spectral plots that there is increasingly higher out of channel energy to make this jump in power and drop in signal quality, pushing the limits of the PA.   

Joe AE6XE
 

 

AA7AU
AA7AU's picture
Is there any progress on

Is there any progress on allowing DtD to be present on LAN0 rather than having to use a second ethernet cable to LAN1 secondary connection to get DtD? Not carrying DtD traffic over the same interface/cable really impairs a more full use of this device.

TIA,
- Don - AA7AU

BTW: I finally un-bricked my CPE220v3 (back at 3.19.3.0 again) but it's a long story full of strange occurances along the way. Will never again turn off RF (for AP) when a node its acting up on other interfaces. WIll try to post more details if/when I get the chance.

Theme by Danetsoft and Danang Probo Sayekti inspired by Maksimer