CPE 510 still low tx power

[mokkin]

The CPE 510 tx problem still exists, even in my first try with LEDE-based gluon. dB is about 20 dbm less.
I'm still using the old firmware at this point: https://github.com/freifunk-gluon/gluon/tree/ffd1f0b3a5b2950d8c626a2de26fb1dc8e384408

[neocturne]

The last update on this matter was https://git.lede-project.org/?p=source.git;a=commit;h=c2e0c41842895ba47819fa98b785c76a2524628b .

[mokkin]

I think it was reverted due to usage of an upper channel. @dangowrt could you add your experience here?
On lower (indoor) channels (36-48) with less tx power there is absolutely no problem and we have a couple of CPE510 working fine and stable since month ago.

[2tata]

I can confirm this problem. I have some CPE510 in combination with CPE210 and all clients tend to use cpe210 because the CPE510 tx power seems very low. state is v2016.2.x.

[mokkin]

I manually used the higher power table with an actual LEDE based gluon like NeoRaider suggest and it worked like before. If you stay on the lower 5GHz channels (36-48) it should work without any damage.

[FreifunkUFO]

@mokkin we could do a gluon- PR/patch!? and include it at least to our gluon-repo

[mokkin]

@FreifunkUFO Yes. I think we should include it at least to our https://github.com/freifunk-leipzig/freifunk-gluon-leipzig as there aren't as many 5 GHz users around here.
@NeoRaider Or could we revert the patch for the lower indoor channels (36-48)?

[neocturne]

This must be fixed in OpenWrt. If a fix is not acceptable for OpenWrt, it is not for Gluon.

Note that #1329 will allow to enable DFS channels on many 5GHz devices by disabling meshing. This will become the default on all outdoor 5GHz devices supported by Gluon, including the CPE510.

[rotanid]

this upstream issue will be tracked in Gluon over there: #1156
closing.

[mokkin]

Do you have any idea how can I revert the patch in the actual gluon master or OpenWRT 21.02? Until 19.07 I can find the postion of the file to revert it, but since migration to ath79 it has changed and I couldn't find it anymore.

[neocturne]

@mokkin The calibration data offset is now defined here: https://git.openwrt.org/?p=openwrt/openwrt.git;a=blob;f=target/linux/ath79/dts/ar9344_tplink_cpe.dtsi;h=9d8873ddaf1b3232aee81e59f3f522ac9b3a9b73;hb=HEAD#l86

The following should work to override it only for the CPE510 v1:

diff --git a/target/linux/ath79/dts/ar9344_tplink_cpe510-v1.dts b/target/linux/ath79/dts/ar9344_tplink_cpe510-v1.dts
index b1fe2364cd65..b728f255f137 100644
--- a/target/linux/ath79/dts/ar9344_tplink_cpe510-v1.dts
+++ b/target/linux/ath79/dts/ar9344_tplink_cpe510-v1.dts
@@ -10,3 +10,7 @@
 &led_link4 {
 	gpios = <&gpio 16 GPIO_ACTIVE_LOW>;
 };
+
+&wmac {
+	mtd-cal-data = <&art 0x5000>;
+};

[mokkin]

@mokkin The calibration data offset is now defined here: https://git.openwrt.org/?p=openwrt/openwrt.git;a=blob;f=target/linux/ath79/dts/ar9344_tplink_cpe.dtsi;h=9d8873ddaf1b3232aee81e59f3f522ac9b3a9b73;hb=HEAD#l86

The following should work to override it only for the CPE510 v1:

diff --git a/target/linux/ath79/dts/ar9344_tplink_cpe510-v1.dts b/target/linux/ath79/dts/ar9344_tplink_cpe510-v1.dts
index b1fe2364cd65..b728f255f137 100644
--- a/target/linux/ath79/dts/ar9344_tplink_cpe510-v1.dts
+++ b/target/linux/ath79/dts/ar9344_tplink_cpe510-v1.dts
@@ -10,3 +10,7 @@
 &led_link4 {
 	gpios = <&gpio 16 GPIO_ACTIVE_LOW>;
 };
+
+&wmac {
+	mtd-cal-data = <&art 0x5000>;
+};

Thanks a lot for you patch @NeoRaider !

Those threee lines worked perefectly on my recent builds with OpenWRT and TX power is back. I guess they will work with gluon as well.

[ddreamer75]

is it possible to change settings on running gluon fw ?
iw phy0 info shows 23dB on channels but all tx-power settings will be ignored (lower/higher).. only 10db