Add CLI control to LoRa's fem LNA.

[Quency-D]

It is currently being tested in heltec-v4.3 and can be configured remotely.
The Python script used is https://github.com/Quency-D/meshcore-cli/tree/lora-lna-cli, which will be further improved later.

A416A94E|*> /9791b78edd4a/get radio.fem.rxgain
Heltec Repeater (D): > on
A416A94E|*> /9791b78edd4a/set radio.fem.rxgain on
Heltec Repeater (D): OK - LoRa FEM RX gain on
A416A94E|*> /9791b78edd4a/set radio.fem.rxgain off
Heltec Repeater (D): OK - LoRa FEM RX gain off

[Quency-D]

Solve this problem #2128.

[towerviewcams]

@Quency-D @IoTThinks I flashed the firmware from the pre compiled firmware.

I'm not sure why Quency was telling me to run a python script? anyways, yes, thank you @IoTThinks, the cli commands work! I was going to try this first for sure.

Powered on 4.3 board, LNA on and repeater working normal. noise floor -103, transmit power +28.3 dbm.....working normally.

here we go-

"set radio.fem.rxgain off" give me a reply "OK - LoRa FEM RX gain off"
I can confirm that the LNA is OFF, noise floor dropped to -118
power draw dropped from 12.3mA down to 5.8mA (this is with powersaving on of course, its older version as we all know)
Transmit power remained at +28.3 :) very very nice Quency!!

now to turn back on LNA-

"set radio.fem.rxgain on" give me a reply "OK LoRa FEM RX gain on"
LNA is now on, power draw increase back to 12.3mA
Noise floor clamped instantly -120......tried agc rest command at 4, 8, 16.....no luck
So then I try reboot and success! noise floor -102

Conclusion-
This works!
Reboot is needed when turning the LNA back on, otherwise clamps to -120 even with agc at 4,8,16
Transmit power remains +28 both with LNA on and off
@weebl2000 @AI7NC

Update
I've also learned that you can turn off the LNA, then, flash Powersaving 14 back into the 4.3 board and the LNA remains OFF. I have rebooted and removed power. Still holds true - LNA off. This is huge progress forward!

[Quency-D]

I'm not sure why Quency was telling me to run a python script? anyways, yes, thank you @IoTThinks, the cli commands work! I was going to try this first for sure.

Python scripts can be used for remote control, but it's also fine to control it directly like this.

[towerviewcams]

I'm not sure why Quency was telling me to run a python script? anyways, yes, thank you @IoTThinks, the cli commands work! I was going to try this first for sure.

Python scripts can be used for remote control, but it's also fine to control it directly like this.

It worked great @Quency-D and thats what 99% of people will do. Things are coming along nicely.

UPDATE I also found out that the LNA will stay in bypass even if I flash Powersaving 14 (no erase) after loading your test firmware. This way sleep is instant. I've been testing this all day since this morning when you contacted me on the other issue that was closed for this PR. Thank you for all your work on this sir!

[IoTThinks]

@towerviewcams So it will best fit you to have LNA on or off?
I am thinking if LNA off benefits more then LNA should be off by default.

[towerviewcams]

@towerviewcams So it will best fit you to have LNA on or off? I am thinking if LNA off benefits more then LNA should be off by default.

@IoTThinks that depends on the repeater site. Most sites you can have it on, "IF" you have proper filtering and shielding. Having the ability to turn this on or off is SO PERFECT i cannot tell you how this will change everything deploying this board.

I'm 60/40 that the LNA should be off. This will pretty much setup people to have a nice working repeater. Then, if they have the expertees on how to check for noise, have the proper equipment or just want to try it, then yes, turn it on.

I understanding more of what your saying Kevin, I am. More people will benefit initially with it off, then on for a turn key build default.

[IoTThinks]

@towerviewcams Most of repeaters do not have proper filtering and shielding I guess.
Nevermind, enjoy.

[towerviewcams]

@towerviewcams Most of repeaters do not have proper filtering and shielding I guess. Nevermind, enjoy.

@IoTThinks I agree with you. I do, look up ^. I just really needed to think about what your saying on a more simplified safe scale for users. This would force more learning for those that don't understand what a LNA can truly do.

[IoTThinks]

Hi,
I'm a bit confused.
what is the difference between set radio.fem.rxgain on/off vs. set radio.rxgain on/off?

• This controls the external LoRa FEM receive-path LNA where the board supports it.
• This is separate from radio.rxgain, which controls the radio chip receive gain mode.

#1653

[IoTThinks]

@Quency-D I hit "Error: unsupported by this board" for Heltec v4.3.
https://github.com/meshcore-dev/MeshCore/pull/2140/changes#diff-1dd5aa67ab63293d7e31b9b7e8e424cf33f77662d3507d37149a058e4f2e1c7d

Over here, in order to set fem.rxgain, canControlLoRaFemLna() must return true.

Then over here
https://github.com/meshcore-dev/MeshCore/pull/1867/changes#diff-dfc1af77dba320128020094eb5a1d150690ba9451c0ae35b84858b8f9c023b9cR27

bool lna_can_control=false; => Can not set fem.rxgain.

Am I missing something?
Thanks a lot.

[IoTThinks]

When I modify to bool lna_can_control=true, then I can set the fem.rxgrain in my Heltec v4.2.
However, I don't see any clear reduced on power consumption (may be 1mA less) and higher noise floor when set radio.fem.rxgain off and even radio.rxgain off.

[IoTThinks]

Hardcoded lna_can_control=true works. PowerSaving at 6.5mA reported by @towerviewcams
Do we need to set lna_can_control=false for Heltec 4.2 and lna_can_control=true for Heltec 4.3?

[Quency-D]

Hi, I'm a bit confused. what is the difference between set radio.fem.rxgain on/off vs. set radio.rxgain on/off?

• This controls the external LoRa FEM receive-path LNA where the board supports it.
• This is separate from radio.rxgain, which controls the radio chip receive gain mode.

#1653

radio.fem.rxgain is used to control the gain of external amplifiers. radio.rxgain is used to control the internal amplification gain of the SX1262.

[Quency-D]

Hardcoded lna_can_control=true works. PowerSaving at 6.5mA reported by @towerviewcams Do we need to set lna_can_control=false for Heltec 4.2 and lna_can_control=true for Heltec 4.3?

Yes, only version 4.3 supports it; version 4.2 does not. I will improve this aspect.

[IoTThinks]

Yes, only version 4.3 supports it; version 4.2 does not. I will improve this aspect.

Yah, thanks a lot.
Looking forward to the improvement.
If Heltec v4.3 can sleep at 6.5mA when radio.fem.rxgain=off then Heltec v4.3 is awesome.
The new board is truely good for low power.

[Quency-D]

@IoTThinks I checked again on version 4.3 and found that this problem does not exist. You can try it again.

[IoTThinks]

@Quency-D Ah, I see.
Then the current code is correct as it returns not supported for my Heltec v4.2.
Let me set back lna_can_control=false and let friends to test on Heltec v4.3 again.

Thanks a lot.

[towerviewcams]

I confirm that the 4.3 board is working great. Have full control of the LNA and the internal Semtec SX12xx small boost. And independent of each other

Tomorrow testing at very noisy broadcast site

[towerviewcams]

I have been running the 4.3 board for many days controling the FEM (LNA) and and also small rxgain in the semtec sx1262....All working great on 4+ repeaters and growing.

suggest this be merged into dev @liamcottle @Socalix @Quency-D @weebl2000

[IoTThinks]

Confirm this PR works for Heltec v4.2 and v4.3.
Many friends have tested this PR fine on PowerSaving 14.1.

The nice thing is Heltec v4.3 can reach 5.8mA when FEM LNA off.
Solid number for low power.

I guess Heltec v4.3 with deepsleep can reach much lower mA.

[etienn01]

Very nice!
Tried a few trace path on a v4.3 without cavity filter:

• LNA OFF -> RX +8dB SNR (noise floor -116dB)
• LNA ON -> RX 0dB SNR (noise floor -87dB)

It works 👍 (also running with PR #1687)

[fizzyfuzzle]

@Quency-D thanks for this PR, its working perfectly 👌

Would it be possible for you to resolve the branch conflicts it currently has? Maybe it can then be merged into the 'dev' branch soonish 😅 🔜

[grantsavill]

Hi @Quency-D. Great work!

How do I get the repeater firmware containing this PR please?

[Quency-D]

Hi @Quency-D. Great work!

How do I get the repeater firmware containing this PR please?

You can simply take this branch https://github.com/Quency-D/MeshCore and compile it.

[Quency-D]

@Quency-D thanks for this PR, its working perfectly 👌

Would it be possible for you to resolve the branch conflicts it currently has? Maybe it can then be merged into the 'dev' branch soonish 😅 🔜

Okay, I'll handle it as soon as possible.

[grantsavill]

Hi @Quency-D. Great work!
How do I get the repeater firmware containing this PR please?

You can simply take this branch https://github.com/Quency-D/MeshCore and compile it.

Using the Power Saving firmware for now with it included, thanks @IoTThinks.

[liamcottle]

Hey, thanks for the PR. Have submitted some review comments. Looks like it needs some merge conflicts resolving too :)

[Quency-D]

Hey, thanks for the PR. Have submitted some review comments. Looks like it needs some merge conflicts resolving too :)

Hi @liamcottle, thank you for your valuable suggestions. I have made all the changes.

[grantsavill]

Not sure if it's the Power Saving 15 firmware or the updated code since Liam's review, but the setting no longer appears to persist across reboots.

[IoTThinks]

@grantsavill PowerSaving 15 uses the code before Quency-D updated yesterday.

For setting across reboot, I will check tomorrow.

By right, the code changes do not change the original behavior.

[IoTThinks]

@Quency-D: There is one bug in the companion radio.
@liamcottle Do I need some permission to add a review?

[IoTThinks]

Not sure if it's the Power Saving 15 firmware or the updated code since Liam's review, but the setting no longer appears to persist across reboots.

@grantsavill Confirmed missing get radio.fem.rxgain and radio.fem.rxgain is not persisted accross reboots.
We are getting patch from this PR, fixed one bug in this PR and rebuild for Heltec v4 shorttly.

[liamcottle]

Do I need some permission to add a review?

@IoTThinks click the green "submit review" button in top right

[IoTThinks]

There is 1 typo in DataStore.cpp of companion_radio.

[IoTThinks]

@grantsavill Applied latest changes in this PR.
Fixed missing get radio.fem.rxgrain and not persisted radio.fem.rxgrain accross reboots.
https://github.com/IoTThinks/EasySkyMesh/releases/tag/PowerSaving15

There are quite significant changes compared to the first commit.
Please help to test.