FR: Add T-Echo-Lite Variant

[SoulOfNoob]

Lillygo has a smaller and newer version of the T-Echo: the T-Echo-Lite

It has a different display, pinout and (optional) GPS
https://github.com/Xinyuan-LilyGO/T-Echo-Lite

I own one and would like to add support for MeshCore since Meshtastic already supports this version.

But i'm new to creating forks and PRs across repos.. so no idea if i'm doing it right.

• fork and add new display to GxEPD2
• fork and add new variant to Meshcore PR (Work in Progress)
  • Fix GPS

[SoulOfNoob]

The TechoBoard.h file was put into the t-echo folder in a recent refactor so i needed to copy it also to the lite folder. does anyone have a good idea how to reuse the code properly?

[SoulOfNoob]

Display setup was a pain but works now. Dialing in the scaling and offsets needs feedback from more users to be verified.

[fdlamotte]

The TechoBoard.h file was put into the t-echo folder in a recent refactor so i needed to copy it also to the lite folder. does anyone have a good idea how to reuse the code properly?

That's a thought I had the other day, might not be a good idea to move board files in variant folders as we can't put two variant folders in the includepath (or it will conflict ;))

In your case, it might be better to have your own board file, we can consider its a totally different board ;)

[fdlamotte]

I'm always glad to see someone starting a port for a new device ;)

I'll probably make a full review tomorrow (I don't own the device but I'll check for some things, such as modifications on the display class)

It would be better if the PR was done against the official meshcore source ;) because in the end it's there that the code will be pulled ...

I have a first remark on your PR, is that you removed code that was used by other variants (there are several devices relying on GxEPD2Display.h, Heltec MeshPocket and Seeed tracker L1 and they both use 213_B74 display ;) ), please make sure the code at least compiles with these two variants

Also you replace initialization that works for two other devices by your own ... we rely on the initial 128x128 size that we then scale if you want to do in another way, it's up to you but you should find a way to have other targets still working as intended (this morning I discovered that techo would not build because someone had introduced a new symbol for gps init ... but the worse is that the symbol already existed with another name ...).

So my word of advice would be, please try not to break other ports when doing yours ;)

[SoulOfNoob]

Ah ok then i'll create a new PR toward the Meshcore source, i'm new to this kind of colaboration.

I Removed the explicit inclusion of the 213_B74 because all of them are included by the GxEPD2_BW.h.

The only reason why i left the GxEPD2_150_BN.h in is because it is explicitly referenced on line 30 but actually that's also not needed.

I also removed the explicit inclusion i used during testing for my GxEPD2_122_T61 and everything works fine. So i assumed this was an unused inclusion that wasn't cleaned up.

I'll check again that everything else compiles and revert the changes accordingly!

Sadly i couldn't find any automated testing :/

Compilation Test:

• Heltec MeshPocket
• Seeed tracker L1
• T-Echo

[SoulOfNoob]

@fdlamotte can you explain the display scaling a bit more? why is it 128x128 and not the true display size of 128x250 as defined in the GxEPD file?

I want to avoid seperate ifdefs for just this part.. so either i find a way to use the 128 as well or i add that as a build var to the platformio.ini or i get it from the display file.
Maybe take the ::WIDTH for both values? would eval to 128x128 for the 213_B74 and 192x192 for mine.

[fdlamotte]

Hi, You should make your PR against dev branch  Regards -- Florent Le 2025-09-24T14:29:33.000+02:00, Jan R. ***@***.***> a écrit :

…

 SOULOFNOOB left a comment (meshcore-dev/MeshCore#824)  [#824 (comment)]  And also i have no idea why my PR wants to merge 27 commits. i only  made 8 and my forked dev branch where i branched off is up to date  with the upstream dev branch...    —  Reply to this email directly, view it on GitHub  [#824 (comment)],  or unsubscribe  [https://github.com/notifications/unsubscribe-auth/ABCB5JAMUYXPL74WC4VHW5T3UKFCXAVCNFSM6AAAAACHJVT6U2VHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZTGMRYGE4DANZVGM].  You are receiving this because you were mentioned.Message ID: ***@***.***>

-- F. de Lamotte

[fdlamotte]

@fdlamotte can you explain the display scaling a bit more? why is it 128x128 and not the true display size of 128x250 as defined in the GxEPD file?

I'm not really fan of what has been done, but that's how it works for now ;)

We did this because users complained about size compared to let's say heltec v3 which have different pixel density

The scaling affects mostly the pixmaps and text position in the end for text size we use different fonts.

The size given when creating the display class is not the actual size of the display but the size for the user, we then scale things to make them look as the user wants.

I want to avoid seperate ifdefs for just this part.. so either i find a way to use the 128 as well or i add that as a build var to the platformio.ini or i get it from the display file. Maybe take the ::WIDTH for both values? would eval to 128x128 for the 213_B74 and 192x192 for mine.

I think you're close to what is done on first TEcho ... so maybe you can just try with the same values ?

[SoulOfNoob]

ok i reverted to the 128x128 and adapted the scaling. it cuts of long node names in the top but everything else is displayed properly. new PR is here and i compiled against all other targets which define a custom eInk al all of them succeed.

[SoulOfNoob]

Merged!