Thanks for the tip. Good to know I am not the only one having issues with it.
Ironically, I came to the same solution after pulling the kit a part after submitting an issue on GiitHub (Issue):
Slightly unplugging the 3.5mm cable from the output device. You want to pull it out just enough so the plug is still in the jack, but not all the way.
I have some extra cables, so I’ll try your other suggestions as well to see how those work out. I really only need them for TTS or short audio clips, so as long as I can get rid of most of the white noise it should be fine.
Where I had the most issues, was with my soundbar which has a Voice Assistant 3.5mm AUX-In port. I needed to get the audio cleaned up enough to not trigger the soundbar when no audio was being sent. Works like a charm, the sound bar handles the lowering of the media volume automatically.
Hey tim.plas, how did you get your Atom Echo working/recognised in HA? Did you use the yaml code from above via ESPHome or did you use Romkabouter’s ESP-Rhasspy-Satellite code? Wanting to have my Echo above my bed as satellite for bedtime/morning automations
@tome88 - I really would not recommend the Atom Echo as a production device for any audio purpose; there’s just not enough speaker volume to hear it.
You MIGHT be able to hear it well enough in a VERY quiet environment like a bedroom; highly unlikely to wake anyone up with it.
It’s a great toy / educational learning device, in terms of audio. And as an Atom Lite, it is fully functional, and is quite usable for anything the Atom Lite can do.
If you need to hear audio, and you like the Atom series, I’d recommend the Atom Speaker. Also pretty low volume, but at least a real speaker. And it has 3.5mm jack, to plug in a powered speaker or whatever.
I typically build my own YAML for all my Atom devices – I’ve got Atom Lite, Atom Echo, Atom Speaker, & AtomU, I especially like the AtomU devices – I buy thin-format USB wall adapters, & plug the AtomU directly into that; no cables, & everything lies flat against the wall. Great for Bluetooth Proxy and similar applications.
Mostly I use AtomU or Atom Lite. Besides other applications, the Atom Lite works really well with TailBat device from M5Stack – Rechargeable battery for portability (no AC power), provides pass-thru of all connections to Atom Lite. Not terribly long battery time, but sometimes very useful to be portable.
I don’t recall where I got my base starting YAML code for Atom; I’ve got a couple different ones that I use as base templates. I can share if you need one.
@tim.plas if you’re able to share that code here, that’s be excellent. It’s more for the listening for my voice for HA commands via Rhasspy in a bedroom, if I can barely hear the acknowledge audio though and it still does the action it should should still meet my needs.
Ouch; not sure it’ll be suitable for that. Unless things have changed since when I last worked w that stuff (last year), There was NO code support for the onboard microphone. I would love to use it in that way. Maybe there’ll be more support coming, in this “Year of Voice”. But last I checked, I couldn’t find support for I2S microphones in ESPHome.
Update: Support for I2S microphones in ESPHome has been added as of 2023 April update of ESPHome. --in preparation for “Voice Assistant”, coming in 2023 May update of Home Assistant.
So yes, Atom Speaker should indeed now be usable for voice input thru its I2S microphone. I’ll be interested to see code for anyone doing that.
Thanks so much @tim.plas - I’ve been trying to have ChatGPT write some code for me for this with not much luck, so I’ll certainly give this a go this weekend. Appreciate it
Seeing as the M5 atom echo appears to be out of stock everywhere, is there a guide to making one DIY style? I have few esp32’s kicking around, is it just a case of connecting a microphone and say a bluetooth speaker? Do i need a special microphone?
Hi,
I’m looking for a speaker solution that can work with H.A ,without help of Google/Alexa, and found the M5Atom Echo. It seems really small, do you know other solutions ? how is it powered?
As it’s for a solution with NFC tags envolved, a tag reader with a speaker might make the deal… I want to being able to say something like “Hello Paul” or “Good by Paul” when tags as scanned… but a single ESP32 speaker can make the deal, I just need something that don’t look too much “home made” as it’ll be used in a customer office… . any ideas?
I am trying to config my Atom Speaker Kit too. All I can found from google is this post, but I didn’t make it out based on the existing discussion here.
My speaker can play TTS sound, can play radio, but just can’t play mp3. When I select a mp3 file, it shows up playing but no sound at all.
No Error found from the log. Configuraion yaml is done as suggested in Post#2. I don’t know why it just can’t play mp3.
By the way, this media_player based on ESP Home doesn’t support Music Assistant, right? I can’t add it as a Player Provider in Music Assistant.
Yesterday after I posted I found something new. It is not all mp3 cannot be played, it’s just some mp3 cannot.
I used to try with some of my daughter’e audio book without success. But accidentally I tried to play a music mp3 it worked without issue.
Then I tried a few diifferent mp3, music fine, audio books failed.
Now it’s getting really strange…
Question about M5Stack Atom Echo with SPK speaker kit.
It looks like you have this working. I bought a couple Atom Echo devices and was successful in setting one up and it responded correctly to voice commands. Subsequently I bought the SPK speaker modules but I find that when the Atom Echo is stacked on the the SPK module, it stops responding to voice commands. Otherwise it appears in HA and responds to switch changes from the voice assistant setup page.
Did you have to change the firmware file to accommodate the SPK module?