EEVblog Electronics Community Forum
Products => Test Equipment => Topic started by: frankvh2 on February 23, 2022, 05:15:58 am
-
Wasn't sure where the best place to post this was - "Test Equipment" seemed right, but feel free to move or delete it if not.
(https://www.frankvh.com/wp-content/uploads/2022/02/stop-closed-housing-crop.jpg)
This is a module to provide voice control for the Rigol MSO5000 family of scopes. Tested on my MSO5072, but should work on any of the family. If both your hands are full holding scope probes, you can call "stop" for example to stop the scope. Or "single" or "run". Simple and surprisingly useful. It works by recognizing voice commands, then sending the appropriate SCPI command to the scope via its network interface.
It's a simple DIY build. Uses three cheap little off-the-shelf boards you can buy on Aliexpress or eBay. If you have a 3D printer you can also print a nice enough little housing for it. Complete details are here:
https://www.frankvh.com/oscilloscope-voice-control-using-esp32/ (https://www.frankvh.com/oscilloscope-voice-control-using-esp32/)
-
Wooo very nice device :-)
Was looking for something like this. Something which not needs a PI...
Another approach was:
https://github.com/patricksebastien/speech2scpi
Maybe i will build your device as well..
-
Cool device.
-
Nicely done. Looks like a tidy solution with nicely integrated modules.
For anyone with a Rigol scope other than the 5000 series, or with a Keysight 1000 - 4000 series, an alternate approach is provided here: https://www.eevblog.com/forum/testgear/diy-voice-control-for-keysight-1000-x-scopes/ (https://www.eevblog.com/forum/testgear/diy-voice-control-for-keysight-1000-x-scopes/) .
One nice thing about Jim's design [which I have built and use frequently] is that it supports a flexible grammar so that any of its 15 or so supported functions can be invoked by several different command utterances. So, for example "time base 10 seconds per div" is the same as "horizontal scale 10 seconds per div". Neat trick. I wonder how complicated it would be to train the elechouse voice recognizer to handle that kind of flexible grammar?
-
Very cool.
The Voice module is about $22. I think voice recognition could be done within the ESP32 as well, but it would take significant work: https://www.survivingwithandroid.com/esp32-tensorflow-micro-speech-i2s-external-microphone/ (https://www.survivingwithandroid.com/esp32-tensorflow-micro-speech-i2s-external-microphone/)
https://www.aliexpress.com/item/1005003782550871.html (https://www.aliexpress.com/item/1005003782550871.html)
These are cheap but seem to have fixed commands?
https://www.aliexpress.com/item/1005003129261277.html (https://www.aliexpress.com/item/1005003129261277.html)