Damaged audio output when attempting headphone detection?

Status
Not open for further replies.

zirafa

Member
Hello,

I have a Teensy 3.2 + Audioshield. I hooked up a switched headphone jack and I am wondering if my headphone detection circuit has fried the audio output. I followed the schematic found here:

http://www.eetimes.com/document.asp?doc_id=1274924

I am wondering if connecting Vcc through a 1k resistor to the audio left output (as shown in the schematic) may have caused permanent damage?

maxim_comparator_detect_fig1.jpg
 
If you used the headphone output of the audio shield, which is at a floating 1.65V potential, then yes, you fried it most probably. => Study data sheets and schematics thoroughly before doing such mods...
But if you used the line output which is decoupled by 2.2uF capacitors, then no, no harm possible with that.
 
Thanks, that is what I was afraid of. It seems that there is no easy way to do headphone detection using a GPIO pin without some additional protection circuitry.

Alternatively, since there *is* a floating voltage -- now I'm thinking I could use one of the analog inputs on the Teensy to detect a change in voltage. With no headphone jack, the analog input should just read the 1.65v bias, but when a jack is inserted the connection is broken and the analog input would read 0. Would there be any danger with this method?

teensy-headphone-detect.png
 
Last edited:
No danger, but this “new” method won’t work though. The headphone GND is virtual and always at 1.65V above nominal ground. The audio outputs are too and the audio signal goes around these 1.65V, so that even with an audio signal, the average voltage at both audio outputs and the Headphone GND which is isolated from common GND remains always at 1.65V. This is to allow symmetric voltage swing of the audio signals, positive and negative without the need for huge decoupling capacitors.
 
Should be safe to put a very light 100K load on the output to GND (not VGND, but real GND).

So you could probably look at that pin for 1.65V (plus audio) versus zero volts.
 
If you used the headphone output of the audio shield, which is at a floating 1.65V potential, then yes, you fried it most probably.
Why? I don't see any specs for voltages applied to outputs, but I would expect 3.3V / VDDA through a 1kOhm resistor to be safe.

Connecting the headphone virtual ground to the real ground could of course fry things.
 
Many thanks for the suggestion Paul. I wired up a new audio board and am able to detect a headphone jack using a 100k resistor with an analog input.

headphone-detect.png
 
Just to confirm (before I break something):

Assuming it's only connected to headphones, and not some grounded device, I can safely connect the external jack's left/right/ground connections to left/right/VGND on the audio board, and can connect one of the jack's switch legs to real ground through a 100k resistor.... right?
 
Status
Not open for further replies.
Back
Top