Maximum signal level on audio inputs.

Status
Not open for further replies.
Kuba0040

Kuba0040

Well-known member
Hello,
I'd like to connect some analog circuitry to the audio board's (Rev D with Teensy 4.0) line level inputs. And I have a couple of questions. First of all. What is the voltage range for the accepted line in signals (From Teensy GND)? So, they get converted to values from 0 to 65,535 (entire 16bit audio library signal range). (I want to cover all of the range for maximum quality). If that isn't exactly an option, what is the recommended voltage range? And most importantly. Does the audio board have overvoltage protection on the line level inputs or do I need to implement it myself?
Thanks for the help.
 
What is the voltage range for the accepted line in signals
Click to expand...
According to the SGTL5000 datasheet, the typical LineIn voltage is 1.0VRMS:

SGTL5000.PNG
(on page 6, it states maximum 2.83VPP)

As far as I can tell there is no overvoltage protection on the LineIn pins.

Regards,
Paul
 
The voltages are converted to signed values (the range -32768..32768) with the Audio library. This is normal for
audio ADCs and DACs (but not for non-audio ADCs and DACs).

The SGTL can output upto 24 bits, the Audio library only takes the 16 most significant bits.

The SGTL5000 may well be fried if you over-voltage it, though the series capacitors provide some protection.
Its unclear if there's any on-chip protection, probably some anti-static protection diodes, but I doubt they'd
survive anything like a 10Vrms incoming signal from an audio mixer for instance(!)
 
Perhaps nitpicking, but setting 0: 3.12Vpp does formally violate the SGTL5000 spec.

SGTL5000table4.PNG

Paul Stoffregen apparently measured these voltages to be OK [see control_sgtl5000.cpp] and the chip will survive this voltage but distortion figures might not be guaranteed.

@MarkT: any thoughts about this?

Paul
 
Its a 3.3V chip, the inputs can go upto the rails in CMOS. The max input is likely the sigma-delta clipping
point (sigma delta converters often clip somewhat before the rails).

The datasheet not being self-consistent is not exactly rare! Hoever it might be something to do with gain
variability - unlike other converter types the gain of a sigma-delta converter can be subject to process
variation and temperature dependence. Some of those figures may be maximums, some typical.
 
Status
Not open for further replies.
Back
Top