Converting Line level audio into something easier to work with?

Thread Starter

MachineHum

Joined Nov 3, 2014
74


Does anyone have any examples some circuits that can do this with minimal components? I'm looking at the bright red wave, so shifting up by 1.414V so I can sample into an ADC and it doesn't go below zero, I don't want to need a negative power supply.:)
 

Papabravo

Joined Feb 24, 2006
21,159
How about a unity gain buffer with 2 resistors as a voltage divider and a coupling capacitor
. Is that minimal enough?
 

Mike33

Joined Feb 4, 2005
349
Hi, first of all..what you're looking for is a very small increase in amplitude - you know that, right? *shrug*

A basic single-supply opamp circuit would easily allow you to adjust the amplitude of an audio signal. You can use a potentiometer to make it variable, too. You don't need a neg. supply ; you use a biasing network (called "Vref" or "VCC/2"). Input and output caps will keep you zero-crossing without DC showing up, and your signal would vary between a max. and min. near -4.5V and +4.5V, for a 9V supply.

For what you are trying to do, of course you'd be nowhere near those numbers! You'll need a multimeter at min. to know what level you're at, and an oscilloscope would be great, too, but not required. Read/save this document: https://courses.cit.cornell.edu/bionb440/datasheets/SingleSupply.pdf You'll want one of the circuits on page 6...inverting or non-inverting.

"VCC/2" is just 2 equal value resistors, their junction goes to that point, also called "VRef" in many places...this biases the opamp to pretend that point is "zero", thus allowing you to use only your positive supply voltage. 10k is a very common value in audio work...a 1 to 10uf cap is placed from that junction to ground to smooth the reference voltage.

Does this make sense, I hope? Depends on how much building you've done, I think. I'd use a TL071 for the opamp...good performance, cheap and everywhere. Get a little box at Radio Shack and the right jacks, and you're in business.

What's this for?
 

Mike33

Joined Feb 4, 2005
349
Here is a schematic for you. Make R2 a pot wired as a variable resistor, and you get an adjustable output. A 100k pot will give you up to 11x amplification. The 'stuff' on the left is the power supply, with Vref (Vr). If the sound is 'thin', increase C1 to 1u or more. Same for C3.

Didn't see your post, Papabravo - your solution is simpler. Mine allows for 'tweaking' to level but may be more than is needed.

 

Thread Starter

MachineHum

Joined Nov 3, 2014
74
Here is a schematic for you. Make R2 a pot wired as a variable resistor, and you get an adjustable output. A 100k pot will give you up to 11x amplification. The 'stuff' on the left is the power supply, with Vref (Vr). If the sound is 'thin', increase C1 to 1u or more. Same for C3.

Didn't see your post, Papabravo - your solution is simpler. Mine allows for 'tweaking' to level but may be more than is needed.

First off, thanks for your reply. I'm not sure I worded the question properly.

I'm looking to add a DC bias so the signal never passes through zero vdc. This looks like the signal output would be (1+ (R2/R1))... effectively adding a gain?


This is what I would use if I was running a split supply (no V3 or R3) ...

After reading the sec. 2.3 of the doc you linked, it looks like the theory holds with all single supply bias circuits as they do with traditional split supply circuits.

Another thing I need to do is shift it back to being zero centered... in this case would I just use my com for my output as vref? (ie if I was hooking it into an amp, I would just use Vcc/2 for the common pin?)
 

shteii01

Joined Feb 19, 2010
4,644
First off, thanks for your reply. I'm not sure I worded the question properly.

I'm looking to add a DC bias so the signal never passes through zero vdc. This looks like the signal output would be (1+ (R2/R1))... effectively adding a gain?


This is what I would use if I was running a split supply (no V3 or R3) ...

After reading the sec. 2.3 of the doc you linked, it looks like the theory holds with all single supply bias circuits as they do with traditional split supply circuits.

Another thing I need to do is shift it back to being zero centered... in this case would I just use my com for my output as vref? (ie if I was hooking it into an amp, I would just use Vcc/2 for the common pin?)
That's summing amplifier. It adds two or more inputs. We use it to perform arithmetic operations, summation to be precise. Are you building a calculator and need summation function?
 

JDT

Joined Feb 12, 2009
657
So, you want to shift your signal so that it never goes below zero?
Simple: Two equal resistors in series across your ADC supply. The junction of these will be half the supply voltage.
Couple your audio signal to the junction of these with a capacitor.
If your ADC has a 5V supply you need to make sure that the audio signal never goes above +- 2.5V peak.
 

alfacliff

Joined Dec 13, 2013
2,458
the reason for line level isnt difficulty, but as a standard level. so that people specing microphnes, amps and such have a standard to use. and if you dont want a signal to go below zero, use a coupling capacitor and bias.
 

Thread Starter

MachineHum

Joined Nov 3, 2014
74
So when it comes to signal attenuation and phase shift...



For the frequency span of "f", 20Hz-20Khz, you want the keep this term as low as possible... so C= as large as possible? Electrolytic okay?

And then what about removing the DC bias? Can you just put a cap in series? ~infinite DC resistance with minimal AC resistance?

Thanks guys!! It's been a while for me...
 

JDT

Joined Feb 12, 2009
657
The capacitor and resistors will form a high-pass (low cut) filter. But if you keep your resistors as high value as possible and the capacitor as large as possible then there will be very little effect in the audio range.

Electrolytic is OK, because there is a constant DC bias. But electrolytics have a fairly large leakage current and their capacitance changes with voltage. So not recommended for quality audio. Easy to get 1 - 2.2uF polypropylene types that will be OK.

How many bits resolution is your ADC? I only ask this because if the ADC is not particularly special, then it's not worth taking too much effort elsewhere.
 

Thread Starter

MachineHum

Joined Nov 3, 2014
74
I'm stilling looking at ADCs, I'm thinking I may actually use a codec... as I will most likely need a DAC on the other end. It's going to be specked for audio. So most of them are I2S, 24+bits, 48ksps, stereo... I'm thinking something like this:

http://ca.mouser.com/ProductDetail/...31SEDS-V/?qs=sGAEpiMZZMthNezIpKscMWquczdXLZMI

I may not even need to add a DC bias, I have to look into the datasheet further... But either way, this is very important info, because I will be dealing with this in the future...
 

JDT

Joined Feb 12, 2009
657
Well that chip has AC coupled line inputs so you don't have to do anything. Writing the software to control the chip might be more difficult...

Interesting project though.
 

Thread Starter

MachineHum

Joined Nov 3, 2014
74
Typically I prowl through the digital/firmware/software domain, so I'm not majorly concerned with the firmware...

When I was going to school I was always better with digital stuff, but now I would like to get more into the analog side again. That's while I made this account, and so far I love this forum!
 
Top