Validatation of an audio codec

Hello Ramya,

ramya sri said:

what way we can evaluate these parameters if my audio codec is interfaced with FPGA

The best way to evaluate whether a device works for your system is to buy an evaluation module (EVM) from ti.com. Please find a list of TI's audio CODEC profile Audio CODECs | TI.com. The EVM page and user 's guide can be found by going to a device's product page and selecting the "design and development" tab. Additional documentation and application references can be found under the "technical documentation" tab of the product page. 

ramya sri said:

SNR, THD, frequency response, harmonic distortion, sensitivity are few of the parameters by which we can validate the audio codec.

We take these measurements using audio test and measurement equipment such as the Analog Precision. Some oscilloscopes can take these measurements, but in general, how the measurements mentioned above are performed is highly dependent on the equipment available to you. Please consult your own equipment manuals and references to see if you can take these measurements. 

Best,
Andrew

Hello Ramya,

Thank you for your patience.

ramya sri said:

audio response shall be in between 300Hz to 3.3Khz with +/-3dB with an audio input of 0.125-2.1vrms at 150ohm differential

Our audio codecs support the full audio signal range with a 1Vrms (2.8284Vpp) full scale signal. They also have the option to include a programable amplifier (PGA). 

ramya sri said:

audio output of 2.5-20vpp at 600ohm differential

Many of our codecs are available with an integrated Class-D speaker amplifier. For example, the TLV320AIC3120 has an integrated Class-D amplifier capable of driving 1.4W load.

The EVM options for this device can be found under the "Design and Development" tab or here: TLV320AIC3120EVM-U Evaluation board. If more power is needed then condsider an external power amplifier such as TAS2555YZ(EVM).  

ramya sri said:

how do I know the acceptable values for SNR, THD

The answer to this depends on what your application is. Different audio applications need to meet varying requirements. For example, telephone and mobile applications have a much smaller frequency range and do not need the highest quality while professional audio and most listening applications require 1% THD across the whole audible frequency range (20-20kHz).  

I hope this helps.

Best,
Andrew

Hello Ramya,

I believe I have answered this in the previous post:

ramya sri said:

Also, please provide if any references or information "in what way we can evaluate these parameters if my audio codec is interfaced with FPGA" i.e, what are the equipment's that can be used to validate these parameters for audio codec and as well as audio devices, if I am using the CODEC part which you have suggested i.e., TLV320AIC3120

Andrew Jackiw said:

We take these measurements using audio test and measurement equipment such as the Analog Precision. Some oscilloscopes can take these measurements, but in general, how the measurements mentioned above are performed is highly dependent on the equipment available to you. Please consult your own equipment manuals and references to see if you can take these measurements. 

When we take these measurements in lab, we use a piece of equipment (and accompanying software package) provided by Analog Precision (AP), the APx555. The AP test and measurement system takes these measurements for us. If you do not have access to such equipment, then it is possible to perform these measurements with a spectrometer or on some oscilloscopes. Most lab equipment capable of making an FFT plot will be able to measure these specs.

In general, I do not know if TI has dedicated app notes to taking these measurements, however there are many online resources provided on how SNR, SINAD and THD measurements can be taken. Please note that searching these measurements on TI.com do show relevant app notes, references and devices that may be helpful for understanding these measurements and methods on how to improve these specs (see a few examples below)

• ADC Driver Ref Design Optimizing THD, Noise, and SNR for High Dynamic Range
• High-Res, High-SNR True Raw Data Conversion Ref Design for Ultrasound CW Doppler (Rev. B) (ti.com)
• TPS92314 THD design consideration
• 18-Bit, 2-MSPS Isolated Data Acquisition Reference Design for Maximum SNR and Sampling Rate (Rev. A)

In closing, how these measurements are taken is up to the equipment available to you. Most test and measurement manufactures have trainings and references how to take advanced measurements. Please refer to the manufacturer trainings on the equipment that is available to you. 

Best,
Andrew

Hi Ramya,

Let me clear up where you can find the specs you're mentioning, using the AIC3120 as an example

Audio Input: This is the full scale input to the codec, meaning the highest voltage that can be provided before the ADC saturates. This is typically found in ADC characteristics. 

Audio Output: This is the max voltage the device can output across the given load. You can convert this to Vpp if needed. There's also a section on the line driver but the info is a little more limited for this part.

Audio Response: Use the graphs found in the following section to evaluate frequency response. It's very typical for our audio parts to have a flat/close to flat response between 20-20kHz. As for your volume control concern, that is a feature that will have to be found elsewhere in the datasheet.

Distortion: We don't typically document SINAD since it's only used in telecomm applications, but with identical bandwidths, SINAD = THD+N. The different THD+Ns can be found for the different inputs and outputs.

Regarding the impedances you listed, 150ohm is a typical output impedance from a source. The input impedance of the codec is much higher, so there should be no issue with the specs here. The 600ohm raises questions for me. While this device does have a line out driver, this is starting to sound like a telecomm application with a 600ohm transmission line. I doubt this device's ability to transmit meaningful power over that line. You may need a more powerful line driver to take the output from the codec and boost it to a more reasonable level.

I hope this answers your questions,
Jeff McPherson

Hello Ramya,

Please allow me to differentiate the two terms:

Maximum Input Voltage:  this is the highest voltage that the pin can take without causing damage to the device. Please see 5.1 Absolute Maximum Ratings for more information. Voltages higher than the maximum input voltage risk damaging the device.

Output voltage: In the datasheet the value 2.1 Vrms is used to describe the output if the DAC to the Class-D input.

Input signal level (0dB): as Jeff said, this is the full-scale audio input. At 0.707Vrms (~2Vpp) the digital value of the ADC will be all 1's (0dB). Any signal higher than this will be "clipped" in the audio domain.

I hope this helps clarify the specks above. 

Best,
Andrew