WiFi Balanced Audio Volume Control

Description

Active (powered) speakers can sound very nice, but consumer audio gear doesn't typically have good support for this kind of speaker. I don't have a receiver, and I wanted a way to control the volume of my LSR-305 powered speakers.

This is a simple widget which provides ~0-30dB of attenuation. It exposes a simple web page over WiFi for control. It uses an MCP4661 digital potentiometer to adjust a voltage divider on each channel.

Project Logs

Collapse

Frequency response testing
Adam Demuri • 05/12/2020 at 03:04 • 0 comments

Ideally, this device would be "transparent" - it would reduce the volume equally, without changing the frequency response. I designed the PCB roughly following best practices for analog signals. However, I am not an expert in PCB design. So, to validate the design, I have performed some measurements with an oscilloscope.

I generated sine wave sweeps using this tool. I captured the output signal on my oscilloscope. For the test, the device was between a DacMagic DAC and LSR-305 speakers. I used a volume in the tool of 40, and an exponential sweep. I did three tests - 20-200hz, 200-2000hz, and 2000-20,000hz. I performed the tests on the left and right channels, with no attenuation, and half (~20dB) attenuation.

Since this device is for balanced audio, there are three signals in the scope trace. The blue and yellow are two raw balanced audio signals. The red is the difference, which is what we care about.

Left channel, no attenuation

20-200hz:

200-2000hz:

2000-20,000hz:

Left channel, half attenuation

20-200hz:

200-2000hz:

2000-20,000hz:

Right channel, no attenuation

20-200hz:

200-2000hz:

2000-20,000hz:

Right channel, half attenuation

20-200hz:

200-2000hz:

2000-20,000hz:

Right channel, full (~30dB) attenuation

I also did a test with full attenuation, on the right channel only. You can see that there is more noise here.

20-200hz:

200-2000hz:

2000-20,00hz:

Analysis
The frequency response appears to be pretty flat. Theoretically, all of the sine waves should have equal height if the devices is perfect. The variation with no attenuation is minimal. The variation with higher attenuation appears to be within the noise present. I am not sure if the noise is coming from this device or the other components. When using the device, I don't hear any additional noise. My testing included using the device at max attenuation, with the speaker gain turned all of the way up. Even at these settings, I didn't hear any additional noise.