This is the best Volume Control for DIY I know of.

0 Members and 1 Guest are viewing this topic. Read 17035 times.

Warpspeed CE

  • Jr. Member
  • Posts: 95
Re: This is the best Volume Control for DIY I know of.
« Reply #20 on: 19 Sep 2013, 05:29 pm »
Please note that this dB vs. step curve is NOT the dB vs. step curve of the LDRx. We pick points along this curve to build the 70 step audio taper curve that gets programmed into each LDRx with each step being close to ~1 dB. See the last pic for our typical audio taper curve.

I come in peace...and some observations:)

Seeing your curves and your post above, the way I understand it that would be the original 250 steps getting compacted into 70 steps on your final curve? At 1 dB per step all the curves on your pics would now have curves that are very much steeper with less points (70) to connect, right? This would be similar to an analog signal getting digitized with 70 sampling rate...a rather low rate for Nyquist.

And on using LDRs for input switching...wouldn't that be like adding another 100 ohms to your Source's output impedance?

All in all good design albeit very complex.


tortugaranger

  • Facilitator
  • Posts: 1746
  • Handcrafted high performance audio
    • Tortuga Audio
Re: This is the best Volume Control for DIY I know of.
« Reply #21 on: 19 Sep 2013, 06:26 pm »
Seeing your curves and your post above, the way I understand it that would be the original 250 steps getting compacted into 70 steps on your final curve? At 1 dB per step all the curves on your pics would now have curves that are very much steeper with less points (70) to connect, right? This would be similar to an analog signal getting digitized with 70 sampling rate...a rather low rate for Nyquist.

I'd describe it as "sampled" rather than "compacted". You don't need to break up ~60 dB of attenuation into 250 steps to have smooth (to the ear) attenuation. But since we had 8 bits to divide up the steps it's our baseline data. We settled on 70 steps of approximately 1 dB/step to arrive at a logarithmically flat audio taper. You could argue it's somewhat arbitrary and I wouldn't disagree. But 70 steps gets the job done quite nicely. And yes there's some quantitization in this sampling process but it's at most a few 1/10ths of a dB.

Quote
And on using LDRs for input switching...wouldn't that be like adding another 100 ohms to your Source's output impedance?
Yep. But it's immaterial relative to the overall impedance of the attenuator. Plus that gets baked into the attenuation test data you saw in the curves so it's just another ~100 ohms of series resistance.