I recently built an LM3886 amp, which I posted here. I read that some prefer the sound without the feedback capacitor (Ci on the datasheet) and was wondering why.
I know the DC offset will be higher without the capacitor and I plan on trying that the next time I open it up as I built mine with the feedback capacitor.
My guess would be different bass response but my training is in repair not engineering.
My question is: Can anyone describe the sound difference with and without the feedback capacitor?
Thanks.
I know the DC offset will be higher without the capacitor and I plan on trying that the next time I open it up as I built mine with the feedback capacitor.
My guess would be different bass response but my training is in repair not engineering.
My question is: Can anyone describe the sound difference with and without the feedback capacitor?
Thanks.
Try C1 = 100uF, then it won't affect the bottom octave as much. 20uF is a bit anaemic for 1k, with cutoff at 8Hz (or worse as spread in electrolytics is wide).
DC-coupling in that circuit would give upto 0.23V offset absolute worst case (input pot at 100%, worse case 1uA bias currents in the LM3886. Typical LM3886 bias is 200nA, giving upto ~50mV offset which is more realistic and not horrible.
Either way it would be good to AC-couple the input before the volume pot so DC can't be imposed on the amp, 22uF would be about the right value for that position as the load is 10k.
DC-coupling in that circuit would give upto 0.23V offset absolute worst case (input pot at 100%, worse case 1uA bias currents in the LM3886. Typical LM3886 bias is 200nA, giving upto ~50mV offset which is more realistic and not horrible.
Either way it would be good to AC-couple the input before the volume pot so DC can't be imposed on the amp, 22uF would be about the right value for that position as the load is 10k.
I used 47uF
I know, if I remove it I will check the DC offset.
I didn't use a volume pot. I have it paired with a 6N6P tube preamp I built, which has capacitors at the output.
Yep. And given that the LM3886 itself can have as much as 10 mV of DC offset at its input, you could end up with 200 mV at the output of the amp. That's not exactly stellar.
If you remove C1 you really should add a DC servo and some would argue also a speaker protection circuit that disconnects the output if there's excessive DC at the output.
I'd design for 2 Hz or lower for the lower pole. So C1 ≥ 1/(2*pi*2*1000) -> C1 ≥ 79.6 µF.
If you use a modern electrolytic capacitor it won't have any sonic impact. It can even be a low-voltage type (say 6.3-10 V) as it only ever sees the DC offset.
Tom
In the past I posted:
https://www.diyaudio.com/community/...citor-and-hardly-no-dc-offset-voltage.217712/
Both amplifiers still work and there is still (hardly) no DC offset voltage.
I didn't compare the sonic quality with and without electrolytic feedback capacitor.
One person replied that he can hear the difference.
But recent studies report that although an electrolytic capacitor may cause distortions, it isn't audible.
https://www.we-online.com/components/media/o763278v410 ANP125a Audio Capacitor THD_EN.pdf
https://www.audiosciencereview.com/forum/index.php?threads/capacitor-distortion.40917/
The choice is yours.
Marc.
https://www.diyaudio.com/community/...citor-and-hardly-no-dc-offset-voltage.217712/
Both amplifiers still work and there is still (hardly) no DC offset voltage.
I didn't compare the sonic quality with and without electrolytic feedback capacitor.
One person replied that he can hear the difference.
But recent studies report that although an electrolytic capacitor may cause distortions, it isn't audible.
https://www.we-online.com/components/media/o763278v410 ANP125a Audio Capacitor THD_EN.pdf
https://www.audiosciencereview.com/forum/index.php?threads/capacitor-distortion.40917/
The choice is yours.
Marc.