Hybrid ZEN Amplifier + LM317 = Efficient and simple

OK on moving the LM.

Max current through an LM317 is 1.5 amp. Dissipation across 0.5 ohms is 1.5x1x1.5 = 1.65 W. Four or five watts would be better. I will make the change ( 2 x 2 ohm per leg) and place text on the board about mounting well above the board surface.

ray
 
  • Like
Reactions: 1 user
To choose R5 and R6, I considered that the amplifier had to be useful as a desktop one, due to its limited power. As today it is very common to have a laptop, notebook or desktop PC, I considered it appropriate to operate it from any of them, through the headphone output. This is why the relatively low input impedance should not be a problem, as many outputs are capable of handling 32 ohms with ease (whereas this amplifier has an input impedance of around 350 ohms).

In any case, for those who wish to modify it, they should consider that the frequency response (mainly the high frequency) could be significantly altered. The distortion profile will probably be altered, so it should be checked if it is still useful in how it is altered.

I show you how the frequency response would be altered by taking other values for R5 and R6, maintaining similar gain within the passband.

Sin título.jpg
 
Ok, thanks for the explanation on choosing the low input resistance.
For my personal taste I would prefer driving your nice hybrid Zen amp from a normal unbalanced output from example a DAC or similar source, not from the output of an headphone amp unless you know it’s great sounding.
The headphones output jack on regular laptop or similar computer are normally not really Hi-Fi, your amplifier deserves a much better source.

What if R6 is 3Kohm, what would be the upper -3db frequency ?

Thanks
 
Last edited:
  • Like
Reactions: 1 users
Here is the latest version. I will wait a couple of days for suggested corrections or improvements and then post the Gerbers. I will have my laptop so I can make changes.

I am going to Japan later today for two weeks. It's my first time in Japan and I am hoping to visit a couple of Kissas as well as the electronics shops in Akihabara. And ride a lot of trains and eat plenty of ramen.

ray

1696629053436.png
 
  • Like
Reactions: 2 users
A simple scheme that includes a transformer, rectifier and filter will be more than enough. For the filtering capacitor I recommend a value of 6800 uF up to a power of approximately 6 W per channel (one 6800 uF capacitor for each channel), so that the ripple does not exceed 2.44 Vpp. This requirement will be necessary to be aligned with the minimum resolutions for 16-bit content.

For resolutions up to 16 bits it will be enough with a simple scheme consisting of transformer, rectifier and capacitor. The rectifier bridge could be the MB1510 (one per channel, if possible). The capacitor could be 6800 uF x 35 V. I suggest one transformer per channel, if possible.

To be aligned with 20-bit resolutions, I recommend interleaving voltage regulators with power ripple rejections of at least 24 dB at the mains frequency.

To be aligned with 24-bit resolutions, I recommend voltage regulators with power ripple rejections of at least 48 dB.

For both the 20-bit and 24-bit options, I am assuming the value of the main filter capacitor is kept at 6800 uF (prior to voltaje regulator).

Best regards
 
Last edited:
Member
Joined 2019
Paid Member
^ The bit-depth defines the dynamic range. They're recommending that the PSU be measurably / quantifiably 'quieter' if someone were to want to distinguish a signal from noise at a certain dynamic range. The PSU typically sets the "noise floor".

@Vunce = Jumping in just to see if I actually understand any of this ... I had to look it up. Sorry if I hopped on to your question.

Diego, I'm also curious. Am I close?
 
  • Like
Reactions: 1 users