The plan is to develop two boards, at the beginning only the Headphone amplifier, and then will prepare the DAC board 
In this way I can try many combination of amplifier and DAC
The layout is finished, it is a 4 layer board.
First layer: Signal trace
Second Layer: Ground
Third Layer: Supply plane (it is splitted in two half pcb VCC:L and half PCB VCC_R
Fourth Layer: Some signal and ground.
The 2 and 3 plane form also a capacitor that help to filter the noise.
In addition all unused copper it is put to ground and connected in the 4 layer with the vias to reduce the noise.
In this way I can try many combination of amplifier and DAC
The layout is finished, it is a 4 layer board.
First layer: Signal trace
Second Layer: Ground
Third Layer: Supply plane (it is splitted in two half pcb VCC:L and half PCB VCC_R
Fourth Layer: Some signal and ground.
The 2 and 3 plane form also a capacitor that help to filter the noise.
In addition all unused copper it is put to ground and connected in the 4 layer with the vias to reduce the noise.
Hi xrk971,
thank you very much.
Yes probably having volume knob in the rear is not comfortable. I have rotated it of 90° by eschanging the input jack and the potentiometer position.
If your are "scared" about the distance"between the input power connector and the output stage, it is not a problem, because there is the supply plan, therefore the impedance is very very low.
I have never haved problem also with the electronics that I develop for work.
In addition I have removed the GND under the PWR Jack to avoid parasitic coupling before the common-mode coke.
thank you very much.
Yes probably having volume knob in the rear is not comfortable. I have rotated it of 90° by eschanging the input jack and the potentiometer position.
If your are "scared" about the distance"between the input power connector and the output stage, it is not a problem, because there is the supply plan, therefore the impedance is very very low.
I have never haved problem also with the electronics that I develop for work.
In addition I have removed the GND under the PWR Jack to avoid parasitic coupling before the common-mode coke.
I’m not worried about power jack location. I would have power and audio in on back panel and volume pot and headphone out on front panel. The orientation is important from usability standpoint. But if it’s how you like it, that’s fine. I would not put pot 90deg from headphone out. That requires 3 sides to be drilled and fitting a board into 3 fixed walls is almost impossible.
As attached you can find the simulation file with the current generator instead of R3.
The THD it is reduced by factor 10 (0.02% instead of 0.2%).
View attachment Hakuin-v3.1_BambinzMod.asc
The THD it is reduced by factor 10 (0.02% instead of 0.2%).
View attachment Hakuin-v3.1_BambinzMod.asc
If you want to drive 32ohm load, the bias current needs to be increased to about 125mA. Sufficiently large heatsinks will be needed to cool the part. Not sure if DZT5401 can cool 1.5W via SMT thermal pads. I would recommend changing to TTA004B and larger heatsink. It will take some manipulations of the setpoints. If you can post an image of your schematic that would be helpful so I don’t have to open up LTspice as I am not going to be at a computer.
You would get something closer to the daughter board for the Yarra/M2X.
You would get something closer to the daughter board for the Yarra/M2X.
What is point of having ultra low distortion? No headphone known to mankind which is lower than 0.2% THD? Their is no microphone i know with dynamic range higher than 85db yet dac designer runs after above 120db SNR. As a guitarist and musician , i make music for people to enjoy not to archive ultra low distortion which takes away all texture from effects i added to it like guitar pedals , tube microphone pre-amp etc. All ultra low noise stuff like objective 2 takes away fluidity from vocals , makes highs too sharp.
Its my request to people here , go to a musical concert , you will understand what is natural sound.
Or do me a favor by recording your speech and play it through a High noise tube amp and ultra low noise op amp headphone amp, tell me which sounds more like your own voice. Record your piano and guitar , play it through a class A single ended amp and another Opamp based amp , tell me which sounds more like real one.
So called Hi-Fi industry runs after ultra low noise which makes zero sense as Speaker and headphones have far higher noise.
Its a request please don't take away life from music for sake what you see in your oscilloscope.
Sorry if i sounded like a jerk.
Its my request to people here , go to a musical concert , you will understand what is natural sound.
Or do me a favor by recording your speech and play it through a High noise tube amp and ultra low noise op amp headphone amp, tell me which sounds more like your own voice. Record your piano and guitar , play it through a class A single ended amp and another Opamp based amp , tell me which sounds more like real one.
So called Hi-Fi industry runs after ultra low noise which makes zero sense as Speaker and headphones have far higher noise.
Its a request please don't take away life from music for sake what you see in your oscilloscope.
Sorry if i sounded like a jerk.
Last edited:
Hi
I have used 32 Ohm but in reality my headphones are 150Ohm and the other 55 Ohm.
The modified schematic is:
Hi Dibya,
I agree with you
I am quite shure that I will not hear differences.
At the end, sometimes, the objectiv of diy an amplifier/preamplifier/dac/... is not just only listen music. It is a hobby and it is funny to try to improve the "numerical" performance of somethings though not results in a real improvement in the music listen or put a little contribution/modification in a schematic produced by another diyer.
Excuse my bad English
Hi Bambinz and Dibya,
You both make excellent points, I fall right in the middle
I don’t listen to music and judge how much I enjoy it based on oscilloscope measurements or distortion figures. I like what I like.
But, as a hobbyist there are certain “hang ups” I have when building projects, I like to call one of them ‘Regulitus’. Most of the time I need to use low noise regulators when cheapo LM78/79xx will work just fine. But the noise numbers are ultra low, can we hear it? That’s another topic
You both make excellent points, I fall right in the middle
I don’t listen to music and judge how much I enjoy it based on oscilloscope measurements or distortion figures. I like what I like.
But, as a hobbyist there are certain “hang ups” I have when building projects, I like to call one of them ‘Regulitus’. Most of the time I need to use low noise regulators when cheapo LM78/79xx will work just fine. But the noise numbers are ultra low, can we hear it? That’s another topic
Hi Vunce,
MC78xx/MC79xx are drop in better alternative to their lm counterpart.
But I highly doubt it matters for opamp related circuits as supply rejection ratio of even good old NE5532 is really good , newer opamp are overkill in that regard.
I like LM317/337 , more of a middle ground and they are affordable.
As for discreet circuits I will definitely prefer lowest noise possible LDO.
It just my opinion , I might be wrong.
MC78xx/MC79xx are drop in better alternative to their lm counterpart.
But I highly doubt it matters for opamp related circuits as supply rejection ratio of even good old NE5532 is really good , newer opamp are overkill in that regard.
I like LM317/337 , more of a middle ground and they are affordable.
As for discreet circuits I will definitely prefer lowest noise possible LDO.
It just my opinion , I might be wrong.
Here is my modification of Haikun as Preamp for bigger amp.
Stray ".ends" card: .ends
Circuit: * D:\2SA1587_LTspice_20190322\Hakuin-2sk209gr-2sa1837\Hakuin-2sk209gr-2sa1587.asc
Direct Newton iteration for .op point succeeded.
Changing Tseed to 1e-010
Changing Tseed to 1e-012
Changing Tseed to 1e-009
N-Period=1
Fourier components of V(out)
DC component:-0.000232184
Harmonic Frequency Fourier Normalized Phase Normalized
Number [Hz] Component Component [degree] Phase [deg]
1 1.000e+03 4.791e-01 1.000e+00 0.09° 0.00°
2 2.000e+03 6.046e-05 1.262e-04 -90.22° -90.31°
3 3.000e+03 7.552e-07 1.576e-06 0.36° 0.27°
4 4.000e+03 2.157e-07 4.501e-07 174.05° 173.96°
5 5.000e+03 1.719e-07 3.587e-07 -179.99° -180.08°
6 6.000e+03 1.426e-07 2.977e-07 -179.98° -180.07°
7 7.000e+03 1.221e-07 2.548e-07 -179.99° -180.08°
8 8.000e+03 1.067e-07 2.226e-07 -179.99° -180.08°
9 9.000e+03 9.465e-08 1.975e-07 -179.99° -180.08°
Total Harmonic Distortion: 0.012619%(0.012547%)
Its for 15V Supply , I want to use it in future as output buffer for a RIAA Phono Preamp. Either i need to buy a new cartridge for my grandfather's Technics SL1200 (1985) or get a new turntable. I want to know why people so crazy about Vinyl.
Stray ".ends" card: .ends
Circuit: * D:\2SA1587_LTspice_20190322\Hakuin-2sk209gr-2sa1837\Hakuin-2sk209gr-2sa1587.asc
Direct Newton iteration for .op point succeeded.
Changing Tseed to 1e-010
Changing Tseed to 1e-012
Changing Tseed to 1e-009
N-Period=1
Fourier components of V(out)
DC component:-0.000232184
Harmonic Frequency Fourier Normalized Phase Normalized
Number [Hz] Component Component [degree] Phase [deg]
1 1.000e+03 4.791e-01 1.000e+00 0.09° 0.00°
2 2.000e+03 6.046e-05 1.262e-04 -90.22° -90.31°
3 3.000e+03 7.552e-07 1.576e-06 0.36° 0.27°
4 4.000e+03 2.157e-07 4.501e-07 174.05° 173.96°
5 5.000e+03 1.719e-07 3.587e-07 -179.99° -180.08°
6 6.000e+03 1.426e-07 2.977e-07 -179.98° -180.07°
7 7.000e+03 1.221e-07 2.548e-07 -179.99° -180.08°
8 8.000e+03 1.067e-07 2.226e-07 -179.99° -180.08°
9 9.000e+03 9.465e-08 1.975e-07 -179.99° -180.08°
Total Harmonic Distortion: 0.012619%(0.012547%)
Its for 15V Supply , I want to use it in future as output buffer for a RIAA Phono Preamp. Either i need to buy a new cartridge for my grandfather's Technics SL1200 (1985) or get a new turntable. I want to know why people so crazy about Vinyl.
Last edited:
I implemented both a 7818 as well as a TPS7A4701 LDO and there wasn't an appreciable change in the measured characteristics of the amp between the two.
But the LDO did sound $20 better.....
Yes, more $ parts must sound better, always.