dac I/V convertion with very low distortion

[smms73]

In my experience all buffers behave very good as long the output load is high, the problems arise when we need to drive low impedance ( less than 1K ), in that case bipolar transistors do the job better.

Just for curiosity this is images of distortion of a buffer that i going to use in other project that i am working, it uses a mosfet BS170 and the objective is create 2º harmonic distortion, first is with 10k load and second is with 1k , the output voltage is 1,25v rms. As you see only 2º harmonic is created.

This is real measurements not simulations.

[smms73]

Yesterday I have been working in the power supply for this circuit.
First is the power supply based on 7915/7815, second is the modifications at the power supply, as you can see the noise and ripple is extremly low, and almost the noise floor of the EMU1212m.

One tip: Use ferrite beads in your power supply, there are some high frequency spikes that only ferrite beads can supress.

first 7915 , second positive w/mod, last negative w/mod.

measurements made in the power rails , with 40ma load.

[Joachim Gerhard]

It´s not a big deal to make a good buffer anyway.

[Joachim Gerhard]

That looks very clean.

[staccatiss]

Good work, Smms73!

Is there a way to measure how much noise from the PS you are "injecting" into the IV converter?

Very interesting measurement you did about THD vs load! Might be that yuur bufferis leaving class A, and that is why THD increases so much at 600 Ohms load?

To make a buffer that is a good match THD wise for your IV converter can be quiete a challenge



/S

Originally Posted by smms73

Yesterday I have been working in the power supply for this circuit.
First is the power supply based on 7915/7815, second is the modifications at the power supply, as you can see the noise and ripple is extremly low, and almost the noise floor of the EMU1212m.

One tip: Use ferrite beads in your power supply, there are some high frequency spikes that only ferrite beads can supress.

first 7915 , second positive w/mod, last negative w/mod.

measurements made in the power rails , with 40ma load.

[staccatiss]

If you would connect the caps across the Zener diodes in the buffer with teh Zeners to the output, the Zener noise would go to the music, right?
Better to send the noise to the rails.

Like Erno Borbely for ex is doing in the mic amp design below.

Originally Posted by smms73

staccatiss, in the buffers that you present in post#649 the capacitors should be parallel with the zeners not with the ccs, .

[Shinja]

thank you smms.
your measurements are really informative.

[smms73]

Originally Posted by staccatiss

If you would connect the caps across the Zener diodes in the buffer with teh Zeners to the output, the Zener noise would go to the music, right?
Better to send the noise to the rails.

Like Erno Borbely for ex is doing in the mic amp design below.

If you put large caps like that , the noise from the power supply go directly to your output . And the worst is that it will short the output at A.C.
The capacitors in the schematic of E.g. Borbely are for shore small caps for stability and not for noise steering.

[smms73]

Originally Posted by staccatiss

Good work, Smms73!

Is there a way to measure how much noise from the PS you are "injecting" into the IV converter?

Very interesting measurement you did about THD vs load! Might be that yuur bufferis leaving class A, and that is why THD increases so much at 600 Ohms load?

To make a buffer that is a good match THD wise for your IV converter can be quiete a challenge



/S

To measure power supply rejection ratio one inject a signal in the power rail , then measure the amplitude at the rails and then in output, the psrr is the difference between the two . I have already done that and post the results somewhere in this thread, if I remember it is 80db at 50 Hz .

That buffer that I present is intend to distort , is not to be use in this project. Is for a test that I am preparing, to understand the influence of harmonics in the sound , more of a psychoacoustics test.

Just by looking to the fft graph you can see that the buffer is in class a , class b produce lots of high order harmonics.

If you take a good look to the tagus schematic you will find a suitable buffer .

[smms73]

Originally Posted by Shinja

thank you smms.
your measurements are really informative.

Thank you for bring that issue to my attention. non linearity of parasitic caps can be a problem if we are not careful.