By formatting the caps before use you can reduce this to almost zero: take a Vb of the rating of the cap, connect the cap with a 100k or maybe 10k resistor, you will see the voltage go up very slowly, and at the end, maybe a day later, the voltage remains stable.
Formatting configures the layers in the cap. And ensures a smoothing of the boundary layer (giving less shot-gun noise also aggregately seen as leakage).
Hi
I think 680R was intended for use with each individual DAC. The ESS9018 contains eight. I use it in stereo mode, four DACs per channel (Twisted Pear Buffalo II), and so, with four times the maximum current, a value of 150R is more appropriate- to prevent possible overload, and set the output to standard level
Paul N
I think 680R was intended for use with each individual DAC. The ESS9018 contains eight. I use it in stereo mode, four DACs per channel (Twisted Pear Buffalo II), and so, with four times the maximum current, a value of 150R is more appropriate- to prevent possible overload, and set the output to standard level
Paul N
Thanks for the answer Paul.
Looking at the schematics in this post I interpret it as the 680R is for four paralleled dacs. Anyway, I have four 300R Caddock mk132 (I cut the 500R resistors off the board thus destroying them) that I could try instead.
Another question, does anyone know how to calculate the cutoff frequency of a parallel RC filter? I was very satisfied with the R_iv/C_iv combination I had so I would like to calculate what C_iv to use when changing R_iv to get the same cutoff point. Have googled it but only found loads of information on serial RC filters.
Hi
With the greatest respect to Patrick, I think he also took the resistor value ESS intended for use with a single DAC. (Though I’d normally bet on Patrick over me in matters like this!) After all, he originally had no interest in the 9018, and only provided the schematics as a courtesy after much pleading.
The output impedance of the (paralleled) DAC(s) is equal to Riv, and Civ acts together with the resistor as a bog standard low pass RC filter. So if you half the resistor value, you need to double the capacitor value to keep the cut off frequency the same
Paul N
With the greatest respect to Patrick, I think he also took the resistor value ESS intended for use with a single DAC. (Though I’d normally bet on Patrick over me in matters like this!) After all, he originally had no interest in the 9018, and only provided the schematics as a courtesy after much pleading.
The output impedance of the (paralleled) DAC(s) is equal to Riv, and Civ acts together with the resistor as a bog standard low pass RC filter. So if you half the resistor value, you need to double the capacitor value to keep the cut off frequency the same
Paul N
That is correct, Paul.
The product of Riv*Civ should remain constant for the same cut off frequency.
Thank you for answering for me, and glad you are still keeping an eye on things.
On another note, one member of the team is (only, you would say) now building a SEN V18 for PCM1794, with Vref = Gnd.
We'll publish our implementation when it is ready. Please be patient.
Also we are working on the idea of using a modified version of SEN IV for its original purpose as proposed by Prof. Leach,
i.e. as a Moving Coil Cartridge Headamp.
It looks great at least in Spice, but as usual we shall only publish after thorough testing.
So development has not stopped totally ......
Patrick
The product of Riv*Civ should remain constant for the same cut off frequency.
Thank you for answering for me, and glad you are still keeping an eye on things.
On another note, one member of the team is (only, you would say) now building a SEN V18 for PCM1794, with Vref = Gnd.
We'll publish our implementation when it is ready. Please be patient.
Also we are working on the idea of using a modified version of SEN IV for its original purpose as proposed by Prof. Leach,
i.e. as a Moving Coil Cartridge Headamp.
It looks great at least in Spice, but as usual we shall only publish after thorough testing.
So development has not stopped totally ......
Patrick
Thank you again Paul and also Patrick for confirming. I will use the 330R Caddocks and adjust C_iv accordingly. I think most people use the es9018 internal volume control, which I do not, and in that case they will not run into clipping with the 680R.
All that's left now is to solve my hum issue. I am using batteries and LC filter but can still hear a hum if I put my ear around 15cm (6 inch) from the speaker. I will try shielding the batteries and I/V better and that it helps.
All that's left now is to solve my hum issue. I am using batteries and LC filter but can still hear a hum if I put my ear around 15cm (6 inch) from the speaker. I will try shielding the batteries and I/V better and that it helps.
You probably refer to electrolytic (re)forming which is needed to (re)form the chemical electrolytic layer after long periods of being without voltage.
This is not a mechanism in film capacitors.
Jan
I can never understand this hum thing because it just has not happened to me.
But there are enough examples here to show that it is solveable.
Patrick
The thing is that I can not recall having this hum when I first powered it up. I noticed it after a couple of months. I will take some time reading the whole thread through again to see if I can find any good tips.