Introduction

My thesis for the development of the DDDAC1794-MK3 was to have a as low output impedance and low noise for the on-board regulator as possible (the LT1763). Especially in the low frequencies! In my introduction post on the newly designed DDDAC1794-MK3 (link here) I already showed some measurements from my AP-System. But not in depth on the LT1763. But that is solved today!

Several people asked me why I used this regulator and also why I decided for a bypass capacitor of 4.7uF where the manufacturer recommends 10nF as sufficient (in the datasheet). I already answered, that the LT1763 was widely used in digital clock and audio circuits, so I just followed that route for the first designs. The use of the 4.7uF was a tip I got from Yuri. It produces a lower pole in the voltage reference of the regulator, and this should have a positive effect on the output impedance AND the 1/f noise he claimed… Later listening tests confirmed it was a good indeed a good choice to use this setup with a 4.7uF MKP capacitor. I did listening tests versus LT3045 and Tentlabs Shunt regulators.

But the following question kept coming up… and honestly, I asked also my self:

What is the real effect of the bypass Cap?

A few weeks ago, I built this DDDAC Clock Source Board (based on a design on DIY-Audio from Mark – link here)

Mark also uses a LT1763 with 10nF (datasheet recommendation)

Based on my positive experience, I actually used the same on-board regulation as in my DDDAC1794-MK3

I decide to use this little board to do the experiments…

so far so good, but how does it measure?

Measurements…

I measured both output impedance and output noise in 3 different configurations:

• no bypass capacitor
• bypass capacitor 4.7 nF (I had a nice MKP of this value, seems close enough to 10nF)
• bypass capacitor 4,7 uF (the totally out of the box value 4.7uF MKP)

Noise Measurement

Output impedance measurement

Conclusions

• I think the graphs speak for itself. No further comments – Thesis are proved by this as far as I am concerned.
• The 4.7uF creates a very low frequency pole. This has the extra benefit that of a smooth start. It takes like 10 seconds for the power to come fully up. Used in the analog side of the DAC, the DAC starts smooth and start playing quietly and arrives at full output after like 10 seconds.
• Again, I was impressed by the fact, that datasheet recommendations are not always leading to the best results. Thinking and going out of the box is still something one should consider!

Useful Links related to this Post