Voltagr Reugulator NOISE

[binatang]

ALW:

Nice! ... Where can I get the pcb from?

Bina

[carlmart]

Originally posted by janneman
Just for the record: the Jung regulator article series was a joint effort of Walt Jung, Gary Galo and myself. The introduction of the feedback from the load point (known as remote sensing) to these regulators was from the undersigned.

If you look for an updated pcb, ALW has an excellent implementation, which I have tested and found to be superior to the original ones.

Nice to know that Jan introduced the remote sensing concept. I wasn't aware of that and credit should go where it's due.

Jan also had great experience with the Sulzer regulator, as he also designed a very simple and very smart DAC that used that regulator. On that design he was also one of the first I saw that starred separate ground lines (and power if I remember well) on the pcb.

Another interesting concept which ALW mentioned on past threads was the importance of impedance and other curves being low but also flat, instead of being only low at certain frequencies and then going up. The Pooge regulator does that.

About the three pin regulator chips, implementation is usually wrong on many DIY applications. E.g.: 7XXX regulators can (and should) have large caps at the outpjut for better specs. Variable regulators should have up to 150uF at the output only and large caps from the regulating pin to ground, in parallel with adjust resistor.


Carlos

[jean-paul]

About the three pin regulator chips, implementation is usually wrong on many DIY applications. E.g.: 7XXX regulators can (and should) have large caps at the outpjut for better specs.

If so I guess the manufacturers of the chips all made mistakes in their datasheets because they generally advise to use small caps.

[carlmart]

Originally posted by jean-paul


If so I guess the manufacturers of the chips all made mistakes in their datasheets because they generally advise to use small caps.

Not necessarily mistakes, but they think different from us audio DIYers.

They recommend tantalum and ceramic, which is usually rejected by most audio designers or enthusiasts.

Audio applications are mostly a bit different from others, and this Forum is a place where such different implementations are suggested and applied.

Though I doubt manufacturer's engineers stop by here or pay any attention to our suggestions. That has been widely mentioned on articles on CD player mods and why they used the audio chips they used.

In any case, on the variable regulators datasheets I remember to have read warnings on not going over 150uF at the output and using protection diodes on the regulating pin if you use large caps, with no maximum warning. No warnings on maximum caps on 7XXX types that I remember of though.


Carlos

[janneman]

Most data sheets for integrated regs mention MINIMUM cap - not for low Z, but for stable operation. They often oscillate if there is too low a cap at the output. The cap is an integral part of the loop compensation.

Because these caps cost money & pcb real estate, there is a drive to make them stable with much less cap. One manufacturer uses the slogan "AnyCap (tm)" regulators, because they are stable with any cap, meaning whatever the ESL is. Very high quality film caps at the reg output may sound (no pun intended) like a good idea for audio, but because of the very low ESL they actually can also cause oscillations.

Look up a data sheet for those modern, low-dropout (LDO) regs by LTI, they describe this very well.

Jan Didden

[ALW]

Another interesting concept which ALW mentioned on past threads was the importance of impedance and other curves being low but also flat, instead of being only low at certain frequencies and then going up. The Pooge regulator does that.

Actually only the POOGE in unmodified form does that, the additional cap I added, which lowers o/p impedance by an order of magnitude, results in the classic rising impedance with increasing frequency.

The reg above that Jan tested has o/p impedance circa 50-60uOHm at LF to about 1k, rising to 1mOhm at 20k, 3mOhm at 60k.

Andy.

[randytsuch]

Originally posted by carlmart



Go here and you will find the Sulzer regulator:

http://hitechnetworks.net/bwaldron/a..._regulator.htm

That sophisticated regulator is also very good, still measured quite well on Walt Jung's test and it's easier to build, because it uses a one-sided board.

W.Jung did a revision on his superregulator, which was published in in Audio Electronics, to solve some problems the first design had. Do find out if the pcb Audioxpress sells now is modified for those revision changes. If it is go for it, if it's not you would have to do further wirings and modifications to do those changes. Certainly not worth the money you pay.

Warning: none of these regulators (or any discrete one) is "small enough". Certainly not to replace a chip regulator. But they are worth finding the space to put them in.


Carlos

Hi Carlos,
Thanks for the reply. I will check it out.

Randy

[fmak]

They recommend tantalum and ceramic, which is usually rejected by most audio designers or enthusiasts.

Audio applications are mostly a bit different from others
------------------------------------------------

There is nothing wrong with tantalum and ceramic in power supplies, although I prefer OSCons when I am not feeling mean.

In fact, ceramics are mandatory in digital circuits. There is no hope in hell of using RTXs in a dac for decoupling and ceramic SMTs do improve sound. All high end tweaking companies use them, and the first Sony Esprit series use aerospace tantalums for low impedance in the power supply.