Bear Symphony #1

[mikeks]

Hi Bear,

looks interesting....

http://www.bearlabs.com/NEXT/SYMPHONY.html

Perhaps you might enlighten us further in respect of your design philosophy...shematics...etc....?

[Eva]

After '2,5K VA' and 'half farad' I decided not to spend my time reading further... My mains line peak voltage drops by nearly 25% when loaded with 2,5KW of rectified DC load [about 15KW peak power consumption!!]

People should stop wasting their money in 2K VA toroidal transfoemers and huge capacitors and build regulated SMPS [or linear regulators]. There is no other way to get good regulation since mains line itself is pretty unregulated

Active PFC is also a very interesting option since it reduces peak mains current consumption by up to 70%. It appears as a pure resistive load to mains, produces almost pure 100Hz ripple on its output [much less harmonics than standard rectification] and provides less than half the peak to peak ripple voltage with the same filter capacitance. And it offers average output regulation against mains fluctuations too

[PMA]

Exactly, Eva. Huge capacitors do only increase RF interference and overload the diodes.

[Elso Kwak]

Originally posted by PMA
Exactly, Eva. Huge capacitors do only increase RF interference and overload the diodes.

NOT if these are after the choke in a PI-filter!

[Swedish Chef]

These amps would be awesome to weigh down your car with on an icy mountain road.

[Eva]

In order for a pi filter to not resonate at audio frequencies, capacitor ESR / inductor DCR / capacitance values should be big

Further increasing inductor value in order to get a low enough cutoff causes the DCR/ESR/capacitance values required to tame the resulting third order resonant system to be huge

So pi filters damp ripple and reduce peak mains current value but allways at the expense of additional voltage drop on its output [ie: worse regulation] or resonance at audio frequencies

Pi filters were very useful for tube PSUs since inductor DCR in the range of 100 ohms was practical at values like 500V and 200mA [resulting in no filter resonance], but this is no longer true for solid state values like 50V and 5A

[Elso Kwak]

Originally posted by Eva
In order for a pi filter to not resonate at audio frequencies, capacitor ESR / inductor DCR / capacitance values should be big

Further increasing inductor value in order to get a low enough cutoff causes the DCR/ESR/capacitance values required to tame the resulting third order resonant system to be huge

So pi filters damp ripple and reduce peak mains current value but allways at the expense of additional voltage drop on its output [ie: worse regulation] or resonance at audio frequencies

Pi filters were very useful for tube PSUs since inductor DCR in the range of 100 ohms was practical at values like 500V and 200mA [resulting in no filter resonance], but this is no longer true for solid state values like 50V and 5A

Eva,
This is totally wrong.
Some Pass model has indeed a working PI-filter. Maybe Nelson Pass himself can persuade you....
I am sorry I don't remember which Pass model.
For your information I have 2 x15mH and 4x 12000µF in my Ampzilla and it is a working circuit even at full power into a 4 ohm load. The DC resistance of the coil is 0.57 Ohm
I still remember Nelson's post: Pi-filter multo bene.

[PMA]

Elso - she is right and argueing in technical terms (it is very easy to calculate what happens in the LC filter), contrary to you, who argue in terms of audio gods .

[Jocko Homo]

Oh........so no one has ever built a commercial SS amp with a Pi-input filter.

Wrong!

Jocko

[PMA]

OK - let's speak about values and behavior of the defined filter.