CRCRC vs. LCLCLC which is better and why?

[Clarkcr]

Can anyone tell me why inductors should or should not be used in an Aleph PSU? Someone told me once (I read it here I think) that only resistors should be used to drop voltage for the rails and that inductors were a no no. Why is this? I can't find the link anywhere.

Don't inductors posses a lot of great qualities for dropping voltage, quietly without affecting current.

Or give me a link to the post and I'll be happy.


C

[analog_sa]

Don't inductors posses a lot of great qualities for dropping voltage, quietly without affecting current.

No. Sadly, they quietly obey Ohm's law when dropping voltage.

[mikelm]

Yes but an inductor can have a much lower DC resistance than high frequency impedance - they can be usefully used to filter out HF noise without dropping too much DC potential.

But great care has to be taken with the design otherwise you can get 2 x Vcc voltgae at turn on which could well destroy your amp.

I would recomend modelling in spice before you build.

The basic principal I have fould is that if you want to avoid this initial overshoot is to have for instance 10kuF L 20kuF L 40kuF and the resistance of the coils should not be to low perhaps 0.1 - 0.2 ohms minimum.

PLEASE NOTE THAT I AM NOT GIVING YOU A DESIGN HERE JUST SOME IDEAS TO START WITH AND TEST VERY CAREFULLY.

good luck - I find that choke supplies sound great

mike

[Greg Erskine]

Hi Clarkcr,

I just happened to reading about this today so I'll quote you a few points from the text. Please be aware that I have no personal experiences with these filters.

Generally speaking L-C filters are preferred.

L-C filter
======
"The choke input filter allows a continuous current from the rectifier diodes, rather than the pulsating current demand by the capacitor input filter, and because of the, it has better voltage regulation. Maximum current is available to the load with this type of operatoin, so this filter is more suited to higher load-current demand.

On the debit side, the output voltage of the filter is equal to the average of the AC voltage delivered to the rectifier, rather than the peak value. It is also very costly, because a physically large, low resitsance inductance is required."

C-L-C filter
========
"The pi filter is more efficient.....is used on medium-current equipment, it does produce a higher output voltage then the L-C filter, but the regulation is not as good."

C-R filter
======
"this type of filter is less expensive. The filtering action is not as good, and here is some loss of DC voltage at the output because of the voltage drop across ther resistor. It is limited to low-current equipment..."

Reference
=======
Understanding DC Power Supplies - barry Davis

As I said earlier I am only quoting from a text and can't confirm the information is correct, but I hope it is useful.

[Clarkcr]

Thanks you guys, that's a great start. I hav a copy of SPICE that I havn't installed yet....maybe now is the time. I was told by a neighbor of mine that the toroidial inductor will will also bring the capacitors up slowly so as to limit the in rish of current that plays havok with my house circuit breaker.

It seems from Greg's text that the LC is preferred....Why not an LCLCC or even an LCLCLC. I'm using 6 72,000 uF capacitors in my filter supply per amp. I'm also using a 1500va Piltron toroid. I don't mind the heat of the CR filter but, am very intrigued by the other advantages of the choke.


Chris

[Clarkcr]

Originally posted by analog_sa


No. Sadly, they quietly obey Ohm's law when dropping voltage.

I meant with less heat, slower in rush of current. Greg's text book quote even talks of the differences in current loads. It also adds to the ultimate goal of lowering the ripple current.

Will somebody please speak up for the lowly reisitor in a CR???????




Chris

[Clarkcr]

This is probably a forum no no, but I posted the same thread on the solid state board to see what response I'd get there as well.

here is a response I got. Very interesting...does Single Ended Class A make this situation any better or any worse?


Quote:
A LC cell provides 12dB/oct filtering but the L and C values have to ve carefully selected to create a damped non-resonant system, depending on inductor Rdc and capacitor ESR

The LC filter design goes more and more complex as more LC cells are added in series, it will definitely require custom-wound inductors with custom inductance and Rdc values and some minimum load resistance on the output to avoid resonance

In the other hand, a RC cell provides only 6dB/oct filtering but it will never resonate, even in case of crazy R and C values, and you can easily place several in series




Thanks to you all again.


Chris

[mikelm]

the first LC section of an LCLC... behaves completely differently from subsequent stages. It means that your amp will create less havoc with the mains supply. It is quite an elegant design that requires very careful design and is only really suitable for amps that runs in class A. The more the bias current he easier it is to make this design work properly. ( Properly means that there is always current flowing in the transformer coils )

have fun with the spice

mike

[Clarkcr]

I'm building a pair of Aleph 1.2s BTW. 60 volt rails.

C

[Nelson Pass]

Originally posted by Clarkcr
This is probably a forum no no, but I posted the same thread on the solid state board to see what response I'd get there as well.

Depends on who you want an answer from.

Originally posted by Clarkcr
an LC cell provides 12dB/oct filtering but the L and C values have to ve carefully selected to create a damped non-resonant system, depending on inductor Rdc and capacitor ESR

The LC filter design goes more and more complex as more LC cells are added in series, it will definitely require custom-wound inductors with custom inductance and Rdc values and some minimum load resistance on the output to avoid resonance

In the other hand, a RC cell provides only 6dB/oct filtering but it will never resonate, even in case of crazy R and C values, and you can easily place several in series

All quite true, but easily covered as long as you use big enough
capacitors - which is what you want to do anyway. Resistors
across inductors also help damp out such systems - values
like 10 ohms or less.

And you know what I mean when I say big. I mean porno BIG.



/pass/: gets lots of amplifier ideas from porn movies