POWER SUPPLY

As you say, you're designing it so ask yourself.

--DF

Using an inductor will give better output voltage regulation versus load current. If you have a constant load current then don't bother with the inductor.

For given ripple volts across the capacitor, if you don't use an inductor then you will need a much bigger capacitor.

If all you want is rectified AC then don't bother with either.

--
Reg.

I was thinking more about noise. How does an inductor do with noise versus a capicator?

Really ?

How about the inductor DC resistance ?

Graham

Noise ?

You're certainly no designer.

This is a *basics* question not design.

Graham

A damn lot better than an equivalent resistor, at any rate.

I wonder how the relations work out for ripple vs. transformer regulation. Metric shitloads of capacitance don't make transformers happy.

Tim

-- Deep Fryer: a very philosophical monk. Website:

you're an audio guy Graham, go read the Langford-Smith, its all in there :)

an inductor (if > 3%) will also knock the input harmonics down, a lot.

Cheers Terry

Jimbo, A couple of interesting things happen with an inductor input filter when the inductance is high enough to ensure continuous inductor current (This is called the critical inductance).

1) The load regulation improves dramatically. 2) The ripple voltage is independent of load current. (as the load current increases beyond the critical current, the ripple voltage no longer increases. Also, the capacitor peak current is much lower, since the current is spread out over a larger portion of the power cycle. The power factor will be larger for the same reason. Disadvantages: 1) Lower output voltage for a given capacitor value. 2) The inductor must have an air gap to prevent saturation from the DC current. This means that for a given inductance, the inductor must be physically large.

but hey, it can be made from sheet steel - McKenzie Holland used to do this for big battery chargers!

Cheers Terry

Oh, very true.

When IEC1000-3-2 came along I did some simulations using inductors to reduce the harmonic currents.

The regulation compared to a simple capacitor input filter was utterly miserable though ( and I assumed DC R to be near zero ! ).

Graham

There is no such thing as an 'equivalent resistor'.

Just keep the DC R low and they don't actually mind !

Graham

In article ,

The regulation from no-load to full-load is certainly worse, but the regulation from a defined minimum-load to full-load is better. The final dc output is always lower than the RC filter, but the choke input filter makes better use of the available VA.

AFAIR the harmonic currents are better, but there is a nasty shorting effect when the diodes switch over, (commutate), effectively shorting the leakage inductance of the transformer.

--
Tony Williams.

I used to have a dc arc welder, powered by a 24V lead acid battery. It had a curious (over)current limiter in series with the supply.

It was a 'spring' of heavy gauge galvanised steel wire, about 2" diameter, 7" long, tightly wound so that adjacent turns touched (and had continuity).

It was claimed that under a transient overcurrent the magnetic field between adjacent turns extended the spring, breaking the inter-turn connections.

--
Tony Williams.

How do you reconcile that with your first statement ?

I found the regulation to be much worse than a capacitor input filter under any practical conditions.

I simply reduce the ( typically toroidal ) transformer's DC R to avoid any such problems. VA isn't determined by core size in reality when the possibility of 'special winding' is there.

p.s *what* RC filter ?

Graham

Download the software PSU designer from

and check it out for yourself. It isn't highly accurate as Spice might be, but it's certainly close enough for a basic PS design. It allows you to compare RC vs. LC filters at constant R loads, constant I loads, or stepped I loads. Pretty versatile for the hobbyist.

--
Dave M
MasonDG44 at comcast dot net  (Just substitute the appropriate characters in the