Re: [CAD_CAM_EDM_DRO] Re: Capacitor Sizing in Power Supplies

jmw@... wrote:

> Jan, Jon, Mariss: Thanks very much for your replies to my cap sizing
> question. I am indeed attempting to make a power supply for formerly
> series wound, and now independently engergized DC motors.
>
> I found a nice stout-looking transformer rated 120 VAC in and 12 VAC
> @ 30 A out. There were 40 turns of 1/8" x 1/8" square wire on the
> secondary which I replaced with 2 turns of #4 copper wire. I don't
> know if such big wire was required, but it looks very impressive.
> Anyway, I fired it up and 1.2 volts, nice and steady, everything
> running cool, no fuses blowing--all at no load.
>
> Surley this thing would not be capable of 1.2 VAC @ 300 A?

Pretty close! The 2 turns don't properly fill the 'window', and so
leakage inductance (which signifies poor coupling between primary
and secondary) will be a little higher than the previous winding, but
it would come close.

>
> I next connected one of C&H's magic 600V @ 40A $4.95 butter pat
> rectifiers and a 43K uF cap which happened to be on hand and read
> very close to 1.2 * 1.4 VDC.

Now, note that when the bridge rectifier is conducting, there are two diodes
in series, giving a loss of about 1.4 V, so the rectifier will lose about as
much energy as there is delivered to the load.

> I doubt the motors' simultaneous amp draw would ever exceed the
> nameplate rating of the rectifier, but can full wave lumps be
> connected in parallel, or does one just hunt down a bigger unit?

They do not parallel well, and the lower a voltage you go to, the worse
it gets.

> Jon, I'm concerned re your comments about current surges in the
> secondary--what would cause such surges and how could they be
> controlled?

Every time the line voltage passes through zero (120 times a second)
the diodes stop conducting, and the capacitor voltage starts to drop
(due to the load continuing to draw current). As the line voltage comes
back toward a peak, the secondary voltage reaches a point where it is
higher than the capacitor voltage plus the 2 diode voltage drops, and the
diodes begin to conduct again. The line voltage continues to rise for a bit,
and the current flow from the secondary, through the diodes to the
capacitor can be quite large, depending on the series resistance of
the whole circuit and the transformer leakage inductance. These pulse
currents repeat 120 times a second, and can be very large, sometimes
10 to 100 TIMES the steady state current draw from the capacitor.

> I hadn't hought about connecting the field coils of the axes motors
> in series. Suppose just X any Y were connected in series (Z is
> another installment)--how might this affect the voltage requirements?
> Maybe a couple more turns of #4?

Well, two of them in series need 2 x the voltage at 1/2 the current.
3 of them need 3 x voltage, at 1/3rd the current.

> Jan, would you say the 1k uF per amp independent of voltage is a
> working rule of thumb most practical power supply designers would
> agree with?

Hmmm, it totally depends on the character of the load, and how sensitive
it is to voltage fluctuation. In this case, you should have a LOT of
leeway.

Jon