LM338 regulator InputMax voltage..

[Roushon]

I just purchased toroidal trnasformer for a LM338 regulated power supply for LM3886. I ordered 2 x 30v sec. (300VA). But the transformer giving 2 x 32v. So after the bridge rectifier I will get around 45v approx. Is this supply safe for LM338? The datasheet of the regulator says max input voltage is 40v.
I want 35volt output from the regulator.

Thanks in advance for any suggestion...

[Roushon]

i got my answer. sorry for starting a new thread. just read, several people are using with even 60v input for LM338. so i guess, i can too with 45v.

[Tweeker]

Vin-Vout <35V.

The regulator can be used at hundreds of volts provided its protected and doesnt see greater than 35V across it.

[Roushon]

thanks Tweeker. i read about it in the datasheet, but never saw an example with large input voltage till i searched the net like mad.

i have another concern. the regulator can supply around 5 amp current. is this independent of the input voltage as along as Vin-Vout<35v?

[FastEddy]

Just because the spec sheet says "up to 35 DVC" does not mean success in all cases. The Absolute Maximum is just that = absolute. If there is any ripple or surges coming from upstream then = no joy

http://www.national.com/ds/LM/LM138.pdf ...

Here is a neat trick however ... you can use a daisy chain of diodes to "knock down" the incoming voltage = each diode = 0.6 VDC throw away = 10 diodes = 6 VDC down.

You can also do the same daisy chain on the ground leg 'tween the adjustable resistor and ground ...

[pinkmouse]

Originally posted by Roushon
The regulator can supply around 5 amp current. is this independent of the input voltage as along as Vin-Vout<35v?

You need to check the load/SOA lines in the datasheet for your application.

[Roushon]

thanks for the responses. using series of diodes to drop the voltage is really a nice idea. to extend this idea, i think it will be even better to use more diodes in series in the rectifier stage itself. although it will be more expensive and cumbersome.

[FastEddy]

Roushon: " ... using series of diodes to drop the voltage is really a nice idea. i think it will be even better to use more diodes in series in the rectifier stage itself. ..."

Warning: the use of diodes in a daisy chain (stacked / chained in series) from a pure AC source (like right outta the transformer to the rectifier diode bridge) is NOT advised. This is how higher voltages can be generated, not lower voltages !!! (the AC peaks get "pumped up" in the diode series ... This daisy chain of diodes is sometimes used in old radio circuits to increase peak voltage (!!!!) = a bad idea for AC circuits, but fine for well filtered DC only.

Try something else !

(The use of a daisy chain of diodes to "drop" DC voltage is an old solar power supply trick as published in Popular Electronics circa 1989 ... I know 'cause I wrote it up)

[wine&dine]

>This is how higher voltages can be generated, not lower
>voltages !!! (the AC peaks get "pumped up" in the diode series

There's no "pumping" to a higher voltage unless you specifically connect up diodes and capacitors in a multiplier (cascaDe) circuit, like those used in CRT HV supplies.

Diode drops are fine when you need a small voltage drop. In Roushon's case, an extra transistor is much safer. See http://www.national.com/an/LB/LB-47.pdf for a circuit.

[FastEddy]

" ... There's no "pumping" to a higher voltage unless you specifically connect up diodes and capacitors in a multiplier (cascaDe) circuit, like those used in CRT HV supplies. ..."

If poor wiring practices are used, the stray capacitance could conceivably cause the voltage to pump up, possibly much more than the 0.6 volts "lost" to the diode barrier. ... I would not recommend diodes in series passing AC = too unpredictable ... (Diodes in parallel are another matter.)