dumb Shunt Zener question

[volki]

sorry, dumb question probably:

for a psu, i want the voltage from the rectifier bridge to drop 10v or so before the v.reg, so i doesn't get beyond max. pwr rating at high current levels. i've seen schem's with dropping R's or even transistors for that purpose, but never a simple shuint zener. is there a major disadvantage in that?

say, my xformer gives me 38 dc after the rectifier, and want that to drop down to 28 to 30 dc before the v.reg, for 24v output at .8 to 1A current. can't i simply place a stack of zeners (28 or 30v in total) to ground before the v.reg?

thanks for enlightening me!
cheers,
volker

 

[owel]

Why don't you just use a LM317 regulator. As long as the input-output voltage differential is less than 40V, you'll be fine.

For your app, that's just a 14volt difference... well within the specs.

Just use a heatsink to be safe.

 

[volki]

my point is that i don't want the v.reg to do all the regulation from 38 to 24 or 20v respectively (multi supply psu). a 14 to 18v drop at .8 to 1A would probably be close or even exeed the max. pwr ratings of the lm317 (to220 version) i intend to use...? just trying to be on the safe side and leave some extra margin, not to get too close to the abs. max. ratings.

you're welcome to prove me wrong, since that'll simplify things :green:

 

[owel]

According to the PDF specs, the load current for the TO-220 is 1.5A... same for the TO-3 package, 1.5A.

If Vin-Vout <= 15V, the LM317 has a maximum current limit of 3.4A for the TO-220 and TO-3 package.

If I'm wrong, somebody correct me. But I think you'll be fine.

 

[CJ]

That's not a very efficient/pratical way to do it.
Once your transfomer delivers rectified dc at a certain voltage, there is really only one way to get the voltage back down to where you want it, which is chew up some volts and amps and therefor power.
This can be done with transistors resistors, zeners, whatever, but it is still simple heat generating voltage dropping any way you slice it.

The zeners would have to be able to handle, lets say, 1 amp at 20 volts, which is 20 watts. I haven't seen too many 20 watt zeners around lately.
Also, this might overheat your transformer. And you would also need a power resistor in series with the zeners so they won't cook. This will generate a lot of heat also.

The only way to kill the voltage without heat is to use reactive elements.
A choke after the rectifiers would help.
A cap feeding the primary has been done at low power levels.

A series resistor in the primary of the power transformer would have to
chew up as much power as a series resistor in the secondary due to the laws of consevation.

Best way:
different transformer
choke input
20 watt pwr resistor in front of the reg.
big heatsink on the reg

 

[DrFrankencopter]

One more way to do this is to use a tracking pre-regulator, which knocks down say 5 of the 15 volts, a series resistor, which takes out a little, and 2nd regulator to do the fine tuning. The advantage of this approach is that ripple is really low. I built a phantom supply like this, and had something like 100 microvolts ripple.

I was quite happy with the results...and so were my mics!

K

 

[volki]

yeah, i thought about the double regulated approach before, too. probably, a simple transistor preregulation would work best for me.

is there an advantage if you use a resistor (small value, i figure?) btwn pre-reg. and main reg.?

so, can anyone recommend me a pair of transistors (pos. and neg.) to drop down voltages from +/-38 to +/- 24, as well as from +72 to +52v? should be suitable for 1A current (abs. max) each. - would i place an resistor btwn collector and emitter, and float the base with some zener, maybe with a cap parallel to ground?

or would it be even easier to just get some standard L78xx/79xx v.regs and float'em on zeners? :roll:

thanks,
volker