Regulated Negative Power Supply

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
You need to address the transformer before you think about the caps. The transformer is just not adequate in this circuit unless you dispense with the regulator.

I can do just fine with +-25 to +-27 voltage output at 5A.
AHsJFJhC1JVdTiwNFHjIzQRYghQooATggIBTTTRlyNVL0Z4COQLoGuCnQgAduFxBEWgoDaMibGzFWSPmU4csj+wM1PKj1okIIUjp6WEESB+BGGKhQHUrj5U6CaosGTVmjgxZfDhtogIpxgAJAgBgmZTpC5kNAxkcmaIq1B1PgSotWCgQzyIagHx4afCnFZqFdF5IKWCKgGENQxAxKdGi1yBCpWrMCCDCAC0TAdrsuGRDyAcYFIJ0QSDAAIIEiQBEYpWEh5JXTySZAQBmxYADD8KQaOQCTi8cXFZ5cGRBDYRUZd6AuiFnIIhOW1QQucCBFxJMLM7QFTgiioNZheYuAgwIADs=
 
What part of this are you not reading. Your transformer is too small to feed a regulated supply.

It would be suitable for a 24V regulated supply at its VERY BEST. That would require massive pre-reg capacitors in order to keep the pre-reg ripple under the Vin requirement of the regulator.
 
Last edited:
I can do just fine with +-25 to +-27 voltage output at 5A.
AHsJFJhC1JVdTiwNFHjIzQRYghQooATggIBTTTRlyNVL0Z4COQLoGuCnQgAduFxBEWgoDaMibGzFWSPmU4csj+wM1PKj1okIIUjp6WEESB+BGGKhQHUrj5U6CaosGTVmjgxZfDhtogIpxgAJAgBgmZTpC5kNAxkcmaIq1B1PgSotWCgQzyIagHx4afCnFZqFdF5IKWCKgGENQxAxKdGi1yBCpWrMCCDCAC0TAdrsuGRDyAcYFIJ0QSDAAIIEiQBEYpWEh5JXTySZAQBmxYADD8KQaOQCTi8cXFZ5cGRBDYRUZd6AuiFnIIhOW1QQucCBFxJMLM7QFTgiioNZheYuAgwIADs=

What part of this are you not reading. Your transformer is too small to feed a regulated supply.

It would be suitable for a 24V regulated supply at its VERY BEST.

What part of this are you not understanding? I repeatedly stated that i don't need a FULL 30v, and about the 24v or the 18v that's you guessing right?
 
You have 2*24VAC at 480VA
You have therefore 2*36VDC at no load condition.
This drops to 2*33VDC at full transformer power.

For each secondary the output voltage = 24*1.41 - 1.41 = 32.43v,
input that to your normal run-of-the-mill lm317 and considering a drop of about 2-3v across the regulator gives approximately 30v,

Is there anything wrong with these calculations?:)
 
You have 2*24VAC at 480VA
You have therefore 2*36VDC at no load condition.
This drops to 2*33VDC at full transformer power.

No, at no load, the voltage of the transformer should be higher than rated. The voltage at full load should be around the rated voltage. (Though in a rectified circuit you can't count on the peaks being at 1.4142 * RMS any longer).
 
For each secondary the output voltage = 24*1.41 - 1.41 = 32.43v,
input that to your normal run-of-the-mill lm317 and considering a drop of about 2-3v across the regulator gives approximately 30v,

Is there anything wrong with these calculations?:)

LM337 specs say to expect 2 to 2.5V dropout voltage at full load. Add to that the Vbe of about 1 V for your pass transistor. So your total dropout will be at least 3.5 V, expect a little more.

Your rectified voltage will have ripple. Your Vout + dropout (around 4V) must be higher than the minimum value of the rectified voltage (minimum point including ripple). Measuring with a standard voltmeter will only give you the average, not the minimum.

To answer another question, yes, you can use almost any high power NPN/PNP transistor in place of the 2N3055/2955 in this application.
 
Your Vout + dropout (around 4V) must be higher than the minimum value of the rectified voltage (minimum point including ripple).

Thanks for the clarification anything higher than 25v will be good. :)
any advice on the capacitance values especially on the load side taking into consideration that it will be used to power a chip amp which actually was my original question. My info so far is that powering chip amps with a regulated supply that have a large capacitance on the load side will yield a muddy mid range. On the other hand there is the output impedance of the regulator and datasheet that recommends adding some output capacitance.

Some test results in a good (or better) performance with 220uf on the output...
Simple Voltage Regulators Part 1: Noise - [English]

In this very forum a while back when members built the Carlosfm regulated power supply for the GC commented about having better sonic results with no output capacitance or as little as 47nf.

You see my main problem is not an inverted PNP or head room for the regulator it is essentially this: CAPACITANCE DISTRIBUTION AND VALUE TO OBTAIN BETTER SONIC RESULTS. :)

Though of course i appreciate your advice on the first circuit posted and the link you provided.

Regards
 
No, at no load, the voltage of the transformer should be higher than rated. The voltage at full load should be around the rated voltage. (Though in a rectified circuit you can't count on the peaks being at 1.4142 * RMS any longer).

Sorry, I can not change my original statement...

Lets assume the stated 4V additional voltage drop for the regulator stage.
He therefore has 29V+/- to his amplifiers. 29V to 8ohm load gives him 37W of power/ amplifier, minus losses...
We can easily assume 37W music power including all losses, because music is not a constant sinusoidal signal.
He is therefore taking only a maximum of about 80W from the transformer to both amplifier channels.
Which means the actual rectified voltage output from the over kill 480VA transformer including the transformer regulation is about +/-35V, which means plenty of head room for a +/-29V regulated output...

IMHO...

Capacitors?
Im a big fan of a massive capacitor bank, even if regulated power supply...
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.