0-30 VDC Stabilized Power Supply

Thanks a lot for the explanation.

The ripples should disappear if you use a 10A rectifier bridge module instead of the tiny low current 1N4148 diodes as the rectifier.
Maybe you SIM program will still show ripples because it did not read the datasheets for the parts.

The inefficiency of all linear ac-dc 50Hz must be reduced otherwise the series drop V * I represents a huge bypass loss for lower V out.

I dont know what transformer you plan or have or budget limitations, but let's shoot for 30V 3A supply with a 24V 100VA transformer

Thus if your requirement is 90W for 30V @3A which represents a 10 Ohm for a linear load, the diode peak current will be 5x or 15A with 20% ripple.
Using 4 of these SMT Schottky diodes

MBR745 = 45Vpiv 840mV @ 15A but can withstand 100A for startup cycles to charge up Cap.

Note from VI curve these diodes have an ESR loss from the slope above from 10 to 20A = 0.1V/10A = 0.01 Ohm which is far superior to what you selected.

Diodes drop in voltage when they get hot but a good design limits this to 85'C not 125'C even though they withstand this.

Simulate your Transformer with a series resistance.
If P=V^/R = 100 VA and efficiency is 95% then 5W loss represents 1.67V at 3A.

Then simulate this unregulated design first with a variable load to see Vavg, Vmin vs I or R from 10 to 1K Ohms. Vmin will be the absolute highest Vreg you can achieve with 2N3055's or even better a <=100mOhm P-type MOSFET mounted on a CPU cooler heatsink.

Then you can rethink how to regulate it simpler. using single supply Op Amps and NPN driver to P-MOSFET, YOu can add CC current limiter adjustment and V regulator adjustment without having to use a simple LDO which drops 2.5V Instead the only drop is past the low RdsON resistance of the MOSFET , and of course the diode bridge and minimum ripple peak Vmin.

From your schematic I see 10mF so a 10 Ohm load will produce an RC= 10*10= 100ms time constant. which for the half cycling in 50Hz rectified is 10ms giving only 10% ripple. . This would suggest, you can use a bigger power transformer or smaller Cap if you can tolerate 20% ripple and get 30Vmin out of the bridge with 10 OHm load.

Try that first or adjust parameters to suit your budget and goals.

For a load that uses 30V at 3A then its power is 90W. Many parts have a voltage loss so I think the transformer should be 28V, not 24V.
The transformer gets warm, the opamps get warm, the rectifiers get hot, the driver and output transistors get hot, some resistors get hot and the 0.47 ohm current sensing resistor heats with 3A x 0.47 ohms= 1.41W. Then I think the transformer should be rated at 28V/118VA.

If you use "typical" spec's to calculate the voltage losses then only the projects made with "typical or better" parts will work properly. Projects made with lower spec (but still passing) parts will not work properly. When you buy the parts you do not know if their spec's are excellent, typical or minimum so I think the design should be based on using minimum spec parts for ALL projects made to work properly.

so here is my pre final schematic for 20W 28 V power supply.

i have to make this power supply cost effective so any suggestion will be highly appreciated.
Thanks Audioguru for your explanations.

- - - Updated - - -

i am thinking of using
1. 50VA, 2A transformer
2. 50V, 10A Bridge Diode Rectifier.
going to remove 10 Volt ZENER.

I made my suggestion of a 100W 30V supply for you to compare.
Adjustable 3~30V ( but a little noisy with wrong Q1)

A MOSFET 50mOhm 75V would be much better for Q1.
As would hysteretic PWM control.

You can play with SIM here in java.

Your pre-final circuit is the same as the one I fixed years ago at Electronics-Lab and I posted here, EXCEPT it is missing the LED that warns that the current regulator is reducing the output voltage, it is missing the current calibration trimpot R13 so its max output current will be about 2.6A and it is missing the R20 trimpot to adjust the input offset voltage of the voltage regulation opamp so that the output can go down to zero volts. A 50VA transformer will be overloaded when the output voltage is low or shorted and the current is set to maximum.

Why do you say 20W? An output of 28V at 2A is 56W and many of the parts dissipate another 23W for a total of 79W from the transformer.

you're right...
actually i need a 20-25W power supply with 0-28V regulation.
so i thought 50VA transformer will be okay.
is 50VA okay for 20W power supply. or this ckt is not efficient for what i need.

- - - Updated - - -

i have seen there are other cheap power supply circuits .
i m not getting what should i do.

If you have a peak input of 35V from a 25V RMS transformer then a maximum output voltage of 28VDC might be possible when the output current is 2A.
But if you use a 50VA transformer then the maximum output current must be reduced to 50VA/35V= 1.43A.

I have never seen a power rating for a power supply. Maybe you are talking about the maximum output power which will be 28V X 1.43A= 40W.
If you want a maximum output of only 20W then the maximum output current is 20W/28V= 0.71A. Then you do not need two output transistors because only one will be fine.

yes of course i have removed that transistor from my list..
i am gonna use only one 2N3055.
but i am bit confused about transformer and rectifier rating for my circuit.

The transformer in the original defective project was only 24V but then the project could not produce 30V at 3A, instead it produced a max of only about 25V at 3A. The original project also had many overloaded parts including an overloaded transformer rated at 24V RMS / 3A.

So I replaced the transformer with a 28V RMS / 118VA one and upgraded many parts. Your circuit might not produce 30V at your max current if your transformer is only 25V RMS.
I do not know what max current you want.

The rectifier bridge charges the huge 12,000uF filter capacitor. It conduct only at the peak voltage of the rectified sine wave so its peak current is very high, maybe 30A(!) so it gets hot and needs to be bolted to the metal chassis. Your circuit has less max current so your rectifier bridge might conduct "only" 10A to 15A and might survive without a heatsink.

i need a maximum current of 1A.

root2hell said:

i need a maximum current of 1A.

Click to expand...

Are you going to build the power supply and test it or just simulate it?

i have sent it for PCB designing but i am modifying my design for 1A.
i want to build same unit but cheap with 1A current limit.

- - - Updated - - -

this circuit looks less complicated and cheap...

The new circuit you found seems to have many pins on the LM723 mixed up.

i cant find anywhere simulation model of LM723 Voltage Regulator so cant simulate...and i cant write Pspice model for it... anyone can suggest me something.

The LM723 adj. volt. reg. was designed long before simulation programs were available.
Its circuit is very simple so you can make a model of it with a voltage reference IC, an opamp, a current limiter transistor and a power transistor.

My PCB Design for Previous Circuit is about to Ready.
and i am thinking of build the same supply using LM723.

The first circuit has an adjustable voltage down to zero volts. The Minimum output of the LM723 is about 2V.
The first circuit has adjustable accurate current regulation. the LM723 has a fixed rough current limit that changes a little when the temperature changes.

minimum output of LM723 is 2V which will be down to 0 using circuit as shown in schematic and how much current fluctuate if temp changes from 40 Degree to 5 Degree?