It is heating that causes a linear supply to be inefficient. Voltage across circuit parts times the current causes the heating.

A transformer voltage is RMS, but the bridge rectifier charges the main filter capacitor to the peak voltage (1.414 times higher) minus the peak rectifier voltage drop of about 2V. Since any more voltage than you need produces wasteful heating in the high current parts you should calculate how much voltage is needed and do not exceed it.

The transformer must provide enough voltage for the circuit and enough power (current) for the circuit plus for the heating.

On www.electronic-lab.com, there is a 0V-30V from milliamps to 3A laboratory power supply project that has been there for 15 years and is/was a kit. I helped fix its many overloaded parts and recently some Chinese companies copied its original problems. The original defective circuit is still at Electronics-lab and my fixes are in its notes somewhere.
For its 30V/3A (90W) output I spec a 28VAC/4.2A (120VA) transformer. Its supply is +37.6V and -1.3V. If its output is shorted or is set to almost 0V at 3A then the heating is 37.6V x 3A= 112.8W plus the 2V x 3A= 6W in its bridge rectifier.

The positive voltage is +37.6V because its output transistors and their emitter transistors need about 34V, the driver transistor needs about 35V and the opamp needs about 36.6V to produce 30VDC at 3A.

 

It is heating that causes a linear supply to be inefficient. Voltage across circuit parts times the current causes the heating.

A transformer voltage is RMS, but the bridge rectifier charges the main filter capacitor to the peak voltage (1.414 times higher) minus the peak rectifier voltage drop of about 2V. Since any more voltage than you need produces wasteful heating in the high current parts you should calculate how much voltage is needed and do not exceed it.

The transformer must provide enough voltage for the circuit and enough power (current) for the circuit plus for the heating.

On www.electronic-lab.com, there is a 0V-30V from milliamps to 3A laboratory power supply project that has been there for 15 years and is/was a kit. I helped fix its many overloaded parts and recently some Chinese companies copied its original problems. The original defective circuit is still at Electronics-lab and my fixes are in its notes somewhere.
For its 30V/3A (90W) output I spec a 28VAC/4.2A (120VA) transformer. Its supply is +37.6V and -1.3V. If its output is shorted or is set to almost 0V at 3A then the heating is 37.6V x 3A= 112.8W plus the 2V x 3A= 6W in its bridge rectifier.

The positive voltage is +37.6V because its output transistors and their emitter transistors need about 34V, the driver transistor needs about 35V and the opamp needs about 36.6V to produce 30VDC at 3A.

Thanks for your input sir, much appreciated.

I think I know the project you are referring to, is it "the hiland circuit" originally produced by a Greek copied by the chinese?, I remember reading about it and your long comments about its shortcomings and the improvements needed. One being to change the op-amp, as I did in this circuit.

I bought 3 from China on Ebay and implemented you alterations, I built two, worked fine but only work at two amps load.
Neither worked initially on closer inspection the .1 ohm resistor was 10 ohm.

I have pondered over it many times, the reference voltage seems to be set by an op-amp?.
Has anybody ever written a full enthusiast step by step explanation of how it works?, there is a lot of interest online in this circuit.
It's obviously way above me at present but a lot could be learned from it.

 

The Hiland power supply circuit is very simple and is similar to the one you are making now. It uses a 5.6V zener diode that has a voltage that barely changes when the temperature changes and doubles it to 11.2V (its reference voltage) with an opamp because the modern reference IC you are using now was not invented yet.
It has adjustable voltage regulation and adjustable current regulation.

 

The Hiland power supply circuit is very simple and is similar to the one you are making now. It uses a 5.6V zener diode that has a voltage that barely changes when the temperature changes and doubles it to 11.2V (its reference voltage) with an opamp because the modern reference IC you are using now was not invented yet.
It has adjustable voltage regulation and adjustable current regulation.

Can I just ask you one question about the Hiland Power Supply?

The op-amp in the centre provides the ref voltage, the op-amp on the right actually sets the output voltage and the op-amp on the left is the current limiting op-amp? is that correct, sorry three questions

 

Can I just ask you one question about the Hiland Power Supply?

The op-amp in the centre provides the ref voltage, the op-amp on the right actually sets the output voltage and the op-amp on the left is the current limiting op-amp?

Yes, except the opamp for current produces adjustable accurate current regulation, not simple current limiting.

 

Yes, except the opamp for current produces adjustable accurate current regulation, not simple current limiting.

Thanks to all who helped me so far, I've learned a lot, I could and have wasted a lot of time online looking for answers to questions which I know are elementary to you but not to me.
The internet is also full of bad misleading information, which is very frustrating.
I don't know if anybody is still available or willing to help me, a novice, understand this circuit further, or is fed up with it and me at this stage, but I'm going to ask anyway.

How or can I power the op amp circuit "A" at a low voltage?, say 12v and power the "B" power transistor stage at 23v.
If I decrease "A" to 12v and leave the collector of Q2 and Q3 connected to +23v I will lower the max output voltage.

Basically my question is how can I run A at 12v or lower and run B at the full 23v to increase the voltage range. I know the minimum voltage is limited to ZD ref voltage (5.1v). The reason I am asking is most op amps seem to work at low voltages, I used his one (TLE2141) because it will operate at a higher voltage.

 

How or can I power the op amp circuit "A" at a low voltage?, say 12v

If you do that the max voltage output would be less then 10 volts with the present design.

 

most op amps seem to work at low voltages

There are still a lot of op amps available that work at more than 30V across there supply terminals.

So what's wrong with using the TLE2141?

 

If you do that the max voltage output would be less then 10 volts with the present design.

So it could work, but not in it's present design?, it would require a more complex design

 

So it could work, but not in it's present design?, it would require a more complex design

One of the reasons I'm asking is, I have a working Altai unit which is very similar in design to the one I'm making.

The Altai unit has two output ac voltages from the transformer, none centre tapped.
One output over 20v and the other 6v.
Soldered to posts on the board, it looks like this 6v is on a doubler circuit powering the op amp (quad)

 

One of the reasons I'm asking is, I have a working Altai unit which is very similar in design to the one I'm making.

The Altai unit has two output ac voltages from the transformer, none centre tapped.
One output over 20v and the other 6v.
Soldered to posts on the board, it looks like this 6v is on a doubler circuit powering the op amp (quad)

I am still learning about op-amps, and trying to get my head around the powering of these devices, ie. dual/single rail supply, centre tapped transformers etc.

 

I am still learning about op-amps, and trying to get my head around the powering of these devices, ie. dual/single rail supply, centre tapped transformers etc.

OK here is one thing that confuses me, it's a straight question.

Some variable power supplies using an op amp as a comparator, use a single rail supply.

Some designs incorporate a dual supply, and are a lot more complex.

Is using a dual supply an advantage over a more straightforward single rail supply.

Is feeding a dual supply only necessary for audio amplification?.

 

Some designs incorporate a dual supply, and are a lot more complex.

Those designs may have more sophisticated regulation plus current limits, overload protection, etc.
What is your ultimate goal?

 

Those designs may have more sophisticated regulation plus current limits, overload protection, etc.
What is your ultimate goal?

Those designs may have more sophisticated regulation plus current limits, overload protection, etc.
What is your ultimate goal?

I have no complaints at all about the circuit in its final design and modification, regulation, and your suggestion to combine the two pots, even running it at the high voltage, I'm trying to fully understand whats really going on in the circuit.
My questions are rhetorical (love that big word), I have to ask in order to learn.
It's easy to copy someone else's design without understanding it fully.
It's much harder to alter it to your own end's.

 

I have no complaints at all about the circuit in its final design and modification, regulation, and your suggestion to combine the two pots, even running it at the high voltage, I'm trying to fully understand whats really going on in the circuit.
My questions are rhetorical (love that big word), I have to ask in order to learn.
It's easy to copy someone else's design without understanding it fully.
It's much harder to alter it to your own end's.

Humour me for a moment, I remember the first computer I had, it would only use whole numbers, nothing to the right of the decimal point here thank you, if you wanted to play a new game you had to type in pages of basic code in a program.
Computer magazines had pages of code in basic format, anyway I one day decided to peek into it's power supply (big black exterior unit) and was surprised to see how complex it was, it's basic curiosity, same today.

Building a good stable power supply is surely the first brick down after the foundation, I cannot hope to make anything good if the supply is bad.

 

1) If you power the TLE2141 opamp with 12V then its max output will be about +10.5V. The transistors are emitter-followers with no voltage gain and when they are fed the 10.5V their maximum output voltage will be only about +7.5V and the emitter resistors will further reduce the maximum output voltage of the project when fully loaded at 2.5A each to +7.25V.

2) The simple circuit has no current limiting and no current regulation so an overload might blow a fuse if you add one or cause the transformer to catch on fire.

3) The datasheet of the TLE2141 does not show your capacitor between pin 1 and pin 2, it might cause oscillation.