DC Power Supply 0..40V 0..5A

[kekon]

I design Regulated DC Power Supply with parameters shown above. I designed the whole circuit but I want to see any other solutions. Does anybody know where I can find them ? The main problem connected with designing such power supplies (with wide output voltage range) is with error amplifier which must drive power transistor at high voltage above 40V. Typical operational amplifiers have output voltage range below 30V. Because of that I added additional transistors to operational amplifier to achieve 0..40V output. However, it is sometimes unstable (oscillations appear at the output)

 

[Johnson]

Did you consder the linear regulation, if so waht about power dissiaption on this wide range power. Let the output is 2.5V @5A so you have more than 40*5=200W! power on outout transistor which must be dissipated.

You may need to do more control over this wattage, or go toward switching mode devices.

 

[IanP]

Two remarks at this stage:

1. Consider deviding the whole range into two, with automatically operating relay for lower and upper halves. What I mean is using a transformer with 2 taps 9or 2 transformer with secondary connected in series) so if you operate with output voltages below 20V you use the lower tap, and if the adjusted voltage is above 20V you the relay will switch to upper tap (with some histeresis to prevent the relay to switch from one to the other if working around 20V).
This will dramatically reduce the amount of generated heat...

2. Publish your circuit diagram. It will be easier to find out what causes the unstabile work of your PS..

 

[kekon]

I can't use any switching mode of regulating the voltage. This will introduce high noise levels. The power supply must be of linear type. It will have two additional connectors to sense the output voltage when using long wires.
The cooling will be made by using fans. However I considered to mix switching and linear mode. It would look like that: mains 230V AC voltage would be stepped down through switching regulator (smps) to, say, 50V (to eliminate heavy 50Hz transformer). The voltage from smps would be further stepped down by additional regulated step-down regulator (second stage). The second stage would power linear unit. Voltage value at the output ot the second stage would be so regulated to achieve low voltage drop on the power transistor (in linear unt) to minimize power losses. For instance, if one wishes to set 30V 5A at the output so the second stage must be set to, say, 35V. So the power loss would be only 25W. But there is one serious problem: I'm not any good at designing forward smps. I can only make flyback converters up to 150W

Added after 33 minutes:

Well, using two ranges with a relay is a quite good idea. But maybe it's better to use triacs to switch between transformer windings ? (because all the circuit will be controlled by PC which will be selecting different voltages from full 0..40V range at 1..100 Hz so the relay is not a suitable solution). Power losses in triacs wouldn't be high and utterly acceptable.

 

[platonas]

I do suggest to use a Dimmer like control of your transformer's primary. Take feedback from the final output voltage and affect the firing angle of a TRIAC. This will reduce the output voltage, hence the exessive heat on control transistor.

 

[kekon]

The transformer can't work properly when powered from triac with controlled firing angle - this will cause huge interferences.

 

[VVV]

The first thing that comes to mind is the use of a transistor in a "cascode" configuration, driven by your opamp.

I do not see a schematic, so I cannot tell how you connected those extra transistors. It is possible you introduced some additional phase lag that causes instability. You need to do a gain-phase analysis. Can you post the schematic, or at least the part that is your concern ?

 

[kekon]

Well, I've jusr changed the circuit a bit so the problems with stability seem to decline. The oscillations at the output disappear when larger capacitor is connected to it. However, I want the capacitor to be as small value as it's possible and ensure stability. Now I'm working over it in my laboratory.

 

[kekon]

Here I publish the diagram of my circuit. One must emphasize that this is not the final circuit yet. Components values are being changed and are not the final values. The diagram presents mainly an idea of dc power supply 0..40V 0..5A. However, some parameters were estabilished. Voltage and load regulation are excellent. There is concern about step response (applying a load to the output) which must be improved. Does someone have an idea how to do it ?
Components C1, C2, C4, R6 ensure stability. Calculating their values is not easy and some experiments were carried out to choose proper values. R20 resistor was introduced to measure output current and to create current limiting unit (it is not introduced to the circuit yet). Unfortunately I'm unable to show screen shots at the moment from my oscilloscope as it doesn't have the possibility to send them to PC. I understand that one can use special high voltage operational amlifier instead of U2 amplifier and additional componets (T1..T4 etc.) but I don't have one at the moment. VREG voltage controls the 0..40V output voltage. It must be negative in reference to AGND ground. Please note: there are two different grounds AGND and PGND. Don't confuse them !