HP power supply clone problem

[electros6]

HI guys,
            I tried to do power supply using the exact schematic using HP E3620A . I used LM358 for the control opamp and it works fine. Now I tried to do the same using LF353 every thing works fine expect a problem. under no load the output is good (attached the screen shot of scope) but when loaded above 500mA the ouput is oscillatory about 200mV peak (attached the screen shot of scope) I tried to used caps from 10pf to 470nf across the feedback loop of the opamp but the same oscillation persist. any suggestions. I used IRFP150N . the power supply designed for 42V ,5A.

[kjs]

Try a series resistor in series with the feedback capacitor you implemented. That normally cures that problem. It looks like your modified circuit doesn't have sufficient phase margin.

Juergen

[jman]

can you post the Schem?

[prasimix]

can you post the Schem?

+1

Question: are you entering CC mode of operation with such load?

[diyaudio]

HI guys,
            I tried to do power supply using the exact schematic using HP E3620A . I used LM358 for the control opamp and it works fine. Now I tried to do the same using LF353 every thing works fine expect a problem. under no load the output is good (attached the screen shot of scope) but when loaded above 500mA the ouput is oscillatory about 200mV peak (attached the screen shot of scope) I tried to used caps from 10pf to 470nf across the feedback loop of the opamp but the same oscillation persist. any suggestions. I used IRFP150N . the power supply designed for 42V ,5A.

Great someone cloned it. +1 do post a schematic.
LF353 is a General Purpose Amplifier,

The original design uses an LF422 for both CC and CV loops. To begin with, the LF422 op-amp offset is 5.0mV(Max) where the LF353 is 10mV(Max) see attachments and compare the differences.
see attachments

Another thing, why a FET?, the pass through power transistor uses a BJT 2n6056 (Darlington Power Transistor). you have deviated tremendously. honour the design by supplying the original components first, then proceed to mod.

Also note, the power Mosfet gate channel being driven from the *non-ideal* op-amp thanks to that low output impedance forms a  nice RC loop, this creates phase shift and positive feedback  makes a nice oscillator.

The difference between this design, and other power supply design attempts floating is this..., we know this power supply works! an its a pretty simple design (all the hard work has been done thanks to Keysight) all you have to do is use the original parts.

[bdivi]

Some experimental results with replication of exactly E3620A from my side:

Opamp - first tested TL072 with no issies and then leplaced with OPA2134 in the final schematics - higher speed and lower offset
Power transistor - two 2SC5200 in paralel, I wanted fast transistors in order to keep phase control within the opamp capacitive feedback

Overall very stable with different loads.