Electronics > Projects, Designs, and Technical Stuff
0-20V 1A Benchtop PSU - Need some advice
(1/2) > >>
Vu3vwr:
Hello Folks,

While looking for a variable bench psu design, I came across two designs ; adesign from Everyday Practical Electronics Magazine. (Jan’02), and one from KD1JV, Steve” melt solder” webber .


Both these design seem to suit my requirements, 15+V, Upto an Amp, with cc-cv modes for those gotcha! Moments ( some casual tinkering around the bench, testing out radios and digital circuits) . I do have a kit on hand based on the lm723 but after some reading around I figured that the current limiting feature maynot be the best on the 723. Therefore, im leaning towards a discrete solution involving opamps.

Im looking for feedback on these circuits,  any recommended modifications or any noticiable flaws. My plan is to build one and replicate them if succesful in the local radio club.

TIA!



Kleinstein:
I see 3 weak points in this design:
1) the current control loop could be tricky to get stable, as the extra transistor and OP in the loop add gain. So I could imagine I has to be really slow to be stable.

2) The current limiting may not bring the voltage all the way to zero, because of the CE saturation voltage -  so it may not work with a dead short. If at all use a MOSFET here.

3) Using a MOSFET with a slow current control could alow high current spikes in case of a short, as the trans-conductance tends to go up with current. BJTs on the other side have their gain to go down when the current is really high. A MOSFET also needs more extra voltage to drive - so a higher drop out.
David Hess:
The first one is terrible.  All of those stages cascaded in the current control loop will yield poor performance at best.  The first example below shows how that type of current control loop should be done.

The second one is much better albeit complex for its intended function.
Kleinstein:
The second design is really much better - it's a version of the floating regulator. The way it is drawn it looks a little complicated - the number of parts is not that large.

It still has an odd range switching for the current range - I would not recommend this detail.

It also follows the old way of adjusting the voltage by changing the feedback divider - the better way is to adjust the reference value. In newer supplies with digital control the better solution is chosen by chance, as the DAC usually sets the voltage.

The floating type regulator has slightly different properties than the "classical" regulator with a emitter follower type (low output impedance) power stage. The floating regulator needs more capacitance at the output, tends to be a little slower in voltage regulation but faster/better with current regulation. The compensation is a little more critical (I have some doubt the details are right in the shown 2 nd plan, but it's hard to tell without a simulation).

The floating regulator is quite flexible in the voltage range and can especially easy be used for a high voltage, while the regulator with a follower gets difficult when the voltage range exceeds the swing of the OPs.
xavier60:
Actually the first design has a P channel MOSFET, therefore is current outputting rather than voltage following. Meaning that the compensation method is also wrong for the voltage loop as well as for the current loop.
The circuit may only outwardly appear to be stable because of the massive output capacitor.
Navigation
Message Index
Next page
There was an error while thanking
Thanking...

Go to full version
Powered by SMFPacks Advanced Attachments Uploader Mod