|
Experimental, untested....best not to lower the load resistor below 5 ohm ...use any variable higher resistor.
The pi filter settings may be unnecessarily high (earlier component wear due to current rushes)...If only mA loads are reqd. then they could be reduced.
You could also try increasing those caps & inductor in order to lower the Vpp ripple if you wish...it is roughly about 200mV at the moment.
For a basic single cap instead of the pi filter ....you could use the formula to find the ripple Vpp:
V = I/fC where I is the load current amps, f is the output frequency (ie 2x50 =100 here), C is capacitor value farads.....main aim is to get this V pp ripple down as far as possible eg to 100mV or perhaps 5% of initial dc peak volts....THEN THE FINAL DC OUTPUT from the filter= output transformer max give or take nothing much apart from 1.4v less due to the two diode path.... then comes the final section to keep that voltage as per the zener regardless of load current changes:
Mosfet is Nmos LL with a standard 1.5 VTO ....you will need a good heat sink on it for heavy loads, and consider transformer/inductor heat then too as well as the load....you may need to fiddle with resistors for individual alterations...this setup uses negative feedback via the op-amp after the pi filter with respect to the zener diode output and is quite sensitive and is not a good efficiency design.
The EC op-amp gets its supply as per its settings....you could probably supply the op-amp on a board using the filtered output from the pi filter?
It's a Linear Regulator as opposed to the more efficient Switched Mode power supply, and its usage may be better used for lower loads and/or when input voltage is very close to reqd. output voltage, but it is an interesting option. I normally advise keeping the Source of a Nmos connected directly to the -ve supply, but
we are not fully switching here, so you can see that the load here is between the mosfet Source and -ve. (more like a mosfet linear mode just above the gate opening stage...reminiscent of a bjt amplifier mode)....giving the advantage of less switching noise cp a SMPS.
|
