You don’t
necessarily need a bridge, but it should be a full wave
rectifier, otherwise it is very inefficient. In a high current
supply, it would also cause a great deal of power dissipation
in the filter capacitors. Half wave rectifiers are best left
for very low current applications.
Most battery
charger use a full wave center tap rectifier , which requires
only two diodes. It is slightly more efficient than a bridge,
since there are two fewer diode drops (0.6V @ 100 A would mean
60W dissipated per diode). Dynamotors do not like raw
rectified AC and will run hot. Obviously, filaments don’t
care, but the RMS value needs to be correct for proper
emission. So, really, the KISS method is the minimum
acceptable of a stiff, filtered, full wave supply. If it is
sufficiently sized, load regulation will probably be
acceptable.
Man....all
this info is great!
And with regards to building an xfmr 'brute force' supply,
do you really need a bridge rectifier or will half-wave do?
Do dynamotors and filaments really care if there's a bit of
ripple? I have some large diodes and caps I could put to
experiment with. Kinda the KISS method.
de kn7sfz