i'm thinking of building a buck dc dc power supply, e.g. using perhaps a LM2596 module available from ebay e.g.
http://www.ebay.com/itm/1PCS-DC-DC-...nverter-step-down-module-NEW-CA-/142147326064
what i actually wanted to use it for is actually to build a hot wire foam cutter
while the LM2596 has specs rated for 3A output current, i looked at the various smt components from photos provided on ebay e.g. a small 100uf input capacitor,
a tiny 3A smt diode (possible for the output) vs a 5A diode suggested in the specs e.g. from ti
http://www.ti.com/lit/ds/symlink/lm2596.pdf
i'm doubting if the ebay module is safe to run up to the rated 3A. in addition, for the hot wire foam cutter, based on other estimates, i'd likely need to run the device at 3V across 2 ohms (the wire) (around 4.5 watts) say for minutes to tens of minutes at a run

to prevent driving too much current hence power downstream e.g. into the LM2596 buck module and the nichrome hot wire, i'd limit the downstream load current to roughly 1-1.5A (around 4-5 watts on the hot wire) and on the upstream feeding the LM2596 module, i estimate that it would be less than 1a and 12v dc (less than 12 watts) from a full wave rectifier
-------
i'm thinking to feed the buck dc dc module, i'd build a full wave rectifier
http://www.electronics-tutorials.ws/diode/diode_6.html
but to prevent too much current from being drawn downstream i want to put a 1A fuse in the rectifier itself.

for this full wave rectifier, i'd put in additional electrolyte smoothing capacitors bumping that up to around 1000uf to reduce the ripple currents/voltage on an otherwise 'over stressed' 100uf input capacitor

the question then is
1) should i put that 1A fuse after the full bridge rectifier (the diodes) or put it before the rectifier(diodes)? would it help / make it better to put it prior the rectifier (diodes) i.e. the ac part of it
2) the transformer that i'd select is likely a 12v ac feeding into the rectifier (diodes), would this 1A fuse be adequate or would it blow even if say i stick within the normal operating conditions (e.g. downstream at the hot wire 3v x 1.5a (4.5 watts) and supplying the buck dc dc module at around 12v x less than 1a (i.e. around 12watts or less)? i'm asking this as i actually intend to use 3A diodes for my rectifier itself but accordingly there may be 'surge currents' e.g. at power up when the 1000uf smoothing capacitor charges?

the idea is that i'd limit input power into the circuit itself to say less than around 12 watts (12 V x 1 amp) so that if an overload condition happens, the fuse blows and possbly prevent damage to the buck dc dc module etc.

thanks much in advance

 

I would suggest using a Constant Current source instead of a regulated voltage, as the wire is practically a 2 ohms short circuit, that will give you a better control.

Like one of these...

http://m.ebay.co.uk/itm/DC-Constant-Current-Step-Down-Voltage-Regulator-LED-Power-Supply-Battery-Charger-/282249786024?hash=item41b765f2a8:g:ZOYAAOSwtnpXkzyp&_trkparms=pageci%3Ab83fc18b-2c3e-11e7-a964-74dbd180e9b1%7Cparentrq%3Ab5c6c6dd15b0a88b6be1e60effff85ff%7Ciid%3A3

As for the psu, go with a switchmode psu like an atx, or laptop psu type, they can provide more current, otherwise you are buying an expensive transformer, or use one from an old car battery charger.

 

i've been thinking about using one of those linear regulators e.g. LM317
http://www.ti.com/lit/ds/symlink/lm317.pdf

however, i'm put off in the sense that the transformer outputs would probably deliver voltages in excess of 5-6 volts or perhaps 12 volts rms.
this could mean that at say around 1-1.5A, a large amount of heat would be generated at the power regulator itself and say only 2 watts (1 amp x 2 ohm) is used up at the load (the wire). i end up doing a lot of searches and it seemed a buck converter may possibly help in this case

----
edit thanks, i see an update above

 

Yes a Lm2596 ready-made one with current limit plenty on Ebay. Linear regulators are no good they will get hot and waste too much power.