Power Supply Problem - diodes get really hot

Power Supply Problem

I have a problem with a power supply.
The circuit is basically a DC-DC converter. It is supplied from an 18V transformer, rectified and fed into a switching regulator to provide 13.8V DC.
My problem is that as I try to drive the circuit to 4A output, the input bridge diodes, feedback diodes and steering diodes all get extremely hot.
The circuit is based on an Allegro Semiconductor design with their 7A switch regulator.
Can anyone help with the reasons for this and some solutions too, if possible.

Re: Power Supply Problem

What diodes you use?

Re: Power Supply Problem

S5D surface mount diodes for most (5A).

The circuit is based on A8697 from Allegro Microsystems.

There is also a single schottky of 5A rating.

The circuit gets hot at 1.6A load (>60degC) and is supposed to go to 3A basic.

Power Supply Problem

Hali tmd,
I think, thats fully normal! ...
Refer pls to the data sheet; it has A thermal resistance J-A of 47Cels/W, but with coppers fo 16x16mmm!(note 1) Did you have these "build-in" heatshink pls? For me seems you missed it, or have to small, than you has ca 1.5W (at 3A) to dissipate, with 47degre/W>>makes ~60degre delta, also wit room temp, it can be very normal with a temp. of 70-80 Celsius...
Even, these is one time more example for that, SMD -or miniaturizing- isnt problemless, & often will forgeted to apply proper coolings copper areas_ clear I did assumed it only in your case as treason for "overheating", but I think Im not falsh with my assuming...
k.

Re: Power Supply Problem

I did think that might be a problem, so I changed the input bridge diodes for SOD device diodes (I cant remember the actual diodes but 1N5401 types). And these had the same temperature on them.

Re: Power Supply Problem

I don't know what you mean with "feedback diodes and steering diodes". The behaviour of a rectifier bridge build with standard
silicon rectifiers in SMC package should be clear. These diodes are capable of carrying 5A each, about the threefold of the intended
power supply capacity at 18V input voltage. But they need respective PCB area with copper pour to handle the dissipated power. If
you mounted the four bridge diodes densely packed, so that they have only their package surface as a heatsink, the current capacity
will be considerably lower. A possible solution is in using schottky diodes for the bridge rectifier as well. Due to the lower forward
voltage, you get lower losses. But schottky diodes are more sensitive to inrush currents and voltage transients.

Power Supply Problem

Frank,
he spoke over a switcher application too, maybe it has higher input "pulse" currents (obendrein) as measures/calculate!?
Basically is the cooling surly insufficiently realized...
@tmd,
I dont understand you, is it same package as 1N5401 or what pls?
The 5401 has RthJ-A of 10degre/W, & by 1.2V/3A>>ca 3.5W, also you will become a temp. delta of min +35 Celsius, but only than not more if the diode contacts are minimum 2X 9mm long! Refer pls to data sheet.
Is it so pls on your bread board? I can not assume it-sorry...
K.

Re: Power Supply Problem

Sorry for it being unclear.
Design was originally using S5D devices. They got very hot drawing 1.6A.
So I replaced the S5D diodes with 1N5401 diodes. These have legs bent to fit on the pads for S5D. Both diodes had approx 60degC rise on the same load.
The bridge feeds two large storage caps then through a diode then on to the A8697 Allegro Stepdown Switcher regulator, This has a foldback schottky diode MBT type which goes on to coil. The output then feeds a lead-acid battery trickle charger and output current sense. The battery also feeds into the output and three S5D diodes feed the currents to the correct points (diode into the charger, diode direct to output and diode from battery to output). All the diodes (except the charger diode and battery to output) get to 60degC.
I hope this clarifies