Connecting Multiple Buck Converters in Parallel

You could do that, Chief, and you'd find that only one of them would be supplying any current - the one with the lowest voltage setting. Sounds a bit daft to me but, hey, it's your bank balance.

Sounds like you have exceeded the maximum output design of the buck converter?

I would not parallel them as mentioned by Crabtree.

Look for a unit with larger output capabilities, note drooping battery voltage due to usage will cause converter to output maximum amps depending on differential untill voltages balance up again.

In some cases the poorly designed output voltage control circuitry in this type of application requires a minimum impedance on the load side in order to stabilize and control the IC gain. (Rfb/Rin)

Without load side impedance present the controller can/will go to infinity gain and the output voltage will go to maximum source value which can be very destructive.

If the machine worked well when attached to a load and failed when "open circuited" with no load/impedance, I would assume that the machine you were using suffered failure from this design flaw.

Parralleling multiple IC's will not correct the problem however, adding a resistor with a high enough ohmic value and proper wattage rating across the output terminals can prevent circuit overvoltage failures from happening.

Proper resistor sizing requires analyzing the circuitry and applying IC gain formulae (or SWAG experimentation if you are inclined to do so but I wouldn't recommend it.)

This type of circuit design flaw is becoming common in today's "git-er-done", lowest-cost production business environment.

This and many other circuit design flaws are manifesting their ugly presence in products from major manufacturers that used to avoid them at all cost.

The simple answer to your question is "NO".

As a minimum, you need high current blocking diodes on each converter output to prevent feedback.

It is also possible, depending upon the design, to also need a resistor across each output to draw some small load from each unit, when none is being made, e.g.open circuited as you mentioned......the value of the resistor needs to be calculated KNOWING the design in use.....but is also a waste of power.

I personally use switched mode designs or Step-Down Switching Regulator chips like the LT1074 for example....they can be run in parallel when correctly configured, are cheap and almost "bomb-proof". Plus they can be very, very efficient.....many other designs are not.

I use those chips (x2) in an Elektor design that can go from 0 amps to 10 amps at any voltage up to 40 volts.....are short circuit and over temperature proof, at around 85% efficiency!!!

i aint gonna generate another web user id
so if you already have an BA EBei whatever account i suggest you send a product quality feedback (perhaps you'll receive a free not 4 replacement) , see http://www.ti.com/lit/ds/symlink/lm2596.pdf for "protection" - it tells the incomplete ESD implemented "handle cautiously" - may be the seller "didn't notice"
also there might be better(higgher voltage) lytes and non paced in inductors that may "advance this particular design" + the chip doesnot show much cooler space there must be alot for 15W op (by my experience some 50x more than this)