Question about stepping down voltage using LM2596

[Jay112]

I'm trying to power a 3.6v wireless module with my golf cart batteries. The cart has 6 x 6-volt batteries. I'm planning on using a step-down converter based on the LM2596, like this one: http://www.ebay.com/itm/281901421649

Is it more efficient to feed 6v into the LM2596 module when converting to 3.6v, or would it be better to feed 12v into the module?

Thanks!

[xyrtek]

check the datasheet page 6 for the info you want.

http://www.ti.com/lit/ds/symlink/lm2596.pdf

EDIT: In case you can't access the datasheet it seems that for 3.3V at 3A load the efficiency is almost 75% between 20 and 25 (input voltage)

[Jay112]

check the datasheet page 6 for the info you want.

http://www.ti.com/lit/ds/symlink/lm2596.pdf

EDIT: In case you can't access the datasheet it seems that for 3.3V at 3A load the efficiency is almost 75% between 20 and 25 (input voltage)

Thanks! I should've thought to check the datasheet, but also I didn't know if the other components in the module would affect the efficiency much.

So it looks like from that graph on page 6 that using 12v would be about 5-6% more efficient than 6v when there's a 3.3v 3A load? I wonder if that's significant at all, especially when my device will probably be using less than 50mA. Also I wonder if the efficiency changes a lot when the load is much lighter than 3A.

It will be easier for me to just use 12v as the input, since I already have some 12v wires nearby. Maybe I'll just do that, unless others here have a different recommendation.

[xyrtek]

check other parameters that can affect the overall circuit performance, but on a quick look it seems that there are no advantages to use 6V over 12V.

If nothing in the circuit will be affected by the 12V operation then you should be all set.

Anyways check all your requirements.

ex:
For battery operated the Quiescent Current could be important for battery life.

Outdoor operation can affect the temp ranges.

You get the idea.

[newbrain]

check the datasheet page 6 for the info you want.

http://www.ti.com/lit/ds/symlink/lm2596.pdf

EDIT: In case you can't access the datasheet it seems that for 3.3V at 3A load the efficiency is almost 75% between 20 and 25 (input voltage)

Or, since the modules come from eBay, it would be better to check this datasheet

It's quite common to have fake 2596 (just rebadged 2576, when you are lucky).

The modules will work, but, as the 2576 has a lower working frequency, there will be problems at higher currents and more ripple with the provided inductor.

[Jay112]

Or, since the modules come from eBay, it would be better to check this datasheet

It's quite common to have fake 2596 (just rebadged 2576, when you are lucky).

The modules will work, but, as the 2576 has a lower working frequency, there will be problems at higher currents and more ripple with the provided inductor.

That's interesting. I had never heard that before. So the fake 2596's will look exactly the same as the 2576's, even the lettering?

[newbrain]

That's interesting. I had never heard that before. So the fake 2596's will look exactly the same as the 2576's, even the lettering?

I'd say except for the lettering, that will read 2596 
See e.g. here (the comments from Dan and Dave).

I took the attached image with a 1.75A load: 54kHz, about 1V of ripple and a hot chip (input 19V, output 5V).
Yellow track: output (AC coupled), cyan: switch output; spikes are due most probably to the jumble of wires.
No way to get it on 150kHz under any load/Vin/Vout combination.

And it didn't even come -directly- from eBay or Aliexpress, but from a well known Swedish distributor of small and home electronic: they took it back with no problems and "started an investigation" but the item is still on sale.

Re-labelling inferior chips is quite a common practice, plenty of examples on the net and on this forum.

Edit: added where to look in the linked article.

[Jay112]

That's interesting. I had never heard that before. So the fake 2596's will look exactly the same as the 2576's, even the lettering?

I'd say except for the lettering, that will read 2596 
See e.g. here (the comments from Dan and Dave).

I took the attached image with a 1.75A load: 54kHz, about 1V of ripple and a hot chip (input 19V, output 5V).
Yellow track: output (AC coupled), cyan: switch output; spikes are due most probably to the jumble of wires.
No way to get it on 150kHz under any load/Vin/Vout combination.

And it didn't even come -directly- from eBay or Aliexpress, but from a well known Swedish distributor of small and home electronic: they took it back with no problems and "started an investigation" but the item is still on sale.

Re-labelling inferior chips is quite a common practice, plenty of examples on the net and on this forum.

Edit: added where to look in the linked article.

This is all very interesting. Thanks for the info.

Do you guys think these knockoff power regulators are significantly hazardous to the items they're powering? Also, where should one go to find the good components?