LM3596 switching regulator modules

[mrcrud5]

Recently got a bunch of these LM3596 switching regulator modules for super cheap on ebay. Did a little blog thing on my website so check it out if you are thinking of buying these. I didn't go super in depth with the testing. Only tested and compared it using one load and voltage value. But it was enough to get a rough idea of their performance. Sorry if this topic has already been covered to death on here.


https://www.mrelectronics.net/single-post/2017/04/19/LM2596-switching-regulator-board

[wraper]

There is one issue with those modules. LM2596 is fake. And if switching frequency different from 150 kHz, which usually is a case, that would be a dead giveaway about it.

[mariush]

Like others say... it's supposed to be LM2596 (the 150kHz or something like that) version, but the chip is not LM2596. If you're lucky, the chip is a LM2576 running at 50-60kHz made by TI or some Chinese factory. They just erase what's on them and print LM2596 because that chip was more expensive.
Some use the XLSEMI XL6009 IC which is probably much cheaper for the chinese buyers:



Anyway... check the datasheet for LM2576 and then see what ranges (voltage and current) the inductor and diodes they use on the circuit boards are most suited for. You'll probably see that what they put is not really suited for a wide output voltage and current range, and that you may get very poor efficiency (in the 70-75% range) at low currents... almost making it cheaper to use LDOs (linear regulators with low dropout voltage)

[mrcrud5]

There is one issue with those modules. LM2596 is fake. And if switching frequency different from 150 kHz, which usually is a case, that would be a dead giveaway about it.

Yes! I found it strange that the data sheet said the switching frequency is 150KHz but I was getting 50K(see attached). I am not building this into any products or anything though so for my use it should be fine. Would you know if these fake ones have a history or failing or anything?

[james_s]

Anyway... check the datasheet for LM2576 and then see what ranges (voltage and current) the inductor and diodes they use on the circuit boards are most suited for. You'll probably see that what they put is not really suited for a wide output voltage and current range, and that you may get very poor efficiency (in the 70-75% range) at low currents... almost making it cheaper to use LDOs (linear regulators with low dropout voltage)

It may be cheaper, but you'll have trouble using a linear regulator to make a boost converter.

[mariush]

It depends. The circuit takes so much space in a product that often it may be cheaper to just use more batteries or a different battery and use a ldo if needed.

[Sredni]

Each and every one of the LM2596 modules I've bought from vendors on the Amazon marketplace were using fake LM2596 chips, running at 51-52 kHz. See scopeshots (BTW, I believe the little bump before the spike to be an artifact due to too many averages - it did not show up in other acquisition modes).

I hope some day someone like Ken Shiriff will decap one of those to see what's inside (LM2576 or much worse?)...

[james_s]

I wish they would just print the real part number on the devices, if they're knockoffs, sell them under the Chinese brand and whatever they actually are. I'd still buy them if I needed a cheap regulator and I wouldn't feel cheated the way I do when I get something fake.

[rrinker]

 Only problem I've had is the batch of them that I bought had the input and output caps arranged horizontally instead of vertically, and on 1 out of 5, the output cap was completely broken off the board (cut the static bag open, out come 2 pieces - the board, and the cap) and on a DIFFERENT 2 out of the 5, one leg of the input cap wasn't soldered on. I believe that was because the pad is RIGHT at the edge where the boards were snapped off their panels. The edges were not straight AND the holes for the input and output connections, the copper was visible at the edge.