National LM2575 Buck Converter

lm2575

Hi friends.

I have made a 12V to 5V dc-dc converter using LM2575. As the datasheet says it is a 1A step down Buck regulator. In my circuit the load's current is about 980 mA and the efficiency of the regulator is about 80 %. Nonetheless the dissipated power emerging as heat in the regulator (which is mount on a suitable heatsink)
seems to be much and harmful to my circuit.
Now I want to know if there is or not any other device to use instead of LM2575-5 to decrease the produced heat ? or may I use any other topology for this purpose?

I will be really grateful to any help.
Yashar

lm2575 circuit

Newer regulators have higher efficiency, e. g. LM26xx family, that utilizes MOSFET switches instead of bipolar with LM25xx. With 12V input, you can further improve the efficiency by using a synchronous buck converter. See the rich choice of recent chips from National, TI, Linear and other manufacturers.

lm25xx

Hello again dear EDAboard friends,

Actually I consulted with other switching regulator designers and they told me that the heat is not too much and it seems to be natural. so I didn't do any changes and I used it the same way and fortunately it has been working without any problem since that time.

I'm planning on using a fixed 2575 5v output IC with a unregulated 24vdc input. Is it ok to lift the ground pin of the 2575 with a couple IN4007 diodes to get 6.3v? I've done this successfully with linear 7805 ICs.

Hi, I can understand the seriousness of your circuit and I would suggest you to use a LM2576 5V regulator in this case because the heat level of this is very low moreover your load current is not high for which the above will be most suitable

analognow: That's not generally advisable, but you can try if you'd like. The Feedback pin expects to see a DC voltage, but the voltage across the diodes would pulsate up and down with current surges through the chip... so if you want the output voltage represented correctly, you'd need to bypass the diodes with capacitors. Then the GND pin would see >1.2V, while the "output" switch pin is permitted to swing to -0.7V. I wouldn't be surprised if the chip is destroyed (probably latchup).

It's a better idea to use the right tool for the job: a switching regulator with an adjustable output voltage (this one, for example).

- - - Updated - - -

analognow: That's not generally advisable, but you can try if you'd like. The Feedback pin expects to see a DC voltage, but the voltage across the diodes would pulsate up and down with current surges through the chip... so if you want the output voltage represented correctly, you'd need to bypass the diodes with capacitors. Then the GND pin would see >1.2V, while the "output" switch pin is permitted to swing to -0.7V. I wouldn't be surprised if the chip is destroyed (probably latchup).

It's a better idea to use the right tool for the job: a switching regulator with an adjustable output voltage (this one, for example).