Power Supply using LM2576HV for 5V regulation from 48V input.

Thread Starter

Saiteja chinthalapati

Joined Oct 25, 2018
97
I am using LM2576HV for 5V regulation from 48V input. out put pin is connected to SS36 diode after running for some time the diode getting shorted and regulator is heating up... what might be the issue any idea
 

ScottWang

Joined Aug 23, 2012
7,003
From 48V converting to 5V is waste too much power on LM2576HV and the power dissipation will turn to heat.
You can buy a 9V/2A or 12V/2A adapter or in series with a 40Ω/100W power resistor between 48V and LM2576HV, let all the heat focus on the resistor to protects the LM2576HV.
 

dendad

Joined Feb 20, 2016
3,810
What is the inductor you are using? The wrong inductor can easily damage the circuit. It must be the correct core material for the frequency, inductance and power handling for the application.the classic "Fred's inductor" just does not work. If you don't know what that is, Google it, along with "Jim Williams" ;)
And yes, please post the schematic.
 

MisterBill2

Joined Jan 23, 2018
8,405
There are many circuits available on several semiconductor manufacturers websites, and most of the circuits are very well documented. The general category is "5 volts from telecom power", because 48 volts is a standard voltage in many big phone systems. AND, the propeties of the inductor in a switchmode power supply are very important.
 

ebp

Joined Feb 8, 2018
2,332
That diode should work quite well for the circuit. It is well below its rated current and shouldn't overheat even with the smallest mounting pads at a current of up to 1 ampere. It is always best to try to use reasonably large copper areas for heat dissipation. In a buck converter, the anode of the diode is at "ground" so that pad can be make large without concerns for problems due to capacitance or noise radiation. The cathode is the "switching node" of the supply and often it is best not to make that area too large.

There may be some possibility that it is being subjected to excessive voltage due to ringing in the circuit. This is usually due to a circuit layout that has inductance where you don't want want it. If the input supply is not regulated, perhaps its voltage is higher than nominal at times, bringing the diode very close to its rated peak reverse voltage.
 

Thread Starter

Saiteja chinthalapati

Joined Oct 25, 2018
97
Thank you for all your inputs.... i didn't get exact solution but the same circuit if i am regulating from 12v to 5v it is working fine but components i am using has to sustain for 60V
 

Thread Starter

Saiteja chinthalapati

Joined Oct 25, 2018
97
What is the inductor you are using? The wrong inductor can easily damage the circuit. It must be the correct core material for the frequency, inductance and power handling for the application.the classic "Fred's inductor" just does not work. If you don't know what that is, Google it, along with "Jim Williams" ;)
And yes, please post the schematic.
100uh
 

Thread Starter

Saiteja chinthalapati

Joined Oct 25, 2018
97
What is the inductor you are using? The wrong inductor can easily damage the circuit. It must be the correct core material for the frequency, inductance and power handling for the application.the classic "Fred's inductor" just does not work. If you don't know what that is, Google it, along with "Jim Williams" ;)
And yes, please post the schematic.
Find the attachment for circuit
 

Attachments

ArakelTheDragon

Joined Nov 18, 2016
1,353
I am using LM2576HV for 5V regulation from 48V input. out put pin is connected to SS36 diode after running for some time the diode getting shorted and regulator is heating up... what might be the issue any idea
Your diode is burning I guess. Try to use "1n4007 and 1n4148". One of them should work.

Also please post the circuit so we can check the connections and other components. I think the coil or capacitors are not placed properly.
 

ebp

Joined Feb 8, 2018
2,332
Your diode is burning I guess. Try to use "1n4007 and 1n4148". One of them should work.
No, do not do that. Neither of those diodes is even remotely suitable as the freewheeling diode. The 1N4007 is grossly too slow and the 1N4148, while extremely fast, is much too small.

If you do have input voltage up to 60, it would be worth trying a diode rated for higher voltage. Unfortunately, 60 volts is a common upper limit for Schottky diodes. There are certainly 80 V and 100 V parts available, but they are considerably less common. The also usually are slightly higher forward voltage. An ultrafast diode could be used, but a Schottky diode is better.

With 48 V input and 5 V output, the diode will be conducting almost 90% of the time, whereas with 12 V in and 5 V out it would be conducting less than 60% of the time. For the same output current, this would mean it would run hotter with the higher input voltage. Still, I don't think diode temperature is likely to be a problem unless the operating environment is very hot.
 

ArakelTheDragon

Joined Nov 18, 2016
1,353
Do you have the same components?
No, do not do that. Neither of those diodes is even remotely suitable as the freewheeling diode. The 1N4007 is grossly too slow and the 1N4148, while extremely fast, is much too small.

If you do have input voltage up to 60, it would be worth trying a diode rated for higher voltage. Unfortunately, 60 volts is a common upper limit for Schottky diodes. There are certainly 80 V and 100 V parts available, but they are considerably less common. The also usually are slightly higher forward voltage. An ultrafast diode could be used, but a Schottky diode is better.

With 48 V input and 5 V output, the diode will be conducting almost 90% of the time, whereas with 12 V in and 5 V out it would be conducting less than 60% of the time. For the same output current, this would mean it would run hotter with the higher input voltage. Still, I don't think diode temperature is likely to be a problem unless the operating environment is very hot.
The diode is given in the circuit.
 
Top