DC DC Converter : LM73605 Synchronous Step-Down DC DC

Thread Starter

Mahantesh Teli

Joined Apr 2, 2016
21
Hello,

Can you please me to understand on the below queries;

While designing DC DC converter using LM3605, as Data sheet says Vout max can support upto 34V and o/p current 5A from Vin 36V ; how to judge whether LM3605 can support o/p : 24V@5A. from Vin: 36V ??

Rgds/
Mahantesh
 
Last edited:

ebp

Joined Feb 8, 2018
2,332
That voltage relationship requires a high duty cycle (34/36 = 94.4%), ignoring losses. You must evaluate the maximum duty cycle, the minimum OFF time for the switch and any switch-rectifier drive non-overlap time (when a conventional diode acts as the rectifier until the synchronous rectifier FET turns on; this is part of the switch OFF time, but if the non-overlap time is not very short the rectifier FET may only be on for a very short time, which reduces efficiency) very carefully, along with all of the losses (efficiency). It may work, but it is extremely close. You may have to use a relatively low switching frequency. Unless you need the absolute maximum efficiency, at 34 volts output and with such high duty cycle, there is probably almost zero benefit in using synchronous rectification. Sync rectification is highly advantageous at lower duty cycles and/or low output voltage. In your circuit, the rectifier will only be conducting for 5.6% of the time (again ignoring overall losses that influence duty cycle), so the loss due to the rectifier is very small. Think about how you would calculate the power lost in the diode (5A, conducting for 5.6% of the time, diode forward voltage compared with output voltage).

If it always operates at those voltages a linear regulator might be worth considering. Efficiency of a linear will be the same as the maximum duty cycle for the switcher - 94.4%, ignoring some small losses for the control circuit of the regulator. At 5 A & 2 V drop, the total loss is 10 W for 180 W input. Again, you would have to evaluate the efficiency of the switcher very carefully to decide if it has any merit for your application. The linear regulator is much simpler. Care is necessary to design a linear regulator that will work at 5 A with only 2 volts of input-output differential (usually called "drop-out voltage").
 

Thread Starter

Mahantesh Teli

Joined Apr 2, 2016
21
That voltage relationship requires a high duty cycle (34/36 = 94.4%), ignoring losses. You must evaluate the maximum duty cycle, the minimum OFF time for the switch and any switch-rectifier drive non-overlap time (when a conventional diode acts as the rectifier until the synchronous rectifier FET turns on; this is part of the switch OFF time, but if the non-overlap time is not very short the rectifier FET may only be on for a very short time, which reduces efficiency) very carefully, along with all of the losses (efficiency). It may work, but it is extremely close. You may have to use a relatively low switching frequency. Unless you need the absolute maximum efficiency, at 34 volts output and with such high duty cycle, there is probably almost zero benefit in using synchronous rectification. Sync rectification is highly advantageous at lower duty cycles and/or low output voltage. In your circuit, the rectifier will only be conducting for 5.6% of the time (again ignoring overall losses that influence duty cycle), so the loss due to the rectifier is very small. Think about how you would calculate the power lost in the diode (5A, conducting for 5.6% of the time, diode forward voltage compared with output voltage).

If it always operates at those voltages a linear regulator might be worth considering. Efficiency of a linear will be the same as the maximum duty cycle for the switcher - 94.4%, ignoring some small losses for the control circuit of the regulator. At 5 A & 2 V drop, the total loss is 10 W for 180 W input. Again, you would have to evaluate the efficiency of the switcher very carefully to decide if it has any merit for your application. The linear regulator is much simpler. Care is necessary to design a linear regulator that will work at 5 A with only 2 volts of input-output differential (usually called "drop-out voltage").
HI Ebp,
I was just curious about the how it works with high Vout which is nearly close to Vin, thanks for your your explanation . but real scenario i want understand does this device support 24V @5A. ? if YES - How ? and If NO - How ?
 

ebp

Joined Feb 8, 2018
2,332
At 24 volts and 5 amps output it should work well with no issues except for all of usual issues with switch mode power supplies. At 24 volts output the efficiency improvement with synchronous rectification won't be very much compared with using a Schottky diode.

If you don't already have LTSpice, I recommend getting it (free) and using it to simulate some circuits. Other people at AAC have used it much more than I have and will be able to offer some help using it, if you need it. It takes some time to learn to use simulation tools, but they can really help you understand what happens in circuits.
 
Top