# Boost Converter

Discussion in 'Homework Help' started by saturnsakini, Jul 10, 2015.

1. ### saturnsakini Thread Starter New Member

Jul 10, 2015
8
0
Hi i am trying to design a boost converter.

Vinput=1-3.8V
Iinput max = 12.5A
i choosed to duty cycle 0.75.
and 10kHZ

i think like that; max Power input; 3.8V * 12.5A = 47.5, i assume that worst case, my converter be lossless and transmit to power out side. Poutput=Pinput

Pout= 47.5Watt, that means my Output current is Poutput/Voutput = 11.878 A
For this condution my output resistance is V=I*R 4=11.878*R R=0.33676ohm
also i want to max current ripple on inductor 0.5 A for this i choose the indctor
L= [Vinput*(Voutput-Vinput)]/ inductor ripple current*freuency*Vout
L= 0.00015

and choosed very high capacity for eleminate output voltage ripple like 100F.

but in simulation this is not work. i took a 0.17 V output. 9A input current ?

2. ### DickCappels Moderator

Aug 21, 2008
5,752
1,801
Even simulators can suffer from real-world problems, like nonzero saturation voltage in semiconductors.

Please post the schematic of your simulation in a common image format. Maybe that will get you some more specific answers.

3. ### Roderick Young Member

Feb 22, 2015
408
169
A schematic would help. If you meant 13.8 volts input, then the output should be at least 13.8 volts, even if the converter is not "boosting." Try changing that 100 F capacitor to 100 μF instead, and see if the results are different. It could be that your circuit is just taking a long time to charge the 100 F capacitor. Roughly speaking, if you can put 10 amps into a 100 Farad capacitor, the voltage will rise only 0.1 volt per second.

Feb 17, 2009
4,923
1,381
5. ### saturnsakini Thread Starter New Member

Jul 10, 2015
8
0
thank a lot. especially web site was helpfull for me.

6. ### saturnsakini Thread Starter New Member

Jul 10, 2015
8
0
Hi
i managed to chosing component but my simulation don't work exactly. i did it with same values for proteus but took same terrible result? why it can be?

Vinput= 1V and 12.5A input
for 4V and 10.56 A output?

but i take 132 mV and -354 mA?

L=786.3nH
C=100uF
Routput = 4V/10.56 = 0.378

File size:
1.1 KB
Views:
8
7. ### Alec_t Expert

Sep 17, 2013
10,267
2,511
You need to select a specific MOS device for the sim. The generic one does nothing useful.
Using a supply of only 1V is going to be problematic. Can you find a FET which will turn on fully at a hoped-for 12.5A current, with a Vgs of only 1V?
You show a gate drive voltage (0-12V) which is suited to a N-MOS device, not a P-MOS device.
Think about how much voltage D1 will drop and select a suitable diode accordingly.
What real-world voltage source do you envisage that will provide 12.5A at 1V?
Stop and think about power in and power out!

8. ### saturnsakini Thread Starter New Member

Jul 10, 2015
8
0

Actually, my source is 62.5W and olso limited with max 12.5A output current. i trid to step up input voltage step by step. like series boundend boost converter. If Vgs increase over 4 V(like between 4V-16V), it jumped directly second stage with a control loop.

stage 1: 1V-4V input to step up 4V,
stage 2: 4V input to step up 16V
stage 3: 16V input to step up 64V
(first stage's output will be second stage's input)

let's turn my first goal. i am trying to design first stage now. i solved the previous problem with syncronous trigerring instead of diode. but still i cant see on my desired output voltage(4V) on output.
and current desired values input (12.5A) and output 10.56A are dosent exist. ????

File size:
1.3 KB
Views:
6
9. ### Alec_t Expert

Sep 17, 2013
10,267
2,511
I think a problem with that arrangement is current through the M1 body diode.

10. ### crutschow Expert

Mar 14, 2008
23,130
6,860
Edit: Here is the corrected circuit from my previous incarnation.

Below is the simulation of your circuit slightly modified.

I reversed M2 so that the drain is connected to the output and the body diode will be reverse biased when M2 is OFF and M1 is ON (MOSFETS conduct equally well in either direction when ON).
I changed the timing slightly so that there is no overlap that would cause the two transistors to be on at the same time, generating shoot-through currents.
I also increased the value of L1 to reduce the current ripple.

• ###### Draft1a.asc
File size:
1.5 KB
Views:
2
Last edited: Jul 15, 2015
11. ### crutschow Expert

Mar 14, 2008
23,130
6,860
Bump to note that I changed the circuit in my previous post.

12. ### saturnsakini Thread Starter New Member

Jul 10, 2015
8
0
thank you so much crutschow. i am working on your simulating. But i can't see where i did wrong. our values are so close to each other but why my simulate dosn't work as required?

also why we dosn't make output current about 10 amper when input is 12.5A? is my calculating wrong?

13. ### Alec_t Expert

Sep 17, 2013
10,267
2,511
Note in Crutschow's sim that the load resistor is now 1 Ohm.
Your calculations fail to take into account the duty cycle of the FETs and the rate of rise/fall of the inductor current.

14. ### crutschow Expert

Mar 14, 2008
23,130
6,860
Close is not always sufficient in circuit design.
The circuit won't work properly with the N-MOSFET oriented with the source to the output since its parasitic body-diode will conduct when it turns off, draining off some of the charge on the output capacitor.

If the average input current is 10A @ 1V (10W) then the maximum output, with 100% efficiency, would be 2.5A @ 4V (also 10W).