Buck Regulator help!

Thread Starter

cgha20@yahoo.com

Joined Oct 21, 2009
82
So I am having some trouble coming up with a circuit that accepts an input voltage of +5V and achieving a stable output of 3.3V. The datasheet shows an example that accepts 8-12V input and achieves a 3.3V output, but not sure how to get this output with 5V input. Can you guys please help me please. I have to use Part number TPS40200QDRQ1. Please help!

http://www.ti.com/lit/ds/symlink/tps40200-q1.pdf
 

THE_RB

Joined Feb 11, 2008
5,438
5v to 3.3v can be done with a standard buck circuit, no problems at all.

It works slightly better using a buck that has synchronous rectification (which removes the buck diode Vf drop). All that does is increase the efficiency a small amount.
 

Thread Starter

cgha20@yahoo.com

Joined Oct 21, 2009
82
I am not sure what the circuit is supposed to look like to achieve this 3.3v out from 5 v. I am a newbie and need some help on coming up with the circuit. or how do i even tweak the sample circuits within the datasheet to achieve this
 

ronv

Joined Nov 12, 2008
3,770
There are IC's that are much easier to use than this one, but if you make it just like Fig. 35 it should work.
 

Thread Starter

cgha20@yahoo.com

Joined Oct 21, 2009
82
Yea originally I used LT3470a, but the man above wanted to use this part. I am not familiar with it at all. What can I strip down from this example on figure 35. Can you help explain the few sections on there. For example what is the purpose of section near q1, The section near J1 and J2. Thanks
 

ian field

Joined Oct 27, 2012
6,536
There are IC's that are much easier to use than this one, but if you make it just like Fig. 35 it should work.
The MC34063 is a great example, its an old chip so nothing if not a mature technology. The great thing is, there's a couple of online calculators that you just "plug in" the known values and it gives you the missing ones along with a suggested circuit.
 

ronv

Joined Nov 12, 2008
3,770
The 'stuff" to the left of R5,C5 is there if you want to sync it to some other clock in the system and can be removed.
C3, R1, C2 & c1 are decoupling and noise filters. You could probably loose R1 & C2 if noise is not a factor.
 

ronv

Joined Nov 12, 2008
3,770
You might need a logic level FET to run it from 5 volts. Other than that it can have a duty cycle of 90% so 3.3 from 5 should be easy. I think the FET might have been why they made it 8 volts in the example. It's fussy. Beg to use an easier part.
 

THE_RB

Joined Feb 11, 2008
5,438
There are dedicated 5v to 3.3v buck SMPS chips that just need an external inductor and diode. Some don't even need the diode.

Why fuss with a chip that needs an external FET? That's just nuts.

What current do you need at 3.3v? You might be best served with any cheap 3.3v linear regulator. Efficiency won't be that much different.
 

Thread Starter

cgha20@yahoo.com

Joined Oct 21, 2009
82
Well let me say I share everyone's frustration. I myself do not want to use this part, but it works faithfully at 185 degrees Celsius which is a requirement of this application (nevermind what datasheet about temp., i assure you it works). With this temp restriction, I only have limited choices. I previously was using the LT3470a, and it was working very well (and easy circuit), but this part was not easy due to solderability issues due to pad below IC. Brownout and Ian field, i totally understand your frustration!

Ronv I really appreciate your suggestions per this application. so far you have been spot-on. FDC654P appears to already be a logic level part. For my application I was thinking of using FDC5614P (https://www.fairchildsemi.com/ds/FD/FDC5614P.pdf). Do you think this will work? Lastly, do I need to change the zener diodes to get this to work from 5V in to achieve a 3.3v out or do you think they will work as is?
 
Top