Will this design work well?

[king.oslo]

Hello there,

I am making a switch mode power supply. I am finding it challenging. It has a LM25116 buck controlled by PWM from a PIC18F4550.

I understand from the datasheet of the LM25116, that a voltage higher or lower than 1.215V on the FB pin alters the the duty cycle of the buck either up or down. As a result Vout is altered.

I thought that the best way control this by MCU was to use an op-amp with Unity Gain Differential Amplifier configuration where I input 1.5-4.2 voltage to correspond to Vout of 1.5-4.2. The inverting input has to be 1.215 lower than the non-inverting: f(Vout) = .199 * Vout.

I just wonder if my solution is good, or if it can be improved. I find at least one problem. But I am not sure how to solve it. I mostly need help with the feedback circuit inside the blue ellipse at the moment. The rest of the circuit is nowhere near completion.

Any critique is much appreciated. When I listed to Dave talk about his design, I am often surprised about things that I did not know.

Thank you very much for your time! 

Kind regards,
Marius

PS: This is the LM25116 datasheet. On page 13, the Error Amplifier is discussed. http://www.ti.com/general/docs/lit/getliterature.tsp?literatureNumber=snvs509c&fileType=pdf

[amspire]

Any reason for using that regulator?

The reason I am asking is that it gives you no access to the internal 1.25V reference. If the micro could use the same reference as the power controller, then it could use the PWM to do purely relative adjustments.  I suppose you are assuming that you can correct for errors between the two references via the ADC input, but this adds another control loop and it can affect the stability.
It could work, but I suspect you may be able to find a friendlier regulator chip.

It looks like that regulator needs external compensation, so it could be a fair bit of work getting it stable over the the whole voltage range.

100 ohms resistors in the amp circuit is way to low. 10K would be better.

Richard.

[king.oslo]

Thanks for that.

Which regulator do you recommend?

If I increase from 100R to 10K, the voltage divider Vout will be altered. Correct? How can I ovcome this?

Thank you for your time

M