Hi, I want to use THIS switching regulator in adjustable output voltage operation, using 2 external resistors as shown below in the first picture. The 2nd picture shows an optional filter I can add, which I would like to. So the question is does the filter go before or after the feedback resistors? I could not find anything in the datasheet. My guess is the filter goes before the resistors because then the feedback is a little more accurate.

Please correct me if I am wrong about that.

(It may not be a huge deal because the resistance of the inductor is small but I imagine it does have an effect)

Thanks

 

My guess is the filter goes before the resistors because then the feedback is a little more accurate.

That certainly is a valid reason to put the filter before the resistors, but unfortunately the added phase shift from the second filter will likely cause instability or oscillation of the circuit.
That's the reason the connected it as shown.
Unless you are willing and able to add an appropriate correction network in the feedback loop, you should keep the feedback connection in front of the added filter.
Just pick an inductor with a low enough resistance that the output load regulation is not adversely affected.

 

Hi, I want to use THIS switching regulator in adjustable output voltage operation, using 2 external resistors as shown below in the first picture. The 2nd picture shows an optional filter I can add, which I would like to. So the question is does the filter go before or after the feedback resistors? I could not find anything in the datasheet. My guess is the filter goes before the resistors because then the feedback is a little more accurate.

Please correct me if I am wrong about that.

(It may not be a huge deal because the resistance of the inductor is small but I imagine it does have an effect)

Thanks
View attachment 180283


View attachment 180284

Hi,

It is often called a "post filter" because it is after everything else.

It amazing how much difference this kind of filter can make. Even 2uH makes a big difference at typical switching speeds.

 

That certainly is a valid reason to put the filter before the resistors, but unfortunately the added phase shift from the second filter will likely cause instability or oscillation of the circuit.
That's the reason the connected it as shown.
Unless you are willing and able to add an appropriate correction network in the feedback loop, you should keep the feedback connection in front of the added filter.
Just pick an inductor with a low enough resistance that the output load regulation is not adversely affected.

So ignoring the values in my picture, this is how the adjustable configuration would be set up right?

 

So ignoring the values in my picture, this is how the adjustable configuration would be set up right?

Yes.
But I will have to mention that if the two inductors are the same value as you show, then there will still be some additional adverse phase shift at the feedback network.
That's why, in the example circuit, the second inductor had less than a 10% value of the first.

 

Yes, but build the circuit as shown in the datasheet then you know it works, and then modify it.

 

So ignoring the values in my picture, this is how the adjustable configuration would be set up right?
View attachment 180405

L1 looks too small for the chip you had selecred previously. 50uH min 100uH better.
L1=10uH is more for 500kHz converters.

 

See

 

See
View attachment 180415

Hi,

Sorry to say but there is a lot more to testing a power supply than just the constant resistive load test at one input voltage.
There is load step change, input line change, stuff like that, that needs to be tested to ensure stability. There is much much written on this subject and for good reason.
I especially note that 0.1uf cap. Offhand that's rather large for a feedback capacitor which could affect transient response greatly.

Do some more tests see what you can turn up.

 

Thank you for all the feedback. Yes, as I said, ignoring the values in the picture, I want to be sure that makes sense.

How do I simulate a switching regulator in OrCAD Capture?

 

See
View attachment 180415

How did you simulate with the LM2596_ADJ? Is this part built into PSPICE?

If I want to simulate an IC that may not be built in, how can I do that?

 

I simulated the scheme in LTspiceIV. To simulate in LTspiceXVII it is necessary to correct the model. But I didn't like how the model works. You can take the models from the site of the chip manufacturer and attach them to the pspice.