# LLC Resonant DC-DC Converter Feedback

Discussion in 'The Projects Forum' started by RedZeppelin, Jun 13, 2014.

1. ### RedZeppelin Thread Starter New Member

Jun 13, 2014
8
0
Hi everyone,

Here's a challenge for you, I'm designing an LLC resonant converter (12VDC output) and currently have a working PSIM simulation. However, my design seems to be prone to overshoot and instability at light loads. I have a feeling the problem lies in my feedback network because it's the section I understand the least. To be honest, without a reference voltage it seems like magic that it 'knows' to regulate at 12V.

The control IC I'm planning on using is the Fairchild FAN7631 (Datasheet).

I get the the basic operation of how it works. When your output voltage is too high, it increases the switching frequency (fs), lowering the gain etc. This is archived by using an optocoupler operating in the linear region to sink current. The more current it sinks, the faster the converter switches (page 11 of datasheet).

However, the part I think I'm having trouble with is the optocoupler's diode side of the feedback circuit.

The method I used was:
1. Find the maximum current to sink through the transistor side of the optocoupler (for max fs). I then use the CDR to refer it to the diode side (Id = 241.72uA).

2. Choose Vout = 11.5V as the value that the optocoupler will start conducting under (gain starts to lower).

3. If the threshold voltage of the optocoupler is 1.1V, Vz = 11.5 - 1.1 = 10.4V (where Vz is the voltage over the zener regulator).

4. Biasing resister is set so the transistor saturates at 12.5V (converter at minimum gain).

5. Voltage over biasing resister is Vr = 12.5 - 1.1 - 10.4 = 1V

6. Therefore, biasing resistance is Rr = Vr/Id = 1/241.72uA = 4.137kOhm

Anyone got any ideas? Let me know if I can make anything clearer for you.

• ###### feedback circuit.bmp
File size:
33.4 KB
Views:
72
Last edited: Jun 13, 2014
2. ### Dodgydave Distinguished Member

Jun 22, 2012
4,986
745
You need to use a KA431 zener diode to set the output voltage, diagram 12 on page 7 is the circuit you need, for an output voltage of 11.5V R1 is 3.6K, R2 is 1K(you need a R1/R2 ratio of 3.6)

3. ### RedZeppelin Thread Starter New Member

Jun 13, 2014
8
0
Thanks for your help Dodgydave, I'm a bit confused as to what I use the KA431 for? Do I use it with the optocoupler in the same way shown in the attachment of my first post?

edit: Also, just to be clearer, output is 12V.

4. ### Dodgydave Distinguished Member

Jun 22, 2012
4,986
745
so now its 12V ok, R1/R2 ratio is 3.8, you need the zener to switch on the opto coupler so it operates the feedback circuit, see the Fairchild datasheet page 16 you posted.

Last edited: Jun 14, 2014
5. ### RedZeppelin Thread Starter New Member

Jun 13, 2014
8
0
That might be the problem I'm having, I set the optocoupler to start conducting at 11.5V to avoid overshoot, so the gain is lower once it reaches 12V.

So if I set the zener to 12V and the output of the converter is >12V current flows through the optocoupler. But what happens when the converter is <12V? gain is max?

6. ### Dodgydave Distinguished Member

Jun 22, 2012
4,986
745
The ic increases and decreases the voltage to suit the zener strike voltage,its an isolated feedback regulator circuit, so you need the opto to switch pin 2 low to lower the output,then the zener shuts off the opto and the ic increase the output,thus maintaining a steady output.

7. ### RedZeppelin Thread Starter New Member

Jun 13, 2014
8
0
Thanks, that makes a lot of sense, and it explains why my original transformer turns ratio didn't work! (I changed it from 8:1 to 6:1 because the output was regulating too low, now i know why ).

I just ran a quick simulation, it looks pretty promising but it still has a pretty big transient and overshoot on start up with light loads. I'll have a closer look at what's happening and run some more simulations tomorrow and report back.

Also, was the method I used to find the resistor value to limit the current through the diode side of the optocoupler correct?

8. ### RedZeppelin Thread Starter New Member

Jun 13, 2014
8
0
I'm back, sorry some other matters took priority but this is my main focus again! Ah, the joys of dodgy resonant converter design...

I ran some basic simulations (attached) to test the output for different load and line conditions.

Simulation 1 which is of minimum input voltage and max load has low ripple but is regulating at about 10.8V. As far as I understand this condition is where the maximum gain of the converter is required? Is this result a sign that the maximum gain is not high enough? Or does my controller require integral action to add a dc bias?

Simulation 2 is the opposite to 1, I think the max input voltage and min load will require minimum gain? It reaches 12V like it should but the ripple is far too high. I'm thinking this is most likely still an issue with the controllers feedback? The controller need to be more aggressive?

Simulation 3 seems pretty spot on, regulating very close to 12V at max input voltage and load condition.

Simulation 4 I just included this for completeness, it seems to have all the problems of the above simulations put together, below 12V and high ripple!

Any idea of where I could go from here? I'm still not convinced that the resistor in series with the opto-diode and zener works how I think it does.

File size:
9.4 KB
Views:
35
File size:
10.2 KB
Views:
37
File size:
8.6 KB
Views:
26
File size:
10.6 KB
Views:
29
9. ### RedZeppelin Thread Starter New Member

Jun 13, 2014
8
0
Maybe I should try and make my question more specific, can anyone suggest a way that I can remove the DC offset error where the converter regulates below 12V?

I thought this might be fixed using integral control, but I can't see anything about that in the Fairchild FAN7631 datasheet?