UCC25660: 1 Hz oscillation at the output

Yes, it is clearly trying to re-start every 1s. Is your output negative? There isn't much we can do without the schematic. 

I'd also suggest to zoom in the waveform at start up and capture more than one node at once to see what went wrong.

Try capture HS, HO, LO, ISENSE at the same time.

I have added the schematic. Presently, the LLC converter is powered from a 400V power supply. The 12V rail for the opto-coupler is provided by a DC-DC converter which is also powered by the 48V rail. And no, the output is positive, I probably reversed the probes. 

LLC SMPS.pdf

I probed the OVP/OTP pin. It's not constant and has a pulsing signal with 1Hz frequency. I'll look at this circuitry.

Hi Jonathan,

could you let us know at what load condition you are facing issue.

Regards,

Yamini

There's no load connected as of now. 

Hi Jonathan,

From the ISNS waveform that you have shared, it looks like there is huge noise. Could you increase the noise coupling capacitor at the ISNS pin (C44) from 560pF to 1nF.

And also make sure that R30, C43 and C44 should be as close as possible to ISNS pin.

Also please share the following items,

• Design calculator excel sheet with your design specifications
• Waveforms LO, ISNS, Vout  in one scope (clearly zoomed in form and zoomed out form) when fault is occurring.

Regards,

Yamini.

Hi Yamini,

Here's an image of the layout. C43 is close to the resonant capacitor, C44 and R30 are close to the IC. 

The following are the waveform (Yellow - across C44, green - LO, blue - Vout). ISNS and LO were measured using non-isolated 10x probes. Vout was measured using a differential probe. This is measured after changing C44 to 1nF.  

Here's the excel sheet.

 UCC25660x_CALC_1.0.xlsx

Hi Jonathan, we are looking into your schematic and design calculator. We will get back to you .

Meanwhile , please try increasing C44 to 1nF.

Regards,

Yamini.

Hi Yamini,

The scope captures were done using C44 as 1nF. However, I noticed that the converter stops switching when the VDC is beyond 360V. It was switching at the rated voltage of 400V when C44 was 560pF. In both the cases, it always switches with the 1 second fault idle time. If I increase the capacitor further, it doesn't start.

Jonathan, 

I don't see any relation that noise coupling capacitor change across the ISNS pin that affecting the input voltage at this moment. There might be something else causing issue in the circuit.

Can you please, provide the waveforms of LO, ISNS, resonant current and switch node voltage.

Regards,

Yamini.

Hi Yamini,

I am short of one probe, I have attached waveforms of switch node, resonant current (using clamp meter) and ISNS. The below image has all three together:

I removed the single ended probe from the ISNS pin and captured the waveforms again.

The current seems to be affected by the probe, I suspect it to be the capacitance. 

Also let me know if there are mistakes in the schematic/calculator.

Hi Yamini,

Can we continue the conversation over email to move things a little faster?

Hi Jonathan,

Thank you for your patience.

Can you please check and tell the resonant current peak and ISNS peak, from your waveforms as I cannot see the values. 

I am suspecting that since your Ln value is very low, in no load. high magnetizing current is flowing, which causing your ISNS to go beyond 3.5V causing OCP fault.

In your design calculator, you operating frequency is 400kHz, but with your chosen magnetics min. switching frequency is 472.6kHz.

Also, can you please try reducing the ISNS to 100ohmn and test.

Regards,

Yamini.

Hi Yamini,

I have tried a few things. You are right about the ISNS voltage going beyond 3.5V. The inductances are low because I am using planar magnetics. I increased the TSET option to number 15 from 12 which increased the switching frequency at no load to ~500kHz from ~400kHz. This reduced the peak to peak ripple at the output, however this still didn't result in the set output voltage of 48V. It seems to be hovering around 20V.

How do I set the TSET option? Does it come from the inductance value? Also, why is the converter switching at 500kHz? I was expecting it to switch at higher frequencies as it's operating on no load. 

Hi Jonathan,

Converter is operating at 500kHz because of chosen option 15.

From the above table, if you choose option 15, converter operates with IPPC at around 500kHz. That's why you are getting around 500kHz. For Option 12, it is 304kHz, you will get a frequency higher than this. Since your minimum operating frequency is 472.6kHz you can not choose option 15.

With our new version of LLC controller, UCC25661x series choosing TSET option is easier than this. We decoupled over current and overload protection thresholds, by changing TSET options.

We recommend you to use UCC25661x series.

Thanks & regards,

Yamini.

Hi Yamini,

So I should choose a TSET option that is lower than the operating frequency but above the resonant frequency so as to avoid capacitive region? I understand that ucc25661x is a drop in replacement, however I need to get through this design using ucc25660x.

Hi Jonathan,

choose a TSET option that is lower than minimum operating frequency. And please note that while choosing TSET , check your Vfb_replica Vs Pin graph.

you need to choose a TSET option for which V_fb replica should be less than 4.75V  for rated power other wise OLP(Over load protection) fault may occur at or below your rated power.

Regards,

Yamini.

Got it. I am trying to resolve the noise that's present on the ISNS pin. Can you confirm whether the schematic is fine? 

Hi Jonathan,

Will get back to you.

Regards,

Yamini.

Hi Jonathan,

Could you provide the part number of the shunt regulator that your using, it is not mentioned in your schematic.

Regards,

Yamini.

I am using this part : TLVH431ACDBZR. By the way, the 10k resistor was changed to 5k as the selected part had a reference voltage of 1.25V. I checked the working of this circuit by providing 48V using an external supply and measuring voltage across pins 1 and 2 of the optocoupler. 

Hi Jonathan,

Thankyou for the confirmation. Are there any other changes done ?

That's the only change.

Hi Jonathan,

Can you try reducing RISNS by 20-30% of the recommended value  and test.

Regards,

Yamini.