WEBENCH® Tools/LM51551: Inquiry of abnormal output from FLYBACK CONVERTER using LM51551.

6663.8V TO 30V DC DC 15V_2A Converter.pdf

flyback_200805.pdf

Hi Kim JaeHyung,

Thank you for reaching out with your question. I am sorry to hear that you are having issues with the LM51551.

Can you please post a higher resolution schematic? I am having trouble reading the component values. 

Thanks,

Garrett

Dear Garrett,

Thank you very much for your interest.

There are drawings and waveform data in the attached file.

Thanks,

JaeHyung Kim,

4682.ISO_SMPS.zip

Hi JaeHyung Kim,

Thank you for providing the waveforms and the higher resolution schematics. 

Based on the waveforms that are being provided it looks like the LM51551 is entering hiccup mode current limiting. The device will enter this when the peak current limit is exceeded for more than 64 cycles.

A few suggestions are

1. Solder the transformer to the board. In the provided presentation the transformer is soldered through a number of wires to the PCB. This can cause a number of issues in the circuit and can possibly cause the peak current limit to trigger.

2. Increase the value of R14 to 50kOhm. This will increase the time of the secondary soft-start and should help keep the device from trigger hiccup mode operation.

Please let me know if you have any questions.

Thanks,

Garrett

 Hi, Garrett

We will send you back the measurement waveform.

Please review.

Thanks,

Garrett

30W_15V Flyback waveform_200814.pptx

Hi Kim JaeHyung,

It looks like the device is running into hiccup mode current protection. This is why it looks like the device is turning on and off.

Try increasing the value of R14. I would try changing the value to 100 kOhm to see if that helps.

Thanks,

Garrett

Hi, Garrett Roecker,

30W_15V Flyback waveform_200901.pptx

As you told me, I experimented with changing R14 to 100kohm, but the hiccup phenomenon continues to occur.

I suspected CS PIN, removed the shunt resistance of MOSFET and shorted it to test, but the results are the same.

When we simulated whether the transformer's inductance value is appropriate, a stable output was obtained at 1uH.

Request a technical review on this.

Best Regards,

jaehyung Kim,

Hello Kim,

Sorry to hear that you have trouble. Could you please add which signals you are measuring in the attached presentation. I was not able to identify them.

Jaehyung,

Can you please provide the turns ratio of the transformer? This will help me understand if the transformer is design correctly.

-Garrett

Hi, Garrett,

Transformer Turns ratio is 1:3 (Np:Ns).

A noisy peak voltage was found in CS input pin (CH3).

Please consider whether this is the cause of the problem.

Best Regards,Flyback waveform 200915.zip

Jae Hyung Kim,

Hi Jae Hyung Kim,

Can you please try changing the loop compensation to the following values?

R17 = R18 = 1.2kOhm

C26 = C23 = 220nF

This will lower the crossover frequency and should help to stablize the control loop,

Can you also please zoom in on the current sense voltage (CS pin) that you probed previously? I want to see if the peak current limit is being triggered.

Thanks,

Garrett

Hello. Garrett

As advised, the parts were changed and tested.

[Change point]

R17=R18=1.2kOhm

C26=C23=220nF

[Test result]
The waveforms of CH2(Snubber) and CH3 (CS) are improved from the test result on September 15th.
CH3 is CS PIN, and 3.8V pulsed noise is observed.
I doubt this. Due to this noise, it seems to be mistaken for overcurrent FAULT and stop PWM.
Is there any solution to this?

Thanks,

Jae Hyung Kim

30W_15V Flyback waveform_200927.pptx

Hello Jae Hyung,

Garrett is out of office this week.  I took a look at your waveforms.  They are typical waveform due to pulse skipping under light or no load.  What is actually the problem?   Sorry I don't see the 3.8V pulse noise that you mentioned. 

Thanks,

Youhao Xi, Applications Engineering