Tool/software:
Hi,
I designed an isolated flyback converter using the LM5156 controller with the following specifications:
- Input Voltage (VIN): 22-36V
- Output Voltage (VOUT): 40-125V
- Power Output: Up to 125W
The output reference signal is isolated from the input using a high-stability isolated error amplifier (ADUM3190).
To allow for adjustable output frequency, I implemented a current injection method.
The system operates well under light load conditions (e.g., 1 kΩ). However, I encounter issues at higher loads (e.g., 150 or 50 Ω).
Specifically, when I attempt to increase the output voltage, it becomes non-regulatable and remains fixed at a specific value for the same load.
Interestingly, I've noticed that the microcontroller's switching frequency also changes; for instance, if the design frequency is set to 300 kHz, it drops to 150 kHz under load.
I also experimented with increasing the switching frequency to 800 kHz to see if that would resolve the issue, but the same behavior persists.
For example, these are images with 800kHz set (27kOhm on RT pin)
Attached is an image showing the output at 30 V:
- Fsw is correctly set to approximately 800 kHz.
- CH1: Gate signal
- CH2: Compensation signal
- CH3: Input current on the winding
- CH4: Output current on the winding
I've observed considerable jitter in the gate signal.
Moreover, when I try to increase the output using the current injection method, the COMP signal tends to rise.
However, at a certain point, as mentioned earlier, the frequency changes unexpectedly.
In the attached image, you can see this frequency shift: at an output voltage of 45 V, the switching frequency drops to 380 kHz.
I can further increase the voltage, as illustrated in another image where the output voltage reaches 80 V. At this level, the gate signal exhibits unusual behavior, making it challenging to fine-tune the output voltage.
The COMP voltage rises gradually but ultimately stabilizes at 2.8 V, as depicted in the final image.
Do you have any suggestions?
Thank you.
Francesco
