UCC256304: Limiting the minimum switching frequency.

Thank you for your answer.

Isn't it risky not to limit the lowest working frequeny. No matter how carefully the LLC converter has been designed, there might be always extreme situations that makes the controller IC decrease working frequency more and more -- until it might reach the ZCS-region. As far as I know limiting the working frequncy is the most simple and the only way to prevent this.

Paul

Thank you again.
Actually I overlooked this part of the datasheet. However, it is only loosely connected to our problem.

Our problem is that we asked a company to develop a high power LLC-transformer for our needs. They did it and gave us guidelines concerning working frequency, especially lower working frequency, to prevent core saturation and overheating.

This is quiet normal and as being familiar with the UCC25600 I did not think about it very much. As TI recommends their new series for new designs (and there are some nice features) I was very surprized and could not believe that there does not seems to be a feature for frequency limitations. I have no idea how to accomplish working frequency demands of the transformer with the UCC256304. And - as my simulations show - it is not possible just by the design of the LLC-part.

Are there any solution for frequency limitiations with the UCC25630x?

Sincerely
Paul

Paul,

UCC25630x monitors the amplitude of the resonant current via its ISNS pin. Core saturation can be avoided by proper selection of the ISNS parameters to trigger OCP before the core saturation point of the transformer.

Switching frequency is dictated by the FB signal and the VCR pin waveform (VCR is the sampled resonant capacitor voltage). The VCR pin has a 3V DC offset. The amplitude of the feedback signal sets two two switching thresholds symmetric around 3V on VCR.

VTH controls the gate turn off of the high side FET. VTL controls the turn off of the low side FET. The turn on of both high side and low side FETs are controlled by an adaptive deadtime circuit. As the FB signal increases in amplitude, VTH and VTL move farther away from 3V. This decreases the switching frequency. 

If you can share your design spec and transformer details, I can help you select the appropriate IC parameters to satisfy the requirements given to you by your transformer manufacturer.

 Best Regards,

Ben Lough

Dear Ben,

thank you again for your effort.

Using the OCP features for frequency control might be an option, but I am very uncertain about dynamic aspects of this approach. To implement a appropriate frequency limit, relatively strict ocp 3 programming is necessary. But this might prevent the PS to provide enough current for shorter time periods. Ocp1 and Ocp 2 does not help because time periods are much to short. Dynamic behaviour is an important factor of the PS and the transformer.

To clarify my problem why I am thinking of lower frequency limitation, I calculated the transfer function of the resonant circuit for medium and high load (by matlab). Vertical lines are min. and max. gain requirements, right vertical line is f0 and the black line in the middle is the recommended minimum working frequency of the transformer.

 

While there is no problem for medium loads, high load condition has a special problem. High gain requirements could be fulfilled, when working frequency would be regulated to about 55kHz. This is still above ZCS but very much below recommdended low frequency limit of the transformer.

Unfortunately transformer power capabilites decrease a lot with lower frequencies. Therefore, there might be an additional  big voltage loss due to the transformer and the controll mechanism would decreas frequency even more. There migh be a lot negative effects because many "bad things" come together.

If I could limit the lower working frequency directly - ok, I will have a voltage drop because of insufficient gain of about 10 to 20%. But that's all.

----

Thank you for offering help. I would really appreciate that.

I have datasheets and test results for the transformers, schematics of the PS and board layouts. I have not tested them yet because I am still waiting for the boards. Actually I made two board versions for comparison, one for the UCC256304 and one for the old UCC25600. Of course the boards include a PFC. There is also a paper of what the PS is intended for.

Unfortunately I cannot provide all these documents to a public forum. Is it possible to share the documents without making them available to the public?

Thank you .

Paul

Hi Paul,

Please feel free to email these documents to me at ben.lough@ti.com

Best Regards,
Ben Lough