TPSF12C1: CM choke model parameter

Hi Hegarty-san

Thank you for your reply.

I have checked the following document.

www.how2power.com/newsletters/2403/articles/H2PToday2403_design_TexasInstruments_part3.pdf

Also, when generating an equivalent model in the calculation tool, it seems that the LR circuit is modeled three times, but the impedance characteristics of the CM choke coil used by my customer are as follows. It seems to me that three circuits are insufficient to model this. What do you think?
In particular, it is unable to express a wide range above 1 MHz.

Best regards,

Higa

Hi Hegarty-san

Thank you for your reply.

My customer is currently preparing to evaluate this product.
However, he is unable to model the CM choke coil and is about to give up on the evaluation.
Would you be able to support me in modeling it?
I have actually tried using a tool, but I just can't get it to model at frequency bands above 1 MHz.

Best regards,

Higa

HI Hegarty-san

Thank you for your kind support.

I attach quickstart calculator file.

Best regards,

Higa1884.TPSF12Cx-Q1_quickstart_calculator_A9_PFC4kW.xlsm

Thank you, Higa. I'll take a look and reply asap.

Hello Hegarty-san

Our customers are in little bit hurry.
It would be very helpful if you could update the information.

Best regards,

Higa

1884.TPSF12Cx-Q1_quickstart_calculator_A9_PFC4kW_solved.xlsm

Hi Higa,

Please find attached the excel sheet in the model that we are able to come up with for the customer's choke data. 

comparing the model and the choke data you provided - they closely match.

could you please ask the customer to use this

Thanks and Regards,

Naresh

HI Naresh-san

I have a question.
In the document below, the formula for calculating Ri is as follows.

From this formula, Ri ≠ Ri + 1.
The result you presented was R2 = R3, so it doesn't fit the above formula, but is it possible that Ri = Ri + 1 in some cases?
What are the conditions for this?
It's very difficult to understand, so my customer doesn't seem to be able to figure it out.

Also, after setting the optimal values ​​for the other parameters, is it correct to understand that
INJ pin voltage swing at Fsw < VDD supply voltage?

Best regards,

Higa

Hi Naresh-san

Do you have any update?

Best regards,

Higa

Hello Diasuke Higa

I will get back on this

Hi Adepu-san

This is a request from a customer. Could you please help us derive other optimal filter settings using the equivalent model of the CM choke coil you created?
Since it has taken a considerable amount of time since we first contacted you about this, there is a possibility that we may lose this opportunity.
We would appreciate your support.

Best regards,

Higa

Hello Higa,

When you mention about document, what document are you referring to? (In the document below, the formula for calculating Ri is as follows.)

Is the customer planning to use the model for any simulation? could you check with him on which tool he uses?

The model which is attached here is the closet to accurate that we could get at this moment. Could you please ask the customer to use this?

Thanks and Regards

Naresh

Hi Naresh,

This is Conor, a colleague of Higa. Thank you for always following up on EMI filters!

I downloaded the Excel file you attached and checked the contents. This is very helpful for introduction before the customer evaluates it. So I have a few questions about the Excel file.

[Question 1]
I think it is fine to leave LGRID and RGRID at 0 in Step 3. However, I entered LGRID = 1µH and RGRID = 50mΩ as realistic initial values, referring to the User’s Guide. Is it appropriate to modify only the above from the Excel items you sent me and run the macro to verify it? If there are any other items that need to be modified based on the customer’s specifications, please comment. As far as I can see from the Excel parameters, each parameter seems to be reasonable as an initial value.

[Question 2]
When I ran the macro based on the contents of Question 1, an error occurred in the INJ pin voltage amplitude. The INJ pin voltage amplitude is about 50Vpp, which is a significant error value for VDD = 12V. The INJ pin voltage amplitude should be kept at 7.5V (=12-4.5V). Possible solutions include lowering the CINJ, but do you have any advice?

[Question 3]
I will repeat what Higa asked.
In the following document, equations (8) and (9) show a method of combining multiple RL parallel circuits in series when modeling nanocrystalline chokes. In this equation, Ri ≒ Ri+1, but in your answer, R2 = R3. Could you please comment the reason for this?

H2PToday2403_design_TexasInstruments_part3.pdf

A three-stage RL circuit is used to model the impedance characteristics of nanocrystalline chokes. ​The purpose of this model is to reproduce the decrease in inductance and the increase in loss components (resistance) in the high frequency band. ​Since changes in reactance have a significant effect on the phase characteristics, especially in the high frequency band, is it correct to think that the overall impedance characteristics are optimized by adjusting the inductance while keeping the resistance constant?

Thanks,

Conor

Hi Naresh,

Do you have any update?

I will also add a supplement to [Question 2]. If the INJ waveform is saturated, the following measures can be considered.
• Increase the capacity of the Y capacitor (regulator side) to a value larger than the value currently used.
• Improve saturation by lowering the gain of the AEF loop (increasing the RG resistor).

See "Any tips for circuit optimization and debug?" in the FAQ URL.
[FAQ] TPSF12C1 and TPSF12C3 power-supply filter IC FAQs

Please comment if there are any other measures.

Thanks,

Conor