TPSF12C1: TPSF12C1 debugging issues

mac he

Part Number: TPSF12C1

Tool/software:

Hi Team,

I currently am debugging 3200w server power supply with TPSF12C1, the topology of power supply is ac switching interleaved PFC plus single-phase full bridge LLC, the parameters of power supply is shown as below.

Input Voltage=230Vac
Output Voltage/Current=54V/60A(full load)
PFC operating frequency=65KHz
PFC Bulk voltage=380V@3240W(full load)
LLC operating frequency=>150KHz
Vdd of TPSF12C1=15V

My questions are below in debugging:

1.Please help check whether the quick start calculator of attachment(zip format) is ok or notTPSF12Cx-Q1 quickstart calculator B1 nanocrystalline（16T) 1 and 2 work for 3200w.zip

2.Are you how to measure the equivalent circuit of Ycap? I have found that the equivalent R of Ycap from download TI's TPSF12Cx QuickStart calculator Rev A3 format is about 5ohm or 10ohm(see figure1), but I used RLC equivalent model in series of Agilent 4294A PRECISION IMPEDANCE ANALYZER to measure that it is about 0.5ohm(see figure 2), Is my measurement method correct?

3.The INJ voltage(pin13) of TPSF12C1 is unstable (see CH1 signal), please give some suggestions how to make it output stable.

peak location nearby peak location zero across location

CH1:Vinj(pin 13 of IC)

CH2:Vsenl(pin 4 and 5 of IC)

CH3:Vsenn(pin 6 and 7 of IC)

CH4:Iin(ac input current)

figure 1 figure 2

Best Regards!

Mac He

2 months ago

0 Naresh Adepu 1 month ago

TI__Expert 5835 points

Hi Mac

we will look into it and get back to you

thanks

0 Naresh Adepu 1 month ago in reply to Naresh Adepu

TI__Expert 5835 points

Hi Mac,

What is the peak-to-peak voltage levels of the SENSE pin and INJ pin waveforms you are seeing? from the waveform it appears to me that it is saturating.

Also, ESR of Y-cap may not significantly change the performance on quick start calculator.

But as quick start is a model, in the real system you might have to do further tuning.

Could you please try increasing regulator side Y-caps and check as well and try tuning component values in step 5 and step 6 in real measurement system?

for immediate support, the documentation on the following link helps - [FAQ] TPSF12C1 and TPSF12C3 power-supply filter IC FAQs - Power management forum - Power management - TI E2E support forums

thanks and regards,

Naresh

0 mac he 1 month ago in reply to Naresh Adepu

Prodigy 20 points

Hi Naresh,

To simplify the experiment, I only enabled the PFC function of the product and the operation frequency of PFC is fixed at 65KHz,

Figure 1 is shown testing waveform of Vinj,Vsenl,Vsenn and Iin。Vinj signal seen from figure 1 is almost within output voltage range(see Figure 2) of Vinj that is shown on page 4 of TPSF12C1 datasheet except zero-cross location, but the conduction test is almost same(see figure3) when the AEF is ON or OFF, could you help give me some suggestions how to adjust effectively? Thanks a lot!

Figure 1:

Remark:

CH1:Vinj

CH2:Vsenl

CH3:Vsenn

CH4:Iin

Figure 2:

Figure 3:

Best Regards!

Mac He

0 Naresh Adepu 1 month ago in reply to mac he

TI__Expert 5835 points

Hi Mac he

Thanks for your question. we will get back to you

Thanks and Regards

0 Naresh Adepu 1 month ago in reply to Naresh Adepu

TI__Expert 5835 points

Hi Mac He,

Could you please confirm if you checked by increasing regulator side Y-caps and try tuning component values in step 5 and step 6 in real measurement system? In the real system, you might have to do few iterations of EMI testing by tuning component values to see the improvement.

for immediate support, the documentation on the following link helps - [FAQ] TPSF12C1 and TPSF12C3 power-supply filter IC FAQs - Power management forum - Power management - TI E2E support forums

0 mac he 1 month ago in reply to Naresh Adepu

Prodigy 20 points

Hi Naresh,

I have made experiments on 1600w product, by reducing the number of turns of common-mode inductor from 9T to 4T and decreasing regulator side Y-caps from 2.2nF to 0.22nF, I could see changes in the conduction emissions testing. however, there were no changes at the low frequency of 213KHz.

Best Regards!

Mac He

0 Naresh Adepu 1 month ago in reply to mac he

TI__Expert 5835 points

I received your question. I will get back to you

0 Naresh Adepu 1 month ago in reply to Naresh Adepu

TI__Expert 5835 points

HI mac he,

As the device is sensitive, trying to tune the components to improve the performance at 213kHz can affect the performance at other frequencies. You can try tuning some of the components in step 5 and step 6 of the quick start calculator but avoid iterations if it doesn't help improve performance.

Thanks and Regards,

naresh