UCC38086: Less efficiency

Hi Huang,

The input current is increasing gradually even in a constant load, constant input voltage and output voltage. That's the problem

Okay. I will check for leakage current. Do i need to tune CS or CTRL related components? Used a 10K between 12VDD and CTRL. Rsense is 45mohms & Rf is 1K with Capacitor 940pf

I have used 50V capacitors at the input(15-21V) & 250V capacitors at the output(150V).Is it okay?

Hi,

My input capacitor is heating too much after loading for some time.(15 minutes)

Used a ceramic capacitor across 220uf al- poly type capacitor to minimise esr. Still it is heating

Replace it with a new one, then also after some time that capacitor is heating too much

Hi,

I have replaced the capacitor with  very less ESR type capacitor, my input current is still increasing gradually. Is there any other reasons for it, other than leakage?

Mosfets are also heating after keeping it on for more than 15 minutes. Input current is gradually increasing. Getting an efficiency of 74% only.

Also, when i Increase the output inductance value & sense resistor value, then I am not able to fully load the circuit. The output voltage is dipping. I think, all these are interrelated. Can you please advise?

I would like to share vds wave form to you. How can I send it

Please see the waveform attached. One of the mosfet is heating little higher. & input current is gradually increasing. Can you please check & advice on this. Your inputs will be a real help!!

Hi,

Where did you measure the Vds in your circuit? Can you label your measurement points of your circuit? Also please measure Vgs in the same screen.

yellow and blue are vds1& vds2 waveforms respectively. Waveforms looks okay I think. But could not figure out the reason for MOSFET and input current increase.

I will share the vgs and vds waveform in same screen to know if there is any turn on- off delay. Meanwhile, can you pls advice, how can I decrease the delay? No gate resistor or capacitor is there. Also, gate charge of the MOSFET is 42nC only. Switching frequency is 100khz

I will share the waveform asap. Do you know, what might be the reasons for input current increase?

okay. Why the losses are gradually increasing, which makes the input current gradually increasing? Please see below Vgs and Vds waveform

Do you think, the output inductor value has any impact on the input current increase problem?

Pls see my observations below

Test-1

Load- 35W

Inductor value- 210uH

Rsense-45mohms

Swiching frequency- 100khz

Transformer- etd29 core, 40/45 turns

Efficiency- 86%

During these conditions, the input current is stable. No MOSFET or other components heat is there. But, using this I couldn't load more than 35w. After 35w, the regulation is out.

Test-2

Load- 95W

Inductor value- 500uH

Rsense-20mohms

Swiching frequency- 100khz

Transformer- etd29 core, 40/45 turns

Efficiency- 86%

During these conditions, the input is slowly increasing for 1 minute. After 1 min, it will increase suddnely, MOSFET is heating too much. Transformer- and inductor heating reasonably. If I decrease the inductor, then current increase problem is not there. But couldnt load to 95w.

Pls help me.

Can you provide your schematics to help understand your Vgs and Vds?

I have shared the Vgs & Vds curves to you. Do you have any comments on the waveforms? PFA- my scheme & photo of a damaged unit. Driver track is burned here, 

Input voltage is 15-21Vdc & output is 150Vdc

Hi,

What was the operation condition when the driver U16 damaged (such as Vin, load current, etc)? As U16 damaged, anything else were damaged, such as Q2, Q3, or any other components?

Can you measure the current sense resistor (R15? the resistor below Q2 and Q3) waveform together with Q2, or Q3, Vds and Vgs?

It looks your snubber R26, R34, C29 and C39, or (R27, R35, C30, C40) are too much. How did you get these values? 

Also, when you say "inductor", is it L6? Please confirm or clarify.

Operating conditions- 18.5vdc input, 0.95w load, input current started increasing after 1 min, then damge happend.

Along with driver, Q2,Q3 are also damaged.

Snubber was removed, during the test. Because, after inserting the snubber, the power loss was higher. Do I need to reduce or increase snubber capacitor/ resistor value to overcome this power loss issue?do you have any suggestions on the values?

The inductor I meant is l6. As informed earlier, the inductor value used for loading greater than 35w is 500uh.(when I used ti recommended 1.5 mh, then no load current itself was 1•1 A.

I think, if I tune the inductor and sense resistor value, then I can overcome the input current increase problem. As per the datasheet calculation, I am getting inductor values in the range of mH, which is not suiting my application. Can you please suggest?

Hi,

Can you provide your circuit after remove your snubber? I need see what is your circuit like after remove the snubber. Then I can make suggestions after review your current circuit after the snubber removed.

Can you measure the waveform of current sensing resistor (R15?) so to see what is the current waveform and see if the current waveform shows something in concern. You need to measure the current sensing resistor waveform with different L6 to help find why you need to change L6 to smaller to get full power. 

I will share the modified circuit as well as waveform across sense resistor.

What might be the reason for MOSFET damage? MOSFET is rated for 100v, 80a. So, it shall not get damaged beacuase of high V or I. What might be the other reasons for MOSFET damge?

Hi,

MOSFET damages are due to over current, over voltage (Vgs or Vds)  or over temperature. You would have to measure the waveforms of its Vds, Vgs, Id (drain current), and temperature close to damage condition to find out. Over temperature can happen from MOSFET high switching losses. 

Hi,

Did you send to me your R sensing current yet? I have not received it.

Hi,

You need to measure your driver Vcc and other pins' voltage to find out why it is damaged. If you see all pins showing voltages within specs, then you need contact your local TI sales office to start an FA process to help find why the driver operation within specs can be damaged. 

Also, if you see all voltages and currents on your MOSFET are correct within its specs, while the MOSFET still damage, then you need to contact your MOSFET vendor to ask them why the MOSFET shows damage while its operation within specs.