top243p failure analysis

I received your message and I will try very hard to find the root of the problem.

In order for me to start digging into the problem I need you post the following information.

Schematic of your power supply,  

Transformer design documents. Including number of turns on each winding.

What I understand when you say

“whenever the unit power is ON ,the failed units first going to thermal shutdown with in 10Minutes,and after it will try 3or4 times simmilar way and then the device is failing.”

Is that the power supply has been running in steady state for 10 minutes. Then the controller goes into thermal shut down. Then , after the controller cools down, the controller starts to switch again and runs for 10 minutes and goes into thermal shutdown again.  After three or four cycles, the device is damaged.   Am I right in my description?

I have few questions about your design

1 Who designed the transformer?

2 Is the failure happening when  Vin = 390VDC and 3 watts output? 

3 What is your Output Voltage?

4 What is your ambient temperature when the power supply fails?

5 Is there another component that fails when the TOPswitch fails? 

6 If you have a current probe, can you send me a picture of your DRAIN current? 

Once I have the information I need I will start digging in the documentation to try to find why the failure is happening

mailed

Attached you will find the spreadsheet for your transformer design.

As it is shown in row 59 (file 1), your transformer is getting into saturation.  This mean the DRAIN current of the TOPswitch is going too high( higher than its ILIMIT 0.75A) and you are putting stress in the controller

I suggest you take a picture of the DRAIN current with your actual transformer design. The waveforms of VDRAIN you sent to me  do not tell me anything. The longer ringing just indicate the secondary is finishing its conduction time little earlier. 

But if catch a picture of the DRAIN current. Preferable the bad unit. Then the picture is a proof about the saturation conditions.

After you take the IDRAIN waveform, you need to add to the circuit a resistor from the X pin to the SOURCE pin to make your KI (row 17 of the spreadsheet #2) equal to 0.7. To see how to adjust your KI see data sheet figure 54B. 

This extra resistor will reduce the TOPswitch ILIMIT and will prevent the transformer to go into saturation.

The you can take a picture of your IDRAIN and you will see the new ILIMIT much lower that with KI=1.0

Talking about the way you are turning ON/OFF the controller.  Keep in mind that sw1 bounces every time you turn it ON/OFF. This bouncing is not desirable for the TOPswitch.   You need to add some solid state circuit to eliminate the bouncing  

thank you for your sugessions.this can be implemented in next batch  production.but present field units

the change of T/F design / I limit which is prefered.the ON/Off switch we already modified for next batch.

but implementing I limit and Undervoltage for top243p is it possible ,if yes pl suggest the same.

regarding failure we still not get correct root cause .

the failed units we replaced the TOP switch & diode MUR160  the units are working from last one month.

wehther the mur160 is prefered or mur1100 ?

we changed the t/f inductance  to 800uH  and 2.5mH  and we tried no of times ON/OFF the unit is working ok.

the unit is powered with 230V ac then the topswitch input is 300vdc only.after the topswicth on then only 400V DC will be the input to TOPSWITCH .so the peak flux dencity  calaculation whether it 300v or 400v ?.

we are not having current probe, we have  taken the waveform across the 1ohm resistor in the source path.pl check

and give ur comments

Thanks&Regards 

I read your message and I do not understand when you said “regarding failure we still not get correct root cause  “  

As I showed to you in the spreadsheets I sent to you, your transformer is not properly designed with the actual ILIMIT you have. You need to add the resistor in the X pin to reduce the flux density in your design.

Now if this the main cause for the failure,  I can not be 100% sure because I am far from the physical design but the flux density needs to be fixed

About your output rectifier,  If you see row 86 of the spreadsheets I sent to you,  This is the calculated reverse voltage on the output diode (52 volts for file 1)   As long as your output rectifier is rated about 20% more of this voltage you should be ok,  Of course you need to measure the reverse voltage with the oscilloscope under start up and over load conditions to see that the reverse voltage is lower that the rated voltage. MUR160 I think it is not good for your design. You need a diode with larger reverse voltage capability

I see in your message you made some changes to your design but it is hard for me to understand the whole picture. 

I suggest you use the spreadsheets I sent to you to make changes to your deign. This program is a PI expert type and you can download it for PI support page.

About the pictures you sent to me,  They do not tell me anything .  The sampling of the oscilloscope is very low and the shape looks like is not the right waveform.  

Dear Sir,

pl guide how to use both Input Undervoltage sense & I limit in TOP243P.for Top243p  RCD snubber the diode is MUR160(PIV600V) / MUR1100(PIV1000V) whcih is preferred. 

the unit is powered with 230V ac then the topswitch input is 330vdc only.after the topswicth 'ON' then only 400V DC will be the input to TOPSWITCH .so the peak flux dencity  calaculation whether it 300v or 400v ?.

Thanks&Regards 

Please read the data sheet of the controller. You can find examples how to adjust your input overvoltage protection and your ILIMT (Xpin) resistor value

There are many good pictures easy to follow. If you read the document it would be much easier for you to undertand than if I try to explain you

http://www.powerint.com/sites/default/files/product-docs/top242.250.pdf

in continuation with above said subject,in place of I-limit reduction ,shall we go for TOP242P  the peak flux density will be ok.is it recomanded ?

regarding top243p  as per ur datasheet ,we can't implement  both Undervoltage & cuurent limit is it correct / not?

 pl clarify

check figure 35 od data sheet to see how to adjust Input overvoltage and ILIMIT   "Figure 35. Current Limit Reduction with Line Voltage (Not Normally   Required-See M Pin Operation Description).",

Dear Sir,

regarding t/f peak flux density ,instead of TOP243p with I limit ,can we use  TOP242p for field units.because we provide the switch at control to M pin we can't implement the I limit. (refer our sch & T/f)

pl find the current waveform detials ,the waveform is 1mV/1mA

It is very clear than TOP244 saturates.  I am not sure about TOP243.  TOP242 is very clean waveform

Please check the attached spreadsheet. Check it for your transformer specifications (number of turns, core size, )and your design requirements (Vin, VOUT, POUT) .   This is the same file I sent to you before. The only difference is that in row 15  I enter TOP242P. The flux density is in acceptable levels now. right?   

So aftre you check the complete spreadsheet to fit your design (include thermal test to make your the controler delivers full power under worts hot conditions)  you do not find any possible issue, Then I would say it is safe to change the device.  Of course you need to fully test the supply aftre you changed the device

Dear Sir,

In TOP242P Datasheet the  self current limit protection the internal di/dt is 90mA/uS ,but our current waveform the peak current is 0.4A with in 2uS,but what is the meaning of  internal di/dt is 90mA/uS,how it effects the design,pl explain.

Thanks&Regards 

There is delay from the time IDRAIN reaches data sheet  ILIMIT to the time the controller shuts down the power mosfet. This delay is called the current limit delay.  It is equal to 100nS.  So when the power supply is at high line the slop of the IDRAIN will be larger. With a larger slope, the IDRAIN will reach larger IPEAK before the mosfet shuts down. Larger than when Vin is at low line.

If you are having a low flux density under worst condition, then you do not need to worry about it