TOP265 troubleshooting guide

Hi,

Following are some of the reasons why the power supply may be unable to regulate:

1. As described in the TOPSWitch datasheet, the part will start switching once the control pin voltage is above 5.8V and then the control pin voltage will drop as the part starts switching. It is expected that the feedback loop will close before the control pin voltage drops to 4.8V (nominal). If this does not happen, the part will stop switching and then attempts startup again after a delay. This phenomenon is described as auto-restart.

This condition is easy to find. If you verify the drain voltage of the TOPSwith and find that it intermittently switches for a short period of time and the interval between to bursts of switching is about 2 seconds, the the part is going into auto-restart

The failure to close the loop can occur if:

1. The transformer is wound incorrectly.

If the transformer is wound incorrectly, and the start and end of the windings are not terminated as shown in the schematic, the converter will not function as a "Flyback". This condition typically leads to high current in the MOSFET for each turn-on of the MOSFET and leads to premature termination of the pulse. This is again easy to find using a current probe connected to the drain connection to the transformer.

2. The bias winding is producing inadequate voltage.

The control loop is isolated using an optocoupler. The transistor side of the optocoupler is fed from the bias winding. If the bias winding output voltage is low, the optocoupler  is unable to supply the current required for the bias winding. This can be easily verified by measuring the bias winding rectified output voltage. THis voltage should be at least 9V.

3. Feedback loop is not functional:

If the feedback loop is not functioning (such as optocoupler is defective or is connected incorrectly) this would also prevent the feedback loop from closing and result in auto-restart. This requires verification of all connections and components in the feedback circuit.

 The power supply may also fail to start if there is a short circuit on one of the outputs. This could also be due to reversal of polarity of any of the output capacitors.

Please see page 6 through 10 of the datasheet that shows control pin voltage charge and discharge and the switching of the MOSFET and the auto-restarts behavior.

Hope this helps.

Regards,

PI-Sarek

My (x) input is tied to SOURCE.

Would this disable the part?

Thank you,

HN 

Hi,

As shown in figure 37 of the datasheet, the resistor connected to this pin sets the current limit of the part. If no resistor is used and the X-pin is connected to the Source pin, the part will operate at full current limit. The part should still work satisfactorily.

Many of the pins have multiple functions/features (For example the X-pin can be used to remotely turn the TOPSwitch ON and OFF as shown in figures 20 through 22 of the datasheet)  and as the data sheet indicates, once the X-pin is connected to source pin directly, some of these features are unavailable.

Regards

PI-Sarek

Thank you for your guidance.

Is there any point that it should be functional based on my schematic?

Anything that I can measure and be assured that the step is working and I move to next?

I am attaching a copy of my schematic. 

Hi,

Your circuit as shown will not work. The "start" end of the primary is pin 4 of the transformer and that would make pin 2 the "end" or the "non-dot" end. Your schematic shows a dot on pin 2. If your primary is split in two parts with the first part from pin 4 to pin 3 then pin 4 will need to be the start  or the "dot" end and pin 3 will be the end. similarly for the section of the winding from pin 3 to pin 2, pin 3 will be the "start" end and pin 2 will be the end or the "non-dot" end.

The negative output from pin 9 is also incorrect. this will need to be a separate winding from pin 11 to pin 9 with the start or the "dot end" at pin 11 and the "non-dot" or the end of the winding at pin 9.

Unless the dot conventions or the "starts" and the "ends" are correctly connected, the converter is not a "Flyback" and will not operate as expected.

Please see AN-22 for additional information:

http://www.powerint.com/sites/default/files/product-docs/an22.pdf

You may also find a TOPSwitch RDK (Reference Design) board useful to get familiar with the part. These are available from our distributors such as Digikey.(Digikey#  596-1312-ND)

RDK boards are fully functional  boards that are supplied with parts, circuit boards, application note etc.

You may also contact one of our sales offices in your area. A field applications engineer may be able to support you directly and offer more help beyond this forum.

 http://www.powerint.com/en/company/sales-information/worldwide-sales-support-locations

Regards,

PI-Sarek

Hello PI,

I updated my schematic. You are correct about the dots but it was my mistake on schematic symbol design.

I am showing a new schematic.

I also had an extra trace on the schematic that I removed from my board.

Removed the diode D8 for now.

Comparing my design with the eval board I am missing a few zener diodes and other circuits.

I copied my design from earlier fourms.

Do I need to add anything else to my design?

Thank you,

HN

Hi,

The schematic is still incorrect. For this to be a "Flyback", the dot end of the primary needs to be the drain of the primary side MOSFET which is Pin-4 of the transformer. your schematic still shows this to be pin-2 which is incorrect.

The way you have drawn the schematic, this will be a "forward" topology and the converter will not work correctly so please correct your design for this to be a "Flyback".

The negative output is still incorrect. Please see the previous reply regarding the negative output. For the negative output there needs to be a seperate winding from pin 11 of the transformer to pin 9 and the dot end of the winding needs to be connected to pin 11. 

There is significant amount of capacitance used in this design. You have a 1000uF capacitor at each output  however that may be considerably oversized. some of your outputs are relatively low power and hence will not need so much capacitance. Kindly review your design again. The power supply may need a soft finish circuit to start correctly if so much capacitance is used.

R23, D6 and C19 in the schematic shown in the link below form the soft finish circuit. 

 https://www.powerint.com/en/soft-finish

If you plan to use large capacitances in the output, you will need to add this soft finish circuit and fine tune it to enable the converter to start. 

Regards

PI-Sarek

Hello PI,

Thank you for checking my schematic.

Please explain how to change to to FLYBACK. I thought I had a flyback schematic.

1- I will flip the transformer input pins 2 & 4.

 2- For now I am ignoring the output on pin9 by cutting the D8 diode.

     I need to order a new transformer with isolated coil for negative supply.

     The -15V is a low priority on my design and not used yet.

 3-How much of the capacitance should I remove to get the supply working

     before the SoftFinish is added to the board.

I have 9 more boards that are already assembled like this.

I can remove the caps easier than adding the SoftFinish.

I will add the SoftFinish on my revised PCB.

This is my prototype but I have to make them work before going for

transformer and PCB modifications. 

Thank you,

HN 

Hi,

The topology becomes a "Forward" or a "Flyback" depending on the dot conventions (start and end terminations) of the primary and secondary. You have to follow the dot conventions correctly to make it a Flyback other than achieving the required primary inductance as calculated by the design tool.

For help with capacitor value selection, please use our design tool PIExpert.

Once all connections are correctly made and then if you find that the converter struggles to startup and it goes into auto-restart mode wherein it will attempt to start again and again and the outputs appear to rise but fail to reach regulation , then the soft finish circuit will be necessary. You will need to test the design at lowest and highest input voltage and with full rated load. 

Regards PI-Sarek

Hello PI,

I swapped the pins 2 with 4 on transforme T2.

The dot is on Drain now.

Also removed all the D1, D2, D4, D7 & D8 to reduce any other issues for now.

So there are no capacitors on other outputs.

The remaining capacitors are on feedback line and I reduced it to 1100uF.

Is this too much?

What should I look for next?

Please give me pin names.

Thank you,

HN 

Hello PI,

I swapped the pins 2 with 4 on transforme T2.

The dot is on Drain now.

Also removed all the D1, D2, D4, D7 & D8 to reduce any other issues for now.

So there are no capacitors on other outputs.

The remaining capacitors are on feedback line and I reduced it to 1100uF.

Is this too much?

What should I look for next?

Please give me pin names.

Thank you,

HN 

Hello,

Thank you for the good corrections.

I would like to get a copy of the expected waveforms on a working powersupply just like the

evaluation board before my order arrives.

Thank you,

HN 

Hello,

Thank you for the good corrections.

I would like to get a copy of the expected waveforms on a working powersupply just like the

evaluation board before my order arrives.

Thank you,

HN 

Expected waveforms

Hello,

We have several reference designs on our website for which we have detailed reports available for download. These reports have several waveforms and test data. Additional information is also provided in the application note and datasheet.

Kindly review the links below.

http://ac-dc.power.com/design-support/reference-designs/design-examples?...

http://ac-dc.power.com/ac-dc/design-support/product-documents/applicatio...

Regards,
PI-Sarek