Solution Finder Get Tech Support

How can I test flyback transformer ?

Posted by: Tugay on

Hi frends, I am trying to do 32V 2A power supply. I used top259 and designed the circuit in PıExpert. But circuit is not working. I see 32 volts at the output but when I load 500ma the top259 explodes. I've searched for this issue on the low power forum and I'm guessing the flyback transformer is faulty. I designed the transformer using pıexpert transformer parameters. So I have some questions about transformer. 1. How can I test after wrapping flyback transformer? 2. I guess leakage inductance is very high. How to measure leakage inductance ? 3. What is the LP_nom(Nominal Primary Inductance) and how can I mesaure ? 4.It is said that the leakage inductance should be below 10%, but based on which parameter it is 10% ? 5. Are there any other parameters I need to measure to test transformer ? I have a LCR Meter to measurement. I've searched in google these question and found some answers but I am not sure. Thank You.

Files

Attachment Size
Real Schematic design 91.66 KB

Comments

Submitted by PI-Cylon on 04/15/2021

Hi Tugay,


it is sad to hear that you are having problems with your design.


Thank you for attaching the PI Expert file and schematic. So far, I do not see any issues on them. PI Expert does warn you that you might not meet the safety isolation standard for your transformer. This is not related to your issue, but you might want to double-check that your safety requirements are being met. Do you have any additional transformer drawing or just what was provided by PI Expert?


You are correct to think that high leakage on the transformer can cause issues for the device. However, the primary clamp and the TVS D3, in particular, should prevent damage to the device. You might want to look to verify that those are working correctly.


1. I will look into it more and let you know what are the best tests to run.


2. The standard method to measure leakage is to remove the transformer from the board, short-circuit the other windings, and measure inductance of the primary using an inductance meter.


3. LP_nom is the nominal primary inductance. From your PI Expert file's Design Results report, I see that it is 1414mH. The tolerance is given as 10% in the report. You can measure it by connecting an inductance meter to the primary winding with the rest of the wires not connected.


4. Where does it say so? The % is from LP_nom of 1414mH. In PI Expert design report, you can see what PI Expert assumes is your leakage as L_LKG (35.35uH in your report). PI Expert uses this value to design the clamp that prevents damage to the device. In general, the smaller the leakage for the flyback design, the less work needs to be done by the clamp and the higher your design efficiency.

Submitted by Tugay on 04/20/2021

In reply to by PI-Cylon

Hi PI-Cylon,

I forgot to add something in the PI-Expert file I uploaded earlier. Now I added it. I used 23AWG cable right now because of I don't have any other cable sizes. When the 24AWG cable comes in I'll fix it but I think this issue not related to cable size.

I attach the measured values and winding directions of transformer. I used only LCR meter to measure.

1. The measurement in figure1 gives the value of leakage inductance, right?
2. Are Magnetization inductance and LP_nom the same thing ?
3. If they are not same, how to measure Magnetization inductance and LP_nom ?
4. Sorry, I couldn't understand what is the meaning of "with the rest of the wires not connected". Did you say other pins must be open circuit or should I measure the primary winding before wrapping other windings(secondary winding and primary bias winding) to find the lp_nom value?
5. Is the measured value in figure2 LP_nom or Magnetization inductance ?
6. Are directions of windings correct ? If they are not correct, can you describe it on the photo I took in which direction I should wrap with using colors?

I know that the answer to most of the questions I asked is in PiExpert, but I could not operate the power supply using PIExpert information.I probably getting some things wrong

Attachment Size
Measurements and Transformer photos 335.59 KB
PıExpert File 533 KB
Submitted by PI-Cylon on 04/30/2021

Hi Tugay,



1. Yes, the measurement in figure 1 should be the leakage inductance.

2. Yes, they are the same.

3. I believe that you measured it correctly in figure 2.

4. Yes, I meant open circuit.

5. LP_nom is the nominal magnetizing inductance.

6. As the transformer construction tab in PI Expert says, you should be winding all windings in the same direction for this design. Clockwise left-to-right (top to bottom in your pictures). What really matters is that the direction the currents will flow will be consistent with the schematic. The simplest way to ensure that is to follow the transformer construction instructions from PI Expert. If you swap a winding direction you also need to swap its pins on the schematic. However, winding direction may affect the EM noise generated by the transformer.



Based on your drawings, it looks like you did the measurements correctly. In your table on page 1, the first row contains the leakage inductances for each winding, while the second row has the magnetizing inductances for each winding. The ratio between the inductances of two windings should be equal to the ratio of their turns, squared. So, it is 16 between your primary and secondary and 64 between your primary and bias. Based on that, I think you accidentally swapped the values for the Secondary and Bias windings in the second row.



Comparing your values to PI Expert, your primary inductance (57.68mH) is completely different from what is expected in PI Expert (1.4mH) and will not work in this design. The only way I can explain the difference is that you didn't gap the core. The AL for your ungapped core is 4750 nH/Turn^2, while the ALG for the gapped core in this design is 196 nH/Turn^2, as you can see in the Design Report tab. The difference of more than 20x helps explain the difference in the inductance value.



PI Expert estimates that the gap size will be 1.439 mm to achieve the necessary inductance. The actual value of the gap to achive LP_nom might differ, but as long as your primary inductance is within the LP_tol of 10% (seen in design report), it should work.



If my guess about the core is correct, you can maybe still salvage the transformer by removing the core and putting a gapped core in it. If the core can't be removed then you will have to build a new transformer.



In addition, your leakage (58.57uH) is a bit higher than expected by PI Expert (35.35uH). This is an issue that will make your design have higher stress on the topswitch and be less efficient. In general, it depends on how well the transformer is wound. You can re-measure it once you fix the primary inductance.