The TNY277 designed 15V 1A output switching power supply can work normally, but it is very hot. How to solve it?

selfcap_mcu

The TNY277 designed 15V 1A output switching power supply can work normally, but it is very hot. How to solve it? [Copy link]

 

 

I am a beginner in switching power supply. I have never been exposed to switching power supply design before. However, this time, I had to consider designing a switching power supply myself for a project considering the cost and installation area of the board. So I read a lot of information and copied a circuit according to the schematic designed by the previous engineer. Now the overall project has been completed, and the customer also approves the function. The project control board is as follows:

The project requires two voltage inputs: 110V and 220V, and 15W output (voltage 5V and current 1A).

The problem we are facing now is that when the actual output exceeds 6W, the surface of the TNY277 chip and the transformer are very hot, especially when the TNY277 chip outputs 8W, it is so hot that you dare not touch it with your fingers, as shown in the electronic load test

The following is the original circuit diagram of the switching power supply and the design parameters of the transformer:

Transformer design parameters

Please give me some advice, where should I test the parameters, modify the component parameters or modify the transformer parameters. Thank you very much.

maychang

Guess, just a guess: your transformer core is of poor quality and the iron loss is too high.

selfcap_mcu

maychang posted on 2020-8-26 17:11 Guess, just a guess: the quality of your transformer core is not good, and the iron loss is too large.

This shouldn't happen, because I used two transformers made by two manufacturers, and both had this problem. Is it possible that there are other parameter problems?

selfcap_mcu

maychang posted on 2020-8-26 17:11 Guess, just a guess: the quality of your transformer core is not good, and the iron loss is too large.

Not only does the transformer surface temperature rise, the chip is also very hot when the power exceeds 6W.

maychang

selfcap_mcu posted on 2020-8-26 17:31 This shouldn't happen, because I used transformers made by two manufacturers, and both had this problem. Is it possible that there are other parameter problems? ...

There doesn't seem to be any major problem with the component parameters. The transformer is not sandwich-wound, which increases the leakage inductance, but this doesn't seem to make the chip that hot.

chunyang

The device manual states: The rated power value of TNY277 is only 8W when it is continuously output in a closed space with wide voltage input without air cooling, which obviously cannot meet the needs of 15V1A continuous output, but TNY280 can. However, the transformer should not be too hot. If it is very hot, it means that the transformer has too much loss. You need to pay attention to whether the wire diameter, number of turns, and core material are suitable. However, you must first confirm whether the heat of the transformer is its own heat or the heat conducted from TNY277. It can be determined by short-term heavy load operation.

philips_lu

Take out the oscilloscope and look at the waveform of the seventh pin of the 277 chip under several working conditions. Pay attention to the high-voltage contacts and see the changes in the waveform under light load and heavy load, and then find a specific solution.

selfcap_mcu

philips_lu published on 2020-8-26 23:37 Take out the oscilloscope and look at the waveform of the seventh pin of the 277 chip under several working conditions. Pay attention to the high-voltage contact and see what happens to the waveform under light load and heavy load...

Can you be more specific?

selfcap_mcu

philips_lu published on 2020-8-26 23:37 Take out the oscilloscope and look at the waveform of the seventh pin of the 277 chip under several working conditions. Pay attention to the high-voltage contact and see what happens to the waveform under light load and heavy load...

Pin 7 is grounded. You must be referring to pin 4. Check its peak voltage and waveform under different load conditions.

asimilar

1. Regarding chip heat, someone mentioned it above, and the data sheet also states that the adapted power is 8W, and the maximum can reach 18W. So how to reach 18W? Of course, it is to increase the heat dissipation of the chip, PCB copper plating for heat conduction, and add a heat sink, etc.

2. You said the transformer is hot, but you didn't say whether it is the core or the coil that is hot? I checked your coil and found that even when the output is 15V/1A and the efficiency is 80%, the current density of the wire is no more than 5A per square millimeter, so you need to see if the working environment of your product is thermally poor before adjusting the current density. In addition, the minimum switching frequency of the control chip is 124KHz, so the skin effect must also be considered, and the wire should be selected with a diameter of no more than 0.35mm.

elec921

The minimum switching frequency of the control chip is 124KHz, so the skin effect must also be considered, and the wire should be selected with a diameter not exceeding 0.35mm.

selfcap_mcu

elec921 posted on 2020-9-1 14:29 The minimum switching frequency of the control chip is 124KHz, so the skin effect must also be considered, and the wire should be selected with a diameter not exceeding 0.35mm.

Yes, the default frequency of TNY277 is 132K. I don't understand why you said the wire diameter should not exceed 0.35mm. I used 0.4, so it should be about the same.

Because the second transformer proofing has not been done yet, the effect cannot be seen yet. Can you tell me how the diameter of this wire is selected?

selfcap_mcu

asimilar Published on 2020-8-30 16:43 1. Regarding chip heat, someone mentioned it above, and the data sheet also states that the adapted power is 8W, and the maximum can reach 18W. So how to reach 18W? Of course, give...

Hello, thank you for your answer. The idea is very clear. Another friend copied your reply downstairs.

The minimum switching frequency of the control chip is 124KHz, so the skin effect must also be considered, and the wire diameter should not exceed 0.35mm."

I asked a further question. I am very sorry. I misread it. This also shows that your answer is highly satisfactory.

This is what fever is like.

1. The chip surface heats up; 2. The transformer winding heats up; 3. The components next to the transformer also heat up.

The kind that is very talkative.

-------------------Another question.

Yes, the default frequency of TNY277 is 132K. I don't understand why you said the wire diameter should not exceed 0.35mm. I used 0.4, so it should be about the same.

Because the second transformer proofing has not been done yet, the effect cannot be seen yet. Can you tell me how the diameter of this wire is selected?

asimilar

selfcap_mcu posted on 2020-9-10 09:05 Hello, thank you for your answer. The idea is very clear. A friend downstairs also copied your reply "The minimum switching frequency of the control chip is 124KH...

You search for the skin depth formula and you will get the result.