Specialized engineering for flyback transformer winding

High-frequency transformers play a crucial role in switching power supplies. This project focuses on the winding method of a flyback transformer, which is used in the calculated 240W design.
The theoretical


calculations for the transformer are covered in another project, which you can refer to.


The most important aspects of transformer design and calculation are determining the inductance, turns ratio, and Ap value. However, the most critical data for transformer winding are the window factor, window area, number of turns and strands, and wire diameter.


The data needed for winding the transformer is in the attached specifications. Since my lab didn't have fine enameled wire, I used 0.46mm and 0.54mm wire as a compromise.
In the actual winding process ,


multi-strand wire

is made using a hand drill. First, fix a hook to the drill. Then, prepare the enameled wire and measure its diameter with calipers. This demonstration uses two 0.56mm single-strand wires of the same length. Tie a knot, the length of which depends on the number of turns you'll be winding; remember to leave some slack. Secure the knotted end to a door handle, and fix the other end to either side of the drill bit's hook. Then, as shown in the attached video, turn the drill. The two strands will be wound together. Pay attention to the tightness of the double strands and the winding arrangement. If there are too many strands, the tension will be too high, making it difficult to wind. Here, 10 strands of 0.57mm enameled wire are used, resulting in 5 two-strand 0.57mm wires and 3 two-strand 0.46mm wires. After preparing the double-strand wire, the next step is to wind the primary winding. To improve the coil's induction effect, the primary winding is divided into two layers: the bottom layer is pins 1-2, and the top layer is pins 2-3. At this point, pay attention to the issue of the same-name terminals. A rule of thumb is: for coils wound in the same direction, the input terminals are the same-name terminals. Then begin winding. When the bobbin can't be wound any more times, slightly fix it, wrap it with Mylar tape, and then extend it in the opposite direction and continue winding. I wound 10 turns in the first layer. After winding, as shown in the picture below, remember to wrap it with Mylar tape (I forgot to take a picture of this, but there's a guide on how to wrap Mylar tape in the attachment, and you can probably imagine it after seeing this picture). Bend the multi-strand wire slightly, extend it towards pin 2, and wrap another layer of Mylar tape. Then cut the multi-strand wire, remembering to leave a little length, and put it in a sleeve (I used heat shrink tubing here). Scrape off the enamel from the enameled wire. Here's a crucial point: the enameled wire must be scraped clean in strands. Don't cut corners; ensure every copper wire is thoroughly scraped clean. After scraping, tin each strand to ensure it's completely clean, and also tin the pins for easier soldering. Wind multiple strands of wire onto the pins and solder them in. Make sure the soldering is secure and avoid any gaps. Repeat this process for the other pin. Once this is done, the first layer of the primary winding is complete. Next, wind the auxiliary winding. Since the auxiliary winding only requires one strand of enameled wire and has very few turns, wind it evenly. Pay close attention to the same-name terminals! After winding as shown in the diagram, process the pins using the same method. Next, wind the secondary winding. Because the secondary winding has many strands, use two pins as one: 6 strands for pin 8 and 4 strands for pin 9. Wind it using the same method as before, again paying close attention to the same-name terminals. Then wind the second layer of the primary winding, using the same method as before, so I won't elaborate further. After winding and processing the pins, it should look like the image below. Then, take out the magnetic core, insert it, and secure it together. Wrap it with Mylar tape and test. After testing the transformer primary winding with a bridge circuit, you'll find the inductance is extremely high, indicating this processing is incorrect. Therefore, we need to add some space between the magnetic cores using Mylar tape. Here, I've attached it to the stick-like end, as shown in the image below. Next, adjust the number of Mylar tape pads, adding or removing them little by little until the inductance matches the calculation. Then, apply some glue between the magnetic cores to secure them firmly, and wrap it with Mylar tape again. The winding is now complete. In summary,




my winding wasn't very good. First, there were too many strands, resulting in high wire tension, making it difficult to wind. Second, the evenness of the distribution on the frame is important; ideally, it should be relatively even, but mine wasn't very good, so please forgive me.


Pay close attention to the same-name terminals; for the same winding direction, the input terminals are the same-name terminals.


If there are gaps between the winding layers, remember to add retaining tape; you can buy retaining tape on Taobao.


After the transformer is wound, you'll find there's a noticeable difference compared to a factory-made one. Factory-made transformers aren't too expensive, and their machines can quickly produce the same precision and uniformity that hand winding struggles to match. Factories also have better testing equipment. All these advantages make hand-winding transformers a less desirable option. However, learning the process is still very important. I


have a detailed transformer winding tutorial in my attachment; you can refer to it.
The following are the revised parts. If there are any shortcomings, please point them out in the comments section, and I will put them in this section for reference.
(No
attachments available.)