How to Wind a 15-Turn Transformer (0.6 mm × 3)?

[Smile2016]

For existing transformers, you can determine the winding polarity (dot convention) in several ways:

Visually, if the winding direction is obvious.
With an oscilloscope by applying a pulse or signal to one winding and observing the polarity/phase on the other.
Using a battery and a multimeter: briefly connect and disconnect a battery across one winding. The resulting voltage pulse induced in the other winding will cause a momentary deflection on the multimeter. The polarity of that pulse indicates the relative winding phase (dot orientation). Cheers !

I'm a newbie when it comes to winding transformers. Could you enlarge upon those three options? For example, when you can see the windings are both in the same direction, does that mean that the dots are both at the same end? Or opposite ends?

I think I understand that if a positive pulse is applied to the dot end of one winding, the dot end of the other winding also goes positive. Is that correct?

visual check, windings same direction = dots at the same end
pulse with battery : A quick practical example, better than some sensitive DMMs that might not survive or analog meters that you might no longer have...: take a transformer with a primary and secondary winding (or simply two windings on the same core, such as a common-mode choke). Connect an LED across one winding, the primary or secondary, it doesn't matter for 1 to 1 ratio transformers. .. Use a 1.5 V manganese battery (avoid cells capable of delivering very high current) and briefly connect it to the other winding.

If the LED flashes when you disconnect the battery, place a dot on the LED positive (+) terminal and another dot on the battery negative (–) terminal. Those two dotted terminals have the same instantaneous polarity/phase. If LED does not flash flip battery polarity... apply same logic. dot to LED+ and BATT-

Since you're using DC, do not leave the battery connected. The LED flashes at the moment the battery is disconnected because the collapsing magnetic field releases the energy stored in the core, inducing a voltage in both windings. The polarity of that induced voltage reveals the correct dot convention.

Another simple method is to use an oscilloscope.

Connect one winding across Channel 1 of the oscilloscope. Connect a signal generator to the other winding. For most of the ferrite cores and transformers we've discussed, a 3 Vpp sine wave or triangle wave is the best test signal. A frequency between 20 kHz and 500 kHz works well. (A square wave also works, but you'll see ringing and other parasitic effects that can make the waveforms harder to interpret.)

Connect Channel 2 across the signal generator output,  both oscilloscope probe and generator share the same ground (ground clip to ground clip).

Compare the two waveforms. If they are in phase, mark the two terminals connected to the oscilloscope/probe GND and generator GND with a dot. Those terminals have the same instantaneous polarity, which defines the transformer dot convention.

[ommsiva]

Dear All,

Although the majority of PWM ICs were released more than 50 years ago, why haven't they become as common/widely used as linear ICs?

I came across the two circuits (while discussing in this thread) on the web. Could anyone share a working circuit or guidance for constructing them?

Are there any video Link to understand the construction of Flyback transformer???

[Andy Chee]

Are there any video Link to understand the construction of Flyback transformer???

Watch from about 9:00. You will see transformer deconstruction. For construction, just reverse the process :-)

https://youtube.com/watch?v=Pa-bPlZikfE

[MariuszD]

Although the majority of PWM ICs were released more than 50 years ago, why haven't they become as common/widely used as linear ICs?

They are very popular; in my country, finding a power supply in a store that is not an SMPS is almost impossible. UC3842 is very popular; over half of the small power supplies <100W that I have seen were built on this chip. The rest were built on integrated circuits with an integrated switch.
SMPS of any kind are not popular in hobbyist projects because hobbyists do not want or cannot dedicate as much time to learning as it requires.
Another thing is that hobbyists do not look for well-developed and well-documented projects on the internet. They look for those with few components, meaning flawed projects. That's why search engines will recommend faulty projects to you, like that first diagram. Look for a project with a description, measurement results, and oscilloscope waveforms.

I came across the two circuits (while discussing in this thread) on the web. Could anyone share a working circuit or guidance for constructing them?

Detailed descriptions of the principles of operation and design can only be found for very simple circuits, at the initial stage of learning. SMPS is a more advanced level, and no one will write such a detailed description because it would be very lengthy. It's just that everyone assumes that when you tackle SMPS, you have transistors, resistors, capacitors, and inductors mastered.
Application notes from semiconductor manufacturers are good learning material. They explain the principle of operation and design methods well, even if they do not pertain to the circuit you want to use.

Are there any video Link to understand the construction of Flyback transformer???

Are you asking about flyback or sepic? I'm asking because you presented a non-working flyback circuit and a working SEPIC.
https://en.wikipedia.org/wiki/Single-ended_primary-inductor_converter
https://en.wikipedia.org/wiki/Flyback_converter
These two circuits are different topologies and have different requirements. In a SEPIC, there may be no coupling between the inductors at all. But the coupling is beneficial.  In a flyback, it is important; the better the coupling, the fewer problems with overvoltage on the primary side.

Flyback is simple but has unfavorable parameters, which is why it is not used for high power applications. To achieve 10A at the output with a 1:1 turns ratio and input voltage equal to the input voltage while operating in BCM mode, the peak current will reach 40A, which means significant losses in the transistor, output capacitor, high ripple, and RFI.

Forget about circuits that contain few components and forget about building an SMPS without dedicating several weeks to learning electronics.

[ommsiva]

Hi all,

I don't have oscilloscope,  suppose I am switching  at 1Mhz . What should be bandwidth. Can you suggest the thumb rule before for buying?

[ME]

Is ai stupidity involved in guptas original post, maybe its a part for his free energy harvesting scam/ project?.

[ommsiva]

Is ai stupidity involved in guptas original post, maybe its a part for his free energy harvesting scam/ project?.

Dear All,

I am trying to build a buck converter, but I have found very few designs that are proven to work in practice. One example is this project:



Although there are several circuits available online, most of them are either incomplete, lack detailed design information (such as transformer/inductor design, compensation, PCB layout, or testing results), or have not been fully verified.

For this reason, we are discussing our doubts and design challenges in this forum. The goal is to understand the circuit thoroughly and develop a reliable, working buck converter rather than simply copying an unverified design.

Thank you all

[shabaz]

Is ai stupidity involved in guptas original post.

Who is "Gupta"? Or is that a slang term you're using?

Is ai stupidity involved in guptas original post, maybe its a part for his free energy harvesting scam/ project?.

Someone who demonstrates severe incompetence and laziness in almost every request for help on forums ought to be careful what terms he bandies about to describe other peoples posts.

[Andy Chee]

Is ai stupidity involved in guptas original post.

Who is "Gupta"? Or is that a slang term you're using?

It's a racist trope i.e. all Indians are named "Gupta".

[radiolistener]

epidermal effect

epidermal

skin effect

[MariuszD]

Hi all,

I don't have oscilloscope,  suppose I am switching  at 1Mhz . What should be bandwidth. Can you suggest the thumb rule before for buying?

An oscilloscope with a bandwidth of a 10 MHz is usually sufficient. This is sufficient to determine whether the waveform is rectangular or distorted or the control loop is stable. To assess the rate of rise of the slopes, it may be insufficient. The difficulty lies in the fact that simpler/cheaper oscilloscopes often have serious flaws that make daily use impossible. For example, I used a 20MHz bandwidth oscilloscope with a non-functioning synchronization. 20MHz would be sufficient for simple measurements, but faulty synchronization prevented seeing important segments. That's why when buying a cheap oscilloscope is difficult, you should watch reviews, preferably from people for whom this isn't their first oscilloscope and who can tell if it's suitable for anything.

[ommsiva]

Dear All,

Please check the attached circuit.

I would like to understand how R2 and R3 implement the current limiting function in this design.

From my understanding, these resistors appear to be associated with the current sensing and feedback network, but I am unable to determine exactly how they set the current limit or how their values affect the limiting threshold.

Could someone please explain:

1)How do R2 and R3 determine the current limit?
2)What is the current path involved during current limiting?
3)How would changing the values of R2 or R3 increase or decrease the current limit?
4)Is there a formula relating R2, R3, the current-sense resistor, and the current limit?

The schematic is attached as a PDF for reference.

[Andy Chee]

1)How do R2 and R3 determine the current limit?
2)What is the current path involved during current limiting?

The schematic is attached as a PDF for reference.

The schematic is incorrect.

Please read the datasheet and study the internal block diagram for correct connections.

[ommsiva]

The schematic is incorrect.

Please read the datasheet and study the internal block diagram for correct connections.

" I felt the same"

[MariuszD]

I would like to understand how R2 and R3 implement the current limiting function in this design.

UC3843 is a current mode controller that cannot operate without a voltage ramp at the Isense input. Circuit with R2,R3 provides a waveform from the capacitor, there to avoid using a current sense resistor.

There is no output current limiter here.
I don't see any major errors in this schematic, but it eliminates the most important advantage of the current mode controller.

[ommsiva]

Dear Sir,

What is the Difference between PC Based Oscilloscope and DSO. I went through the below link in amazon:

https://www.amazon.in/Hantek-PC-Based-Oscilloscope-Channel-Computer/dp/B096KV4XZX/ref=sr_1_1?crid=1N5Y5M66Y4LL3&dib=eyJ2IjoiMSJ9.b5neWg8qfjVvsX5ihNXSUQ3MtT64FvTYJWwzroThmGAI6D9a-y_9r1BJJ9v6Y0fzosIWcu2FoZTdx2kifJMM_tOQGFkYmXHDi3mKCtiR1QH4iLUpeYOvFqAEl9j-Vk7sMMy9eHYbOIgDRMNDWx_UDdNA2G6jjqjOM21P9rul_zh9PB84M4encO-PEN2xa8dKTIOTG9tL1txscbLfWOOM0FkU-RNLNGonGeTcroeFc2MfiF4C7f8h4P-j6nNZZOCBfy516VZSgQ6jHLVwtuFG_Hs4nb0vgZ7_8vGE_HiP_-4.1tqZ2wKxgi-_-sG9AvVOFkWDRES0tuDzZwyQ6CbRZM4&dib_tag=se&keywords=hantek+dso+oscilloscope&qid=1786167905&sprefix=hantek+dso+oscilloscope%2Caps%2C238&sr=8-1

Hi all,

I don't have oscilloscope,  suppose I am switching  at 1Mhz . What should be bandwidth. Can you suggest the thumb rule before for buying?

An oscilloscope with a bandwidth of a 10 MHz is usually sufficient. This is sufficient to determine whether the waveform is rectangular or distorted or the control loop is stable. To assess the rate of rise of the slopes, it may be insufficient. The difficulty lies in the fact that simpler/cheaper oscilloscopes often have serious flaws that make daily use impossible. For example, I used a 20MHz bandwidth oscilloscope with a non-functioning synchronization. 20MHz would be sufficient for simple measurements, but faulty synchronization prevented seeing important segments. That's why when buying a cheap oscilloscope is difficult, you should watch reviews, preferably from people for whom this isn't their first oscilloscope and who can tell if it's suitable for anything.

Whether this system could solve my purpose of buying Oscilloscope??

thank you

[MariuszD]

This is exactly the case I was writing about. Hantek 6022BE has terrible software, which makes it useless. There is open-source open Hantek software available that is better, but I haven't had the chance to test it in real conditions.

[shabaz]

This is not related to your original post. It's not fair on others, because the information is unlikely to help many except you, since future others won't notice it in a google search, due to the thread title not matching the information you're now requesting.

This sort of circuit is very easy to simulate, and you'll learn a lot through that. You should be doing that routinely to understand subcircuits, if the theory isn't helping. Try KiCad, or LTspice. You don't need to simulate end-to-end, just the subcircuit you need help with. You could simulate to the right of the rectifiers but minus the BJT, see the signals for that until you understand it, add a load, change the load and so on, and then add the BJT and see the effect.

[ommsiva]

Dear Sir,

Thank you

[ommsiva]

Dear Sir,

I am working on this SMPS topology and I have a doubt regarding the output inductor/coil construction.

Could someone please explain how to practically construct the inductor for this topology?

I would like to know:

1)How to select the core type and size (toroidal / EE / ETD, ferrite or powdered iron)?
2)How to calculate the required inductance value?
3)How to calculate the number of turns?
4)What wire diameter or number of parallel wires should be used for the required output current?
5)How should the air gap be selected if a ferrite EE/ETD core is used?
6)How can I verify the finished coil's inductance and saturation/current capability?

I am particularly looking for practical references that explain the complete process of winding the coil, rather than only calculating the inductance.

Could you please suggest some good application notes, design guides, or video lectures on SMPS inductor construction?

Thanks & Regards

[TimNJ]

The dots on a transformer winding do not indicate the "start" or "end" of the winding. Indicates only the relative polarity/phase between the windings.

Winding a transformer with a single 1.5 mm diameter (~1.77 mm²) magnet wire can be difficult because of its rigidity. Using multiple smaller wires in parallel (multifilar) is sometimes easier. [other reasons too, not important in this case] If you choose a multifilar winding, keep the total copper cross-sectional area approximately the same.

For a prototype, you can start with a ferrite toroid or even reuse a common-mode choke from a scrap power supply, just maintain the correct winding phase (dot convention). See the attached image as an example. You can think of it as a 1:1 transformer where the primary and secondary are identical in your case. Use either the blue or black winding as the primary, I would probably choose the blue one. 🙂

Overall, the schematic looks good.

I mean, the dots *do* indicate the starts of the windings when all windings are applied to the core or bobbin in the same direction. A production transformer winding machine's spindle/arbor always spins in the same direction. It does not reverse direction to reverse phasing. Winding a reverse phased winding just requires swapping the start and finish pins for each other.



Winding a transformer with correct phasing is not more difficult than starting each winding on the pin marked with a phase dot, and finishing on the pin without the dot. Do the same for all windings, all while maintaining the same winding direction.

[ommsiva]

Dear all,

Could any body from India can guide me in buying dso economically as well reliably.

Thank you

[MrAl]

Dear All,

I have a question regarding the transformer used in this schematic.

The transformer is specified as 15 turns, 0.6 mm × 3 bifilar winding. I understand that a bifilar winding means two windings are wound together using parallel wires, but I am confused about the winding polarity and the terminal connections shown in the schematic.

Could anybody please help me understand how to physically construct this transformer?

Specifically, I would appreciate guidance on:

1)How to wind the 15-turn, 0.6 mm × 3 winding.
2)How many individual wires should be used during winding.
3)How to identify the start and finish of each winding.
4)How to connect the terminals so that the polarity (dot convention) matches the schematic.

If anyone has a winding diagram, photographs, or a reference video showing a similar transformer, I would be very grateful.

Hi,

There is something fishy about the way they drew that transformer and the dots.  The dots are supposed to indicate the SAME polarity for each terminal.  For that transformer, if the input dotted end is positive then the output dotted end is also positive, there is no flip of polarity as the seem to be indicating.

As to the physical wires, the two dots would be the two starts of a 2 wire bifilar winding (simplest view) but you could also refer to them as the two 'ends' or 'finishes' it doesn't matter because the polarity will always be DOT POSITIVE, DOT POSITIVE, and if you have three wires all the dots are the same polarity.  You could refer to them as DOT NEGATIVE and DOT NEGATIVE but that doesn't change anything, and thinking about it in terms of dot positive is just easier to think about.  There is no time when you will see DOT POSITIVE and DOT NEGATIVE on the same transformer though as they seem to be indicating.  They either made a mistake or they are trying to indicate some other operational mode, which doesn't make sense though so they probably just made a mistake.

Now as to the operation of the circuit itself, you have to figure out if they reversed the positive/negative labeling or they reversed the dots.  It could be that they made a mistake in labeling the dots positive and negative, or they should have put one of the dots on the bottom terminal not the top.  The circuit depends on the right polarity of the wiring.  If the wiring is not right the circuit probably won't work as intended.  We could look into that also.

So the bottom line is not where the dots are or what the polarity is on the schematic itself, it is what the requirement of the circuit is.  If the output of the transformer has to be in phase with the input then both dots should be drawn on top as shown, but if the output has to be flipped then one dot has to be drawn on the bottom terminal.  The required phase is determined by the circuit operation.

You can wind 3 turns bifilar it's just three wires instead of two.  All of the starts will have the same polarity, and they will all have the opposite polarity as all three of the finishes.

I just took another look at your schematic.  It looks like a flyback topology.
Because of that, the output expects to be REVERSED when the MOSFETs turn off.  That means that the schematic markings (+) and (-) are indicating the period where the output diodes conduct thus producing an output voltage or an increase in output voltage.  That means the dot on the output of the transformer should have been drawn on the BOTTOM not on the top, and the input side the dot is correct being drawn on the TOP terminal.
This works because we don't want an output when the MOSFETs turn on, we want an output when the MOSFETs turn off.  When they turn on, the top input terminal of the transformer goes positive and the bottom negative, and with the output winding reversed the top then goes negative which prevents the diodes from conducting.  Then when the MOSFETs turn off, the primary voltage reverses, meaning the top terminals goes negative, and with the output winding reversed that means the top terminal of the output winding goes positive, and that then produces an output current which increases the output voltage.
If we draw the dot on the output of the transformer on the BOTTOM terminal, the schematic should be correct then.

See new drawing.