How to design a pulse transformer

arturavs wrote:

Note, however, that such luck works casual.

If such welded cores were ONLY in microwaves, then you would be 100% right. Unfortunately; many audio amplifiers also have a welded core. It is only about the fact that the production of such a transformer takes much less time - just put the "E" package on the other "I" and weld what would not fall apart. And that by the way such a way heats up more? In this case, it is secondary - the lower cost overcomes common sense.

Renegat_pol wrote:

Reality of the 1980s

Box. Again, the cost of production is at stake. To this day I have two automatic transformers (flipper) wound without using a wireframe (Made in the early 70's in the USA). Not only that - not even flooded with any resin or other binder. Simply put the coils separately on the machine - and the wire will be tensioned as appropriate (and secured after winding so that it does not unwind like a released spring) faster and cheaper. And no winding protection (sealant or other shellac)? Nobody cared about it - it was faster and cheaper without.

398216 Usunięty wrote:

If such welded cores were ONLY in microwaves, then you would be 100% right. Unfortunately; many audio amplifiers also have a welded core. It is only about the fact that the production of such a transformer takes much less time - just put the "E" package on the other "I" and weld what would not fall apart. And that, by the way, this is more likely to happen
heats? In this case, it is secondary - the lower cost overcomes common sense.

That's right, however MOTs are definitely not for continuous operation and have a very high no-load current. In amplifiers this current is definitely smaller.

arturavs wrote:

That's right, however .........

That is why I wrote about luck in audio to show that the assumption that the welded cores were used only in microwaves due to the fact that he works in an intermittent work regime and that he would not buzz is not 100% true.

PS Transformers with flippers have 600VA each.

398216 Usunięty wrote:

Transformers made of flippers are 600VA each.

Somewhere I am overturned somewhere safety 220/380 / 24V 6.3kVA from the 70s (due to the construction sometimes used as a welder ), where there is no carcass.
Windings wound with a solid copper flat bar (both), and tied with tape (I suspect cotton). In order to stiffen the windings, resotex wedges are hammered on the core.
The whole strongly impregnated with electro-insulating varnish.

arturavs wrote:

Windings wound with a solid copper flat bar (both), and tied with tape (I suspect cotton). In order to stiffen the windings, resotex wedges are hammered on the core.
The whole strongly impregnated with electro-insulating varnish.

At least a copper flat bar with a large cross-section, tying the windings with tape and wedges ... These are not even mine. 2.5mm diameter wire on only one winding (there are several different voltages and currents) and inserted under the box, which is on the top of the windings, a plate made of thicker box with connectors for soldering the windings and output wires. As you look from above, you can see the individual layers of windings and insulation. NOT COVERED ANYTHING.

Winding by hand is something extremely unpleasant, especially when you plan to wind more than a thousand turns of wire 0.1mm ... That's why I decided to make a mini winding station with a tensioner

Next year there will be some topics with practical high voltage transformers, but based on the above calculations. We will see to what extent the practice and theory will agree

_lazor_ wrote:

I decided to make a mini winding position with a tensioner

Beautiful. The only thing missing is automatic wire guidance with feed control This is not necessary in amateur use.

I used the above calculations to design a high voltage transformer:



Using the driver written on stm32f334 I was able to fire it in series resonance using the nightmare of high voltage transformers - capacity between turns of the secondary winding.
It works as a current source, i.e. when there is no load it works in short circuit.

Then maybe you will write something more about this converter because the topic is interesting.

I'm going to make a really extensive article about this construction, but that's how I can describe it for a short time.

Once I bought the book "Resonance circuits in power electronics" Tadeusz Citko, Henryk Tunia, Bolesław Winiarski. A very theoretical book, but that was what I needed - simplified mathematical models.
There is a description of the construction "Serial inverters with a receiver connected in parallel to the capacitor", which construction in the world is called LCC and is a current source.

Powered from 30V, but as it is a half bridge, 15V goes to the resonance circuit itself, the transformer ratio is 14: 850, which does not give much voltage with such a transmission. Due to the fact that the system works in resonance, reactive energy accumulates in the capacitor, which causes a voltage increase on this capacitor, and I decided that the capacitor will be ... the secondary winding. This works very well, although it will not be a cosmic efficiency construction, but you can get rid of the capacity problem on the secondary side to some extent.

In a simple model it works like this:
http://tinyurl.com/wq2s7d2

How to choose the cross-section of the wires? Are there any formulas for this?

Hello

The conductor cross-section depends on the current density, which depends on the operating temperature, cooling conditions, etc.

You have to calculate the resistance of such a wire, assume what effective current will flow through this winding and calculate the power losses. And now the most difficult thing is to count the thermal conductivity of the insulation (wire insulation and tapes insulation together). This already requires knowledge of materials and can be problematic ... You can also assume a fairly low current density, 2-3A / mm ^ 2 and hope a little that it will be a good value.

@_lazor_ And in a converter with a power of several to several watts, do not you think that it can be assumed that 3 A / mm2 will be sufficient? On the featured many times on this forum page 2-3 A / mm2 is recommended, and 1-5A / mm2 in general.

@ mechatron97 On the web you can find a lot of materials that show you how to calculate the thermal resistance and use it to calculate the wire thickness, but as we wrote earlier, it is not trivial and requires some knowledge of thermodynamics. Sample material:
https://www.ti.com/lit/ml/slup126/slup126.pdf