When a purchased imported transceiver is paired with its old, reliable power amplifier (PA), which has served the owner faithfully for many years, a situation often arises when the PA excitation power is reset. The reason is the large input impedance of the PA, which differs from the output impedance of the transceiver.
For example, the input impedance of RA with OS:
on 3-x GU-50 lamps about 85 Ohm; on 4 G-811 lamps about 75 Ohms;
on GK-13 about 375 Ohms;
on GK-71 about 400 Ohms;
ontwo GK-71 about 200 Ohm;
on GU-81 about 200-1000 Ohms.
(Data taken from descriptions of RA designs in amateur radio literature).
TOIn addition, the input impedance RA is not the same across ranges and reacts to changes in the settings of the output circuit. So, for RA on a GU-74B lamp the following data on input resistance is given: 1.9 MHz - 98 Ohms;
3.5 MHz – 77 Ohm;
7 MHz – 128 Ohm;
14 MHz – 102 Ohm;
21 MHz – 54 Ohm;
28 MHz – 88 Ohm.
ExceptIn addition, the input resistance of the RA with feedback changes during the period of HF oscillations from several tens and hundreds of Ohms to several kOhms.
From the given figures it is clear that coordination of the transceiver with the RA is clearly necessary. Typically, such matching is performed using either parallel LC circuits or P-circuits installed at the lamp input. The method is certainly good, it provides matching with an SWR of no worse than 1.5, but it requires 6-9 circuits and two switch bars.
Butthey cannot always be placed in the existing old RA: there is no space and that’s it. Throwing away an old, good RA is a pity, but making a new one is troublesome.
In foreign military, civilian, and amateur radio equipment, broadband HF transformers have long been widely used to match 50-ohm units. They make it possible to coordinate these blocks with other circuits with a resistance that differs from 50 Ohms and lies in the range of 1 - 500 Ohms. Such broadband RF matching transformers can also be used to match transceivers with PA. They are small in size and you can always find a place to place them in the body (in the basement of the chassis) of the old RA.
In Fig. 1a. a diagram of an HF transformer on a toroidal ferrite core with a transformation ratio of
oppositions 1 ׃ │≥ 1…≤ 4 │ , depending on the connection point of the outlet tap.
Fig.1
And in Fig. 1b is a diagram of an HF transformer with a resistance transformation ratio of 1 ׃ │ ≥4…≤9 │ , also depending on the connection point of the outlet tap.
For transceiver output power up to 100 W, two 32 x 16 x 8 ferrite rings with a permeability of about 1000, or a larger diameter, but not with a smaller cross-section of the core, can be used as a toroidal core.
If the input resistance of the PA is less than 200 Ohms, then the transformer is wound according to the circuit in Fig. 1a, and if it is more than 200 Ohms, but less than 450 Ohms, then according to the circuit in Fig. 1b.
If the input impedance of the PA is unknown, a transformer should be made according to the second scheme, which, in case of poor matching, can be switched to the first option. To do this, you will need to turn off the middle winding and connect the outer windings, as in Fig. 1a.
The transformer windings are made simultaneously for the first option with two, and for the second - with three wires, slightly twisted, making 8 turns. In this case, from each turn of one wire a branch is made in the form of a ring (twist). Then the beginning of one winding is connected to the end of the second, and the beginning of the second winding is connected to the end of the third, which has taps. PETV wire with a diameter of 0.72… 0.8 mm. The rings (ring) must first be wrapped with tape made of fluoroplastic or varnished fabric.
Photo No. 1 shows two HF transformers made according to the second option.
Photo No. 1.
One transformer is made without twisted wires (in one row), soldered with taps on the switch strip, the other (smaller) - with twisted wires, both transformers have 9 taps (7 from the winding and plus 2 outer ones).
results transformer testing .
1. Transformer without twisting wires. Input impedance 50 Ohm. The output impedance is transformed into the following values (starting from the connection point of windings 2 and 3) along the 200 Ohm taps; 220 Ohm; 250 Ohm; 270 Ohm; 300 Ohm; 330 Ohm; 360 Ohm; 400 Ohm; 450 Ohm. (Figures are approximate). SWR by range (across all taps): at 3.5 MHz; 7 MHz; 14 MHz no more than 1.3; at 21 MHz no more than 1.5; at 28 MHz - 1.8 (up to 300 Ohms), and then SWR ≥ 2.
When this transformer is turned on according to the first option (with the middle winding turned off), the output resistance is transformed into the following values: 50.70, 80, 90, 100, 120, 140, 170, 200 (Ohm). SWR on all bands (across all taps) is no more than 1.4.
2. The transformer with twisted wires showed the best results. The output resistances are the same as those of the first transformer, but the SWR is much less: on the ranges 3.5; 7: 14 MHz no more than 1.2; at 21 MHz – no more than 1.4; at 28 MHz – 1.5 - 1.65. When the transformer is turned on according to the first scheme, the SWR is even better.
The transformer is connected to the gap between the input connector RA and the transition capacitor going to the lamp (to the cathode). If possible, you need to install a biscuit switch. In this case, you will need to select 2 - 3 positions at which the lowest SWR will be obtained on all bands. If this is not possible, then you will have to look for a compromise; you will need to find one tap from the transformer winding with an acceptable SWR on all ranges. Select a tap and measure the SWR for the RA to operate in operating power mode.
To match the transceiver with the RA, you can use simple matching devices based on a G-filter according to the diagram in Fig. 2, in the form of a separate unit connected between the transceiver and the RA with short sections of RF cables. (Possible with built-in SWR meter).
Fig.2
Frameless coil – 34 turns, wound on a mandrel with a diameter of 22 mm with 1.0 mm wire. Branches from the entrance are made through 2 +.2 + 2 +3 + 3 + 3 + 4 + 4 + 5 and another 6 turns. The coil is bent into a semi-arc and soldered with short taps to the contacts of the biscuit switch.
In switch position 1, the coil is short-circuited (bypass is turned on), and in position 11 the entire coil is connected. Capacitor, doubled from tube receivers. Instead of a variable capacitor, you can select constants for each range, switchable using a second biscuit. Such a control system allows you to match the transceiver and PA with an input impedance of 60 - 300 Ohms. (Photo No. 2).
Photo No. 2
But control systems in the form of a separate block have a significant drawback: in the reception mode, when the “bypass” is turned on in the RA, the output of the control system turns out to be mismatched with the antenna. However, this does not significantly affect the level of the received signal, because Usually the low-resistance antenna resistance is loaded onto the higher-resistance, now (for the antenna) input of the control system.
When setting switch The bib is only necessary when the gear is off!
Literature
1. E. Red.Reference book on high-frequency circuitry. - World. c.10 – 12.
2. WITH. G. Bunin, L. P. Yaylenko, Shortwave Radio Amateur's Handbook. – Kyiv, Tekhnika, 1984. p. 146.
3.B.Semichev. HF transformers on ferrite magnetic cores. – Radio, 2007, No. 3, pp. 68 – 69.
4. A. Tarasov. Do you use a matching device? – HF and VHF, 2003, No. 4, No. 5.
5 .I. S. Lapovok. I am building a HF radio station - Moscow, Patriot, 1992. p. 137, p. 153.
V. Kostychev, UN8CB
Petropavlovsk.
Transformers on ferrite tubes perform several functions at once: they transform the resistance, balance the currents in the antenna arms and suppress the current on the outer surface of the coaxial feeder braid. The best domestic ferrite material for broadband transformers (BCT) is grade 600NN ferrite, but tubular magnetic cores have not been made from it...
Now ferrite tubes from foreign companies with good characteristics have appeared on sale, in particular, FRR-4.5 and FRR-9.5 (Fig. 1), having dimensions dxDxL 4.5x14x27 and 9.5x17.5x35 mm, respectively. The latter tubes were used as noise suppression chokes on cables connecting computer system units with cathode ray tube monitors. Now they are being massively replaced with matrix monitors, and the old ones are thrown away along with the connecting cables.
Rice. 1. Ferrite tubes
Four ferrite tubes, stacked side by side in twos, form the equivalent of “binoculars”, on which transformer windings can be placed, covering all HF bands from 160 to 10 meters. The tubes have rounded edges, which prevents damage to the insulation of the winding wires. It is convenient to fasten them together by wrapping them with wide tape.
Of the various broadband transformer circuits, I used the simplest one, with separate windings, the turns of which have an additional connection due to the tightly twisted conductors among themselves. This makes it possible to reduce leakage inductance and thereby increase the upper limit of the operating frequency band. We will consider one turn to be a wire threaded through the holes of both “binoculars” tubes, and “half a turn” to be a wire threaded through the hole of one “binoculars” tube. The table summarizes the options for transformers that can be used on these tubes. Here N1 is the number of turns of the primary winding; N2 - number of turns of the secondary winding; K U - voltage transformation ratio; K R - resistance transformation coefficient; M - resistance ratio for a source with an output impedance of 50 Ohms.
Table
| K U | ||||
As you can see, a very wide choice of resistance ratios is obtained. A transformer with a 1:1 ratio, like a choke, balances the currents in the antenna arms and suppresses the current on the outer surface of the power cable braid. In addition to this, other transformers also transform resistances. What should you consider when choosing the number of turns? All other things being equal, transformers with a single-turn primary winding have approximately four times the lower limit of the passband compared to a two-turn primary winding, but their upper passband frequency is also much higher. Therefore, for transformers used from the ranges of 160 and 80 meters, it is better to use two-turn options, and from 40 meters and above - single-turn ones. It is preferable to use integer values for the number of turns if it is desirable to maintain symmetry and space the winding terminals on opposite sides of the “binoculars”.
The higher the transformation ratio, the more difficult it is to obtain a wide bandwidth, since the leakage inductance of the windings increases. It can be compensated by connecting a capacitor in parallel with the primary winding, selecting its capacitance to the minimum SWR at the upper operating frequency.
For windings, I usually use MGTF-0.5 wire or thinner if the required number of turns does not fit in the hole. I calculate the required wire length in advance and cut it with some margin. I twist the wire of the primary and secondary windings tightly until it is wound onto the magnetic circuit. If the ferrite hole is not filled with windings, it is better to thread the turns into heat-shrinkable tubes of suitable diameter, cut to the length of the “binoculars”, which, after winding is completed, are shrinked using a hair dryer. Pressing the turns of the windings tightly against each other expands the transformer bandwidth and often eliminates the compensating capacitor.
It should be borne in mind that a step-up transformer can also work as a step-down transformer, with the same transformation ratio, if it is “inverted”. Windings intended for connection to low-resistance resistances must be made of screen “braid” or several wires connected in parallel.
The transformer can be checked with an SWR meter by loading its output onto a non-inductive resistor of the appropriate value. Band boundaries are determined by the permissible SWR level (usually 1.1). The loss introduced by a transformer can be measured by measuring the attenuation introduced by two identical transformers connected in series so that the input and output of the device have a resistance of 50 ohms. Don't forget to divide the result by two.
It is somewhat more difficult to evaluate the power characteristics of a transformer. This will require an amplifier and a load equivalent that can handle the required power. The same circuit with two transformers is used. The measurement is carried out at the lower operating frequency. Gradually raising the CW power and maintaining it for about a minute, we determine the temperature of the ferrite by hand. The level at which the ferrite begins to heat up slightly per minute can be considered the maximum permissible for a given transformer. The fact is that when operating not on an equivalent load, but on a real antenna that has a certain reactive component of the input impedance, the transformer also transmits reactive power, which can saturate the magnetic circuit and cause additional heating.
In Fig. Figure 2 shows a practical design of a transformer having two outputs: 200 ohms and 300 ohms.
Rice. 2. Practical design of a transformer having two outputs
Transformers can be placed on a suitable sized board, protecting it from precipitation in any practical way.
Publication date: 07.12.2016
Readers' opinions
- Petya / 07/31/2018 - 14:23
So, where can I buy tubes?
I opted for a similar design immediately after the first tests, and today I do not know the best way to transform resistances with such weight-dimensional parameters of the transformer itself.
The basis of the device is ferrite tubes from signal cables of computer monitors. The power of such a transformer depends on the cross-section of the tube and their number. For example, a pair of even the smallest cable tubes operate freely at 200 watts. To increase the power of the transformer, the number of tubes can be increased proportionally. Such posts can also be assembled from individual high-permeability rings. In this case, when using ferrites produced in the CIS, be prepared to increase the weight and size indicators due to large losses in them.
This is what a transformer looks like in a power amplifier:
A transformer of this size can operate with an input power of 500 W. It is not difficult to imagine the dimensions of a transformer core for 1 kW - they are relatively small! In reality, I tested such a transformer for strength using power that was clearly too high for it with ACOM-2000. Working in a contest pileup on the 80m band heated it up and after 30 minutes it stopped working (the antenna SWR increased sharply), but after 10 minutes the SWR returned to its previous normal. Now imagine the dimensions of the transformer and the power supplied to it!
The transformation coefficient is calculated as follows:
K=N 2 2 /N 1 2
where N 1 is the number of turns in the primary winding,
N 2 - number of turns in the secondary winding
For example, a transformer with K = 2.25 contains 2 turns in the primary winding and 3 turns in the secondary winding. Such a transformer can be used, for example, to power antennas with Rin of about 100 Ohms.
The transformer is wound with three wires at the same time - we wind 1 turn. Then we wind a turn with the wire of the primary winding and half a turn with the wires of the secondary winding. It is better to use wires of different colors. Connect the two wires of the secondary winding in series. The connection point has zero potential (if the antenna is symmetrical) and must be grounded to drain static. It makes sense to wind the primary winding of such a transformer with a thicker wire.
One turn looks like this:
The entire 1:2.25 transformer is wound like this:
Important note: if the antenna is asymmetrical, then the middle point of the secondary winding cannot be grounded! To drain static, it is better to ground this point through a resistor of the order of tens of kOhms.
For the antenna mentioned above, a 1:2.78 transformer was used, which was wound on 4 tubes like this: 2.5 turns were made with three wires, and then another half turn was added for the primary winding. The secondary was connected in series. The resulting turns ratio was 5:3. Without compensation, I got this graph at a load of 150 Ohms:
Since the antenna only worked in the 1.8 and 3.5 MHz bands, I refused compensation.
Valentin RZ3DK (SK) produced the following graph without using compensation capacity:
When calculating turns, you need to understand that some kind of compromise is needed. On the one hand, the turns need to be made minimally enough for the lowest range, and on the other hand, we cannot obtain large leakage inductance at the highest frequency ranges.
In order to get a decent copy, you need to follow certain “rules”:
1. We must strive to have a minimum but sufficient number of turns in the windings
2. Take the wire with the largest cross-section possible, especially with a low-resistance winding.
3. For a symmetrical secondary winding, use a ready-made cable of two wires (of the type that were previously used in power cords), which we then connect in series. At the same time, they will definitely have the same length and other parameters, which will achieve symmetry. It is more logical to use such a wire if the number of turns of the secondary winding before connecting the ends is a multiple of an integer value.
4. By completely and uniformly filling the core window, you can achieve less “blockage” in the HF ranges.
5. The starting point for the calculation can be taken as the minimum sufficient number of turns at the lowest range. If there are few turns for a given permeability of the tubes, you will get an increase in SWR towards the low-frequency ranges and possible heating.
6. If you want to have more power of the device, you should strive not to increase the number of tubes, but to increase the cross-section of each tube. And the number of tubes should be minimal, i.e. only 2, but “thick”!
In conclusion, it should be noted that the weight and size parameters of transformers directly depend on the quality of the ferrite. I do not rule out that even at 100 watts, your transformer will heat up. There are two options: change the tubes or increase their number. My specimens at 100 watts did not change their temperature at all.
Well, don’t forget that the greater the reactive component in the load, the worse it is for the transformer.
About TDL in three parts:
- #1
Hello Dmitry!
I have a question about ferrous tubes.
The fact is that these tubes have a significant spread in permeability (from 10 to 300 - from those that I came across and were measured). How do you take this point into account and which (in terms of permeability) is better to use?
Currently I use such a trans-r on two tubes to power a vertical delta with a perimeter of 86 m with simultaneous power supply by coaxial cable RD-200. The TRX is next to the TRX. The feeder length is 15 m. The antenna is even built at 1.8 m Hz (hi!), of course its efficiency in this range is like that of a steam locomotive... - #2
Maximum permeability of the tubes is required. 10 and even 300 are not enough. True, it depends on what goals you pursue. I don’t think there is anyone willing to make these transformers to work only at 28 MHz, for example.
- #3
Hello Dmitry!
In what cases is it necessary to do galvanic isolation of the windings, and in what cases not (like yours)? - #4
On antennas, the antennas are always galvanically connected to the ground at least through a high-resistance resistance.
- #5
Hello Dmitry! My 86-meter Delta is powered by a symmetrical line of two 75-ohm cables, their braids are connected together (not connected anywhere). Next is a transformer, made in the form of binoculars from ten tubes. Cross section 5.8 cm2 and then 50 ohm cable (about 10 m). Is it necessary to connect the braids to the ground?
- #6
There is not enough data to assess the whole picture, but it is certain that the braid needs to be grounded!
- #7
Hello Dmitry!
I want to try to power a 1.8 MHz wave dipole approximately 164 meters long using a ferite latch so that I can move the power point along the canvas and find the optimal point for 1.8 and 3.5 MHz. Judging by the mana, the transformer is needed 1 to 2. Tell me how best to do it. house 30 meters at the elevator level.[email protected] Sergey RD0L
- #8
If you move it, then there should be only one turn in the secondary (the blade is passed through the ring once). Since the trans must transform 1:2 and increase the resistance to (as you write) 100 Ohms, then in its primary turns there should be sqr(0.5)=0.7vit, which is technically impossible. Therefore, this method only works with antennas with Rin<=Rкабеля. И то, всего лишь несколько случаев, да еще и на очень высокопроницаемом феррите.
- #9
valentine (Wednesday, 13 September 2017 14:49)
Dmitry, thank you for the wonderful example of a tr-ra, everything turned out 5 works well, power is 500 watts, two tubes are cold, which I’m very happy about, thank you very much
- #10
ps Afterwards, I wound 2 more tr-ra on cable latches - they all work fine, but the output capacitance had to be selected, for each case its own capacitance from 50pf to 30.5 pf at 29.8 MHz max VSWR 1.35 at 330m, but everything works on Windows, although not everyone answers, the power is 100 watts, thanks, everything works, thanks again
- #11
Cheers, Valentin! Yes, the capacity for compensation really depends on the design.
- #12
Hello Dmitry!
I got acquainted with the materials of your article.
The material presented is undoubtedly useful; theory without practice is dead. High power, high currents in stationary radio control units - the efficiency of the transmitter is not particularly relevant. Another thing is portable, small-sized, broadband, linear HF amplifiers with a 12V power supply.
The RPU was built on the basis of transceiver publication schemes from 2011-2014. The sad experience of trial and error led to the conclusion that the ShPT (at k = 1:2 and 1:3) on Amidon binoculars with copper tubes does not allow increasing the efficiency of more than 20-25% in the frequency range up to 30 MHz.
SHPTL, on the same amidon allows you to get an efficiency of about 30-50%, but other problems have emerged: blockages in the lower or upper frequency range (you can still fight this, there are hints) and the most disgusting nonlinear distortion (modulation of 1 kHz distortion from 10 to 35%). Yes, this agrees with the theory.
Therefore, the question is: Which ShPT or ShPTL can you recommend for a portable linear radio control unit? - #13
You neither indicated the Amidon materials (in general, this is Micrometals, and Amidon only sells them) that you used, nor the measurement methodology. I don’t believe that the efficiency ceiling is 35%. And what do you mean by “portable control unit”? Therefore, I do not undertake to give an answer to your question. For my purposes, I don’t know a better way to transform currents than the one described here, and I only use it even on receiving antennas.
- #14
How will a tube transformer work to match a half-wave wire from the end? With a winding ratio of 1/16.
- #15
It will be bad for him. The transformation coefficient is too high and, as one of the consequences, there areaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaremains of remains of interest on this transformation. Use autotransformer connections. Moreover, it is useless to try to galvanically isolate the windings when feeding a half-wave emitter from the end. Generally useless.
- #16
Hello RV9CX!
There are TDK ZCAT3035-1330 filters for signal cables. Do you think such a ferrite will work at least in the switched inductance of an antenna tuner? - #17
Well, where is the link to the datasheet?
I do not recommend putting ferrites into the tuner. Moreover, it is collapsible. It's one thing when you match the purely active component of the impedance. But, as a rule, those who use tuners work on all sorts of random laces - the reactivity there is astronomical and no ferrite can cope with it. No - everything will work, but there won’t be enough power in the antenna, and the ferrite will fall off one fine day. This is like an extreme case. - #18
Thanks, that's what I thought
https://product.tdk.com/info/en/catalog/datasheets/clamp-filter_commercial_zcat_en.pdf
The datasheet is scant and does not reveal the characteristics of ferrite. - #19
From the datasheet it is clear that they are not suitable for use as SMS. Well, as I said, don’t put it in the tuner. And what is the need for ferrite in the tuner. While we are corresponding, we should have tried it a long time ago))) You can simulate a reactive load for it (it’s easier with a capacitor) and see how it behaves.
- #20
I wound up the trance. 1/16 on 4 ferrite tubes from the monitor to match the 21 meter wire, (power supply) from the end to one 7 MHz range. Works fine. But it doesn’t get very hot for long at 400W. If I connect 2 of these, shtpl. Consistently 1/4 + 1/4. Will there be any point? I have not seen such methods on the Internet.
- #21
I won’t write anything about the inappropriate use of a transformer, I’ll say it to the essence of the issue.
Even in this article, the first photo is of consecutive tubes. In the article itself, I wrote that it is better not to increase the number of tubes, but their cross-section. These are two options on what to do!As for your decision... Of course you can do this. Especially after connecting a 1/16 trance to the end of a random snot. Nothing can spoil this decision even more. But if you are interested in my opinion, then I will repeat: you need to increase the power of the trance by cutting it, with an understanding of the intricacies of its work. Namely, that such trances cannot digest reactive chemicals.
- #22
Thanks for the quick response! Apparently, you are right. I only measured the SWR, it was 1.7, but there was nothing to measure the reactance with. With autotransformer winding on a T-200 ring from China. SWR below 3 did not work, and with our other rings too. Adjusting the length of the wire did not help! With a transformer on F. tubes, you can work for a long time at 100 W. But not with 400W. I will look for thick F. tubes. It is not possible to make another antenna like a 20 meter wire from the balcony. Roof. Closed.
- #23
You need to make an L-contour for each range. Not a ferrite transformer at all! Transformers are for other cases. For example, next to me I have an article where I brought the impedance to the same in a 2-band antenna and already transformed it with such a trance. At the same time, the antenna was tuned!
I don’t know what analogy to give, but you’ll probably understand if I say that you went to Alaska on a scooter. You can go, but not far and not for long, and you will not arrive in Alaska.
- #24
- #25
Thanks to your (and not only, but mainly) articles, I built an inclined triangle of 82.7 meters with symmetrical power supply from the corner, the suspension height is 22 m at the top and 12 m at the bottom. But the coordination was done according to the T2FD principle. Those. I inserted a 300 Ohm resistor into the center of the leg opposite to the feed angle (I figured that a higher load resistance would give less current in the antenna panel, and correspondingly less losses). I agreed on your recommendations using ShPT 1:6 on the tubes. Result: The antenna works great on all US bands 3-30 MHz with an SWR of no more than 2! Including WAC and SV! Worked with all continents and collected more than 300 DX with a power of 50 watts!
I built this “monster” from the possibilities of the environment: the city center, the antenna above the yard.
Thanks again and traditional 73! - #26
Well, I will never be able to describe such antennas. But coordination, yes - this option is the most optimal.
2) The SHTL must be loaded at the input and output with ACTIVE loads equal to approximately the characteristic impedance of the lines from which it is made.
Typical example: Our brother, a radio amateur, uses huge ferrite rings near the canvas to “balance” antennas. However, the experiment with active loads described above shows that a ring with a diameter of 10...20 mm can withstand a power of 100 W and does not heat up! So where is the truth? The truth is that the antenna (dipole or loop) has low active resistance ONLY at one single frequency, the frequency of the first harmonic of the antenna. High active resistances, which are present at even harmonics, are not applicable in practice. Low-impedance resonances at odd upper harmonics no longer fall into the amateur radio ranges. And at other frequencies there will ALWAYS be significant reactivity. They cause the ring to heat up greatly and therefore it must have a large cooling surface, i.e. be BIG. For example, imported 100-watt transceivers have microscopic ferrite binoculars at the PA output. AND NOTHING! This is not because they are made of outlandish material. Just one of the requirements for the output load for such transceivers is that it be ACTIVE. (Another requirement is 50 ohms). You should be wary of those publications that recommend winding a strictly defined number of turns for an HF transformer. This is a sign of another “disease of consciousness” - the quasi-resonant use of SPTL. This is where the legs of the legend about the need to use HF ferrites “grow.” But... There is NO broadband anymore!
Now about the mentioned 1:1 and 1:2... In a school physics course, the transformation ratio is the ratio of the turns of the primary and secondary windings. Those. ratio of input and output voltages. Why did radio amateurs turn this parameter “by default” into the resistance transformation coefficient? Yes, because the transformation of resistance is more important in our environment. But one should not go to the point of absurdity! Here is a conversation overheard on the air - two radio amateurs are discussing how to make a transformer from 50 to 75 Ohms. One suggests winding it with a turns ratio of 1:1.5. And when someone timidly objects to them, the only response heard is accusations of technical illiteracy. And this happens at every step! And just - TERMS! It turns out that the great law of conservation of energy does not apply to them and it is possible, with a voltage on the input winding of, say, 1 Volt, applying a power of 20 mW to the 50-ohm input of the transformer, and removing 30 mW at the 75-ohm output. This is what a “perpetual motion machine” looks like! Here you just need to remember that the resistance transformation ratio is a quadratic function of the voltage transformation ratio. In other words, a 1:2 transformer will transform a resistance of 50 ohms into 200 ohms, and a 5:6 transformer will transform a resistance of 50 ohms into 75 ohms. Why did I write 5:6 and not 1:1,2? Here is one step to the design. As already mentioned, the SHPTL should dangle with a line. A line is two or more wires folded together and slightly twisted. The characteristic impedance of such a line depends on the diameter of the wires, the distance between their centers and the twist pitch. To transform 50 Ohms to 75 Ohms, you must use a line of SIX wires and, if there is no requirement for balancing, connect these wires according to the diagram
As you noticed, the circuit is also drawn in a special way, not like a regular transformer. This image better reflects the essence of the design. The usual circuit diagram, Fig. 2, and, accordingly, the “traditional” design of an autotransformer with a single-layer winding and a tap of 0.83 total turns in practical tests “on the table” shows much worse results in terms of broadband.
For design and operational reasons, it is undesirable to make a SHPTL with a shortened section of one of the lines. Fig.3. Despite the fact that this makes it easy to make any, even fractional, transformation coefficients. This solution leads to the appearance of inhomogeneity in the line, as a result of which the broadband deteriorates.
An interesting question: “What are the limiting transformation ratios that can be obtained in SHPTL?” It is especially interesting to find the answer to this question for those who are “sick” with the idea of making a broadband aperiodic tube power amplifier, where it is necessary to transform a resistance of about 1..2 KOhm on the side of the lamp into a resistance of 50 Ohms. The experiment “on the table” gives a rather interesting result. Again, everything depends on the design of the windings. For example, if you make a “traditional” transformer or autotransformer with a transformation ratio of, say, 1:10, load it onto the required active resistance of 5 KOhm and measure the SWR on the fifty-ohm side, then the result can make your hair stand on end! And if, in addition, you remove the frequency response, it will be clear that there is nothing left of the broadband. There is one obvious, rather sharp resonance due to inductance.
This sore subject could be further developed ad infinitum, but... Everything was eclipsed by the design of a broadband balun transformer on a transfluxor (two-hole ferrite core) Fig. 4, which I managed to “spot” in an imported antenna for a “mustache” type TV. The image in the figure is, of course, schematic - in fact, the windings consist of several (3...5) turns. For a long time I looked at its design with bewilderment, trying to understand the winding system. Finally I managed to draw the location of the “windings”. This is an example of using true long lines!
If I didn't know that these were lines, I would think I was crazy! Especially this red short-circuited winding... But why aren’t we surprised in the case when, for example, in a cable U-elbow, it is necessary to connect the braid from the two ends of the coaxial cable at one point. Also, because it’s a LINE! In a benchtop equivalent load experiment, this microtransformer, designed to operate at frequencies in the hundreds of megahertz, showed excellent results at significantly lower frequencies, down to the 40m range and at full transceiver power.
Along the way, we will deal with the legends about symmetry and symmetrization. Let's find out how to very easily determine whether this or that SHPTL is symmetrizing, or the authors only declare this property, but there is no trace of symmetry there. Here “His Majesty – Experiment” and “His Highness – Theoretical Analysis of the Experiment Results” will help us again. First, let's figure out what a symmetrical output is and how it differs from an asymmetrical one. It turns out that everything depends on the design of the transformer. Here, for example, is the simplest case - SHPTL with a transformation ratio of 1:1. Any real or imaginary SHPTL (There are such! And not uncommon!) can be easily checked using your home transceiver. It is enough to connect an active load (equivalent) with a resistance corresponding to the transformation to the transformer output, and check the SWR at the 50-ohm input at maximum transmitter power (maximum accuracy of the SWR meter) in a given frequency range. If the SPTL is real, then the SWR should be close to ideal, i.e. 1.0 and in a WIDEBAND (that’s why it’s a WIDEBAND transformer!) It is advisable to have a transceiver open for transmission with continuous overlap and under no circumstances turn on the internal antenna tuner. The property of symmetry is checked when receiving using a FINGER (not the 21st! Although, you can use it!). Symmetry is the essence of EQUALITY of both load terminals relative to the ground (transceiver body). When receiving any station (possibly a broadcast station, it’s more convenient...) when you touch the ends of the load connected to the SYMMETRICAL output of the SHPTLE with your FINGER or a screwdriver, according to the readings of the S-meter and by ear, everything should be the same. But the signal level should be one point (-6 dB or two times U) less at each single-ended output. (this is in the case of a 1:1 transformation). It is convenient to use a 51 Ohm MLT-2 resistor as a load for a short time, even for 100 W transmission. In this case, an interesting effect is observed - while receiving a signal through a balun, when you hold a FINGER over the body of this resistor, a radio station will be heard from one edge, in the center of the resistor it will not be heard, and from the other edge it will be heard in the same way as from the first . Only under such conditions can the transformer be considered a balun. Try different designs of SPTLs that are published in the literature and on the Internet. The results may surprise you...
Briefly speaking! Make your mixer on any ring with low-frequency ferrite. If you try it, write! Experiment boldly!
Sergey Makarkin, RX3AKT
How to calculate and wind a pulse transformer for a half-bridge power supply?
We will talk about “lazy winding”. This is when you are too lazy to count the turns. https://site/
The most interesting videos on Youtube
 |
 |
 |
 |
Selecting the type of magnetic circuit.
The most universal magnetic cores are W-shaped and cup-shaped armor cores. They can be used in any switching power supply, thanks to the ability to set a gap between the parts of the core. But, we are going to wind a pulse transformer for a push-pull half-bridge converter, the core of which does not need a gap and therefore a ring magnetic circuit is quite suitable. https://site/
For a ring core there is no need to make a frame and make a winding device. The only thing you have to do is make a simple shuttle.
The picture shows a ferrite magnetic core M2000NM.
The standard size of the ring magnetic core can be identified by the following parameters.
D is the outer diameter of the ring.
d – internal diameter of the ring.
Obtaining initial data for simple calculation of a pulse transformer.
Supply voltage.
I remember when our power grids had not yet been privatized by foreigners, I built a switching power supply. The work dragged on until night. During the last tests, it suddenly turned out that the key transistors began to get very hot. It turned out that the network voltage jumped to 256 Volts at night!
Of course, 256 Volts is too much, but you shouldn’t rely on GOST 220 +5% –10% either. If you choose 220 Volts +10% as the maximum network voltage, then:
242 * 1.41 = 341.22V(we count the amplitude value).
341.22 – 0.8 * 2 ≈ 340V(subtract the drop on the rectifier).
Induction.
We determine the approximate value of induction from the table.
Example: M2000NM – 0.39T.
Frequency.
The generation frequency of a self-excited converter depends on many factors, including the size of the load. If you choose 20-30 kHz, you are unlikely to make a big mistake.
Limit frequencies and induction values of widespread ferrites.
Manganese-zinc ferrites.
| Parameter | Ferrite grade | |||||
| 6000NM | 4000NM | 3000NM | 2000NM | 1500NM | 1000NM | |
| 0,005 | 0,1 | 0,2 | 0,45 | 0,6 | 1,0 | |
| 0,35 | 0,36 | 0,38 | 0,39 | 0,35 | 0,35 | |
Nickel-zinc ferrites.
| Parameter | Ferrite grade | |||||
| 200NN | 1000NN | 600NN | 400NN | 200NN | 100NN | |
| Cutoff frequency at tg δ ≤ 0.1, MHz | 0,02 | 0,4 | 1,2 | 2,0 | 3,0 | 30 |
| Magnetic induction B at Hm = 800 A/m, T | 0,25 | 0,32 | 0,31 | 0,23 | 0,17 | 0,44 |
How to choose ferrite ring core?
You can select the approximate size of a ferrite ring using a calculator for calculating pulse transformers and a guide to ferrite magnetic cores. You can find both of them in.
We enter the data of the proposed magnetic core and the data obtained in the previous paragraph into the calculator form to determine the overall power of the core.
You should not choose ring dimensions close to the maximum load power. It is not so convenient to wind small rings, and you will have to wind a lot more turns.
If there is enough free space in the body of the future design, then you can choose a ring with a obviously larger overall power.
I had at my disposal an M2000NM ring of standard size K28x16x9mm. I entered the input data into the calculator form and received an overall power of 87 watts. This is more than enough for my 50 Watt power supply.
Launch the program. Select “Calculation of a half-bridge transformer with a master oscillator.”
To prevent the calculator from “swearing”, fill in the windows not used for calculating the secondary windings with zeros.
How to calculate the number of turns of the primary winding?
We enter the initial data obtained in the previous paragraphs into the calculator form and obtain the number of turns of the primary winding. By changing the size of the ring, the grade of ferrite and the generation frequency of the converter, you can change the number of turns of the primary winding.
It should be noted that this is a very, very simplified calculation of a pulse transformer.
But, the properties of our wonderful self-excited power supply are such that the converter itself adapts to the parameters of the transformer and the load size by changing the generation frequency. So, as the load increases and the transformer tries to enter saturation, the generation frequency increases and the operation returns to normal. Minor errors in our calculations are compensated in the same way. I tried to change the number of turns of the same transformer by more than one and a half times, which is reflected in the examples below, but I could not detect any significant changes in the operation of the power supply, except for a change in the generation frequency.
How to calculate the wire diameter for the primary and secondary windings?
The wire diameter of the primary and secondary windings depends on the power supply parameters entered in the form. The higher the winding current, the larger the wire diameter required. The primary winding current is proportional to the "Transformer Power Used".
Features of winding pulse transformers.
Winding pulse transformers, and especially transformers on ring and toroidal magnetic cores, has some features.
The fact is that if any winding of the transformer is not distributed evenly enough around the perimeter of the magnetic circuit, then individual sections of the magnetic circuit may become saturated, which can lead to a significant reduction in the power of the power supply and even lead to its failure.
We are trying to wind a “lazy winding”. And in this case, the easiest way is to wind a single-layer winding “turn to turn”.
What is needed for this?
It is necessary to select a wire of such a diameter that it fits “turn to turn”, in one layer, into the window of the existing ring core, and even so that the number of turns of the primary winding does not differ much from the calculated one.
If the number of turns obtained in the calculator does not differ by more than 10-20% from the number obtained in the formula for calculating the laying, then you can safely wind the winding without counting the turns.
True, for such winding, most likely, you will need to choose a magnetic circuit with a slightly higher overall power, which I already advised above.
1 – ring core.
2 - gasket.
3 – winding turns.
The picture shows that when winding “turn to turn”, the calculated perimeter will be much smaller than the internal diameter of the ferrite ring. This is due to both the diameter of the wire itself and the thickness of the gasket.
In fact, the actual perimeter that will be filled with wire will be even smaller. This is due to the fact that the winding wire does not adhere to the inner surface of the ring, forming some gap. Moreover, there is a direct relationship between the diameter of the wire and the size of this gap.
You should not increase the tension of the wire when winding in order to reduce this gap, since this can damage the insulation and the wire itself.
Using the empirical formula below, you can calculate the number of turns based on the diameter of the existing wire and the diameter of the core window.
The maximum calculation error is approximately –5% + 10% and depends on the density of the wire.
w = π(D – 10S – 4d) / d, Where:
w– number of turns of the primary winding,
π – 3,1416,
D– internal diameter of the ring magnetic core,
S– thickness of the insulating gasket,
d– diameter of wire with insulation,
/ - fractional line.
How to measure the diameter of a wire and determine the thickness of the insulation - described.
To make calculations easier, check out this link:
Several examples of calculations of real transformers.
● Power – 50 Watt.
Magnetic core – K28 x 16 x 9.
Wire – Ø0.35mm.
w= π (16 – 10*0.1 – 4*0.39) / 0.39 ≈ 108 (turns).
It actually fit - 114 turns.
● Power – 20 Watt.
Magnetic core – K28 x 16 x 9.
Wire – Ø0.23mm.
w = π (16 – 10*0.1 – 4*0.25) / 0.25 ≈ 176 (turns).
It actually fit - 176 turns.
● Power – 200 Watt.
Magnetic core – two rings K38 x 24 x 7.
Wire – Ø1.0mm.
w = π (24 – 10*0.1 – 4*1.07) / 1.07 ≈ 55 (turns).
In reality, 58 turns fit.
In the practice of a radio amateur, it is not often possible to select the diameter of the winding wire with the required accuracy.
If the wire turns out to be too thin for winding “turn to turn”, and this often happens when winding secondary windings, then you can always slightly stretch the winding by moving the turns apart. And if there is not enough wire cross-section, then the winding can be wound into several wires at once.
How to wind a pulse transformer?
First you need to prepare the ferrite ring.
To prevent the wire from cutting through the insulating gasket and damaging itself, it is advisable to dull the sharp edges of the ferrite core. But, this is not necessary, especially if the wire is thin or a reliable gasket is used. True, for some reason I always do this.
Using sandpaper, round the outer sharp edges.
We do the same with the inner faces of the ring.
To prevent breakdown between the primary winding and the core, an insulating gasket should be wound around the ring.
As an insulating material, you can choose varnished cloth, fiberglass cloth, keeper tape, Mylar film or even paper.
When winding large rings using wire thicker than 1-2mm, it is convenient to use keeper tape.
Sometimes, when making homemade pulse transformers, radio amateurs use fluoroplastic tape - FUM, which is used in plumbing.
It is convenient to work with this tape, but fluoroplastic has cold fluidity, and the pressure of the wire in the area of the sharp edges of the ring can be significant.
In any case, if you are going to use FUM tape, then lay a strip of electrical cardboard or plain paper along the edge of the ring.
When winding gaskets onto small rings, it is very convenient to use a mounting hook.
The mounting hook can be made from a piece of steel wire or a bicycle spoke.
Carefully wrap the insulating tape around the ring so that each turn overlaps the previous one on the outside of the ring. Thus, the insulation on the outside of the ring becomes two-layer, and on the inside - four or five layers.
To wind the primary winding we need a shuttle. It can be easily made from two pieces of thick copper wire.
The required winding wire length is quite easy to determine. It is enough to measure the length of one turn and multiply this value by the required number of turns. A small allowance for conclusions and calculation errors will also not hurt.
34 (mm) * 120 (turns) * 1,1 (times) = 4488 (mm)
If a wire thinner than 0.1 mm is used for the winding, then stripping the insulation with a scalpel can reduce the reliability of the transformer. It is better to remove the insulation of such a wire using a soldering iron and an aspirin tablet (acetylsalicylic acid).
Be careful! When acetylsalicylic acid melts, toxic fumes are released!
If a wire with a diameter of less than 0.5 mm is used for any winding, then it is better to make the terminals from stranded wire. We solder a piece of stranded insulated wire to the beginning of the primary winding.
We insulate the soldering area with a small piece of electrical cardboard or ordinary paper with a thickness of 0.05 ... 0.1 mm.
We wind the beginning of the winding so as to securely secure the junction.
We perform the same operations with the output of the end of the winding, only this time we secure the junction with cotton threads. To prevent the tension of the thread from weakening while tying a knot, we secure the ends of the thread with a drop of melted rosin.
If a wire thicker than 0.5 mm is used for the winding, then the conclusions can be made with the same wire. At the ends you need to put pieces of polyvinyl chloride or other tube (cambric).
Then the leads together with the tube need to be secured with cotton thread.
We wrap two layers of varnished cloth or other insulating tape over the primary winding. This interwinding gasket is necessary for reliable isolation of the secondary circuits of the power supply from the lighting network. If you use a wire with a diameter of more than 1 millimeter, then it is a good idea to use keeper tape as a gasket.
If you intend to use it, then you can wind the secondary winding in two wires. This will ensure complete symmetry of the windings. The turns of the secondary windings must also be evenly distributed around the perimeter of the core. This is especially true for the most powerful windings in terms of power take-off. The secondary windings, which take away a small amount of power compared to the total, can be wound at random.
If you don’t have a wire of sufficient cross-section at hand, you can wind the winding with several wires connected in parallel.
The picture shows a secondary winding wound in four wires.
