Two Transformer H-Mode Mixer,The Italian way.

Experiments for Two Transformer High Performance H-Mode DBM.Its good or not.
2 transformer formula H-Mode Mixer of an I7SWX design and a performance are how.

(Ver.1.0 : July 26th 2003)

 

written by Takahiro Kato JH1WBU/JA9TTT, SaitamaJAPAN

http://members.my.home.ne.jp/ja9ttt/myexp/2T-H-Mode-DBM.html


The purpose of this page
Three Trifiler volume RF transformers are used for original H-Mode Mixer. The method of substituting this transformer for two pieces was devised by HAM of Italy, and Mr. I7SWX Giancarlo Moda.
The evaluation result is summarized.


H-Mode Mixer
H-Mode Mixer is one sort of the switching formula DBM (it is also called Commutation mixer). It operates like DBM using Diode. however -- the -- low -- distorted -- an everybody performance is wonderful As a switching element, D-MOS-FET called SD5000 at first was used. Then, it turns out that Chip called Bus-Switch is utilizable, and the way realized cheaply was opened.
About these circumstances High Performance H-Mode Mixer Please look at a page.


2 transformer formula H-Mode mixer (Two Transformer H-Mode Mixer)
As for H-Mode Mixer, improvement progressed by the change to Bus-Switch IC from D-MOS FET. After that, it is a thing that improvement which the examination to the direction to lessen progresses, still minces 0.1dB of conversion losses, and comes out is performed. It seems that the conversion loss of the level of 4dB is obtained now. In order to harness the performance of this mixer in the maximum, high frequency amplification (RF-AMP) is not used. It is because the receiver not more than noise figure skating (NF) =10dB can be made if there are few losses of a mixer. If it is called a NF=10dB receiver, it thinks from the level of visitor noise and HF belt is enough. The dynamic range of the receiver made by such method is supposed that it amounts to no less than 120dB. Thus, while the trial which lessens a loss is continued, the trial in which he wants to simplify is also performed. If the circuit of original H-Mode Mixer is looking-obtained, no less than three pieces have the transformer of a TORIFAIRA (Trifiler) volume, and manufacture is also troublesome and it turns out that the mistake of sushi wiring also tends to occur. HAM of Italy and Mr. I7SWX Giancarlo Moda devised the method of considering that will simplify and substituting for two transformers. If original, although the output of two transformers was compounded in current and the output had been obtained, it is expected to pack the transformer into one piece and to say that it will compound magnetically. Although it is visible also to the somewhat impossible method of existing on structure if it carries out from the view called transmission track type transformer, it will be a good method if a practical performance is obtained. About H-Mode Mixer of 2 transformer formulas, you introduce from Mr. I7SWX Gian of a designer. It is this site.
High Performance H-Mode Mixer After exhibiting the entitled page, mail came from Mr. I7SWX Gian soon. The report which contributed to the bulletin of RSGB and Radio Communication (RadCom) was appended.


Selection of a ferrite core
already
High Performance H-Mode Mixer But although submitted as one of the subjects, there is a problem of the transformer used for the portions of an input and an output. Although the room of a device is also in the structure of a transformer, the quality of the material of the ferrite used for a core is the most important. In the experiment of 2 transformer formula H-Mode Mixer, the kind of core was changed, it was distorted and the characteristic was compared. Core material on hand was three kinds, #43 material, #61 material, and #73 material, when saying by the catalog of a friend boss company (trading company). #43 material was generally good when saying from the conclusion. The form of a core used the glasses type called balun type (Bulun Type). An equivalent performance is obtained also for what put two ferrite beads in order and was made into the shape of glasses. FB-801-#43 which are generally easy to come to hand can be called good selection.


- Attention
A big photograph can be seen if a left-hand side photograph is clicked.
Please return to returning from a big photograph with the back button of a browser.


Manufacture
It is the circuit of 2 transformer formula H-Mode Mixer introduced from Mr. I7SWX Giancarlo Moda. Fundamental Frequency Squarer which builds a rectangle wave required for the drive of Bus-Switch from office right
弦波 is contained. There are also few parts and the mixer itself is easy also parts, although wiring is complex. Bus-Switch -- the original of 3 transformer formulas -- if -- although PI5made from Pericom C3125L was used, QS3126S1 of IDT was used for 2 transformer formulas of I7SWX Although comparison of Bus-Switch should also be carried out, the almost equivalent performance is obtained. A Best thing may be able to be discovered, if Bus-Switch of each company is tried, since it is thought that it has quite a few influences of a material (parts) since it is an easy circuit. VR of many rotations was not used for Fundamental Frequency Squarer. This is because the meaning which adjusts this VR severely was not felt. The office leak improved by adjustment of VR is at most 6dB. Since it is a Mixer use, an abbreviation is also possible and it is omitted also in the circuit to which I had Mr. I7SWX Gian send. It did not adopt, although the measures which prevent excessive current flowing were further taken in Mr. Gian's circuit at 74AC86 at the time of a non-signal. Since self-oscillation frequency is set to no less than 150MHz, the excessive current at the time of a non-signal is the phenomenon in which the consumption current of C-MOS increases. If it is not made no inputting, there is no problem. Moreover, with HF belt, it is 74. AC Not a type but 74 HC A type is also enough, and since there is less this, it can also recommend the trouble of an over-current. It is 74 when using this Mixer by 50MHz. AC What is necessary will be just to use 86. In 144MHz, there is also a report which a loss is large and is said that it is not equal to use.

Although it is an easy circuit, manufacture of a transformer and connection of winding require cautions most. If it mistakes, it will not operate normally with a natural thing. if a performance is poor, first of all,
obtains 結線 of a transformer, and there is necessity


It is the transformer to be used. The left is T2 of a circuit diagram and the right is T1. Probably, there is no necessity for explanation, since T2 is an ordinary TORIFAIRA volume (Trifiler). The problem is T1. The manufacture method is explained below. First of all, please make what twisted four lines. It is the electric wire used for a tetrapod filer volume. A polyurethane electric wire (UREMETTO electric wire : sign UEW) with a diameter of about 0.16mm is suitable for a wire rod. It winds around the hole of a core 4 times through this twisted line, and carries out. And both of holes will be counted with 1 time, if it returns. Then, the winding by the side of secondary is coiled on single track of the same kind. Similarly a number of turns is 4 times. If a secondary side is rolled so that a line may come from an opposite side, it can mitigate a mistake. On the whole, it becomes the transformer of 5 winding called a PENTA filer (Pentafiler) volume. There is only few winding which what thing does not have one than 3 transformer formulas. It is a merit that there are few a cores and they end. However, wiring does not become easy right [ that ]. Hi

Although it seems to be difficult somehow, if it does looking at a photograph and a circuit diagram, I will think that it is not difficult. giving -- being also prodigal -- it makes so that there may be no error in the connection method of winding Black dot mark - currently written to the winding of a circuit diagram is the sign which shows the start of a volume. The direction of winding is connected checking, after it is important. Since no less than five lines are coiled anyway, it is large confusion. He can feel easy, if are done, and there will be no short-circuit between each winding or it will check with the tester. The core of the orange of a photograph is #73 material. The name of a core is BLN-#73-2402. The white core of #43 material (BLN-#43-2402) and a green core looked at by the following photograph is #61 material (BLN-#61-2402).


Manufacture was made for every portion in this way. The portion of a right-hand side universal board is Fundamental Frequency Squarer. A Mixer portion is in left-hand side. The pitch conversion board made from SANHAYATO and ICB-010 are used. The price of that from which the conversion board for 28 pins became three pieces was 400 yen. Since Bus-Switch IC were 14 pins, they were miniaturized by cutting into a half further. If it cuts well, two objects for 14 pins can be made. If it is this, since no less than six circuits can be made for 400 yen, it is economical. Like a photograph, if IC is soldered, wiring between pins, the resistance for a bypass capacitor and bias, etc. will be attached. The copper tape was also used for assistance of a ground. The capacitor also assembled the tipped type in the compact using several pieces. Then, wiring is easy if it unites and manufactures. When attachment and its wiring of a transformer were made into the last, they were easy.


It is done H-Mode Mixer. The thing of an upper left big substrate is the original of 3 transformer formulas. Since this set performance evaluation as the main purpose, it used the big substrate. The thing (two pieces) of a small substrate is H-Mode Mixer of 2 transformer formulas. Since it actually manufactured here for the purpose of inclusion to apparatus, it achieved the miniaturization. Size is 65mmx45mm. Although it does not look well with a photograph, wiring of a transformer portion is complicated and is in a spaghetti state. it is because HAM of Italy thought -- it is Hi


Evaluation of Two Transformer H-Mode Mixer (BLN- #43 - In the Case of 2402 Use

The matter which is common by the following measurement is summarized.

1.     2 signal generators: Each output is 0dBm with the frequency of 7430kHz, and the its original work article which generates 7440kHz. IMD generated in person is less than [ -90dB ] (below a measurement limit).

2.     Signal generator (for Local OSC): To hp3336A etc., it is crystal oscillator use. A level is fixed by 0dBm.

3.     Spectrum analyzer: Advantest TR4171

4.     Power supply: KIKUSUI PDM-35-3. All power supply voltage is measured by 5V.


 17MHz estimates IF frequency from a comparative meaning with original H-Mode Mixer of 3 transformer formulas. The frequency of Local OSC is 10MHz. It was the best characteristic by 3 transformer formulas on this frequency. IP3 (OIP3) seen by the output side is +39dBm. IP3 (IIP3) seen from the input side was about set to +44dBm. Although +47dBm of 3 transformer formulas is not attained to a little, it can be called a sufficiently good performance. Or this difference will not be 2 transformer formula reason, it is unknown whether it is based on the difference in Bus-Switch. Conversion loss is 5.3dB and has become less than 5.9dB of 3 transformer formulas a little.


IF frequency was set to 27MHz, having used frequency of Local OSC as 20MHz. IP3 (OIP3) seen by the output side is +34dBm. Since conversion loss is similarly 5.3dB, IP3 (IIP3) seen by the input side serves as +39dBm. Although it is worse than the time of IF frequency being 17MHz, it is a still sufficiently good performance.


IF frequency was set to 2MHz, having used frequency of Local OSC as 9.74MHz. IP3 (OIP3) seen by the output side is +31dBm. Since conversion loss is 5.8dB, IP3 (IIP3) seen by the input side serves as +37dBm. Although it became a little bad, of course, it is the performance of the practical use range. Diode DBM which was able to improve -- at most -- since it is +20dBm


IF frequency was set to 455kHz, having used frequency of Local OSC as 7.9MHz. IP3 (OIP3) seen by the output side is +14dBm. Conversion loss also increases and it is 6.9dB. IP3 (IIP3) seen by the input side serves as +21dBm. Probably, it will be better for carrying out frequency conversion to stop to 455kHz, although it could not use.


IF frequency was set to 5MHz, having used frequency of Local OSC as 12.7MHz. IP3 (OIP3) seen by the output side is +37dBm. Since conversion loss is 5.8dB, IP3 (IIP3) seen by the input side is +43dBm and a very much good value. Merit that's right [ use / carrying out frequency conversion / for per 5MHz ]. Probably, it will be convenient to reconstruction (?), since the old transceiver also had many machines which have taken the 1st IF in the frequency of this hit.


IF frequency was set to 9MHz, having used frequency of Local OSC as 16.7MHz. IP3 (OIP3) seen by the output side is +35dBm. Since conversion loss is 6.2dB, IP3 (IIP3) seen by the input side is +41dBm and a very much good value. 9MHz is the frequency well used to a single supermarket. A its original work is sufficient and it is used.


It is the carrier leak by IF in case the frequency of Local-OSC is 10MHz. - It is sufficient value, if it is a mixer use, since it is 50dBm.


Two signals of an input are seen how much leak to IF output. Since 0dBm of an input leaks by -27dBm, if the degree of selection of an antenna filter is not improved, the leak by IF will be felt. Probably, cautions will be required also for how to choose IF frequency.


Evaluation of Two Transformer H-Mode Mixer (BLN- #61 - In the Case of 2402 Use

The following is as a result of [ of the transformer using the core of #61 material ] evaluation. A measuring method is the same as others.


17MHz estimates IF frequency from a comparative meaning with original H-Mode Mixer of 3 transformer formulas. The frequency of Local OSC is 10MHz. It was the best characteristic by 3 transformer formulas on this frequency. IP3 (OIP3) seen by the output side is +33dBm. Conversion loss was as excellent as 4.9dB. Therefore, IP3 (IIP3) seen from the input side was about set to +38dBm. # It can be called a tolerable performance although it is less than +44dBm of 43 cores. This is the ability of this transformer, after other parts are the same as that of evaluation of #43 except for a transformer.


IF frequency was set to 27MHz, having used frequency of Local OSC as 20MHz. IP3 (OIP3) seen by the output side is +27dBm. Conversion loss becomes somewhat large and is 5.1dB. IP3 (IIP3) seen by the input side serves as +32dBm. Although it is worse than the time of IF frequency being 17MHz, it is still valid enough as highly efficient Mixer.


IF frequency was set to 2MHz, having used frequency of Local OSC as 9.74MHz. IP3 (OIP3) seen by the output side is +33dBm. Since conversion loss is 5.4dB, IP3 (IIP3) seen by the input side serves as +38dBm. This is quite a performance. It seems that it will also be good to use #61 core if IF frequency is set to 2MHz.


IF frequency was set to 455kHz, having used frequency of Local OSC as 7.9MHz. IP3 (OIP3) seen by the output side is +18dBm. Conversion loss also increases and it is 5.9dB. IP3 (IIP3) seen by the input side serves as +24dBm. # Probably, it will be better for carrying out frequency conversion to stop to 455kHz like the time of 43 cores, if it could do.


IF frequency was set to 5MHz, having used frequency of Local OSC as 12.7MHz. IP3 (OIP3) seen by the output side is +34dBm. Since conversion loss is 5.4dB, IP3 (IIP3) seen by the input side is +40dBm and a very much good value. Merit that's right [ use / carrying out frequency conversion / for per 5MHz ].


IF frequency was set to 9MHz, having used frequency of Local OSC as 16.7MHz. IP3 (OIP3) seen by the output side is +32dBm. Since conversion loss is 5.5dB, IP3 (IIP3) seen by the input side is as tolerable as +37dBm.


It is the carrier leak by IF in case the frequency of Local-OSC is 10MHz. - It is sufficient value, if it is a mixer use, since it is 48dBm.


Two signals of an input are seen how much leak to IF output. 0dBm of an input leaks by -23dBm. It may become good when a float balun (Float Bulun) is put into an input and an output. However, probably, it will be better to cope with it by how to choose IF frequency, since loss also increases, and highly efficient-ization of a front end filter.


Evaluation of Two Transformer H-Mode Mixer (BLN- #73 - In the Case of 2402 Use

The following is as a result of [ of the transformer using the core of #73 material ] evaluation. A measuring method is the same as others.


17MHz estimates IF frequency from a comparative meaning with original H-Mode Mixer of 3 transformer formulas. The frequency of Local OSC is 10MHz. It was the best characteristic by 3 transformer formulas on this frequency. IP3 (OIP3) seen by the output side is +17dBm. Conversion loss became a little large by 6.5dB. IP3 (IIP3) seen from the input side was about set to +24dBm. This is the ability of this transformer (#73 core), after other parts are the same except for a transformer. # I thought that 73 cores had large 透磁率 (mui) and were suitable for the conversion in low frequency from the ability to earn the INDAKU wardrobe of winding. moreover, an B-H curve -- you may see -- although it expected since that was right, it was the result of saying generally that distortion is large like looking at the following


IF frequency was set to 27MHz, having used frequency of Local OSC as 20MHz. IP3 (OIP3) seen by the output side is +16dBm. Since conversion loss is 6.9dB, IP3 (IIP3) seen by the input side serves as +23dBm.


IF frequency was set to 2MHz, having used frequency of Local OSC as 9.74MHz. IP3 (OIP3) seen by the output side is +19dBm. Since conversion loss is 5.3dB, IP3 (IIP3) seen by the input side serves as +24dBm.


IF frequency was set to 455kHz, having used frequency of Local OSC as 7.9MHz. IP3 (OIP3) seen by the output side is +19dBm. Conversion loss also increases and it is 5.3dB. IP3 (IIP3) seen by the input side serves as +24dBm. The difference with frequency also with high conversion in frequency low to be sure will be generally meaningless in a thing with a large distortion, although it was lost. What is necessary will be well, just to use the level of a signal, thinking, after not being necessarily the performance which cannot be used.


IF frequency was set to 5MHz, having used frequency of Local OSC as 12.7MHz. IP3 (OIP3) seen by the output side is +18dBm. Since conversion loss is 5.5dB, IP3 (IIP3) seen by the input side is +24dBm.


IF frequency was set to 9MHz, having used frequency of Local OSC as 16.7MHz. IP3 (OIP3) seen by the output side is +18dBm. Since conversion loss is 5.6dB, IP3 (IIP3) seen by the input side is +24dBm. If #73 core is used like looking, the purpose of lessening characteristic change by frequency can be attained. However, it is troubled by the thing with a large distortion.


It is the carrier leak by IF in case the frequency of Local-OSC is 10MHz. - It is sufficient value, if it is a mixer use, since it is 47dBm.


Two signals of an input are seen how much leak to IF output. 0dBm of an input leaks by -29dBm. What is somewhat better than other transformers will be because combination between winding becomes dense and balance is improved, since 透磁率 of a core is large. It seems that however, this core is essentially distorted since big IMD is observed by even two signals which leak like looking at a spectrum. It may be the core material which can be used at least for a noise filter.


The design for the receiver using H-Mode Mixer
The project which shows the performance of H-Mode Mixer as a performance of a concrete receiver has started focusing on HAM of Britain. GDB2000 Transciever It is. However, faced with the its original work which enjoys leisure, a project full-scale to there cannot be started easily. There are some which are referred to as making the receiver for the low bands of HF belt in the design for me currently warmed for a long time. Although H-Mode Mixer has not adopted the object for 160m, and the object for 80m, the circuit unit was already completed and the test run has also ended. In 40m and 30m band which are severest band, it was just going to think that he wants to face with the receiver of the strongest front end. It is the good opportunity for H-Mode Mixer to appear. The time of realizing a design came by single conversion. Let's use a rudder type advantageous to distortion for a filter. Let's use the crystal with which low frequency is also taken and with which it is thick. That
電界 generated inside crystal becomes small receives distorted, and it is because it is advantageous. By 電界 between 電極, crystal bends mechanically by the piezo-electric effect, and is vibrating. When the amount of bending stops following the law of a hook, distortion occurs. The 2MHz rudder type filter is examined for such the purpose. It is because the crystal of thin high frequency is clearly disadvantageous. Although it is technology [ need / to be highly-efficient-ized / post mixer amplifier ], this is already realized with the transistor for CATV. I want to aim at a highly efficient receiver by 2 transformer formula H-Mode Mixer also as such a design.


Epilogue

How was introduction of Italy style H-Mode Mixer? Although the experiment was begun from Mr. I7SWX Giancarlo Moda's mail, I think that the very much good result was obtained. From 3 transformer formulas, although it is mist and the result of saying that it is inferior, there is no big difference. Therefore, I think that there are merits of enough that can reduce the number of transformers and a circuit also becomes compact. A result with 3 sufficient transformer formulas will be induced if it competes for several dB IP3. However, since performance sufficient also by 2 transformer formulas is obtained, let's use the method of an I7SWX Giancarlo Moda Mr. design for my manufacture.

It seems that the nonlinear nature can be seen if the level of a signal becomes larger and larger, although it is the ferrite core currently conventionally used for DBM casually. It was understood that it needs to be cautious also of selection of a core enough. However, it was happy that #43 material used unconsciously conventionally closely was good results. It is because a result which is said that an old argument reverses altogether for a core did not need to be brought.

        Bibliography and1:"Radio Communication" RSGB, and April 2002

        2: [ Bibliography and ] The core data book of Amidon