Senin, 15 Agustus 2011


Sudah lama saya ingin membuat sebuah radio transceiver sederhana di 80M, tetapi mencakup mode morse dan voice. Ide-nya adalah membuat 80M radio transceiver yang bekerja pada mode CW dan DSB, tetapi dengan komponen yang seminimal mungkin, dan pcb-nya tidak lebih selebar bungkus rokok filter. Direct receiver menjadi pilihan utama, karena tidak memerlukan crystal filter. Untuk mempersempit bandwidth audio (receiver), sebelum input LM386 dipasang low pass filter untuk meloloskan signal audio dibawah 4khz.

Sebetulnya cukup banyak design radio transceiver di internet, tetapi tidak ada yang se-simple Pixie 2 (untuk CW). Untuk voice, yang paling mudah adalah menggunakan mode DSB (double side band suppressed carrier) karena hanya membutuhkan balanced modulator/demodulator.

Ide awalnya menggunakan diode balanced mixer, tetapi komponen ini mempunyai conversion loss yang lumayan. Maka pilihan jatuh pada mixer aktif NE602, mixer ini mudah dibuat sebagaio mixer bilateral (bisa sebagai mixer maupun modulator), dibandingkan dengan mixer aktif lainnya seperi TA7358, TA7310, atau TA7320.

Karena malas menggulung, maka VFO menggunakan koil oscilator MW (merah). Dengan kapasitor yang sudah ditentukan (tune capacitors), maka VFO dapat bekerja pada rentang 3.75-3.85Mhz. VFO ini juga bisa diganti dengan kristal 3.579Mhz. Trafo input receiver menggunakan koil 10.7Mhz (biru/hijau) dengan menambah 560pf.

RF power menggunakan mosfet BS-170 yang mempunyai penguatan tinggi. Dengan input cuma 50mW, 2x BS-170 mampu mengeluarkan RF sebesar 2W dengan tegangan 12V atau 1.2W dengan tegangan 9V.

Penggantian mode dibuat secara otomatis, jika tombol ptt ditekan, maka mic preamp akan mendapat arus dan PA linear akan mendapatkan input dari kaki no 4 NE602 (ouput dari balanced modulator). Jika key down (CW), PA linear akan mendapatkan input dari kaki no 7 NE602 (output dari internal VFO).

NE602 hanya memerlukan arus 8mA dengan tegangan 3.5V-6V, sedang LM386 hanya memerlukan rata2 arus 40mA, jadi cukup efisien. NE602 mempunyai conversion gain sebesar 18-20db, bandingkan dengan diode mixer (-10db).

Skema di atas dapat dimodifikasi untuk bekerja pada band lainnya, seperti 160M hanya dengan mengubah resonansi VFO di 1.750-1.900 Mhz dan mengganti LPF supaya bisa bekerja di band tersebut. Jika anda ingin bekerja di band-band 40M, 30M, 20M, dan 10M, sebaiknya menggunakan crystal oscilator, karena agak sulit membuat VFO yang stabil di atas 5Mhz.

Spesifikasi :
- Bekerja pada band 80M (3.5-3.9Mhz)
- Direct receiver
- mode DSB/CW
- CW full break-in (QSK)
- Receive curent 50mA
- Transmit current 200mA (1.5W)
- Ukuran PCB hanya 6cm x 8 cm

Mudah-mudahan bisa membantu teman2 QRP'er di Indonesia.

17 komentar:

  1. Mas, cari IC NE602 di Surabaya di mana ? harganya berapa sih ?

    Indra S.

  2. bagaimana kalo burung blekok yang datangkan ne602, biar aku tinggal inden 10 biji saja. soalnya aku dah bel eltech di surabaya juga dak ada. salam yd3 fsi lamongan.

  3. Single plated PCB or double plated PCB?

    1. It's using single layer PCB.

  4. Position of Q1-Q7?
    Value of Z1/Z2?

    1. You must see the schematic. Thanks.

  5. Dear Sir, the schematic does not match with the pcb layout. Please advise if you can post the modified version.Thank you

  6. In the schematic there is no part going to earth from the A line. In the layout I see a condensator and another part, marked with a circle.that should be a RFC or something else. there is some mismatch from the predriver MSP10 (BF 199) to the final FETs. Missing 10 nF connection from Collector of first stage predriver 2N3904. 5k6 resistor is around C-B, 10 nF, and L3 is connected to C. From Emitter missing 10 nF an 4w7 resistor. L3 is missing.5K6 resistor should go to line A, and not to collector of stage preamp. Missing decoupling 10 nF from stage one to stage nr. 2. L4: that is in my opinion 10 turn bifilar trafo to raise the power. Maybe the trafo on the right side of the layout does not reflect the situation in the schematic.
    QSk transistor is marked to be between 10uF condensator and 56 mH drossel. In the layout, this is connected direct to pin 3 AFTER the potentiometer. The microphone amplifier: 1 nF to ground - other 100nF to 10k resistor and 22k resistor. common point of C/R/R is goin to 2N3904. Collector gonig to 4K7 transistor, which is going to 10K and 1K and diode. 100nF decoupling condensator is not represented in schematic. I will make a pdf with the corrections. Szilard

  7. Left side down transistors: BC 558 pnp. Left side middel:npn transistor, microphone amplifier. Right side upper: audio preamplfier npn from the SA612(not in the schematic)rx mode, and in tx audio generator for side tone, another npn transistor to the pin 3 LM386: muting transistor npn.

  8. Driver amplifier connection is a little bit different than in the schematics. also power amplifier is a little bit different than in the schematic. One condensator between predriver and driver stage is missing from the pcb. maybe it is not necessary, or it was mounted on the pcb side. Please comment my observations, mister Ekoputro!

  9. Finally, I found out, where the problem is. This pcb and layout is not the good one. If you search on the internet the pcb: kolibri layout2, that is the good one. This one, layout 1, is for information purposes only.I have retro engineerd the pcb and found the component placement order, as the schematic shows.

  10. The PTT PNP transistor 2N3906 is not good at all. It can not support the current of 6 transistors connected to the line A. For this reason, it should be replaced with an 1 A switching transistor like BC 302 or BD 243 or similar, but the 3k3 an the 10 resistors must be changed too, the double value is recomended.

  11. A definitely WRONG design. The PCB doesn't follow the schematic at all.
    Main problems in design:
    1. components not marked properly in schematic nor on PCB,
    2. transistor count on PCB is more then in schematic (schematic not updated???)
    3. missing part some values in schematic
    4. missing voltage measuring points in schematic

    It is not easy to complete the build process this way, therefor to build this little radio is not recommended. It is supposed to work but it wouldn't. Just a time waste...
    Please DO NOT PUBLISH such kind of SHIT on internet. There are beginner HAMs around who are not competent in this field ad want to build a simple radio, but they will be just disappointed...

    p.s.: if somebody ask you a question about the REAL ERRORs in the project it is recommended to answer

  12. SA 612 : pin 6-7: 100 picofarads, pin 7-3: 100 picofarads,pin 6 series with L1: 1 nanofarad.with these values it will work for shure. with the values give in schematic, NO.


    The link above shows a much better muting-audio amplifier stage, which can be adapted to this Kolibri.

    See also the other circuit from the same author:
    Note, that this Kolibri can be adapted to broad band general coverage receiver and extended to digital broadcast receiver with 12.5 khz output for PC sound card.

  14. Audio frequency preamplifier is missing from the schematic of Kolibri 1.

  15. Explication for QSK circuit. The circuit from the antenna input-output is using a method that was developed by W. Hayward, alias W7ZOI. Antenna energy is sampled at the collector of the transmitter PA, routed through a series-resonant circuit (tuned to the resonant frequency), and then to the receiver input port. Two silicon diodes shall be included to protect the receiver during the transmit periods (in our case unfortunately we have only one diode). These diodes conduct during transmit and establish an RF level of 0.7 V RMS maximum, which will not harm the transceivers receiver input circuitry. The loss through the series resonant circuit is minimal. The output filter of the PA is inline with the receiver antenna line during receive period, and this is beneficial in adding additional selectivity at the input to the receiver. The C with a value of 68 pF in the schematic must be absorbed int the filter input circuit. Its value has to be substracted from the normal value of the 1st. capacitor of the LPF (820 pF), since it is effectively in parallel vith that C, when protecting antiparalel diodes conduct. 820 - 68 pF - so 752 pF is the correct value. The reactance of C and a series connected L (which is missing in the schematic, but present in the component placement picture, 18 uH), is 450 ohm. For 40 meters that value shall be around 47 pF. A semi variable trimmer 10-70 pF should be right piece to use there. This information can be read also in the W1FB The QRP Notebook.