Building Your Very Own mcHF
Builders of the mcHF should have access to a quality soldering iron. Good temperature control of the soldering station is essential to working with the small SMT based components which make up the mcHF radio. A good pair of ESD safe tweezers should also be available along with a jewelers loupe or desk mounted magnifier.
Soldering the main CPU and other multipin small form factor components can be accomplished using drag soldering techniques. Chris has a great video he put together specifically on soldering the STM32F407 to the mcHF. Andreas DF8OE also has a video - he shows a different method of soldering. The eevblog has an entire series on SMT soldering techniques with the drag solder technique covered in episode #186.
After the winding of toroids it is necessary to remove the enamel from the wires in order to solder them to the PCB. Here is a howto.
The mcHF is composed of two main boards, the RF board and the UI board. Both are available for sale from Chris (M0NKA) from his webpage. He also has complete kits for order so you can avoid having to source all the components yourself and save on shipping.
The original bill of materials (BOM) references parts from the electronics supplier Farnell, primarily supplying to the European part of the world. Builders in the US can use Farnell part numbers with the electronics supplier Newark. Clint (KA7OEI) has also provided a BOM giving part numbers from Mouser, another major US electronics part supplier.
This section describes the assembly of the mcHF kit bought 2021 from Chris (M0NKA), SMD parts soldered, with metal enclosure. The kit is the actual version 0.6.3 (UI board: 11 March 2019, RF board: 25 March 2019) with green boards.
The mcHF 0.6.3 kit comes well packed and contains:
- Black metal enclosure with loudspeaker, speaker cable, screws, nuts, distance holders, 3 small and 1 larger knob, metal shield plate ...
- Two main boards (UI and RF) with pre-soldered SMD parts
- The tested LCD (HY28B)
- 16 small plastic bags with further parts for the RF board
- 8 small plastic bags with further parts for the UI board
- 2 sheets of paper listing the further parts for the UI and RF board
Now, what else do I need to start assembling my mcHF?
- Necessary parts:
If you want to use an electret microphone with your mcHF you'll have to solder the missing resistor R8 (UI board) used for the microphone bias: --> Resistor SMD 0805 680 Ohm (e.g. Farnell 933-3460)
You should also add the missing temperature sensor U10 (RF board), used for measuring the temperature of U8: --> IC MCP9801 SOIC-8 (e.g. Farnell 143-9486)
The EEPROM U7 is a necessary part. On my UI board, U7 was already soldered. If it is not: --> EEPROM 24LC1025 SOIJ-8 1Mbit I2C (e.g. Farnell 133-1323)
A good idea is to mount the LCD on a socket: --> 2 pcs. samtec CLT12001FD (e.g. Mouser 200-CLT12001FD) ... OR ... --> 2 pcs. female precision headers 2mm 20-pin (e.g. W+P Straight Female Precision Headers)
Later we will have to fasten the voltage regulators U3, U4 and the HF power MOSFETs Q5, Q6 to the metal enclosure. For a better heat dissipation (important!) you'll need: --> Thermal compound (e.g. ARCTIC MX-4)
- Necessary external parts:
Of course you will need a microphone with PTT and optional also an external speaker. Mike/PTT/Speaker: --> Loudspeaker Handheld Microphone (e.g. BAOFENG Emote)
The audio jack on most microphones will not fit the mcHF, so you'll have to build an adapter: --> 2 pcs. 3,5 mm male audio jack, angled (for soldering)
You'll also need a power adapter DC 12 V, up to 4 A. It should be a "low noise" type: --> Power adapter 12 V DC, 4 A, low noise
If this power adapter has another DC power jack than needed for the mcHF: --> DC male coaxial power jack, angled, 5,5 mm x 2,1 mm, mating length >= 12 mm (for soldering)
- External parts for testing and tuning of the mcHF:
After completing the assembly you'll have to test and adjust the transceiver. For this purpose and testing the TX function you'll need: --> Dummy Load 25 W, 50 Ohm (possibly plus a coax adapter between the dummy load and the mcHF's antenna plug P2)
A lot of testing has to be made before mounting the 2 boards in their enclosure. So it is very useful to be able to connect both boards with their P1 headers, but not stacked on top of each other: --> Set of 30-pin male and female header socket strip 2,54 mm and a strip board 2,54 mm with min. 30 parallel copper stripes
- Special parts (not necessary):
After completing the mcHF, it would be possible later to install an internal ATU into the mcHF enclosure. The supplied incremental encoder E3 (UI board) then has to be exchanged for an encoder with BUTTON. If you don't want to desolder this part later in order to install the ATU and the supplied encoders have a D-type axis, then you can solder this encoder for E3: --> ALPS EC12E2424407 (e.g. Farnell 152-0813)
If you are now waiting for some additional parts to arrive, you have plenty of time to read a lot, watch some videos, download some things and plan your work.
Here my recommendation:
- Chris' Homepage: You'll find many useful infos from the creator of the mcHF. HINT: Chris' homepage sadly is down since 2023. Here is the last archived version.
- From his (archived) Download Page you should download the "mcHF_rev_0.6.2" and "mcHF_rev_0.6.3" release files. From the 0.6.2 file you only need to copy the BOM (bill of materials) in the "bom" folder. The files are called "mchf_rf.xls" and "mchf_ui.xls". All other informations can be found in the 0.6.3 release file. Necessary documents in the "assembly" folder are "transformers_winding.pdf" and "LPF Torroid winding guide.pdf".
- Another document we will need is the mcHF_board_modifications_20150825.pdf, which contains coil winding tipps and recommended or necessary mod(ification)s, mostly relevant for the older mcHF versions.
- Watch this very nice mcHF Build Video! It is 2h 26m, but worth every minute.
- Many problems related to the mcHF assembly are discussed in the mcHF Groups.io. You will find answers to nearly every question.
- The German DARC Ortsverband-Sulingen I40 Forum contains many useful informations. This "Ortsverband" also has its own mcHF Project Group.
- ... of course this Wiki ...!
If you are still waiting for necessary parts, you could start working on the metal shield plate, delivered with the mcHF enclosure. It has to be adapted to the version 0.6.3 boards, and will be mounted between them later. The following picture shows the front side of the "ready to use" metal shield.
There are four changes to be made (from left to right):
- The neon lamp DS1 is placed best lying on the front side of the RF board. You have to cut out a rectangle 7,5 x 32 mm (see the picture!).
- For a second distance holder you should drill a new 4 mm hole.
- For the capacitor C74a (UI board) you have to drill a 9 mm hole over an existing 4 mm hole.
- For the capacitors C32 and C32a (UI Board) the existing hole must be expanded about 2 mm to the top.
That's all.
If you like videos, the mentioned mcHF Build Video also shows the working on that metal plate, beginning at 1h 34m, 1h 45m and 2h 03m. The last mentioned part of the video at 2h 03m also contains the mounting of the two 6 mm distance holders (delivered with the mcHF enclosure) and of an optional rubber seal (height 4 mm) on the back side of the metal plate, attached with a thin double-sided adhesive tape (e.g. tesa 5338).
Let's start with the back side. By the way: what is front and back side of the boards? On the FRONT SIDE of each board you'll find a white printed rectangle, within the creator's callsign "M0NKA" and the board version "rev 0.6.3".
After soldering all additional parts your UI board back side should look like the following picture.
This is what you soldered successfully:
- Resistor R8 (SMD 0805 680 Ohm). You soldered this resistor first. Remember: only if you want to use an electret microphone and NOT a dynamic microphone!
- Three 3,5 mm audio jack female sockets J1, J2, J3 (multicomp MX-387GL). These plugs are used for "Line OUT", "Line IN" and "MIC".
- A 3,5 mm audio jack female socket J4 (Lumberg KLBR 4). This plug is used for "PHONE".
- An USB-A female socket J10 (Lumberg 2410 02). This plug is used for USB "HOST".
- A 2-pin female header socket strip 2,54 mm P7. This plug is used as a connection to the RF board, where it has no electrical contacts. The speaker cable will be soldered to the 2 pins of P7 on the back side of the RF board later. HINT: P7 is not easy to solder. Check each connection!
- A 30-pin female header socket strip 2,54 mm P1. This plug is used as the main connection between the UI and RF boards. HINT: P1 is not easy to solder. Check each connection!
So far, so easy. But we have to talk about the two headers. If you read forum threads and this Wiki about P1 and P7, you will find different opinions:
- P1: There is a recommendation to solder a MALE header socket strip to the UI board (and a FEMALE header socket to the RF board). This recommendation is named "AG-xx-N-024" in the Recommended Modifications section in this Wiki. My description here is based on Chris' recommendation (he packs the FEMALE header for P1 in a small plastic bag for use on the UI board).
- P7: The same as for P1 applies to P7. There is no certain recommendation for this header, but it makes sense to solder it the same way as P1.
So you have to decide what to do with P1 and P7 for YOUR mcHF.
If you have populated the above parts, you'll still see some empty soldering footprints on the back side of the UI board (R4d, R43b, R44, C32b, C95, C98a, RFC3, U5, U8, P8). These parts are not used and stay unpopulated.
Let's have a short look at P8: This 2x2-pin header in the middle of both boards seems to make no sense, because it has no electrical connections. But,- you could use P8 for two purposes:
- As another stabilizing "distance holder" between both boards (like P1 and P7).
- For modifications: if you need up to 4 additional electrical connections between UI and RF board, you could lead them over P8.
The jumper P6 is also shown unpopulated on the above picture, but you will need it (once) for installing the UHSDR bootloader with the "P6-Jumper Method" described here. It is possible to solder an angled 2-pin male header socket strip 2,54 mm for P6 and to short-circuit the 2 pins temporarily for the bootloader installation. EA8ARX recommended here to make P6 accessible without separating the UI and RF boards by leading it to the lower edge of the UI board. This recommendation is called "UI-04-N-020" in the Recommended Modifications section in this Wiki. You have the choice ...
HINT: The height between the UI board back side and the metal shield plate is 6 mm. A "normal" angled 2-pin male header is possibly too high. Solution: The pins of a straight 2-pin male header are bent to 90° directly above its black plastic strip. This self-made angled header nevertheless has a height of about 4 mm. So it is a good idea to isolate both pins e.g. with heat shrink tubing after use and before mounting the radio.
The finished UI board front side (still without LCD) should look like this picture.
These are the soldered parts:
- Three incremental encoders E1, E2, E4 (ALPS EC12E24)
- An incremental encoder with button E3 (ALPS EC12E2424407)
- 17 buttons S1..S17
- A green LED LD1
- A red LED LD2
- Two female precision headers 2mm 20-pin (for LCD1 socket)
Soldering the SMD buttons and encoders is easy,- of course you don't have to use a type with button for E3 as shown in the picture, if you don't plan to install the ATU later. There are other possibilities to mount the LCD than with the shown female precision headers, have a look at this section!
The LEDs have to be mounted in a certain height above the UI board. As a rule of thumb there should be 7 mm between the UI board and the LED's plastic body. The best way I found to mount the LEDs fitting into their holes in the enclosure is to mark the 7 mm on the LEDs leads, solder only ONE lead of each LED (longer lead is +) in the marked distance to the UI board, cut the LED leads to 3 mm on the UI board back side, mount both boards with metal shield plate into the enclosure as if they were ready to use (with all screws and nuts). Then close the enclosure's front plate and bend the LEDs to fit in their holes. HINT: While you're at it measure the distance between the mounted UI board and the enclosure's front plate. This will be useful later to mount the LCD. I measured 9,5 mm here. Now unmount the UI board (and all other parts) without touching the LEDs and solder their 2nd leads.
Done (and by the way already learned to mount the whole radio)!
Plug P4 (MCU direct programming/debugging connector) stays unpopulated for now.
You'll find many forum threads dealing with mounting the mcHF LCD. The following picture shows the height of the two male header strips of the HY28B LCD.
So far there are three options to mount the LCD to the UI board,- two of them using a socket (recommendation named "UI-04/05/06-N-003" in the Recommended Modifications section in this Wiki).
Of course you may solder the LCD directly without socket. I cannot recommend this, because exchanging the LCD in case of a malfunction is quite difficult. But the direct soldering is the easiest way to mount the LCD. If you go this way, you should not solder the LCD with the black plastic stripes of the LCD headers lying directly on the UI board. If you mount it with a distance of about 1,5 or 2 mm between the black stripes and the UI board, you could eventually cut the pins one by one for removing a broken LCD. If you have a height of 9,5 mm above the UI board in your enclosure, this kind of mounting should be well possible.
HINT: Be sure how you want to use P4, P6 and P8 BEFORE you mount the LCD without socket!
The samtec CLT12001FD socket is a 2-row 20-pin socket with the advantage of a height of only 2,13 mm. As you see on the above picture, the LCD's height without pins is about 7,5 mm, so LCD and socket together have a height of 9,6 mm, fitting very good. Because the CLT12001FD sockets are 2-row, you have to pull out all contacts of the 2nd rows OR you can carefully cut the sockets lengthwise and use only one half. If you solder them to the UI board, their "empty" rows (or removed halves) have to be positioned to the same (e.g. to the right) side. The LCD's header pins have to be shortened to fit completely into the socket. If you have a height in your enclosure of 9,5 mm for the display, option 2 is the best choice for installing the LCD on a socket.
The W+P Straight Female Precision Headers (1-row 20-pin) are another possibility to mount the mcHF LCD. Their height is 4,2 mm. I filed down the headers a bit to about 4,0 mm. If you look at the above picture of the UI board front side, you'll see these headers. The LCD cannot be mounted "as is" with these headers. But it is possible, to remove the black plastic stripes of the LCD headers. If you heat these stripes from both sides with hot air, they can be pulled off easily. The LCD should be protected with tinfoil against the heat. Then the LCD's height without pins and the plastic stripes is about 5,5 mm, so LCD and socket together have a height of 9,5 mm, fitting perfectly. The LCD's header pins have to be shortened to fit completely into the socket. If you have a height in your enclosure of 9,5 mm for the display, option 3 is the 2nd best choice for installing the LCD on a socket ("2nd best" only because a bit more work is needed to get things done).
HINT 1: For options 2 and 3 you have to shorten the LCD's header pins to fit into the socket. After the shortening sometimes the LCD cannot be pushed into the socket. If you look at the LCD header pins on the above picture, you'll see them a bit sharpened at the top. If you file the shortened pins the same way, the LCD will fit into the socket easily. In addition it can be helpful for the precision headers to push a single wire (about 0,3 mm diameter) once into each female contact. Be very careful to clean the LCD from the swarf after filing, especially between the header pins!
HINT 2: Here I often mentioned 9,5 mm as the height between UI board and the enclosure's front panel. This measured height can only be reached with the newer UI and RF boards delivered, which are 1,0 mm thick. Earlier boards were thicker (about 1,5 mm). With those boards you will probably measure the maximum height for the LCD in the enclosure as 8,0..8,5 mm. In this case the above options 2 and 3 will perhaps not function as described. What to do? It is possible to use option 3 and file the female precision headers down to about 3,2 mm (in total minus 1,0 mm). Then the LCD with socket should also fit into your enclosure. The mcHF Build Video shows this method beginning at 1h 52m.
The mcHF Build Video also shows the complete soldering of the UI board, beginning at 1h 20m.
Now it's a good idea to test the finished UI board stand-alone and to install the actual bootloader and firmware!
It is possible to install bootloader and firmware with just the UI board finished. Please note, these instructions do not apply to the mcHF 0.7 hardware (they apply to all hardware featuring the original 30pin connector between UI and RF board). The mcHF uses 3 PCB and has different connectors.
Approach 1 - For all board revisions:
You will need a reliable source for 5V DC.
- Connect the GND to Pin 30 and 5V to Pin 28 of the 30 pin connector P1.
- Double check you did 30 and 28, and not 1 and 3. Triple check. Remember, if you look on the back PCB side, Pin 30 and 28 are on the right side, if you look from the front side, the pins are left, always assuming the P1 connector is on the top side of the PCB.
- Power on. Check if the connections are stable. Backlight of LCD goes on.
Approach 2 - For board revision 0.6:
You may populate R43b (or it is populated). This connects the 5V from the Mini-USB (J11) to the internal 3,3V regulator and allows to run the board from USB power. That works fine for a standalone UI board test. Remove this resistor before using the UI board and RF board together. Not all PC USB ports will be happy if they receive the 5V from the USB device!
You should see a few milliamps without display, with 2,8" display around 60 to 70mA. These numbers are only valid if no bootloader or firmware is installed and running. In this case you will measure up to 240mA.
Just follow the instructions for first time bootloader installation given here, then install the firmware using your preferred method. See Firmware Upgrades.
Please be aware that due to missing power supply the EEPROM and also the Audio Codec will not be detected, if you just connect 5V. This does not indicate an error, these components get their 3,3V from the RF board. If you don't believe this, connect a 3,3V power source to pin 27. Now the Codec and EEPROM should be recognized during startup.
Because the RF board is not connected, the (normal) startup error messages will be "Local Oscillator NOT detected!", "MCP9801 Temp Sensor NOT detected!" and "SWR Bridge resistor mod NOT completed!".
Let's put the finished and tested UI board aside and start with the RF board!
Let's divide the work on the back side of the RF board into three parts.
The RF board back side (completed Part 1) should look like this picture.
These are the soldered parts of Part 1:
- An I2C temperature sensor IC U10 (SMD MCP9801 SOIC-8)
- A DC female coaxial power plug 5,5 x 2,1 mm J1 (SMD FC68148S). This plug is used for "9-15V DC IN".
- Two 3,5 mm audio jack female sockets J2, J3 (multicomp MX-387GL). These plugs are used for "KEY" and "ACC".
- A coaxial BNC antenna plug P2 (BNC 13-60-2 DGZ). This plug is used for "ANT".
- A balun 1:4 1-500 MHz T1 (SMD CX2074NL)
- A transformer 4:2 T5 (4:1-1T on BN-43-2402 ferrite core)
- A choke 1,4 uH RFC8 (1T on BN-43-2402 ferrite core)
- A transformer 1:1 T6 (2:2T bifilar on BN-43-202 ferrite core)
- A transformer 2:3 T7 (2:3T on BN-43-202 ferrite core)
Since December 2018 U10 (MCP9801) is an optional part of the mcHF kit and the mcHF will work without U10. The task of U10 is to measure the temperature of U8, the local oszillator (LO). The temperature can be used to calculate a temperature-dependant correction value for the LO frequency in order to stabilize it, see this test. This calculation is supported by the firmware, so adding U10 is a good choice. Normally U10 will be detected automatically by the firmware. If not, the LO correction function can be activated in the radio's "Standard Menu" with "TCXO On". The temperature (default: °C) will be shown in the upper left corner of the LCD. If U10 is not populated or doesn't work properly, you will see "SENSOR!" in red in this area. U10 will work best with a thermal coupling to U8. This can be done with a thin piece of copper or aluminium as shown in the 1st post here. Another possibility would be to mount U10 directly on top of U8, see here.
Soldering the 4 plugs J1, J2, J3, P2 should be no problem. Your mcHF kit might contain a female SMA plug instead of the BNC plug P2 shown on the above picture.
T1 (CX2074NL) is a "ready to use" SMD balun (or transformer) 1:4 in the RX mixer section, which is easy to solder.
The winding of T5 is described in chapter "Winding of T5 (rev 0.5 and up)" in the "transformers_winding.pdf" document. T5 has 4 turns on the primary and 2 turns with center tap on the secondary side, wound with enameled copper wire (CuL) 0,315 mm on a BN-43-2402 ferrite core. T5 is the central transformer of the PA driver section. Modifying this transformer T5 could possibly give more TX power. A recommendation for such a mod is named "RF-04-N-015" in the Recommended Modifications section in this Wiki. This mod (originally for the version 0.4 mcHF boards) implements another winding of T5 (5:2-2T with CuL 0,3 mm).
The choke RFC8 is a part of the PA section and has one turn of enameled copper wire (CuL) 0,450 mm on a BN-43-2402 ferrite core, if you look at page 9 of the RF board schematic. The "LPF Torroid winding guide.pdf" document mentions 3-4 turns for RFC8. So with the "standard" PA design anything between 1 and 4 turns should be ok for RFC8. This choke together with C104, C105, C106, C106a is used for DC power supply filtering and RF rejection and there are some mods for it: The suggestions reach from "1 turn" to "as many turns as possible". User "Gab" reported here, that compared to a RFC8 with 1 turn a RFC8 with 3 turns resulted in 0,1 to 0,8 W more TX power. I think there is no certain proof, that RFC8 mods alone (without other PA mods) are useful.
The winding of T6 is described in chapter "Winding of T6 (original mcHF transformer)" in the "transformers_winding.pdf" document. T6 has 2 turns on the primary and 2 turns on the secondary side, wound bifilar with enameled copper wire (CuL) 0,450 mm on a BN-43-202 ferrite core.
The winding of T7 is described in chapter "Winding of T7 (original mcHF transformer)" in the "transformers_winding.pdf" document. T7 has 2 turns on the primary and 3 turns on the secondary side, wound with enameled copper wire (CuL) 0,450 mm on a BN-43-202 ferrite core.
Both transformers T6 and T7 are belonging together and are important parts of the power amplifier (PA) section. Many mods of these transformers were suggested and published over the years (for different mcHF board versions!) in order to reach more TX power and/or to improve signal quality.
If we discuss the options to modify the mcHF PA, we must first clarify what we are talking about. The following picture shows 3 schematic versions of the PA.
The left part of the picture "PA Schematic" is copied from the original version 0.6.3 schematic. This is a "principle schematic" obviously used to route the RF board PCB with T6 and T7 with primary center tap in order to make PA mods easier on this PCB. In the middle of the picture "PA Standard" you see the standard PA schematic with RFC8, T6, T7 realized as described above. This is the fully functional PA you get when you follow this article and Chris' description of the original mcHF. The right part of the picture "PA Symmetrical" is also a "principle schematic" showing the main modification: The PA is built as a symmetrical push-pull version with DC power supply at the center tap of the output transformer's primary side. For this version T6 is no longer needed.
Some links to "PA Symmetrical" modifications:
- IZ6MAF_RF_TX_-_SWR_notes.pdf. Paolo (IZ6MAF) describes his PA mod on pages 3, 4 and 5.
- mcHF_board_modifications_20150825.pdf. Clint (KA7OEI) shows on pages 7, 8 and 9, how Marko (S54AM) made this PA mod.
- T7_Measurements_M1BXF_030116.pdf. Gavin (M1BXF) compares the two above mods of Paolo and Clint.
- Here Karl (DK4ARL) shows his PA mod variant.
- ...
HINT 1: The TX mixer (with T4) and PA driver (with T5) are also parts of the TX chain. Their possible mods will also have an influence on the final PA!
HINT 2: My recommendation is to build all transformers first at all as suggested! If your mcHF is running, adjusted and tested with all functions, you can decide later to make mods, if you are not satisfied with certain aspects. Then every future mod has to prove, whether it implements these aspects better than the "stock mcHF" without worsening other properties of the radio.
The RF board back side (completed Parts 1 and 2) should look like this picture.
These are the soldered parts of Part 2:
- 12 inductors L13..L24 (0,315 mm CuL wire on T37-2 ferrite toroids)
- Two transformers T2, T3 (0,45 mm CuL and RG-316 coax wire on FT-50-43 ferrite toroids)
- A relais K1 (SMD G6KU-2FY_3DC)
The neon lamp DS1 will be soldered later on the RF board front side.
The 12 inductors L13..L24 are parts of the lowpass filters (LPF). There are four LPFs for the 80m, 40m, 20/30m and 10..17m bands. The inductors (3 for each LPF) have to be wound on T37-2 toroids with CuL 0,315 mm. You'll find the number of turns for each inductor on page 5 of the RF board schematic.
L13, L17 and L21 are the inductors for the 80m LPF.
- L13: 2,4 uH - 24T
- L17: 3,0 uH - 27T
- L21: 2,4 uH - 24T
L14, L18 and L22 are the inductors for the 40m LPF.
- L14: 1,4 uH - 19T
- L18: 1,7 uH - 21T
- L22: 1,4 uH - 19T
L15, L19 and L23 are the inductors for the 20/30m LPF.
- L15: 0,58 uH - 12T
- L19: 0,78 uH - 14T
- L23: 0,58 uH - 12T
L16, L20 and L24 are the inductors for the 10..17m LPF.
- L16: 0,26 uH - 8T
- L20: 0,32 uH - 9T
- L24: 0,26 uH - 8T
Now let's start to wind all 12 inductors! Please watch the mentioned mcHF Build Video first. It shows the toroid winding, beginning at 0h 09:44m.
HINT 1: The number of turns of a toroid are all turns to be counted at the INSIDE of the toroid!
HINT 2: Use the number of turns of the above list, NOT the number of turns mentioned in the video!
Many tests and suggestions to improve the LPFs (and also the BPFs) have been made over the years. You'll find a lot of them in the mcHF Groups.io file section, e.g. in the IZ6MAF directory but also in other directories.
If you have the equipment and want to measure the LPFs' function after winding the toroids, then DON'T solder the toroids definitely to the RF board and follow the next section Measuring the LPFs! Else you can solder the toroids as shown in the video and may skip the next section.
First of all: You don't have to test the properties of the LPFs. You can wind all toroids with the turns Chris defined in his documents, e.g. the schematic. Without any modification I found 2 LPFs completely as intended and the other 2 with only slight deviations.
But if you want to be sure, that the LPFs (with your self-wound toroids) are fine, a good way to measure RF filters is to use a VNA (Vektor Network Analyzer). A simple and low-priced VNA is the NanoVNA. Of course: if you own a professional, expensive VNA, you may use it as well. ;-)
I will not write a "how to use a VNA" here, but learning to measure with the NanoVNA is quite easy and a help video can be found on the net.
The picture below shows the setup of the measurements with NanoVNA, example: 80m band LPF.
These are the steps to measure the 80m band LPF:
- Calibrate the NanoVNA for a sweep range of 1 MHz to 10 MHz.
- Solder the NanoVNA measuring wires CH0, CH1 and GND (as short as possible) to the positions for the 80m band on the RF board front side as shown on the lower middle and right side of the picture.
- Solder the toroids L13, L17, L21 to their positions (leads as short as possible) as shown on the left side of the picture.
- Start the S21 measurement.
- Identify the -3dB value on the S21 lowpass curve and read the frequency of this "cutoff point".
- For the 80m LPF the cutoff frequency should be 5 MHz, see the S21 printout (derived from the NanoVNASaver program) on the upper right side of the picture (green marker).
So all LPFs can be measured the same way with these parameters:
- 80m: sweep range 1 MHz to 10 MHz, desired cutoff frequency 5 MHz
- 40m: sweep range 1 MHz to 15 MHz, desired cutoff frequency 8 MHz
- 20/30m: sweep range 10 MHz to 30 MHz, desired cutoff frequency 16 MHz
- 10..17m: sweep range 10 MHz to 70 MHz, desired cutoff frequency 33 MHz
Now what to do, if the measured cutoff frequency is far off the desired cutoff frequency? As I measured my 80m LPF for the first time, the cutoff frequency was about 4,8 MHz (that's not far off). I removed 1 turn of each toroid (L13: 23T, L17: 26T, L21: 23T,- you can count these number of turns on the picture above!) and repeated the measurement. Now the desired cutoff frequency could be reached.
HINT: The LPFs aren't critical and the mcHF will also work with an 80m LPF cutoff frequency of 4,8 MHz.
So if the cutoff frequency is way off, the first thing you can do is to spread or condense the turns on your toroids. If you cannot reach the desired cutoff frequency with this method, you can try to reduce the turns of all 3 toroids by 1 turn, if the measured cutoff freqency was too low. Then repeat the measurement with all turns of each toroid spread again over about 2/3 of the toroid's diameter (default!).
HINT 1: If the measured cutoff frequency is still too low, you can reduce the turns of the 1st and 3rd toroid (80m LPF: L13 and L21) again by 1 turn.
HINT 2: The 1st and 3rd toroid of each LPF always must have the same number of turns!
HINT 3: If the measured cutoff frequency is too high (very rarely), you can of course also increase the turns of the toroids.
This way you should be able to reach the desired value for each LPF. Then solder all toroids to the RF board.
The two transformers T2 and T3 are the main parts of the SWR/PWR Meter. They have to be wound on FT-50-43 ferrite toroids with 0,45 mm CuL and RG-316 coax wire. First have a look at pages 10, 11 of the "mcHF_board_modifications_20150928.pdf" document. The picture on page 10 "Winding 'sense' details of T2 and T3" shows, that T2 and T3 both have 10 "secondary" turns of CuL 0,45 mm, BUT T3 is wound clockwise and T2 counterclockwise. So take care that you wind T2 and T3 exactly as shown on that picture! The "primary" side of both toroids consists of only 1 turn of RG-316 coax wire as shown on page 11 of the mentioned document. Other than shown on the pictures of page 11, I would mount T2 and T3 LYING on the RF board (like L13..L24), because only then the optional ATU will fit into the enclosure below the RF board later. By some users a possible design error of the SWR/PWR Meter, keyword Stockton PWR Meter, was discussed. They recommended a mod named "RF-04/05/06-N-023" in the Recommended Modifications section in this Wiki. Also with the actual SWR/PWR Meter design the SWR readings (tested by TXing into a dummy load 50 Ohm) increase slightly with the frequency. This undesirable behaviour can be improved by replacing the two Schottky diodes D5 and D6 (PMEG6010CEH) with similar diodes with a lower capacitance (e.g. 1N5711W-7-F, BAT41Z).
Now we can solder the relais K1 (G6KU-2FY_3DC) to its position.
The finished RF board back side (completed Parts 1, 2 and 3) should look like this picture.
These are the soldered parts of Part 3:
- A voltage regulator U3 (LM2941CT)
- A voltage regulator U4 (MC7805BTG)
- Two MOSFET transistors Q5, Q6 (RD16HHF1)
The power supply of the mcHF consists of three voltage regulators U3, U4 and U5 (KA78RM33RTF). With these ICs the mcHF is supplied with four voltages: 12V, 8V, 5V and 3,3V,- have a look at the RF board schematic, page 3!
The 8V voltage is supplied by U3 (LM2941CT). This is a variable voltage regulator, which is adjusted to 8V with the voltage divider R11 and R12. U3 has to be fastened to the enclosure.
The 5V voltage is supplied by U4 (MC7805BTG). U4 also has to be fastened to the enclosure. It was discussed here whether the linear voltage regulator 7805 could be replaced by a DC/DC converter OKI-78SR-5/1.5-W36-C in order to reduce the supply current of the whole radio as well as the temperature of the enclosure.
The two MOSFET transistors Q5, Q6 (RD16HHF1) are the active parts of the Power Amplifier (PA). They have to be fastened to the enclosure as well. If you quite accidentally have more than two RD16HHF1 transistors lying around, then you could "pair" two of them for the use in the mcHF. Though they are the same transistors, their properties are varying widely. This may become a problem for the mcHF PA, because both final transistors are driven identically. So if Q5 and Q6 would have very similar properties, the PA would become more efficient. In this thread the possibilities were discussed, how these transistors can be paired.
HINT: There is no stringent necessity to pair the final MOSFETs. The two ones which came with your mcHF kit are certainly ok.
The mcHF Build Video shows the mounting of the voltage regulators U3, U4 and the final MOSFET transistors Q5, Q6 better than I could describe it, beginning at 1h 10:22m. After soldering U3, U4, Q5, Q6 you can again unmount the RF board and start working on the front side.
HINT: It is also possible to solder U3, U4, Q5, Q6 not until finishing the RF board front side, because they are a bit bulky. It's up to you!
The finished RF board front side should look like this picture.
These are the soldered parts:
- A neon lamp DS1 (16/30HB)
- A balun 1:4 1-500 MHz T4 (SMD CX2074NL)
- Three relais K2, K3, K4 (SMD G6KU-2FY_3DC)
- A 2-pin male header socket strip 2,54 mm P7. This plug is used as the speaker connection to the UI board.
- A 30-pin male header socket strip 2,54 mm P1. This plug is used as the main connection between the RF and UI boards.
The 2x2-pin header P8 stays unpopulated, if you don't want to use it.
There is not much to do on this side. The mcHF Build Video shows the soldering of DS1, beginning at 0h 8:48m.
HINT: There is no silkscreen print for DS1 on this side of the PCB. So watch the video and also see the above picture for soldering the neon lamp to the right position. The lamp's leads cannot be soldered "through hole", because the BNC plug P2 obstructs these holes on the RF board back side.
T4 (CX2074NL) is a "ready to use" SMD balun (or transformer) 1:4 in the TX mixer section. Paolo (IZ6MAF) suggested a modification of the TX mixer including T4 in order to reach a better carrier suppression.
After you soldered P7, it is the right time to solder the speaker cable (delivered with the mcHF enclosure) to the two pins of P7 (RF board back side!!) and to the loudspeaker and mount the speaker in the enclosure. The mcHF Build Video shows this, beginning at 1h 14:32m.
If you have come here, the soldering of all parts of the mcHF kit has been done. Congratulation! It's time now for the first Testing and Adjusting.
If you just finished building your new mcHF, the first testing and adjusting is done best without having the PCBs mounted in the enclosure! Be sure to use a good external heatsink for U3, U4, Q5, Q6 as long as they aren't fixated in the mcHF's enclosure!
The mcHF Build Video shows the complete mounting of the radio, beginning at 1h 32:36m, including the works on the metal shield plate and the LCD, as described above.
HINT 1: The three smaller knobs for the incremental encoders E1, E2, E3 are possibly too high (16 mm) to fit the encoder axis. I had to shorten mine by 3 mm.
HINT 2: Before mounting the RF board be sure to apply thermal compound (e.g. ARCTIC MX-4) to the back side of U3, U4, Q5, Q6!
A handheld mike, PTT, speaker isn't a component of the mcHF kit. If you have a good mike/PTT, then you can build an adapter to connect it with the mcHF. If you want to buy a low-cost mike/PTT only for the mcHF, a choice would be the BAOFENG Emote built for the BAOFENG UV-5R transceivers. It is a mass product and quite cheap. The quality isn't great, but can be improved.
HINT: For the BAOFENG Emote with its electret microphone the resistor R8 has to be populated on the UI board back side.
The following picture shows the connections.
You should cut off the 2-pin jack of the BAOFENG Emote and solder the red (MIC), black (PTT) and white (GND) wires (colours may vary!) to a 3,5 mm male audio jack for "MIC" and PTT, like shown in the picture. If you also want to use the speaker, you could split the BAOFENG Emote's cable and solder the green (SP) and white (GND) wires to a second 3,5 mm male audio jack for "PHONE". Done!
HINT: The sound quality of the electret microphone in the BAOFENG Emote can be improved by enlarging the microphone hole in the housing to about 1,5..2 mm. In addition you should build a small tube (length 5-6 mm) of duct tape around the microphone in order to acoustically isolate it a bit from the internal space.
For testing, adjusting and repairs it is necessary to connect the UI and RF boards with the P1 header without being stacked on top of each other. A good possibility to do this is to build an adapter as shown in this picture.
It consists of a 30-pin male and female header socket strip 2,54 mm and a strip board 2,54 mm with min. 30 parallel copper stripes. For installing bootloader and firmware it can be necessary to press and hold the POWER button (S17) constantly. Because this may be buggy, you can add a "PWR S17" jumper like shown in the picture. The jumper connects pin 19 of P1 to GND. With this jumper you can short-circuit the POWER button as needed for the firmware and bootloader upgrade with the DFU method. Instead of "4. Press and hold Power" in the "Starting The TRX in DFU mode" installation instructions you simply close the "PWR S17" jumper and open it after successful flashing.
HINT: If you also want to use the speaker, you will have to connect it to P7 on the UI board or build a two-core adapter cable between the P7 headers of the UI and RF board!
Now we are ready for the first startup. If you don't have already run just the UI board with success, then you should do it now. Connect only the UI board to the P1-adapter and follow the instructions here to launch the UI board and install the actual bootloader and firmware. The supply current (@ 5V DC) should read about 210mA to 230mA.
If everything is fine with the UI board at first glance, remove the 5V power supply and also attach the RF board to the P1-adapter. Connect a 12V DC power supply to J1 "9-15V DC IN" on the RF board. REMEMBER: external heatsink necessary for U3, U4, Q5, Q6! Press the POWER button (S17) until the LCD comes up. The only startup error message should be "PA Bias is 0, TX not possible". The supply current (@ 12V DC) should be about 400mA to 420mA. Switched off, the mcHF draws a current of about 3,4mA.
We have to make some essential settings in the menu, before the mcHF can be used. You'll find informations about the menu itself (structure, handling...) in the Operating Manual Menu!
The Adjustment and Configuration Manual describes the 9 necessary steps. Follow ALL OF THEM carefully!
HINT: For TX settings/adjustments you have to use a 50 Ohm dummy load (> 10W) connected to the BNC "ANT" plug P2!!!
If you plan to adjust all details of the mcHF, that can be adjusted, you will need more measuring equipment than a multimeter. What could be useful:
- Multimeter (current up to 3A, resolution 10mA; voltage up to 40V, resolution 10mV)
- HF Power Meter (up to 25W)
- SWR Meter
- Oscilloscope
- VNA
- Additional SSB Receiver
- Spectrum Analyzer
- Frequency Counter
- Reference Frequency Signal Generator
Of course you will use this equipment, if it is already in your shack. But if not (yet): Here are a few simple advice to adjust at least the essential parameters without most parts of the above measuring equipment.
Measuring HF power is necessary for a proper adjustment of the mcHF. If you have no HF power meter and possibly also no oscilloscope, you can simply use your multimeter to measure HF power. The following picture shows the setup.
A silicon diode 1N4148 and a capacitor 100nF in series are connected parallel to the 50 Ohm dummy load used for the TX tests. A multimeter measures the DC voltage at the capacitor 100nF.
The multimeter in the picture shows a voltage of 21,6V DC. If you look at the second row of the small embedded table, you'll find this value in the 1st column. You have to add the forward voltage (Vf) of the diode (0,7V) to this value and get about 22,3V (second column). This is the rectified peak voltage of the generated HF (@ 50 Ohm). From this peak voltage the RMS voltage 15,8V (3rd column) and the HF power can be calculated. In this example, a HF power of 5W (last column) is generated.
With this simple method you are able to adjust the mcHF's RF power and to calibrate the power coupling of its SWR/Power Meter.
HINT: The HF power is dependent on the temperature (mostly of the PA) and on the power supply voltage. So it is necessary to adjust the HF power at operating temperature and with the power supply voltage you will normally use.
The frequency calibration of your mcHF is of course possible with a precise frequency counter or a reference frequency signal generator. What to do if you don't have such a precise measuring equipment?
There are two options:
- You can use the RX calibration approach. No additional hardware is needed for this option.
- You can use a modified TX calibration approach with additional hardware.
I will describe a possibility to use the 2nd option. You will need an SDR receiver capable of receiving also on SW. There are cheap so called "RTL-SDR" receivers with R820T tuner like the Nooelec NESDR SMArt v4 on the market. They are able to receive from 25 MHz upwards, so adjusting in the 10m band would be possible.
With another good SDR receiver, the Airspy HF+ Discovery, you can receive the whole SW spectrum, so we choose it for this example. The following picture shows 4 screenshots of the SDRSharp software (also called "SDR#", download here) running with the Airspy (HF+ Discovery).
The Airspy (like the NESDR SMArt) receiver has a precise built in 0,5ppm clock, this is 20 times better than the TCXO of the mcHF with (about) 10ppm. So you could already adjust your mcHF's frequency better than 1ppm by simply transmitting a carrier and calibrating the frequency with the help of the Airspy.
But the Airspy can be calibrated too. The upper left screenshot shows the Airspy receiving a DVOR/DME station at 115,2 MHz (DLE) known for a very accurate frequency. The Airspy was calibrated before with this VOR (deviation was only 0,16ppm!). Of course you can use other reference radio stations, DAB+ or e.g. GSM base stations to calibrate your RTL-SDR or Airspy receiver.
On the upper right screenshot the uncalibrated mcHF ("Freq. Calibrate" setting: 0ppm) transmits (FM-N) on 29,7 MHz. You can see, that the Airspy receives this FM spectrum on a lower frequency. On the lower left screenshot the Airspy was tuned to the mcHF's signal at 29.699.813 Hz. This means: the mcHF transmits 187 Hz (6,3ppm) too low.
On the last (lower right) screenshot you see the transmission of the calibrated mcHF ("Freq. Calibrate" setting: -6,3ppm) at 29.700.000 Hz. So it is possible to calibrate the mcHF's frequency without an expensive measuring equipment,- and: to have an SDR receiver in the shack is useful anyway!
HINT: The frequency is dependent on the temperature, mainly of the LO (U8). So it is necessary to adjust the frequency at operating temperature and with installed temperature sensor U10.
For the TX IQ adjustments you will need a second SSB receiver for the 80m and 10m bands or a spectrum analyzer for this frequency range. If you own an Airspy HF+ Discovery (we used this SDR receiver in the Frequency section above!), then you already have both: a "second SSB receiver" and also a simple "spectrum analyzer". The following picture shows two possible settings to graph the spectrum.
The mcHF transmits (FM-N) at 29,7 MHz (as an example). On the left screenshot the spectrum is shown with the SDRSharp (SDR#) program running with the Airspy. The bandwidth is reduced as far as possible. So the FM spectrum can be evaluated quite well.
The second setting is realized with the N1MM-Logger+ program, a known contest logger, and the Airspy. The right screenshot shows the "Spectrum Display Window" of this program. The mcHF's FM spectrum can be observed in this window.
You have built, tested and calibrated your very own mcHF now. Mount it in its enclosure and you are ready for QRP operation!