Digital RF Power Meter with load up to 100W

To run the transmission path of my TRX needed was a RF power meter I had so far no such meter (except borrowed), so itself decided to build such a measure.
Presented here is my version of the meter gauge, designed by OZ2CPU . By building this device was based also on the version described in Świat Radio No. 1/2007 published by SP2MBE.
Itself, for its part decided to make a simple built-in load resistor to 100W making a compact device with a complete instrument or test to run low-power HF transmitters (up to 100W). Due to the simplicity of this solution load should meet the expectations in the frequency range of HF. In practice, it turns out.

The plan of construction equipment

At the outset planned layout of the instrument. The meter fall was driven by a controller board with LCD, sensor board, power supply, transformer, attenuator input and 100W load. The whole was to be placed in a metal casing available in electronics stores. After gathering the basic elements of design have deployed them in the housing plan initially "more or less" at which point each of them should be. The exact principle of operation of such a meter and a chart posted on the websites OZ2CPU.
My version of the meter is to operate so that the transmitter signal is fed to the artificial load. Between the load and the transmitter is turned on muffler symmetric with access to the power meter. The input attenuator decided to apply the solution of the laminate plate placed in a piece of aluminum rail (heat sink). 100W load in my understanding was to be made ​​in the form of resistors placed in the same piece of aluminum rail at the input attenuator. With the deployment of pre-planned and the type used in electronics components before assembly decided to perform mechanical work. At the outset, however, I made the meter circuit board so that it could be dimensioned. PCB have done using the "iron" like for example the synthesis of the DDS board described on my website.


Cover

As I mentioned earlier as the case used the purchased metal casing. I took it disturbing horizontal slot of the fuse. After matching items first turned on the back cover and dimensioned in a hole in the antenna socket and power switch.
Then turned on the front cover. Matched the plate of the display, buttons, and a second signal connector. Since the transmitter PCB had to be soldered directly to the front signal, it had to adjust so that its location does not interfere with the display board. Instead, I decided to use 2 pulser UP and DOWN buttons. Even at this stage did not know where to place it, but in the later assembly of the best place was left of the faceplate. After sizing the front wall started to cut holes.
After excision of the holes in the front and rear plates screwed pre-screen and control buttons on the meter.

Input Attenuator

Implementation started with the input attenuator cut to size on one piece of Copper Coated PCB laminate. Hot wire with a file I have graven keeping her distance from the edge of the copper about 2.7 mm. Then to the chassis section przykręciłem M3 screws. The protruding bolts nuts I made as a basis for the input attenuator PCB laminate. After matching the dimensions of the PCB ran it on the bolts laying around in the mid-height of the profile. The prepared skeleton silencer left to finish off for later. Placing silencer matched so that the input attenuator was close to the aerial socket. The output of the attenuator input provided was to turn the left side in the vicinity of the artificial load.

Dummy load

As an artificial load resistors 44 decided to use metal-2k2 / 2W each connected in parallel. Although the total power of such costs normally is about 88W, but subsequent experience has shown to successfully withstand load capacities greater than 100W. Since this is the amount of 44 pieces of parallel connected resistors, the value inductance should be large, since the parallel combination of inductance decreases. Capacity value of such a solution, fortunately, is not as impact on load as inductance. To minimize both phenomena applied metal resistors. Inductors resistors can be used, but presumably the cost of such solutions will increase. If my solution does not exceed the cost burden 40PLN.


Implementation of the load began to cut two pieces of double sided Copper Clad Laminate in the shape of rectangles. One piece designed as a weight I cut to size roughly corresponding to a vertical wall surface profile. A second similar, but less wide and eventually suspended for the resistors does not touch the edges aluminum profile, which is naturally connected with the mass of the device. Then, on an aluminum profile drew a grid indicating the drill holes for the resistors, and then, after drilling 44 holes I put them in those holes. Inserting resistors tried to be as evenly as possible and move close to the plate, but yes, to be able to be soldered on both sides.
Once all the resistors in three rows of solder them on both sides. I cut the free ends of the resistors in rows, each row of a different length to facilitate imposition of the second plate. After applying the second PCB resistors soldered to it. At the finale I cut off all the ends of the resistors. In two places the plate mass previously drilled holes for screws M3. When inserting a resistor put the cap near one of the holes to be able to Screw the plate to the wall of the aluminum profile.
When assembled, screwed load them to the chassis parallel to the back of the chassis at the same time aligned to the input attenuator.

Finish the input attenuator

Now it's time to finish the connection input and output attenuator. The right side of the plate silencer (input) was a piece of wire soldered 50R linking it to your aerial socket. I have tried to minimize the distance at the solder. I split the cable shield to the two ends to distribute the weight on both sides of the plate. Wire bar hot plate honed in on the ends to about 2.7 mm to ensure that the distance from the edge of the plate in contact with the walls of the input attenuator. In the middle of five brazed resistors connected in series 1K/2W each of the hot wires to the output of the power meter. At the other end was fastened a piece of the laminate, further linking it to both sides of the wire, silver plated with a mass of PCB input attenuator. Between ye and the mass of the brazed plate parallel two 100R resistors. Attenuator output cable connected the 50R with an artificial load. The first end of the cable was soldered to the center of the load as far as providing a steady current in the resistors Spreading. The other end of the input attenuator to the output in a similar way to the entrance.

Transmitter and controller board

PCB before soldering equipment painted rosin dissolved in methylated spirit. When dry, proceeded first to the mounting plate of the converter circuit AD8307. First, was soldered chip itself. Then, resistors and capacitors. Applied at the input resistors of SMD 1W. And so the resistor from the input is set to 33R, while the resistors soldered to ground resistor 39R and 33R one on the "back" of the other. At the end soldered 20uH inductor in the form of three turns with an inside diameter of 3 mm 0.5 mm wire DNE.

Controller PCB done so that everything, ie all the elements and paths were available on one side - from the top. In the case of execution paths across the subsequent soldering LCD controller's impossible to even start because of a short circuit, to which access would be impossible. As usual I used a quartz resonator quartz crystal makes every additional two 10pF capacitors each on its feet to ground quartz.
PIC16F876 chip itself programmed a simple programmer "ludipipo" using IC-PROG . The file to program the PIC16F876 system can be downloaded from OZ2CPU . When setting the timer turned the HS mode to properly choose the type of quartz. By default, the project foresees the use of ceramic oscillator. And I even have the usual quartz.
On the underside of the PCB milled holes for thru-hole components to prevent short circuits to ground. Before connecting anything to the controller board carefully checked the presence of faults and microcracks. So the assembled plate connected the controller with LCD display with pins.

The prepared PCBs screwed to the cabinet. LCD controller for long M3 screws locking nuts. Transmitter PCB soldered directly to the BNC connector in the housing unit. Before the two plates together connected the input voltage of 9V to the stabilizer plate 7805 in the controller. I measured voltage stabilizer and voltage 5V reference voltage is 2.5 V at pin 5 of the PIC. I turned off the power and put the system in the stand. I turned on the power. Unfortunately, the display showed a "rubbish". I turned off the power, took out the plate. I measured again the transition from PIC to display system. One of the tracks have a short to ground. I removed the circuit, I turned on the controller. This time it worked properly. So I combined the two plates together patch cable.
Note: In the draft board from the World Radio changed places A and B inputs of the PIC.

Again I turned on the power, measured the voltage at pin 5 of the value of AD8307 5V. I turned off the power. Buttons soldered to the connector K6. Instead of buttons on pulser applied to the left of the front plate. I included them in this way that one ends joined to the ground, and second respectively to the inputs of the UP and DOWN. Between the input capacitor hooked around 10uF/16V bipolar, which I did with two electrolytic capacitors connected pluses.

The power supply unit

Screwed to the housing unit 10V with PCB transformer rectifier diode. I used for this purpose a transformer from the AC adapter socket. By placing the transformer in the cabinet tried to put it away from the artificial load.

Running the load

The load was hooked to the input attenuator, and the aerial socket. At the launch of the load sensor disconnected from the output of the input attenuator. I measured the resistance of the load to be sure. It was exactly 50R. To the aerial socket plugged my IC-718 transceiver. I set the lowest power. I turned and pushed RTTY mode PTT. I increased power to about 20%. Indicator showed virtually radiated power. And so on I got to 100% power, the 100W. Load without any problems effectively took over the whole energy. Plugged in the antenna tuner. I repeated the test. As it turned out, the setting in automatic antenna tuner does not deviate much from the situation if it was not. To adjust the SWR did not have to make any adjustments. Once connected, without box, with full power of 100W SWR value displayed in the transceiver is less than 1 Thus, the load met my expectations.


Initial start-up and calibration of the meter

Before calibrating the input attenuator plugged into the meter. While running the meter is not calibrated generator has appropriate 0dB, whose use is highly recommended. Therefore, my calibration was limited to a simple set to close. A more accurate calibration at the time I will have a 0dB signal. I used to calibrate the AM my simple generator that gives the output voltage within the limits of 0-1.5 V. First, in the menu mode, set the meter controller LF and-50dB attenuation as my damper with damping resistance of 5K is 1:100. Then I gave the input signal of 3.5 MHz generator by setting the output voltage within the maximum. From the chosen calibration controller. Through trial and error skalibrowałem voltage up and down by turning the knob on the generator. The transceiver set the frequency of 3.7 MHz. After each calibration enclose the maximum power from the transmitter, knowing that it is a 100W I read the meter. Every time I made an adjustment of the generator to get a proper indication. In the end I managed to choose the correct value for 100W of power. At the end I left with a transmitter power in the transceiver for different values ​​down and checked the conviction on the measure. It was surprisingly comparable to the scale on the radio. I repeated the action mode frequencies in the HF and 14MHz radio. Finally, I turned the cover.