A 300 Watt class D Amplifier for 630m (p2)
This page is a work in progress and presents additional information on the 630m amplifier project including a component list and some details on performance. If constructors come across any ‘features’ or make any suggestions for modifications, then those comments will be added to this page.
The component list for the amplifier is presented below. I have attempted to find sources for each component here in Europe, but also in the USA. The Digikey part numbers presented are a ‘best endeavour’ on my part so please check that each suggested part number matches the description before you order. – Should you discover any errors please let me know at g0mrf(at)aol.com
Having said that there are
not many parts that are critical to the design and so you may be able to find
suitable components around your shack / junk box.
Components I would suggest that should not be changed are:
1) The LEDs D8 and D9 used for indicating current and VSWR trips. – These are low current devices and normal 20mA LEDs will not work correctly.
2) The FET driver IC2
3) The CMOS ICs IC1 and IC4. These should be the HEF40xx series from NXP as they have been tested extensively for drive / triggering / and rise and fall times. With other manufacturers, performance may be subtly different.
1) R23 / R24 - Note fig.2 on main page shows these two resistors incorrectly fitted. (positions reversed)
2) 100nF decoupling capacitors changed to 220nF. - Only reason was to reduce the numbers of different values. No difference in performance.
3) The PCB has a small silk screen error. In the row of 5 connections near RV3, the pads marked CUR LMT and REF LMT are reversed. These are the +ve feeds to the LEDs that indicate the cause of a fault. If wired as shown the protection circuits work but the LED for indicating a VSWR trip will illuminate when the current trip has been triggered.
3) Update to circuit diagram. T1 Secondary now 16 turns instead of 15t. / Q4 now changed to read TR4.
On component list above. R23 (12k) and R24 (100k) now correctly listed and consistent with circuit diagram. Digikey part number corrected for TR4
Thanks to John WA3ETD (WG2XKA) and Steve KK7UV (WH2XNV) for these updates
Constructing the directional coupler
three images below show the construction of the directional coupler. Start by
twisting two lengths of 28 SWG enameled wire together. Wind 15 turns onto the
toroid core as shown in the fist image on the left and separate the windings
so you can identify the Start and Finish of each one.
Next, remove the enamel from S2 and F1 with a sharp knife and twist together as shown in the middle photo and solder.
Finally, rotate the core as shown in the final photo on the right and remove the enamel from the remaining wires. Note that S1 emerges from the middle of the core on the right, then runs across the core so it can be connected to the forward power connection (D3). Making sure the windings are correctly oriented is important as the protection circuits need to work on detected reverse power...not forward power. If you follow the method described, all will be fine. Before soldering the completed directional coupler to the board you need to prepare a short length of wire for the primary winding which passes through the middle of the core. I used a short piece of the inner from RG58 coax. Hold the coupler in place then pass the RG58 inner through it. When soldered to the PCB the RG58 inner (apx 4mm dia with its insulation) holds the coupler nicely in place. Then you can trim the coupler wires to size and solder to the 3 pads on the board. If you anticipate using your amplifier portable, you may want to use the smallest amount of epoxy glue to hold it in place to prevent vibration damage when traveling.
The FCC in the USA requires all harmonics to be reduced to a minimum of 55dB below the wanted signal. The image below demonstrates that the second harmonic is around 58dB down while the 3rd harmonic is 66dB below the wanted signal.
How well does it work on-air.
After a lot of bench testing, it's essential to try a new design on air. - It can reveal all sorts of EMC issues that you would never find if you limited testing to bench measurements alone. I took the amplifier along to my local amateur radio club and connected my QRP Labs WSPR Beacon. Picture below.
On the desk you can see the amplifier in the blue case. The U3S beacon generator with the coax on the left connecting to a remote GPS antenna. In the background, A switched mode 18 - 33V 12.5A supply. The red coil with many taps is the ATU and the small rectangular meter is a 0 - 5 Amp RF thermocouple ammeter which is in the feed to the antenna.
Running about 200W to a sloaping wire with the far end at 25m AGL produced the following list on WSPRnet.org. Not bad for early evening. Best DX from the list SV8RV at 2252km. Unfortunately there was no-one further South or further East to test the limits of the signal - and it was still daylight in the USA.
But occasionally, you get lucky. The image below right was taken in March 16 while developing the prototype.
HOME Kits 630m Amplifier