2kW Low Pass Filter for SW Amps

40m and 60m filter

Bill of material:

3 pcs. Amidon T200-6
3 pcs. enammelled copper wire (CuL) diameter 1.5mm, length 0.6m each.
2 pcs. 470 pF 2kV FKP1 capacitor
2 pcs. 680 pF 2kV FKP1 capacitor
2 pcs. 1.0 nF 2kV FKP1 capacitor
2 pcs. 2.2 nF 2kV FKP1 capacitor

L13 and L15:

Wind 10 turns on the core. The inductance will be about 1.28 uH. Check with an LC meter, the tolerance is 1.26 to 1.30 uH. If required, strech or compress the windings on the core.

L14:

Wind 11 turns on the core. The inductance will be about 1.5 uH. Check with an LC meter, the tolerance is 1.48 to 1.52 uH. If required, strech or compress the windings on the core.

C10, C13 ... 470 pF
C31, C34 ... 680 pF
C11, C12 ... 2,2 nF
C32, C33 ... 1 nF

using these values the simulated attenuation looks like this:

       

       

the red curve is the attenuation, the blue curve the reflection (5dB per hor. line). So the reflection is -30dB in the 40m band and -38dB in the 60m band. These are ideal values without any losses. Lets see what this filter really does on our PCB:

This is the attenuation in the 40m band, the passband attenuation with less than 0,1 dB is extremely good, so the filter will not get hot at high power. All the harmonics are damped > 40dB. Together with the harmonics attenuation of the amplifier itself, the results are almost perfect.

           

This is the attenuation in the 60m band. The passband attenuation is also very well. The attenuation at 11,2 MHz is only -18dB. Together with the attenuation of the amplifier itself, we will be better than -40dB which is ok.

           

Finally the reflection attenuation was measured. This is the SWR at the filter input (and output, since the filter is symmetrical) and shows how close the filter input is matched to 50 ohms.

The reflection attenuation is -35 and -33 dB. Converted into the SWR this is an SWR at the filter input of 1,04 : 1. If we transmit 1kW the filter input will reflect on 5.6 MHz or 7 MHz a power of only 0,4 watts back to the amplifier. All harmonics will be reflected by almost 100%. These values are very close to the simulation.

           

These measured values are almost perfect.

Results:

this picture shows the spectrum on a 50 ohms dummy load. The filter was driven by a Helitron.de LDMOS amplifier running at 1kW output power. A HP-attenuator was used to adjust the output to about 0 dB (The value at marker M1 is NOT the passband attenuation, its just the 0-reference for this graph). The readings for the harmonics are the filter attenuation. The legal limit is -40dB (BNetzA 33/2007) and in USA (FCC 97.307)  -43 dB. This requrement is fulfilled.

                   

I will provide the graph for 60m at a later time since I don't have a 60m transmitter available.