Each, who started dealing SO2R or multi operation with the possibility of working on several bands at the same time encountered the issue of receiver protection or the creation of intermodulation products. Why?
The specificity of multi - multi operation is the concentration of several devices in a small space, usually one building. A relatively high voltage is thus induced into the antenna from other bands, which affects the intermodulations in the receiver. When using large powers and an unfortunate coincidence, the input circuits of the receiver may be destroyed.
Mathematical calculation of transmission from one antenna to another antenna
That is, if we take the power of 1kW, so it is +60 dBm. Depending on the distance of the antennas, filming, polarization is reported, that the attenuation between the antennas is -20 to -35 dB. This can be measured quite nicely – send power to one antenna and load the other with an artificial load with a wattmeter (or VF voltmeter or oscilloscope). How much did you measure??
If we take the better value -35 dB, so a level signal arrives at the input of the second RX +60 dBm – 35 dB = +25dBm, which is roughly 0.3W.
The first circuits of the receiver are therefore loaded with a power of 0.3W each can lead not only to the formation of undesirable IMD products, but only to their damage.
The advantage is, that by, that it is a frequency distant signal, so it will be additionally suppressed in the input filters of the receiver (octave, band or preselector).
Frozen HF antennas
Mathematical calculation of the transmission from one antenna to another antenna of the harmonic frequencies produced by the final stage
The second factor, which causes problems are the harmonic products from the transmitter on the lower band. The TCVRs themselves have a fairly decent suppression of harmonic products (at least 50 up to 60dB), it is worse with the end stages.
Let's start again from the power of 1kW (it is easily converted to 100W or 10W). So we have a level +60 dBm. Depending on the quality of the output stage, the suppression of harmonic products can also be around -30 dB. What does the situation look like then?? The attenuation between the antennas is -20 to -35 dB.
+60 dBm – 30 dB – 35 dB = -5 dBm
That is a very strong signal, but it should have been below par, when quality RX starts producing IMD products. You can check from the previous article about S-meters, that it will be a signal of level S9+67 dB.
If, for example, I broadcast to 14,250 MHz, then on 28,500 MHz I would find such a very strong signal, which will certainly not allow me on this frequency (and in its surroundings) work.
Possible filter solutions for SO2R and multi operation
Band filters for the contest are the solution. By connecting such a filter, we narrow the bandwidth of the receiver, thereby eliminating signals from other bands and at the same time suppressing harmonic frequencies emitted by the transmitter.
Making LC filters especially for higher powers is problematic due to the unavailability of suitable capacitors.
For multi - multi workplace MW5W I once made filters from of coaxial stubs according to K2TR. Compared to LC filters, they work differently. They do not create a bandpass, but because they suppress only certain frequencies given by the type and dimensions of the coaxial stubs. By combining them appropriately, satisfactory filtering can be achieved for the desired band. Their advantage is feasibility and efficiency. The excellent results confirm it OK1VWK measurements
The quality of the coaxial cable significantly affects the resulting filter parameters. We prefer to avoid RG-58 - the suppression of unwanted frequencies is low and we also lose some of the power unnecessarily in the cable. For the performance of approx. 1kW PEP complies with RG-213, which is also good to work with. Suppression of unwanted frequency is sufficient and it is also an advantage, that higher band filters suppress not only higher harmonics, but by adding another stump, even lower bands. The resulting suppression is thus very large.
PL T-connectors can be used for outputs of the order of hundreds of watts. All filters should be made from coaxial cable from one manufacturer, preferably also from one drum. The shortening factor may differ slightly from the value declared by the manufacturer. It is advantageous to start with the production of a filter on 160 or 80m and measure the filter.
SO2R TCVR setup
How such a filter will affect the transmission situation from one antenna to another antenna?
Again, the power is 1kW, so +60 dBm and attenuation between antennas -35 dB. In addition, suppression of the filter -25 dB.
A level signal arrives at the input of the second RX +60 dBm – 35 dB -25 dB = 0 dBm
That is still above the S9+70 dB limit, but that's all 1 mW. Such a level should not lead to damage to the receiver or the formation of IMD products.
Example of filter production for the 80m band
We mount a connector on the end of the cable. We calculate the length of the quarter-wave stump according to the formula 75 x V / F (e.g.. 75 x 0.66 / 3.68MHz = 13.45m). We keep 20 cm as a reserve, so that we have something to shorten. So we measure 13.65m from the connector, cut the cable and short the end (we tin properly).
We connect the filter through PSV meter to the device and load it with an artificial load via the wattmeter. In the 80m band, the filter must not affect the PSV value and the wattmeter must show full power. If the device can also broadcast outside the amateur radio bands, when retuning to higher bands, we can see on the wattmeter how the power drops into the artificial load. We measure at reduced power, as PSV is very high (if we have a damping element for the given performance, I recommend using it). So we can find the frequency in the 40m band, which is maximally suppressed, which will also verify the baseband setting. If it differs from the required one, adjust the length of the stump and repeat the measurement. If it differs from the calculated length (13.45m, see. higher), after setting the stub, we will also calculate the shortening factor - we will make it easier to set filters for other bands. We then seal the end of the cable with insulating tape or stretch a heat-shrinkable tube.
The second option is to use a VNA or similar analyzer and measure the filter with it.
We can connect the filter to the final stage directly with a T-connector, or through a short one, about 1.5 m long coaxial cable. The stump can thus lie on the ground. Be sure to mark all the cables well - after making all the filters you will have a lot of cables. A very nice mechanical solution is to roll the cable into a coil and put it in a large paint can.
We used the filters in the contest operation MW5W during CQ WW SSB (six separate stations with PA for each band) this year. I registered only two or three frequencies in the 15m band on which there was a signal from another band, but only in S2 strength. However, the antennas were on a large area and the power limit is lower in the UK.
Filter dimensions for all bands, CW segments, shortening factor of the coaxial cable 0.66
A: suppresses the 20m band, length 3.486m, open end
B: suppresses the 40m and 15m bands, length 6.969m, open end
A: suppresses the 20m and 10m bands, length 6.969m, end shorted
A: suppresses the 40m and 15m bands, length 6.969m, open end
B: suppresses the 10m band, length 3.486m, end shorted
A: suppresses the 20m and 15m bands, length 6.969m, end shorted
B: suppresses the 15m band, length 4.648m, end shorted
C: compensates for the reactance of stub B, length 2.337m, open end
A: suppresses 40m, 20m, 15m and 10m band, length 13.938m, end shorted
A: suppresses 80m, 40m, 20m, 15m and 10m band, length 27.876m, end shorted
Dimensions of filters for bands 160m and 80m, SSB segmenty, shortening factor of the coaxial cable 0.66
A: suppresses 40m, 20m, 15m and 10m band, length 13,451m, end shorted
A: suppresses 80m, 40m, 20m, 15m and 10m band, length 26,470m, end shorted
There are still other ways to improve the resistance between bands, for example, by using a push-pull final stage, which naturally pairs harmoniously (2.harmonic, 4.harmonic,…) suppresses products. In addition, it is possible to use additional bandpass filters at the input of the receiver and the transmitter (do 100W), thereby achieving further improvement.
