Testing different simple input band-pass HF filters (BPF), I decided to focus on very simple two-pole (two L-C resonant circuits) solutions based on ready-made coils and capacitive coupling. Even though filters with Amidon cores coils (with the transformer coupling) seem to perform slightly better (higher Q, better out-of-band attenuation etc.), these simple input filters I mean are quite good and very cheap. However, there are few issues which should be carefully taken into account when designing and assembling such filters, i.e.:
ready-made inductors (chokes) have parasitic parameters (serial resistance Rs, parallel capacitance Cp etc.) and tolerance of L (even up to 20%), which should be strongly taken into account,
the filter parameters (i.e., attenuation and bandwidth) are very, very sensitive to the values of all the applied components (inductors, capacitors) and so the application of the typical (popular) components will lead to unexpected and non-satisfying results.
Simply speaking, the employed capacitors should be carefully chosen and well adapted to the deviation of chokes’ inductances. In addition, the filter characteristics are very sensitive to the deviations of the connected antenna output impedance, so careful design and assembling of HF input filter is not enough to make it work well!!!
Keeping in mind all these things above, I’ve designed the circuit below (don’t bother with the LTSpice commands and two input voltage sources – just for the testing purposes):
The simple one-stage input amplifier is applied only in order to ensure the proper value of the output impedance (of 50Ω) for input of HF filter – just to keep its characteristics fine. The obtained transmittance (with small load of R2=1.5kΩ, typical for inputs of many popular HF chips like NE602/NE612 or MC3362C) looks like that:
One should keep in mind that obtaining such good (exact) characteristics will be quite tough to do using typical solid-state capacitors – even if trying to connect few of them in parallel – just to reach the values like 94pF = 82pF + 12pF (I did it successfuly in about 20 tries!!!) In addition, at least simple HF generator and oscilloscope (or HF voltmeter) should be used to busily scan the filter’s frequency response.
In my opinion, there exist only two good practical solutions for the problem of obtaining exact, good frequency responses of such HF input filters (beside using the proposed Z-matching input amplifier):
(a) usage of variable capacitors in place of C13 and C14,
(b) application of high-quality (low-tolerance) SMD components on a carefully designed PCB board.
I started my hard-way experiments with some kind of old-school „spatial installation” (obtaining quite good measurement results):
However, for this time I finished with option (a), having C12=6.8pF, C13=56pF + variable capacitor 10..50pF and the same for C14:
The frequency response I’ve reached (after several variable capacitors regulations) looked like that (with maximum gain of about 5 dB):
I think it’s not bad but it is time to think also about option (b) – with solid-state accurate SMD components, or about some simple technique of LC-circuits tunning.
To be continued soon …