On 2/24/15 9:25 AM, "José A. Amador Fundora" wrote:
This is very true. Audio bandpass management can improve listening fatigue and general clarity. It is best that the audio band filtering is before any substantial audio gain and the amplifiersIt is not a matter of matching, it is a matter of "filter termination".The response of an LC filter depends critically on its terminations. PSpice is a good tool to evaluate it by means of simulation. Real life may be a little more complicated, a model is better as it resembles better to reality, but a complex model is harder to nearly impossible to simulate.I have used Active Sallen Key low pass filters using gain ~ 1 (actually, 0.999...) using transistors, which does help to combat tirednedness by high frequency audio noise where there is no desired signal, as the crystal filter on the input end has removed them. Any audio bandwidth in excess of the filter bandwidth leads to that problem. I used an extra active audio filter in my old Swan 700 using a Chebycheff LPF with three sections of a TL084, and cutoff on 3.0 kHz. The fourth section was a high impedance input buffer. It did alleviate the HF noise quite a bit.All "simple" SSB radios are plagued by SSB demodulation of desired signals, and DSB demodulation of noise, worsened by AGC. A wideband high gain IF amplifier with a crystal filter at its input end may do it easily. A better solution is using a "simpler" crystal filter at its output end to stop DSB noise, or resorting to an image reject (other sideband noise) demodulator which is fairly more complex. SNR with those solutions may be better, which is what counts for digging out the weak ones, which often are other QRP stations.
after it are quiet.With various radios I've used the tail end crystal filter, audio low pass filter and with no front end change observed an improved MDS and better (S+N)/N. This improved listening fatigue as well. Since at VHF
one tends to do more listening audio quality can be an important factor.The use of a Image reject detector can greatly improve results. First by removing the opposing sideband there is an 3db improvement and then bandwidth shaping helps further. This generally requires a moderate amount for complexity as opposing sideband rejection of 30-40dB is adequate. A straight up image rejecting DC receiver (See KK7B, R2 series) one would want to be in the 40-50db range for acceptable results. I have a single conversion VHF IF 6M transceiver I built for contest conditions using image rejecting mixer at RF(6M), IF filtering for close in performance, and image reject detector to get a receiver that has a very good MDS
and dynamic range without AGC.
My advice is using an active LPF after the demodulator, is the simplest and fairly effective way. It has no termination issues and can be quite small. It is also quite inmune to stray fields.This works very well. There are many types of active filters that apply well.
Allison/Kb1GMX
73, Jose, CO2JA On 23/02/2015 15:26, Franz DX wrote:Hi Minima specialists, I am experimenting with different audio filters directly after the product detector of the Minima (like the three pole lowpass filter in Minima version 2 consisting of a 33mH choke and two capacitors) and different audio amps, but I am having problems with hiss noise. When I use no audio filter and connect the product detector directly into the audio preamp, the hiss is not existent (except for a tiny little bit). But when I use a filter like in Minima version2 with a serial 33mH choke and two capacitors of 100nF connecting to ground, the hiss is getting really loud and I am also observing a high pass effect, although it is constructed and meant like a lowpass! My suspicion is that it has to do with wrong impedance matching, having a low impedance 50ohm output out of the product detector and a high impedance input to the audio preamp. But how can I match these? Using an LC circuit like one at the input of Joe´s IF amp? I am convinced this is a really fundamental and easy to solve problem, so I would really appreciate a link or a hint where to read about a solution. Thanks in advance! Also, maybe the problem would be solved if I use an FST3253 as a switching product detector like Joe planned it? Best wishes, Frank DD4WH50 Aniversario de la Cujae. Inaugurada por Fidel el 2 de diciembre de 1964 http://cujae.edu.cu
