Designing LPF for 30MHz

[D.A.(Tony)Stewart]

With this configuration you can make any Av, Zin, Zout you want. Although not perfect, it is very flexible.

Av can be almost anything but Power gain is dependant on small % of DC power input unless optimized for efficiency.

The Source and load are somewhat matched here for 50% drop in signal.


This can be inverted to NPN as well with AC coupling to output.

 

[volker@muehlhaus]

If you don't know the buzzword look it up, if you care about signal quality, then learn it. If you wish to master design, know it well.

With all due respect ... you are notorious about off-topic stories and not reading the OP's questions. Have you noticed that this thread is about low pass filter, and the only reason we discussed his amplifier was to get the output impedance, as required for filter calculation? Have you noticed that he already has an amplifier that works quite well, and that I have analyzed his amplifier to get the missing data?

And now you start jumping in with funny proposals about transformers to increase impedance to free space impedance (I know what that is and where to use it, and why antenna impedance is different from that, but obviously you don't) and funny PNP amplifier sketches (10pF||2mH !? 10k load!?) that are not even remotely related to the OP's topic. What value do you think this adds to the thread?

 

[D.A.(Tony)Stewart]

If you understand my design, you can scale all components including L down to 50R out by knowing feedback R/Av=~Zo thus 1M becomes 50k for gain of 1000 or 5k for gain of 100. Assuming impedance of L is much greater at low frequency breakpoint.

Filter for 1st order is simply RC, 2nd order RLC using 50 ohms for L and C at breakpoint with ant RLC nomograph or calc. Or online calc. For Nth order rolloff number of LC component sum =N and then you have choices for maximal flat, linear phase etc.

In all cases DC current must be ~ 2x Vpp out into 50R load or if optimized Ic bias > (Vpp+2V)RL

i think this adds a lot of perspective if you have been following OPs threads for last year.

10pF was shown in Sim for stray capacitance reasons, your mileage may vary, depending on SRF. Adding more becomes a capacitance multiplier to assist with fBW breakpoint at low Zout.

 

[volker@muehlhaus]

Filter for 1st order is simply RC, 2nd order RLC using 50 ohms for L and C at breakpoint with ant RLC nomograph or calc. Or online calc. For Nth order rolloff number of LC component sum =N and then you have choices for maximal flat, linear phase etc.

Buzzword bingo again. Sorry, but if you suggest RC then obviously you have no clue about applied RF filter design. Please, stop spamming the threads with trivial non-information.

 

[D.A.(Tony)Stewart]

If in doubt ask, I can back up anything I say with 40yrs of success.

Any filter design must include source impedance, load impedance, pass band ripple and stop band reject.
As a minimum.


For primitive RLC filters using voltage sources with equiv, R series source or current sources with output R , see ...


**broken link removed**


or in this case undefined load, specs req'd as per above.

 

[volker@muehlhaus]

If in doubt ask, I can back up anything I say with 40yrs of success.

Yes, maybe in a management position this buzzword bingo is just the right level. But here we where looking for more detailed, accurate answers.

Any filter design must include source impedance, load impedance, pass band ripple and stop band reject.

Sure, trivial, and that's why this thread started with an investigation of the source impedance.

For primitive RLC filters using current sources with no R source or voltage sources with output R and no load see

Trivial and unrelated to the topic, as we are in RF with finite load and source impedances, and filter that are (ideally) built from lossless L and C. Some way back on the first page, we had already discussed various design/synthesis tools that will help the OP to calculate component values.

 

[D.A.(Tony)Stewart]

The crux of several problems is lack of clarity on specs and vague interpretations of solutions of Class A RF amps. i introduce a general purpose high gain variable Zin, Zout method. There are many more all dependant on specs like Pout, Av, 3IP, P1dB AND filter specs.

Too often questions are myopic and failure to show overall requirements.

Sorry but my acronyms are universal.