increasing dynamic range of receiver

hello I am building a direct conversion HF receiver (like R1 and R2 ones) that uses a mini circuits mixer with high dynamic range. By connecting the mixer directly to the antenna the dynamic range is at maximum.

If I connect an amplifier at the input, to increase sensitivity, this decreases dynamic range.
I am wondering, what if I try to make a logarithmic amplifier? It will have the same effect as the AGC I guess but what is going on with the dynamic range then?

What exactly are you trying to do?

I am trying to avoid ruining the good dynamic range of the receiver, whereas at the same time achieve amplification before the rf mixer.

most log amplifiers only have a dc output, so how is that going to get thru your mixer?

I was unaware of that. So they cannot be used for RF.

Any other ideas about performing rf amplification before the mixer without affecting its dynamic range too much?

They are used as power detectors, or in some cheap receivers as AM detectors.

A very few have an RF output, but it is a highly saturated/clipped waveform, so unless you are doing fsk demodulation it would be unusable.

After you have made your components have as big a compression point as possible, the next step is to add a variable gain (AGC) stage so that you can kill the gain in the amplifier for big signals.

Yes, but at the same time you lose dynamic range. with AGC switched on it is impossible to detect nearby"weak signals in the presence of a strong signal, since the amplifier decreases it's gain. I am talking about dynamic range as far as concern simultaneous distinction of strong signals and weak ones.

DR is two types 1) instantaneous 2) Extended.
All want more DR but is mainly limited by ADC, So have resinable gain and its control based on the max signal to min sensitivity.

yes, the DR should be determined by ADC: by its noise floor and by its p1dB.

If you use a pair of 2n3866s in push pull with WB transformer input and output,you should be able to get a gain of 10dB right across 50KHz->30MHz with superb linearity, able to handle 1v+ input levels with extremely low cross mod. You need ~ 24 Vcc and run them at about 50->100mA. I use to work on stuff like this and its possible!
Frank

Having more gain is nothing but decrease in DR.

From the articles I have read I have concluded to this too, without being very sure.

By the way, what If I add an LC bandpass filter before the mixer? does it affect its dynamic range (for the certain band that is being passed through of course)?

This is a rather bad statement to make, the amplifier that I described, had Xmod below measurement level with a 1V and a micro volt signal. The problems we had was to actual combine the signals without introducing any. We used a pair of E88Cs (?) ultra high mu triodes sharing a common 75 ohm anode resistor. To cope with > 1V rms input a diode attenuator was fitted which caused the core of a ferrite transformer to firstly saturate hence limiting the output voltage from it and after a short delay the core got beyond its curie temperature and the power feed through was even lower. As part of testing you could switch one set to transmit (10W) and touch its aerial to the aerial of a set set receive and no harm came to it. Whoops, that kind of give away where this kit might have been used.
Frank

No.
Probably you ar talking about SFDR. which relates to IMDs. Having BPF willl limit the blocker, Images to certain extent and increase your SFDR. I guess Chuckey is giving information on the linear IF amplifier to increases SFDR from his experience.

In extremis, the dynamic range is determined by the cross mod by signals that fall just outside its notional pass band. I.e. for a SSB Rx (bandpass +- 3dB = 3KHz) would be at say +-10 KHz from its centre frequency, just enough to be outside the crystal filter's bandpass, and not at +- 10MHz where any RF selectivity would have a large effect.
Frank