Regenerative preselection to improve HF dynamic range?

Main forum to discuss RTL-SDR related topics.
Post Reply

How familiar are you with regenerative receivers?

I have built several regenerative receivers, regenerative antennas, and/or regenerative amplifiers of my own design.
I've built one or two regenerative receivers or regenerative antennas following others' designs.
I know what regenerative receivers are, but I've never bulit one.
I am not familiar with regenerative receivers.
No votes
Total votes: 3

Posts: 1
Joined: Sun Aug 06, 2017 9:05 am

Regenerative preselection to improve HF dynamic range?

Post by qrp-gaijin » Sun Aug 06, 2017 9:36 am

I searched this entire forum for the word "regenerative" and found no hits, so I suspect my idea might come off as rather strange, but: I was thinking that a regenerative preselector might be one way of overcoming the limited 8-bit dynamic range of the RTL-SDR devices. My interest is HF reception using an upconverter. The purpose of this post is to solicit feedback on this idea.

Some background: A regenerative preselector is basically a Q-multiplier at RF. A Q-multiplier is a positive-feedback oscillator whose feedback has been throttled back through some user-controllable device (like a potentiometer or a variable capacitor). The feedback is adjusted to be just below the oscillation threshold. In this condition, the Q-multiplier is extremely sensitive to signal energy at its resonant frequency because the amplifier's gain has been adjusted to exactly cancel the LC tank's losses. So any small signal energy at the LC tank's resonant freuqency will be repeatedly amplified in phase (with some unavoidable amplifier noise, of course), leading to greatly amplified signals at the resonant frequency. Entire receivers have been built on this concept, and they are called regenerative receivers.

The motivation to use regeneration with the RTL-SDR is as follows. As we know, the 8-bit ADC resolution limits the dynamic range to be 48 dB. If we set the LNA gain too high, spurious mixing products will rise up out of the noise. This can be fixed by reducing the LNA gain, but then weaker signals sink back into the noise floor.

So... to receive those weak signals, why not have a selective amplifier that amplifies only the frequency of interest, boosting it up out of the noise?

Example: say we've tuned our RTL-SDR to hear from 7 MHz to 9 MHz. There are 2 shortwave bands in this range, and signal levels are high. We turn down the LNA gain to reduce spurious mixing products. Say there is a weak CW signal at 7003 kHz that we want to receive. We place a regenerative preselector (Q-multipler) between the antenna and the HF upconverter input, tune the preselector to 7003 kHz, and set the feedback level to be almost-oscillating. In this condition, a Q-multiplier should be able to deliver up to 40 dB of highly selective (albeit noisy and unstable) gain. Of course, we can also back off the feedback a bit to reduce the gain. That up-to-40-dB gain boost at only the desired frequency should help boost the desired signal back up out of the noise floor, where it is then receivable by the RTL-SDR. Of course, the regenerative preselector (Q-multiplier) itself has a rather poor dynamic range -- strong nearby signals will inhibit regenerative amplification and will introduce spurious mixing products. It is for this reason that better regenerative receivers always have an input attenuator to control the signal levels coming into the regenerative amplifier.

It's certainly not a perfect solution, but it seems that regeneration -- i.e. highly selective gain -- might be one tool that could be useful to help push weak signals back up out of the noise floor and into the usable dynamic range of the RTL-SDR.


Posts: 21
Joined: Fri Jan 19, 2018 10:37 pm
Location: Netherlands, Nuenen

Re: Regenerative preselection to improve HF dynamic range?

Post by Brightnoise » Sun Jan 21, 2018 10:03 pm

This can be used for sure. The basic calculation is as follows: starting with an LC-filter with LOADED (antenna is connected!) Q-factor (quality factor) of 100 makes a -3dB top bandwidth of 70 kHz. The attenuation for off-frequency signals is -3 dB at (+ and -) 35 kHz, -6 dB at 70 kHz offset, -12 dB at 140 kHz, and following that first-order slope further. The feedforward gain can be driven up to a Q of 1000. My experience is that regenerative feedback can not be used far over Q-factor 1000 for stability reasons. So the practical selectivity will be a first-order system with top BW of 7 kHz and an effective selectivity of 35 dB for 7205 kHz; I assume that is your fear-frequency. The problem is that with Q-magnification also the noise in the pass-band goes up: good design is a prerequisite. Finally, also the IM3 is multiplied. Succes when you continue with this idea.

P-40 Warhawk
Posts: 92
Joined: Thu Jun 26, 2014 11:55 pm

Re: Regenerative preselection to improve HF dynamic range?

Post by P-40 Warhawk » Sun Jan 21, 2018 10:45 pm

Mfj sell something called a active antenna model number 1020c . It is a Amplified preselector , the range of the gain is -10db to +10db . it works very well with the V3 dongle . Frequency on it is from 300khz to 40khz . The price is just under $100.00 . You can find them used on Ebay for less .
MFJ1020B.gif (11.66 KiB) Viewed 3272 times

Post Reply