Using 3.579 crystals as crystal oscillator and filter for a PSK31 receiver/transceiver is well known. Getting a decent sideband rejection with a simple circuit is not easy with this design.
Here is a thought : One of the reasons for the lower sideband rejection is the proximity of the 'BFO' frequency to the signal frequency. Possibly oscillator phase noise is contributing to the problems. Here is a trick for a simple QRSS receiver for 3579 :
N4ESS has crystals on 3575.6kHz this is quite a bit further from the signal frequency than is easily done with 3579 crystals. OK, the beat (AF/'IF')then is at about 4kHz, but using SpectrumLab that is not really a problem, since it can 'convert' and filter signals. Of course, the input filter as in Joachim's 30 QRSS Grabber RX is necessary to get some unwanted sideband suppression.
This should make for a simple design for a DC RX Grabber for the color burst frequency on 80m.
Edit : I looked in the 'stock' and I have a 3575.6 crystal (and of course 3579.5 as well, so building this rx for color burst rx is possible. Other projects come first, such as finishing the setup of the 80m SDR from Box73.
The same concept should be usable with a 160 RX at 1843kHz with a 1856kHz Xtal, also available at N4ESS, albeit with a somewhat higher 'IF' of 13 kHz.
"The Art of Electronics" #8 -- Why Not a Simple Emitter Follower as The AF Output Circuit? - Click on the image for an easier read *Back when Dean KK4DAS and I were trying to find a suitable AF amplifier circuit for our High School Direct Conversi...
3 hours ago