21 Mar 2013

Phasing direct conversion receivers

What I think is quite an old paper turned up in my browsing today: http://www.norcalqrp.org/files/austinnc2030presentation.pdf . In it, Dan Tayloe N7VE describes how a direct conversion design with phasing to cancel out the unwanted sideband can be a truly excellent receiver. I have never tried this approach myself, but think it would be worth it for, for example, a WSPR receiver where the phase accuracy has only to be maintained over a 200Hz audio bandwidth. Over such a small audio bandwidth a very simple phase shift network would work.The same applies on TX where a very simple final frequency phasing TX could produce quite good sideband suppression with very few parts.

2 comments:

G1KQH said...

John Hey G3TDZ is the king of phasing RX and TXs. Search for the Heyphone, White Rose reciever etc:

http://bcra.org.uk/creg/heyphone/introduction.html

His designs have been published in Sprat Radcom over the years.

73 Steve

G1KQH


Anonymous said...

OK... This NC2030 design is interesting, but by no means a panacea...

* The direct conversion commutating mixer (other's prefer the term "Tayloe Detector") has excellent attributes, as Mr. Tayloe points out in the presentation. But there are drawbacks, especially when it comes to your antenna match. The receiver gain and operating bandwidth are a function of how well your match is. This is covered in Youngblood's "A Software-Defined Radio for the Masses" four-part series in QEX (2002-2003) which is available for free download on the ARRL site.

* The 90-degree phase-shift (or "All-Pass") networks are notorious for difficult design, tight tolerance requirements and finicky carrier suppression if they are not feed and cared for properly.

* The design employs a Switched Capacitor Analog Filter (SCAF), which also looks like some sort of panacea on paper. But SCAF's are cranky if not fed with relatively stable high level audio. Also, you typically have to be careful with your layout and grounding to keep the SCAF clock from getting into other sensitive parts of your circuit. At first glance this design does not have any AGC/AVC at all - not a good choice when using a SCAF in my opinion unless you want to be a constantly jockeying around with your volume pot. A SCAF can drive you crazy on 80m or 40m on a noisy night unless you have strong finely tuned AGC (granted this design is for 20/30m).

Just my two cents... David