back
10 comments
Ugh WRT NE602 based designs the article says it all "The circuits using diode rings continue to offer the better performance."

Worst thing to ever happen to direct conversion circuit topology was widespread adoption of the NE602 in hobbyist designs in the 80s and 90s leading to the belief that all DC receivers have unusable problems with hum, intermod distortion, miserable strong signal handing in general, etc. Given non-awful components (not a NE602), DC is quite usable.

Hmmm.

I built a couple of 40-meter band direct conversion receivers in high school in the 80's. One of them became the basis of a Westinghouse project. Since all my parts came from either Radio Shack or desoldered from scrapped electronics, I had no IC's to work with. Never thought about it before but I wonder how different my experiences would have been if I had to deal with stuff like that. It was hard enough getting what information I needed to build the projects I did. Thank heaven for easy access to the Brooklyn Public Library!

What would you suggest as an alternative?

Is the SA602 or SA612 any better?

Diode ring, just like the linked article.

The cool part of the gilbert cell ne602 series (and similar) is its one chip with an oscillator and a mixer and the performance is barely good enough for 1980s analog cordless phones at 49 mhz with small antennas over a short range. However intermod and strong signal performance is dismal when connected to a large 80M dipole or whatever in some ham radio application. A diode mixer will require some external (IC?) oscillator of course but performance will be vastly better.

The "new era" of design as seen in some SDRs seems to be a nice DDS synthesizer driving something like a Tayloe mixer (four very fast analog switches in an IC).

I don't have the numbers in front of me but from memory the 3rd order intercept point (higher means less self-generated interference for strong signals) is ridiculously higher for even a cheap analog switch IC as a Tayloe mixer.

Of course the power consumption of a DDS and a fast switch will be a little higher (10x? 100x?) than an analog cordless phone chip. There's always tradeoffs.

more info on the Tayloe mixer for those interested:

https://wparc.us/presentations/SDR-2-19-2013/Tayloe_mixer_x3...

Almost all SDR's are direct conversion (although direct sampling is starting to become popular now that reasonably priced high speed ADC's are becoming available).
True, although the cheaper ones have a pretty significant spike at DC and you end up using a tuning offset and IF in software instead. Not the case across the board, but pretty common.
The significant spike at DC is an inherent issue with Direct conversion receivers, driven by circuit level DC offsets and IP2 contributions. Offset tuning works, but you're stuck with sucky image rejection (30 to 40 dB at best)
I built a simple receiver circuit called the Zeta SDR and then used a 192 KHz sample rate USB audio device to do the A2D... Combined with an antenna tuner I received PSK31 signals from France in Ohio... The antenna was a mains power extension cord I had lying around alligator clipped in and strung up indoors.

Quite possibly the craziest and most satisfying electronics project I ever did and it was like 10 parts or something.

Wes Hayward is an awesome guy. His "Introduction to RF Design" may be of interest: https://www.alibris.com/Introduction-to-Radio-Frequency-Desi...