Column: Embedded design
Third order effects By Myk Dormer, RF Engineer and Director, Smallwireless Limited I
n a sometimes embarrassingly long career in RF engineering, there is one thread of commentary that I heard right at the outset and that I am still hearing (in an updated form) to this day. It goes like this:
When I was a student – and later as a graduate engineer – it was “You just calculate the theoretical circuit parameters, and some technician can build the hardware”. As the years passed by, this axiom
slowly changed to “Design your circuit in the simulator, and it’ll work fi rst time”, leading eventually to (in our new era of increasingly sophisticated AI tools) “Just set the prompt right, and the magic results box will hand you a completed design”. Diff erent words but the message is
the same: Get the numbers right (or get your computer to produce them for you), and the perfect result will follow. But it doesn’t. I still spend most of my working hours at the bench with a soldering iron in my hand, or with a much modifi ed (read “mangled”) snarl of prototype circuits attached by multiple coaxial cable tails to a panoply of exotic test equipment. I’m modifying circuits (that should need no changes), hunting defects (that should not be there) and measuring parameters (that ought to be calculated). Am I some kind of technological
hermit? A Luddite, ignoring the march of progress? No. T is is how I have always worked,
and how engineers like me continue to work, throughout the industry. T ere is one overwhelming reason: T ird order eff ects.
12 March 2026
www.electronicsworld.co.uk
I still spend most of my working hours at the bench with a soldering iron in my hand and a snarl of prototype circuits
What could go wrong? It is entirely possible (considering the imperfections of the human mind) to make genuine circuit errors at a conceptual level. Likewise, simulators are not perfect, PCB layout errors are not that uncommon and simple assembly faults do occur, but all those issues combine to make up a surprisingly small proportion of failed designs and under-par performance results. Far more prevalent (and responsible
for the use of far more engineer-bench- hours) are the eff ects that no simulator or process of calculations can predict. I am referring to eff ects that are caused by infl uences outside the actual circuit under consideration, such as:
• PCB layouts that introduce coupling between input and output or between sub-systems.
• Ground plane currents, inducing noise or spurious signal coupling between circuits.
• “Box modes” where adding a screen introduces coupling between stages.
• Out of band mismatch, producing unpredictable RF eff ects.
• Real world PCB via hole impedances and parasitic capacitances.
• Sensitive circuitry (such as VCOs) picking up noise from… pretty much everywhere. T ere are plenty more, but in every case
the circuit is “correct according to theory”. It may even be a proven sub-system that
Page 1 |
Page 2 |
Page 3 |
Page 4 |
Page 5 |
Page 6 |
Page 7 |
Page 8 |
Page 9 |
Page 10 |
Page 11 |
Page 12 |
Page 13 |
Page 14 |
Page 15 |
Page 16 |
Page 17 |
Page 18 |
Page 19 |
Page 20 |
Page 21 |
Page 22 |
Page 23 |
Page 24 |
Page 25 |
Page 26 |
Page 27 |
Page 28 |
Page 29 |
Page 30 |
Page 31 |
Page 32 |
Page 33 |
Page 34 |
Page 35 |
Page 36 |
Page 37 |
Page 38 |
Page 39 |
Page 40 |
Page 41 |
Page 42 |
Page 43 |
Page 44 |
Page 45 |
Page 46 |
Page 47 |
Page 48
