Question

te

ea

e.

ked > >>>> at the via, so that the return current flowing in the ground-pl anes

were > >> always conical, and couldn't be more that fifteen times deeper th an they

dly, > >> you can't often have your via flipping the trace from one side to the

the > >>capacitors that allow the two ground planes to be at different DC v oltages.

t.

Vias aren't lossy - losses mostly happen in the pcb material - but they do represent impedance discontinuities. You also have to worry about dispersio n in microstrip traces (on the surface of the board). Buried strip-lines ar e non-dispersive.

Not everything that Howard Johnson fusses about is nonsense. Winfiled Hill did recommend his book when it first came out - not because it was good, bu t because it was pretty much the only book available that got into the deta ils of the problems of high speed logic layout.

The "huge" capacitance is between substantial areas of ground plane. If you r return path has to spread itself over a couple of square inches of ground plane to find enough capacitance through which to move between the ground planes you are going to see a significant impedance discontinuity, giving y our brutal active equalisation a bit more work to do.

Probably not a useful observation.

I made a directional fluid / air flow meter, using an interleaved pattern configured as a bridge. Worked really well, although the resistances were a bit low of course :)

I also tried a capacitative sensor for fluid water content, made of PCB tracks. That sort of worked although it turned out the base fluid was conductive which messed up that particular design.

There are umpteen other capacitative things, touch sensors, keys, sliders. I think the digital micrometers work like that too.

All the microwave stuff which looks like a speck of transistor surrounded by weirdly shaped PCB foil patterns, half of them not even connected to anything.

I have seen what can only be spark gaps in commercial offline switcher and telecom designs. Pointy ended tracks cleared of solder resist.

Planar transformers.

I have used them for various essentially mechanical parts, due to familiarity and the low per-square-inch pricing model.

Someone here suggested making front panels out of them. (How is it I can get a beautiful fully finished, double sided PTH board for $20 or $30. Yet a printed plastic overlay or panel can be 10-20 times that, made by what could surely be fundamentally analogous processes. Where are the low cost chinese prototyping shops for these?)

Radio amateurs make RF-tight boxes from double sided PREss.

The top and bottom panels are as long as the long side walls, but about 2 cm wider than the short side walls. The short side walls are soldered about 1 cm from the bottom panel long side and as well as centered along the long axes, leaving about 1 cm on both ends. The short side walls are soldered to the base plate at both sides.

The long side walls are installed against the short walls (extending the whole length of the base plate but slightly less from the sides. The long side panels are soldered to the bottom plate on both side as well as soldering the short side panels to the long side panels on both sides (about 1 cm from the bottom panel end).

Finally after installing the electronics within the box and testing it, the top panel is soldered to the long and short side walls. This is the only seam that is soldered only on a single side. Of course, to improve shielding, fingerstock can be used on the inside to improve conductivity between the inside of the top panel and the inside of the side panels.

To complete the job, put some copper tape over the exposed PCB edges and solder the tape to the top and bottom side of the PCB.

Also various forms of microwave patch antennas can be made with PCBs.

On lower frequencies, double side teflon boards are quite reasonable high voltage capacitors for transmitting magnetic loops.