In article , "Ian White, G3SEK"
writes:

Michael Black wrote:

Perfboard was around long before most hobbyists had heard of wire wrap.

Once transistors came along, perfboard tended to be the method of
construction, unless you were planning to do it point to point on a
chassis like you would with tubes. "Messy" has no relevance since the
wiring was under the board. You'd use component leads to connect the
parts, and if they weren't long enough, bits of wire.

That's right...

Maybe it would help to give an example where plain perfboard was the
best construction method. I recently built a small power distribution
board which involved four 0.1in-pitch connector headers, a voltage
regulator IC and a few passive components.

Why choose plain perfboard for this project? Mostly because the wiring
underneath was going to be fairly complex, so hand-wiring was going to
be the easiest way to make the board reasonably small. The advantage of
hand-wiring is that you can have as many crossovers as you like, using
insulated wire.

What other techniques didn't I choose? Even a double-sided PC board
would have needed some jumpers, and wouldn't have been worth the effort
for a one-off project. A single-sided PC board would have required lots
of jumpers, and copper-strip perfboard (Veroboard/Vectorboard) would
have been even worse.

So bare perfboard it was. The four headers were temporarily superglued
onto the board, so they wouldn't fall out when the board was turned over
to do the wiring. Under the board, I ran a common ground bus of
18SWG/16AWG tinned copper wire around all the headers. The other small
components were pushed through the holes, and anchored by their leads as
the various connections were made.

Most of the wiring was done point-to-point using the bare component
leads. Crossovers were handled using kynar insulated wire - sold for
wire-wrapping, but excellent for point-to-point soldered wiring too. By
the time all the header pins had been soldered to, the headers were well
anchored through the board and are very secure.

The whole thing looks quite tidy from the top. If neatness is important
(and indeed, why not make a good job of it?) the main consideration is
to cut the board cleanly and file off the ragged edges.

Ian, I'm in perfect agreement with what you and Michael Black wrote.

I base that on working prototypes made for over three decades. It
seems as though perfboard has been around longer...:-)

One thing that all should remember: Electrons don't care about "neat"
construction. "Neat" gets something past inspectors, appeals to
customers, looks mighty fine and "professional" in photographs.
Electrons don't care for any of that. Fields and waves only care about
placement of conductors and nearby dielectric material.

And the ones with bits of copper at each hole had the advantage, as
someone pointed out, that you could solder the components to the board
before adding the wires.

People were hesitant to make their own etched circuit boards, and even
after they became fairly common in hobby circles, many would say "I
don't want to bother" and they'd stick with perfboard.

Perf and hand wiring SAVES TIME. Homebrewing is about making
electronic things at home. Few of us have TIME at our disposal.

For relatively simple circuit arrangements, it would take me (somewhat
experienced) more time to sketch out a PCB foil pattern and check it
against a schematic than direct-wiring a perf-board circuit. A PCB
still has to be masked and etched and cleaned.

Printed circuit boards originally were a tremendous advantage in mass
production of electronics. It could physically hold components as well
as perfectly reproduce wiring paths...compared to the longer time needed
to mount (vacuum tube style) components and then hand-wire all the
connections, usually in several stages of wiring by different assemblers.
Production costs dropped considerably once all got through the
capital equipment expenditure for PCB machinery.

SMT got a big following because it was small...but also because it
cut down on production even more. Fewer holes to drill (sometimes
none) and the "baking" ovens didn't need as much maintenance as
the flow soldering machines did with peanut oil, cleaner and degreaser,
solder solutions. SMT took the basic PCB construction technique.

Certainly... and as I just said, it's one of the preferred techniques
for one-off construction.

If perfboard faded from view in recent years, it's likely because
etched circuit boards have tended to take over, if not made by the
hands of the hobbyist then because people were buying premade circuit
boards.

Maybe plain perfboard just hasn't had much publicity. As an author, I
can see why: if it's a fairly basic 'follow these instructions' project,
then at the very least you're expected to design a PC board; but if it's
a project for experienced constructors, you don't need to tell them how
to build it. In either kind of article, plain perfboard probably doesn't
get a mention... but it's still there.

Perf with hand-wired "solder side" connections doesn't photograph
well. Image is important in magazine layout. :-)

As someone pointed out, perfboard is not so great for RF. Better to
use a blank piece of copper circuit board, and build on top of that, so
you have the copper as a good ground point.

Yep. The strengths of plain perfboard are for DC, audio and slow logic
circuits, and in the sheer versatility of hand-wiring underneath the
board.

Right on! But...if one is familiar with RF layout through experience
and a "feel" for lead placement and lengths, perf can be used on
up to 70 MHz. That is NOT recommended for beginners who have
just memorized Ohm's Law.

At RCA EASD in the 1970s we regularly used Douglas Electronics
11-DE-1 boards for prototyping digital logic at equivalent clock rates
up to 10 MHz. Those held 12 DIPs had three-hole pads and #24 or
#26 insulated wire did the interconnects on the component side. A
44-pin edge contact allowed plug-in assembly and extenders. That
was most excellent for churning out 1- to 3-of-a-kind prototypes at up
to 150 cards per system. Techs could just follow a schematic and
hand-wire the cards...sort of "connect the dots" kind of operation.
Was very good with few errors done by anyone.

Douglas is still in business and the 11-DE-1 prototype board is still
available in the USA at about $12 each. A comparable Vector
Electronics prototype board is about the same price. I prefer the
Douglas since it is gold-flashed and a variant is available with a
ground plane on the component side.

But that's likely "messier" than using perfboard with the wiring on the
bottom.

Of course, from time to time I've used copper circuit board as
"perfboard", drilling holes where needed and reaming out the copper
from around the holes so the components go in without shorting. Wire
up on the non-copper side of the board, like perfboard, but all ground
connections go to the copper on the top side.

If all the grounds are to the common top-side groundplane, it doesn't
look messy at all. The top-side looks very neat and the wiring
underneath is simplified by the absence of ground wires. Above all, the
RF performance will probably be very good.

For this kind of project, I very often use PCB design software to work
up the layout, but don't always etch a board. Just as often, I cut a
piece of single-sided board, tape a 1:1 printout of the PCB design onto
the board and use it as a drilling template. Then I hand-wire the
underside using the layout as a guide.

I cheat a bit. Originally an illustrator, I sketch out the foil paths on
vellum 1:1, mark the drill holes and use the vellum as a small
center-punch guide. The paths are then painted in with lacquer,
free-hand, using the vellum as a guide. Lacquer can be peeled off
after etching with a sharp graphics knife (an X-Acto modeller's knife
to olde-tymers) and dissolved in acetone to be reused later.

Lacquer works fine with ferric chloride etchant. Enamel does NOT.
Enamels, now common in aircraft model shops can be undercut in
the etching. One has to really hunt for low-enough-price lacquer
now but it once was standard "dope" for modelers. I'm still using
a half-pint bottle of Testor's model airplane Dope obtained in 1947
(Testor's wisely changed the name slightly after that). Acetone in
pint can sizes is available in U.S. do-it-yourself stores. :-)

Len Anderson
retired (from regular hours) electronic engineer person