Certainly for anything with higher pin counts it is. I've seen untold numbers of folks who hand solder QFN devices on adapter boards and certainly QFPs and similar are no real problem. But with the higher pin counts adapters are just too difficult.
I don't typically do designs where there is a lot of uncertainly in the basic design. If there is much uncertainty I do it a different way, like by wiring eval boards together.
You have to plan on two spins of the board though as there is almost always something that will need changed. Still, PCBs are a lot cheaper than they used to be.
The 276-170 that I use is Perfboard set up for holding a line of DIPs. It h as five pads in a line on each pin, and a VCC/GND trace. Most of what I do is slow, so 74HCT and Dip packed Op-Amps and Micros work just fine with it . I talked them into a discount on a bag of 100 of them a few years ago.
At higher frequencies I run a ground plane down the middle of the dash-170 perfboard using brass shim stock.
Otherwise my prototypes look like Mr Larkin's, only on a lower budget.
I've built avalanche pulse generators and 2 Ghz capacitive sensors dead bug . Thus I sing the praises of hamfest grade thin PCB stock. I've also used a diamond dust based hole saw to cut pads. The diamond drill leaves a nice thick ring of insulation around a 1/8th inch or 1/4th inch pad.
Every once in a while, to test something simple, or a simple circuit. (But I usually regret it, when some wire/connections gets flaky.) Then it turns into live bug air wires over copper clad.
It really depends on what your goals are and what you are doing. Having "early foil" is often a win -- it lets manufacturing take a look at what will (eventually, in some form) be coming down the road. It lets them buy into the design at an earlier stage instead of feeling like it is just *dropped* in their laps while the marketing folks clamor for shipped product to satisfy orders, etc.
It also lets you play with a design in ways that would be too tedious with other implementations: lift foils, add components, connectors, test points, etc. It's one thing to trash a board and "expense" the components that were on it; quite another thing to deal with all the labor lost on a "one off" breadboard!
[I had a colleague "debug" a motor driver with the smoke-n-smell technique. While the boss wasn't pleased with the number of components that gave up the ghost in the process, he was *much* happier knowing a new board would only set him back about an hour in labor costs!]
I haven't seen an inexpensive metal printer. Better might be a photo resist printer if it could handle the fine lines often needed. I'm afraid that a Dremel tool isn't quite the surgeon's scalpel for placing
0.5 mm pitch quad flat packs. lol
Even so, etching a PCB is a PITA and usually the stuff I do needs to be multiple layers. For PCBs one and even two layers is a pretty limiting technology. Then there is the assembly work. I don't know that I can solder anything small anymore. My hand is far from steady these days.
I would always brace my hand on the table, a very solid base is important to a good shot. If you can do that in the air my hat is off to you.
My eyes get steadily worse as well. But then an 88 year old friend has macular degeneration and deals with it very well. He can still read with difficulty and is writing an article on the post Civil War canteens he collects. So I can't feel too bad...
Have you tried 3D printing yet? I've noticed that the most vocal advocates of 3D printing are those that have never tried it. I got my first experience at a local "makers" shop: Mostly cardboard thickness sheets that were assembled into something artsy. Then, my neighbor bought a DaVinci 3D printer version 1.0: Of course, I had to play (after he was done making dinosaur skeletons and robot gears). So, I measured and made a replacement fan rotor for a video card that I broke. The fan eventually ran for about a minute before it flew apart.
The short summary is that 3D printing could probably be used to make a PCB if you don't mind secondary operations and complex CAD drawings. There's no way to get a totally smooth finish. Tiny holes tend to fill in when the plastic gets hot and reflows. It's slower than a snail and makes far too much irritating noise. The final product is rather brittle. Run to slow, and it melts and sags. Run too fast, and it crumbles to dust. However, I suspect all of these problems can be solved with a specialized 3D printer designed for making PC boards.
What would be cool is 3D printing with metal: Multiple nozzles for substrate, conductive traces, and even resistive material. With the ability to lay down very thin layers of insulating plastic, maybe even capacitors. Of course, 3D means I pile components on top of each other vertically, making a truly 3D circuit board. For really thin products, such as cell phones, the component leads would all touch each other, and the PCB would become a skeleton frame that simply supports the component packages.
Oh well... back to the dull and boring reality of bookkeeping and billing.
--
Jeff Liebermann jeffl@cruzio.com
150 Felker St #D http://www.LearnByDestroying.com
Santa Cruz CA 95060 http://802.11junk.com
Skype: JeffLiebermann AE6KS 831-336-2558
Yeah, like I said, I always do PCB from the start. This one just has a few more white wires than usual. Good thing he had all those conveniently located vias. lol
Like the flexible Werner von Braun, according to Tom Lehrer:
"You too can be a big hero / Just learn to count backwards to zero. 'Once the rockets go up / who cares where they come down? That's not my department' / says Werner von Braun"
Cheers
Phil Hobbs
--
Dr Philip C D Hobbs
Principal Consultant
ElectroOptical Innovations LLC
Optics, Electro-optics, Photonics, Analog Electronics
160 North State Road #203
Briarcliff Manor NY 10510
hobbs at electrooptical dot net
http://electrooptical.net
I find that I can do dead bug with 50-mil pitch parts (such as SOICs and SOT23s) by staggering the leads like saw teeth. The main problem is that the leads are so fragile, whereas with DIPs you can just solder a resistor to a pin and bend it anywhere you like without worrying.
Anything smaller than 50 mils is extremely difficult, so breakout boards come in very handy. For devices where those don't work well (e.g. SC70 microwave transistors), it's sometimes possible to wire them up dead-bug fashion with 0603s in midair, but that's very fiddly.
Cheers
Phil Hobbs
--
Dr Philip C D Hobbs
Principal Consultant
ElectroOptical Innovations LLC
Optics, Electro-optics, Photonics, Analog Electronics
160 North State Road #203
Briarcliff Manor NY 10510
hobbs at electrooptical dot net
http://electrooptical.net
Perf board is way too slow, I find--it takes at least twice as long as dead-bug when both methods apply. All that cutting and stripping of wires takes ages, and wire-wrap wire is too easy to nick.
Also the thickness of the board takes up vertical space that's very useful for making connections. With dead bug, you just use clipped-off component leads for hookup wire, and you can easily get three vertical layers of connections without shorts.
One layer is pins bent directly down to the package (e.g. shorting pins
2 and 6 of an op amp follower), the next is where the leads neck down, and the top one arches over that one. You can make the layout pretty tight if you use the bodies of resistors and capacitors to force the wires apart.
So I use perf board only rarely. When I do, it's Vector 8007 (pad-per-via plus ground plane).
Cheers
Phil Hobbs
--
Dr Philip C D Hobbs
Principal Consultant
ElectroOptical Innovations LLC
Optics, Electro-optics, Photonics, Analog Electronics
160 North State Road #203
Briarcliff Manor NY 10510
hobbs at electrooptical dot net
http://electrooptical.net
ElectronDepot website is not affiliated with any of the manufacturers or service providers discussed here.
All logos and trade names are the property of their respective owners.